HANDBOOK OF AIR CONDITIONING AND REFRIGERATION
Shan K. Wang
Second Edition
McGraw-Hill
New York San Francisco Washington, D.C. Auckland Bogot?
Caracas Lisbon London Madrid Mexico City Milan
Montreal New Delhi San Juan Singapore
Sydney Tokyo Toronto
Library of Congress Cataloging-in-Publication Data
Wang, Shan K. (Shan Kuo)
CONTENTS
Preface to Second Edition xi
Preface to First Edition xiii
Chapter 1. Introduction 1.1
Chapter 2. Psychrometrics 2.1
Chapter 3. Heat and Moisture Transfer through Building Envelope 3.1
Chapter 4. Indoor and Outdoor Design Conditions 4.1
Chapter 5. Energy Management and Control Systems 5.1
Chapter 6. Load Calculations 6.1
Chapter 7. Water Systems 7.1
Chapter 8. Heating Systems, Furnaces, and Boilers 8.1
Chapter 9. Refrigerants, Refrigeration Cycles, and Refrigeration
Systems 9.1
Chapter 10. Refrigeration Systems: Components 10.1
Chapter 11. Refrigeration Systems: Reciprocating, Rotary, Scroll, and Screw 11.1 vii
Chapter 12. Heat Pumps, Heat Recovery, Gas Cooling, and Cogeneration
Systems 12.1
Chapter 13. Refrigeration Systems: Centrifugal 13.1
Chapter 14. Refrigeration Systems: Absorption 14.1
Chapter 15. Air Systems: Components—Fans, Coils, Filters, and Humidifiers 15.1
Chapter 16. Air Systems: Equipment—Air-Handling Units and Packaged
Units 16.1
Chapter 17. Air Systems: Air Duct Design 17.1
Chapter 18. Air Systems: Space Air Diffusion 18.1
Chapter 19. Sound Control 19.1
Chapter 20. Air Systems: Basics and Constant-Volume Systems 20.1
Chapter 21. Air Systems: Variable-Air-Volume Systems 21.1
Chapter 22. Air Systems: VAV Systems—Fan Combination, System Pressure,
and Smoke Control 22.1
Chapter 23. Air Systems: Minimum Ventilation and VAV System Controls 23.1
Chapter 24. Improving Indoor Air Quality 24.1
Chapter 25. Energy Management and Global Warming 25.1
Chapter 26. Air Conditioning Systems: System Classification, Selection,
and Individual Systems 26.1
viii CONTENTS
Chapter 27. Air Conditioning Systems: Evaporative Cooling Systems
and Evaporative Coolers 27.1
Chapter 28. Air Conditioning Systems: Space Conditioning Systems 28.1
Chapter 29. Air Conditioning Systems: Packaged Systems
and Desiccant-Based Systems 29.1
Chapter 30. Air Conditioning Systems: Central Systems and Clean-Room
Systems 30.1
Chapter 31. Air Conditioning Systems: Thermal Storage Systems 31.1
Chapter 32. Commissioning and Maintenance 32.1
Appendix A. Nomenclature and Abbreviations A.1
Appendix B. Psychrometric Chart, Tables of Properties, and I-P Units to
SI Units Conversion B.1
Index follows Appendix B
CONTENTS ix
CHAPTER 1
INTRODUCTION
1.1
1.1 AIR CONDITIONING 1.1
1.2 COMFORT AND PROCESSING AIR
CONDITIONING SYSTEMS 1.2
Air Conditioning Systems 1.2
Comfort Air Conditioning Systems 1.2
Process Air Conditioning Systems 1.3
1.3 CLASSIFICATION OF AIR
CONDITIONING SYSTEMS ACCORDING
TO CONSTRUCTION AND OPERATING
CHARACTERISTICS 1.3
Individual Room Air Conditioning
Systems 1.4
Evaporative-Cooling Air Conditioning
Systems 1.4
Desiccant-Based Air Conditioning
Systems 1.4
Thermal Storage Air Conditioning
Systems 1.5
Clean-Room Air Conditioning
Systems 1.5
Space Conditioning Air Conditioning
Systems 1.5
Unitary Packaged Air Conditioning
Systems 1.6
1.4 CENTRAL HYDRONIC AIR
CONDITIONING SYSTEMS 1.6
Air System 1.6
Water System 1.8
Central Plant 1.8
Control System 1.9
Air, Water, Refrigeration, and Heating
Systems 1.10
1.5 DISTRIBUTION OF SYSTEMS
USAGE 1.10
1.6 HISTORICAL DEVELOPMENT 1.11
Central Air Conditioning Systems 1.11
Unitary Packaged Systems 1.12
Refrigeration Systems 1.12
1.7 POTENTIALS AND CHALLENGES 1.13
Providing a Healthy and Comfortable
Indoor Environment 1.13
The Cleanest, Quietest, and Most
Precise and Humid Processing
Environment 1.13
Energy Use and Energy Efficiency 1.13
Environmental Problems—CFCs and
Global Warming 1.15
Air Conditioning or HVAC&R Industry
1.15
1.8 AIR CONDITIONING PROJECT
DEVELOPMENT 1.16
Basic Steps in Development 1.16
Design-Bid and Design-Build 1.17
The Goal—An Environmentally Friendlier,
Energy-Efficient, and Cost-Effective
HVAC&R System 1.17
Major HVAC&R Problems 1.17
1.9 DESIGN FOR AIR CONDITIONING
SYSTEM 1.18
Engineering Responsibilities 1.18
Coordination between Air Conditioning
and Other Trades, Teamwork 1.19
Retrofit, Remodeling, and
Replacement 1.19
Engineer’s Quality Control 1.20
Design of the Control System 1.20
Field Experience 1.21
New Design Technologies 1.21
1.10 DESIGN DOCUMENTS 1.21
Drawings 1.22
Specifications 1.22
1.11 CODES AND STANDARDS 1.23
1.12 COMPUTER-AIDED DESIGN AND
DRAFTING (CADD) 1.25
Features of CADD 1.25
Computer-Aided Design 1.25
Computer-Aided Drafting (CAD) 1.26
Software Requirements 1.26
REFERENCES 1.26
1.1 AIR CONDITIONING
Air conditioning is a combined process that performs many functions simultaneously. It conditions
the air, transports it, and introduces it to the conditioned space. It provides heating and cooling from
its central plant or rooftop units. It also controls and maintains the temperature, humidity, air
movement, air cleanliness, sound level, and pressure differential in a space within predetermined
1.2 CHAPTER ONE
limits for the comfort and health of the occupants of the conditioned space or for the purpose of
product processing.
The term HVAC&R is an abbreviation of heating, ventilating, air conditioning, and refrigerating.
The combination of processes in this commonly adopted term is equivalent to the current definition
of air conditioning. Because all these individual component processes were developed prior to the
more complete concept of air conditioning, the term HVAC&R is often used by the industry.
1.2 COMFORT AND PROCESSING AIR CONDITIONING
SYSTEMS
Air Conditioning Systems
An air conditioning, or HVAC&R, system is composed of components and equipment arranged in
sequence to condition the air, to transport it to the conditioned space, and to control the indoor environmental
parameters of a specific space within required limits.
Most air conditioning systems perform the following functions:
1. Provide the cooling and heating energy required
2. Condition the supply air, that is, heat or cool, humidify or dehumidify, clean and purify, and
attenuate any objectionable noise produced by the HVAC&R equipment
3. Distribute the conditioned air, containing sufficient outdoor air, to the conditioned space
4. Control and maintain the indoor environmental parameters–such as temperature, humidity,
cleanliness, air movement, sound level, and pressure differential between the conditioned space
and surroundings—within predetermined limits
Parameters such as the size and the occupancy of the conditioned space, the indoor environmental
parameters to be controlled, the quality and the effectiveness of control, and the cost involved determine
the various types and arrangements of components used to provide appropriate characteristics.
Air conditioning systems can be classified according to their applications as (1) comfort air
conditioning systems and (2) process air conditioning systems.
Comfort Air Conditioning Systems
Comfort air conditioning systems provide occupants with a comfortable and healthy indoor environment
in which to carry out their activities. The various sectors of the economy using comfort air
conditioning systems are as follows:
1. The commercial sector includes office buildings, supermarkets, department stores, shopping
centers, restaurants, and others. Many high-rise office buildings, including such structures as the
World Trade Center in New York City and the Sears Tower in Chicago, use complicated air conditioning
systems to satisfy multiple-tenant requirements. In light commercial buildings, the air conditioning
system serves the conditioned space of only a single-zone or comparatively smaller area.
For shopping malls and restaurants, air conditioning is necessary to attract customers.
2. The institutional sector includes such applications as schools, colleges, universities, libraries,
museums, indoor stadiums, cinemas, theaters, concert halls, and recreation centers. For example,
one of the large indoor stadiums, the Superdome in New Orleans, Louisiana, can seat 78,000 people.
3. The residential and lodging sector consists of hotels, motels, apartment houses, and private
homes. Many systems serving the lodging industry and apartment houses are operated continuously,
on a 24-hour, 7-day-a-week schedule, since they can be occupied at any time.
4. The health care sector encompasses hospitals, nursing homes, and convalescent care facilities.
Special air filters are generally used in hospitals to remove bacteria and particulates of submicrometer
size from areas such as operating rooms, nurseries, and intensive care units. The relative humidity in a
general clinical area is often maintained at a minimum of 30 percent in winter.
5. The transportation sector includes aircraft, automobiles, railroad cars, buses, and cruising
ships. Passengers increasingly demand ease and environmental comfort, especially for longdistance
travel. Modern airplanes flying at high altitudes may require a pressure differential of
about 5 psi between the cabin and the outside atmosphere. According to the Commercial Buildings
Characteristics (1994), in 1992 in the United States, among 4,806,000 commercial buildings having
67.876 billion ft2 (6.31 billion m2) of floor area, 84.0 percent were cooled, and 91.3 percent
were heated.
Process Air Conditioning Systems
Process air conditioning systems provide needed indoor environmental control for manufacturing,
product storage, or other research and development processes. The following areas are examples of
process air conditioning systems:
1. In textile mills, natural fibers and manufactured fibers are hygroscopic. Proper control of humidity
increases the strength of the yarn and fabric during processing. For many textile manufacturing
processes, too high a value for the space relative humidity can cause problems in the spinning
process. On the other hand, a lower relative humidity may induce static electricity that is harmful
for the production processes.
2. Many electronic products require clean rooms for manufacturing such things as integrated circuits,
since their quality is adversely affected by airborne particles. Relative-humidity control is
also needed to prevent corrosion and condensation and to eliminate static electricity. Temperature
control maintains materials and instruments at stable condition and is also required for workers who
wear dust-free garments. For example, a class 100 clean room in an electronic factory requires a
temperature of 72  2°F (22.2  1.1°C), a relative humidity at 45  5 percent, and a count of dust
particles of 0.5-m (1.97  105 in.) diameter or larger not to exceed 100 particles/ ft3 (3531 particles
/m3).
3. Precision manufacturers always need precise temperature control during production of precision
instruments, tools, and equipment. Bausch and Lomb successfully constructed a constanttemperature
control room of 68  0.1°F (20  0.56°C) to produce light grating products in the
1950s.
4. Pharmaceutical products require temperature, humidity, and air cleanliness control. For instance,
liver extracts require a temperature of 75°F (23.9°C) and a relative humidity of 35 percent.
If the temperature exceeds 80°F (26.7°C), the extracts tend to deteriorate. High-efficiency air filters
must be installed for most of the areas in pharmaceutical factories to prevent contamination.
5. Modern refrigerated warehouses not only store commodities in coolers at temperatures of
27 to 32°F (2.8 to 0°C) and frozen foods at 10 to 20°F (23 to 29°C), but also provide
relative-humidity control for perishable foods between 90 and 100 percent. Refrigerated storage
is used to prevent deterioration. Temperature control can be performed by refrigeration systems
only, but the simultaneous control of both temperature and relative humidity in the space can only
be performed by process air conditioning systems.
1.3 CLASSIFICATION OF AIR CONDITIONING SYSTEMS
ACCORDING TO CONSTRUCTION AND OPERATING
CHARACTERISTICS
Air conditioning systems can also be classified according to their construction and operating
characteristics as follows.
INTRODUCTION 1.3
Individual Room Air Conditioning Systems
Individual room, or simply individual air conditioning systems employ a single, self-contained
room air conditioner, a packaged terminal, a separated indoor-outdoor split unit, or a heat pump. A
heat pump extracts heat from a heat source and rejects heat to air or water at a higher temperature
for heating. Unlike other systems, these systems normally use a totally independent unit or units in
each room. Individual air conditioning systems can be classified into two categories:
 Room air conditioner (window-mounted)
 Packaged terminal air conditioner (PTAC), installed in a sleeve through the outside wall
The major components in a factory-assembled and ready-for-use room air conditioner include
the following: An evaporator fan pressurizes and supplies the conditioned air to the space. In tubeand-
fin coil, the refrigerant evaporates, expands directly inside the tubes, and absorbs the heat energy
from the ambient air during the cooling season; it is called a direct expansion (DX) coil. When
the hot refrigerant releases heat energy to the conditioned space during the heating season, it acts as
a heat pump. An air filter removes airborne particulates. A compressor compresses the refrigerant
from a lower evaporating pressure to a higher condensing pressure. A condenser liquefies refrigerant
from hot gas to liquid and rejects heat through a coil and a condenser fan. A temperature control
system senses the space air temperature (sensor) and starts or stops the compressor to control its
cooling and heating capacity through a thermostat (refer to Chap. 26).
The difference between a room air conditioner and a room heat pump, and a packaged terminal
air conditioner and a packaged terminal heat pump, is that a four-way reversing valve is added to all
room heat pumps. Sometimes room air conditioners are separated into two split units: an outdoor
condensing unit with compressor and condenser, and an indoor air handler in order to have the air
handler in a more advantageous location and to reduce the compressor noise indoors.
Individual air conditioning systems are characterized by the use of a DX coil for a single room.
This is the simplest and most direct way of cooling the air. Most of the individual systems do not
employ connecting ductwork. Outdoor air is introduced through an opening or through a small air
damper. Individual systems are usually used only for the perimeter zone of the building.
Evaporative-Cooling Air Conditioning Systems
Evaporative-cooling air conditioning systems use the cooling effect of the evaporation of liquid
water to cool an airstream directly or indirectly. It could be a factory-assembled packaged unit or a
field-built system. When an evaporative cooler provides only a portion of the cooling effect, then it
becomes a component of a central hydronic or a packaged unit system.
An evaporative-cooling system consists of an intake chamber, filter(s), supply fan, direct-contact
or indirect-contact heat exchanger, exhaust fan, water sprays, recirculating water pump, and water
sump. Evaporative-cooling systems are characterized by low energy use compared with refrigeration
cooling. They produce cool and humid air and are widely used in southwest arid areas in the
United States (refer to Chap. 27).
Desiccant-Based Air Conditioning Systems
A desiccant-based air conditioning system is a system in which latent cooling is performed by
desiccant dehumidification and sensible cooling by evaporative cooling or refrigeration. Thus, a
considerable part of expensive vapor compression refrigeration is replaced by inexpensive evaporative
cooling. A desiccant-based air conditioning system is usually a hybrid system of dehumidification,
evaporative cooling, refrigeration, and regeneration of desiccant (refer to Chap. 29).
There are two airstreams in a desiccant-based air conditioning system: a process airstream and a
regenerative airstream. Process air can be all outdoor air or a mixture of outdoor and recirculating
1.4 CHAPTER ONE
air. Process air is also conditioned air supplied directly to the conditioned space or enclosed manufacturing
process, or to the air-handling unit (AHU), packaged unit (PU), or terminal for further
treatment. Regenerative airstream is a high-temperature airstream used to reactivate the desiccant.
A desiccant-based air conditioned system consists of the following components: rotary desiccant
dehumidifiers, heat pipe heat exchangers, direct or indirect evaporative coolers, DX coils and vapor
compression unit or water cooling coils and chillers, fans, pumps, filters, controls, ducts, and piping.
Thermal Storage Air Conditioning Systems
In a thermal storage air conditioning system or simply thermal storage system, the electricity-driven
refrigeration compressors are operated during off-peak hours. Stored chilled water or stored ice in
tanks is used to provide cooling in buildings during peak hours when high electric demand charges
and electric energy rates are in effect. A thermal storage system reduces high electric demand for
HVAC&R and partially or fully shifts the high electric energy rates from peak hours to off-peak hours.
A thermal storage air conditioning system is always a central air conditioning system using
chilled water as the cooling medium. In addition to the air, water, and refrigeration control systems,
there are chilled-water tanks or ice storage tanks, storage circulating pumps, and controls (refer to
Chap. 31).
Clean-Room Air Conditioning Systems
Clean-room or clean-space air conditioning systems serve spaces where there is a need for critical
control of particulates, temperature, relative humidity, ventilation, noise, vibration, and space pressurization.
In a clean-space air conditioning system, the quality of indoor environmental control
directly affects the quality of the products produced in the clean space.
A clean-space air conditioning system consists of a recirculating air unit and a makeup air
unit—both include dampers, prefilters, coils, fans, high-efficiency particulate air (HEPA) filters,
ductwork, piping work, pumps, refrigeration systems, and related controls except for a humidifier in
the makeup unit (refer to Chap. 30).
Space Conditioning Air Conditioning Systems
Space conditioning air conditioning systems are also called space air conditioning systems. They
have cooling, dehumidification, heating, and filtration performed predominately by fan coils, watersource
heat pumps, or other devices within or above the conditioned space, or very near it. A fan
coil consists of a small fan and a coil. A water-source heat pump usually consists of a fan, a finned
coil to condition the air, and a water coil to reject heat to a water loop during cooling, or to extract
heat from the same water loop during heating. Single or multiple fan coils are always used to serve
a single conditioned room. Usually, a small console water-source heat pump is used for each control
zone in the perimeter zone of a building, and a large water-source heat pump may serve several
rooms with ducts in the core of the building (interior zone, refer to Chap. 28).
Space air conditioning systems normally have only short supply ducts within the conditioned
space, and there are no return ducts except the large core water-source heat pumps. The pressure
drop required for the recirculation of conditioned space air is often equal to or less than 0.6 in. water
column (WC) (150 Pa). Most of the energy needed to transport return and recirculating
air is saved in a space air conditioning system, compared to a unitary packaged or a central
hydronic air conditioning system. Space air conditioning systems are usually employed with a
dedicated (separate) outdoor ventilation air system to provide outdoor air for the occupants in the
conditioned space.
Space air conditioning systems often have comparatively higher noise level and need more
periodic maintenance inside the conditioned space.
INTRODUCTION 1.5
1.6 CHAPTER ONE
Unitary Packaged Air Conditioning Systems
Unitary packaged air conditioning systems can be called, in brief, packaged air conditioning systems
or packaged systems. These systems employ either a single, self-contained packaged unit or
two split units. A single packaged unit contains fans, filters, DX coils, compressors, condensers,
and other accessories. In the split system, the indoor air handler comprises controls and the air system,
containing mainly fans, filters, and DX coils; and the outdoor condensing unit is the refrigeration
system, composed of compressors and condensers. Rooftop packaged systems are most widely
used (refer to Chap. 29).
Packaged air conditioning systems can be used to serve either a single room or multiple rooms.
A supply duct is often installed for the distribution of conditioned air, and a DX coil is used to cool
it. Other components can be added to these systems for operation of a heat pump system; i.e., a centralized
system is used to reject heat during the cooling season and to condense heat for heating
during the heating season. Sometimes perimeter baseboard heaters or unit heaters are added as a
part of a unitary packaged system to provide heating required in the perimeter zone.
Packaged air conditioning systems that employ large unitary packaged units are central systems
by nature because of the centralized air distributing ductwork or centralized heat rejection systems.
Packaged air conditioning systems are characterized by the use of integrated, factory-assembled,
and ready-to-use packaged units as the primary equipment as well as DX coils for cooling, compared
to chilled water in central hydronic air conditioning systems. Modern large rooftop packaged
units have many complicated components and controls which can perform similar functions to the
central hydronic systems in many applications.
1.4 CENTRAL HYDRONIC AIR CONDITIONING SYSTEMS
Central hydronic air conditioning systems are also called central air conditioning systems. In a central
hydronic air conditioning system, air is cooled or heated by coils filled with chilled or hot water
distributed from a central cooling or heating plant. It is mostly applied to large-area buildings with
many zones of conditioned space or to separate buildings.
Water has a far greater heat capacity than air. The following is a comparison of these two media
for carrying heat energy at 68°F (20°C):
The heat capacity per cubic foot (meter) of water is 3466 times greater than that of air. Transporting
heating and cooling energy from a central plant to remote air-handling units in fan rooms is
far more efficient using water than conditioned air in a large air conditioning project. However, an
additional water system lowers the evaporating temperature of the refrigerating system and makes a
small- or medium-size project more complicated and expensive.
A central hydronic system of a high-rise office building, the NBC Tower in Chicago, is illustrated
in Fig. 1.1. A central hydronic air conditioning system consists of an air system, a water
system, a central heating /cooling plant, and a control system.
Air System
An air system is sometimes called the air-handling system. The function of an air system is to
condition, to transport, to distribute the conditioned, recirculating, outdoor, and exhaust air, and to
control the indoor environment according to requirements. The major components of an air system
Air Water
Specific heat, Btu/ lb °F 0.243 1.0
Density, at 68°F, lb/ ft3 0.075 62.4
Heat capacity of fluid at 68°F, Btu / ft3 °F 0.018 62.4
1.7
FIGURE 1.1 Schematic diagram of the central hydronic air conditioning
system in NBC Tower.
1.8 CHAPTER ONE
are the air-handling units, supply/return ductwork, fan-powered boxes, space diffusion devices, and
exhaust systems.
An air-handling unit (AHU) usually consists of supply fan(s), filter(s), a cooling coil, a heating
coil, a mixing box, and other accessories. It is the primary equipment of the air system. An AHU
conditions the outdoor/ recirculating air, supplies the conditioned air to the conditioned space, and
extracts the returned air from the space through ductwork and space diffusion devices.
A fan-powered variable-air-volume (VAV) box, often abbreviated as fan-powered box, employs
a small fan with or without a heating coil. It draws the return air from the ceiling plenum, mixes it
with the conditioned air from the air-handling unit, and supplies the mixture to the conditioned
space.
Space diffusion devices include slot diffusers mounted in the suspended ceiling; their purpose is
to distribute the conditioned air evenly over the entire space according to requirements. The return
air enters the ceiling plenum through many scattered return slots.
Exhaust systems have exhaust fan(s) and ductwork to exhaust air from the lavatories, mechanical
rooms, and electrical rooms.
The NBC Tower in Chicago is a 37-story high-rise office complex constructed in the late 1980s.
It has a total air conditioned area of about 900,000 ft2 (83,600 m2). Of this, 256,840 ft2 (23,870 m2)
is used by NBC studios and other departments, and 626,670 ft2 (58,240 m2) is rental offices located
on upper floors. Special air conditioning systems are employed for NBC studios and departments at
the lower level.
For the rental office floors, four air-handling units are located on the 21st floor. Outdoor air
either is mixed with the recirculating air or enters directly into the air-handling unit as shown in
Fig. 1.2. The mixture is filtrated at the filter and is then cooled and dehumidified at the cooling coil
during cooling season. After that, the conditioned air is supplied to the typical floor through the
supply fan, the riser, and the supply duct; and to the conditioned space through the fan-powered box
and slot diffusers.
Water System
The water system includes chilled and hot water systems, chilled and hot water pumps, condenser
water system, and condenser water pumps. The purpose of the water system is (1) to transport
chilled water and hot water from the central plant to the air-handling units, fan-coil units, and fanpowered
boxes and (2) to transport the condenser water from the cooling tower, well water, or other
sources to the condenser inside the central plant.
In Figs. 1.1 and 1.2, the chilled water is cooled in three centrifugal chillers and then is distributed
to the cooling coils of various air-handling units located on the 21st floor. The temperature of
the chilled water leaving the coil increases after absorbing heat from the airstream flowing over the
coil. Chilled water is then returned to the centrifugal chillers for recooling through the chilled water
pumps.
After the condenser water has been cooled in the cooling tower, it flows back to the condenser of
the centrifugal chillers on lower level 3. The temperature of the condenser water again rises owing
to the absorption of the condensing heat from the refrigerant in the condenser. After that, the condenser
water is pumped to the cooling towers by the condenser water pumps.
Central Plant
The refrigeration system in a central plant is usually in the form of a chiller package. Chiller packages
cool the chilled water and act as a cold source in the central hydronic system. The boiler plant,
consisting of boilers and accessories, is the heat source of the heating system. Either hot water is
heated or steam is generated in the boilers.
In the NBC Tower, the refrigeration system has three centrifugal chillers located in lower level 3
of the basement. Three cooling towers are on the roof of the building. Chilled water cools from 58
to 42°F (14.4 to 5.6°C) in the evaporator when the refrigerant is evaporated. The refrigerant is then
compressed to the condensing pressure in the centrifugal compressor and is condensed in liquid
form in the condenser, ready for evaporation in the evaporator.
There is no boiler in the central plant of the NBC Tower. To compensate heat loss in the perimeter
zone, heat energy is provided by the warm plenum air and the electric heating coils in the fanpowered
boxes.
Control System
Modern air conditioning control systems for the air and water systems and for the central plant consist
of electronic sensors, microprocessor-operated and -controlled modules that can analyze and perform
calculations from both digital and analog input signals, i.e., in the form of a continuous variable.
Control systems using digital signals compatible with the microprocessor are called direct digital
control (DDC) systems. Outputs from the control modules often actuate dampers, valves, and relays
by means of pneumatic actuators in large buildings and by means of electric actuators for small
projects.
In the NBC Tower, the HVAC&R system is monitored and controlled by a microprocessor-based
DDC system. The DDC controllers regulate the air-handling units and the terminals. Both communicate
with the central operating station through interface modules. In case of emergency, the
INTRODUCTION 1.9
FIGURE 1.2 Schematic drawing of air system for a typical floor of offices in the NBC Tower.
1.10 CHAPTER ONE
fire protection system detects alarm conditions. The central operating station gives emergency
directions to the occupants, operates the HVAC&R system in a smoke control mode, and actuates
the sprinkler water system.
Air, Water, Refrigeration, and Heating Systems
Air, water, refrigeration, heating, and control systems are the subsystems of an air conditioning or
HVAC&R system. Air systems are often called secondary systems. Heating and refrigeration systems
are sometimes called primary systems.
Central hydronic and space conditioning air conditioning systems both have air, water, refrigeration,
heating, and control systems. The water system in a space conditioning system may be a
chilled /hot water system. It also could be a centralized water system to absorb heat from the condenser
during cooling, or provide heat for the evaporator during heating.
For a unitary packaged system, it consists of mainly air, refrigeration, and control systems. The
heating system is usually one of the components in the air system. Sometimes a separate baseboard
hot water heating system is employed in the perimeter zone.
A evaporative-cooling system always has an air system, a water system, and a control system. A
separate heating system is often employed for winter heating.
In an individual room air conditioning system, air and refrigeration systems are installed in indoor
and outdoor compartments with their own control systems. The heating system is often a component
of the supply air chamber in the room air conditioner. It can be also a centralized hot water
heating system in a PTAC system.
Air conditioning or HVAC&R systems are therefore often first described and analyzed through
their subsystems and main components: such as air, water, heating, cooling/ refrigeration, and control
systems. Air conditioning system classification, system operating characteristics, and system
selection must take into account the whole system.
Among the air, water, and refrigeration systems, the air system conditions the air, controls and
maintains the required indoor environment, and has direct contact with the occupants and the manufacturing
processes. These are the reasons why the operating characteristics of an air conditioning
system are esssentially represented by its air system.
1.5 DISTRIBUTION OF SYSTEMS USAGE
According to surveys conducted in 1995 by the Department of Energy/Energy Information
Administration (DOE/EIA) of the United States, for a total floor space of 58,772 million ft2 (5462
million m2) in commercial buildings in 1995 and for a total of 96.6 million homes in 1993 (among
these, 74.1 million homes were air conditioned), the floor space, in million square feet, and the
number of homes using various types of air conditioning systems are as follows:
Percent of Million Percent
Million ft2 floor space homes of homes
Individual room systems 12,494 22 33.1 45
Evaporative-cooling systems 2,451 4
Space conditioning systems (estimated) (8)
Unitary packaged systems (including air-source heat
pump as well as desiccant-based systems) 26,628 48 41.0 55
Central hydronic systems (including thermal storage
and clean-room systems) 13,586 24
Others 949 2
Much of the floor space may be included in more than one air conditioning system. Given the
possibility that the floor space may be counted repeatedly, the original data listed in the DOE /EIA
publication were modified. The 8 percent of the space system includes part in central hydronic systems
and part in unitary packaged systems. Among the air conditioned homes in 1993, the unitary
packaged system is the predominate air conditioning system in U.S. homes.
1.6 HISTORICAL DEVELOPMENT
The historical development of air conditioning can be summarized briefly.
Central Air Conditioning Systems
As part of a heating system using fans and coils, the first rudimentary ice system in the United
States, designed by McKin, Mead, and White, was installed in New York City’s Madison Square
Garden in 1880. The system delivered air at openings under the seats. In the 1890s, a leading consulting
engineer in New York City, Alfred R. Wolf, used ice at the outside air intake of the heating
and ventilating system in Carnegie Hall. Another central ice system in the 1890s was installed in
the Auditorium Hotel in Chicago by Buffalo Forge Company of Buffalo, New York. Early central
heating and ventilating systems used steam-engine-driven fans. The mixture of outdoor air and return
air was discharged into a chamber. In the top part of the chamber, pipe coils heat the mixture
with steam. In the bottom part is a bypass passage with damper to mix conditioned air and bypass
air according to the requirements.
Air conditioning was first systematically developed by Willis H. Carrier, who is recognized as
the father of air conditioning. In 1902, Carrier discovered the relationship between temperature and
humidity and how to control them. In 1904, he developed the air washer, a chamber installed with
several banks of water sprays for air humidification and cleaning. His method of temperature and
humidity regulation, achieved by controlling the dew point of supply air, is still used in many industrial
applications, such as lithographic printing plants and textile mills.
Perhaps the first air-conditioned office was the Larkin Administration Building, designed by
Frank L. Wright and completed in 1906. Ducts handled air that was drawn in and exhausted at roof
level. Wright specified a refrigeration plant which distributed 10°C cooling water to air-cooling
coils in air-handling systems.
The U.S. Capitol was air-conditioned by 1929. Conditioned air was supplied from overhead
diffusers to maintain a temperature of 75°F (23.9°C) and a relative humidity of 40 percent during
summer, and 80°F (26.7°C) and 50 percent during winter. The volume of supply air was controlled
by a pressure regulator to prevent cold drafts in the occupied zone.
Perhaps the first fully air conditioned office building was the Milan Building in San Antonio,
Texas, which was designed by George Willis in 1928. This air conditioning system consisted of one
centralized plant to serve the lower floors and many small units to serve the top office floors.
In 1937, Carrier developed the conduit induction system for multiroom buildings, in which recirculation
of space air is induced through a heating/cooling coil by a high-velocity discharging
airstream. This system supplies only a limited amount of outdoor air for the occupants.
The variable-air-volume (VAV) systems reduce the volume flow rate of supply air at reduced loads
instead of varying the supply air temperature as in constant-volume systems. These systems were introduced
in the early 1950s and gained wide acceptance after the energy crisis of 1973 as a result of
their lower energy consumption in comparison with constant-volume systems. With many variations,
VAV systems are in common use for new high-rise office buildings in the United States today.
Because of the rapid development of space technology after the 1960s, air conditioning systems
for clean rooms were developed into sophisticated arrangements with extremely effective air
filters. Central air conditioning systems always will provide a more precisely controlled, healthy,
and safe indoor environment for high-rise buildings, large commercial complexes, and precisionmanufacturing
areas.
INTRODUCTION 1.11
Unitary Packaged Systems
The first room cooler developed by Frigidaire was installed about in 1928 or 1929, and the “Atmospheric
Cabinet” developed by the Carrier Engineering Company was first installed in May 1931.
The first self-contained room air conditioner was developed by General Electric in 1930. It was a
console-type unit with a hermetically sealed motor-compressor (an arrangement in which the motor
and compressor are encased together to reduce the leakage of refrigerant) and water-cooled condenser,
using sulfur dioxide as the refrigerant. Thirty of this type of room air conditioner, were built
and sold in 1931.
Early room air conditioners were rather bulky and heavy. They also required a drainage connection
for the municipal water used for condensing. During the postwar period the air-cooled model
was developed. It used outdoor air to absorb condensing heat, and the size and weight were greatly
reduced. Annual sales of room air conditioners have exceeded 100,000 units since 1950.
Self-contained unitary packages for commercial applications, initially called store coolers, were
introduced by the Airtemp Division of Chrysler Corporation in 1936. The early models had a refrigeration
capacity of 3 to 5 tons and used a water-cooled condenser. Air-cooled unitary packages
gained wide acceptance in the 1950s, and many were split systems incorporating an indoor air handler
and an outdoor condensing unit.
Packaged units have been developed since the 1950s, from indoor to rooftops, from constantvolume
to variable-air-volume, and from few to many functions. Currently, packaged units enjoy
better performance and efficiency with better control of capacity to match the space load. Computerized
direct digital control, one of the important reasons for this improvement, places unitary packaged
systems in a better position to compete with central hydronic systems.
Refrigeration Systems
In 1844, Dr. John Gorrie designed the first commercial reciprocating refrigerating machine in the
United States. The hermetically sealed motor-compressor was first developed by General Electric
Company for domestic refrigerators and sold in 1924.
Carrier invented the first open-type gear-driven factory-assembled, packaged centrifugal chiller
in 1922 in which the compressor was manufactured in Germany; and the hermetic centrifugal
chiller, with a hermetically sealed motor-compressor assembly, in 1934. The direct-driven hermetic
centrifugal chiller was introduced in 1938 by The Trane Company. Up to 1937, the capacity of centrifugal
chillers had increased to 700 tons.
During the 1930s, one of the outstanding developments in refrigeration was the discovery by
Midgely and Hene of the nontoxic, nonflammable, fluorinated hydrocarbon refrigerant family
called Freon in 1931. Refrigerant-11 and refrigerant-12, the chlorofluorocarbons (CFCs), became
widely adopted commercial products in reciprocating and centrifugal compressors. Now, new refrigerants
have been developed by chemical manufacturers such as DuPont to replace CFCs, so as
to prevent the depletion of the ozone layer.
The first aqueous-ammonia absorption refrigeration system was invented in 1815 in Europe. In
1940, Servel introduced a unit using water as refrigerant and lithium bromide as the absorbing solution.
The capacities of these units ranged from 15 to 35 tons (52 to 123 kW). Not until 1945 did
Carrier introduce the first large commercial lithium bromide absorption chillers. These units were
developed with 100 to 700 tons (352 to 2460 kW) of capacity, using low-pressure steam as the heat
source.
Positive-displacement screw compressors have been developed in the United States since the
1950s and scroll compressors since the 1970s because of their higher efficiency and smoother rotary
motion than reciprocating compressors. Now, the scroll compressors gradually replace the reciprocating
compressors in small and medium-size refrigeration systems. Another trend is the development of
more energy-efficient centrifugal and absorption chillers for energy conservation. The energy consumption
per ton of refrigeration of a new centrifugal chiller dropped from 0.80 kW/ton (4.4 COPref)
in the late 1970s to 0.50 kW/ton (7.0 COPref) in the 1990s. A series of rotary motion refrigeration
1.12 CHAPTER ONE
compressors with small, medium, to large capacity and using scroll, screw, or centrifugal compressors
will be manufactured from now on.
1.7 POTENTIALS AND CHALLENGES
Air conditioning or HVAC&R is an industry of many potentials and challenges, including the
following.
Providing a Healthy and Comfortable Indoor Environment
Nowadays, people in the United States spend most of their time indoors. A healthy and comfortable
indoor environment provided by air conditioning is a necessity for people staying indoors, no matter
how hot or cold and dry or humid the outside climate might be. According to the American
Housing Survey conducted by the U.S. Census Bureau in 1991, of 92.3 million homes, 66 million,
or 71 percent of the total, were air conditioned, and 81.9 million, or nearly 89 percent of the total,
were heated. According to the Energy Information Administration (EIA), in 1992, for a total floor
area of 67.8 billion ft2 (6.3 billion m2) of commercial buildings, 84 percent were cooled and 91 percent
were heated.
The Cleanest, Quietest, and Most Precise and Humid
Processing Environment
A class 1 clean room to manufacture integrated circuits in a semiconductor factory may be the
cleanest processing environment that is provided in the semiconductor industry in the 1990s. The
dust particle count does not exceed 1 particle/ ft3 (35 particles /m3) of a size of 0.5m and larger,
with no particle exceeding 5m. A constant-temperature room of 68°F  0.1°F (20  0.56°C) is
always surrounded by other constant-temperature rooms or by a buffer area of lower tolerance to
maintain the fluctuation of its sensed temperature within  0.05°F (0.028°C) during working
hours.
A recording studio in a television broadcasting station often needs a noise criteria (NC) curve
(refer to Chap. 4) of NC 15 to 20, in which a sound level less than that of a buzzing insect can be
heard. A refrigerated warehouse that stores vegetables such as cabbage, carrots, and celery needs a
temperature of 32°F (0°C) and a relative humidity of 98 to 100 percent to prevent deterioration and
loss of water.
Air conditioning or HVAC&R systems will provide not only the cleanest, precisest, quietest, and
most humid environment with fluctuations of the controlled variable within required limits, but also
at optimum energy use and first cost.
Energy Use and Energy Efficiency
Based on the data published in the Annual Energy Review in 1993, the total energy use in 1992 in
the United States was 82.14 quad Btu or 1015 Btu (86.66 EJ, or 1018 J). The United States alone
consumed about one-fourth of the world’s total production. Of the total energy use of 82.14 quad
Btu in the United States, the residential /commercial sector consumed about 36 percent of the total,
the industrial sector consumed another 36 percent, and transportation consumed the rest—28 percent.
Petroleum, natural gas, and coal were the three main sources, providing more than 85 percent
of the energy supply in 1992 in the United States.
According to DOE/EIA energy consumption survey 0321 for residential buildings (1993) and
survey 0318 for commercial buildings (1995), the average annual energy use of HVAC&R systems
INTRODUCTION 1.13
1.14 CHAPTER ONE
in the residential /commercial sector was about 45 percent of the total building energy consumption.
Also assuming that the annual energy use of HVAC&R systems was about 1 percent of the total in
both industrial and transportation sectors, then the estimate of annual energy use of the HVAC&R
systems in 1992 in the United States was about 17 percent of the total national energy use, or onesixth
of the total national energy use.
The world energy resources of petroleum, natural gas, and coal are limited. The population of
the United States in 1992 was only about one-twentieth of the world’s total population; however,
we consumed nearly one-fifth of the world’s total energy produced. Energy use must be reduced.
After the energy crisis in 1973, the U.S. Congress enacted the Energy Policy and Conservation
Act of 1975 and the National Energy Policy Act of 1992. The enactment of energy efficiency legislation
by federal and state governments and the establishment of the Department of Energy (DOE)
in 1977 had a definite impact on the implementation of energy efficiency in United States.
In 1975, the American Society of Heating, Refrigerating and Air Conditioning Engineers
(ASHRAE) published Standard 90-75, Energy Conservation in New Building Design. This standard
was revised and cosponsored by the Illuminating Engineering Society of North America as
ASHRAE/IES Standard 90.1-1989, Energy Efficient Design of New Buildings Except New Low-
Rise Residential Buildings, in 1989; and it was revised again as current ASHRAE/IESNA Standard
90.1-1999, Energy Standard for Buildings Except Low-Rise Residential Buildings, in 1999. Many
other organizations also offered valuable contributions for energy conservation. All these events
started a new era in which energy efficiency has become one of the important goals of HVAC&R
system design and operation.
Because of all these efforts, the increase in net annual energy use for residential and commercial
buildings, i.e., the total energy use minus the electrical system energy loss in the power plant, from
1972 to 1992 in the United States was only 1.3 percent, as shown in Fig. 1.3. However, during the
same period, the increase in floor area of commercial buildings in the United States was about 83
Commercial building
floor space
Energy
use
U. S. commercial buildings, billion ft2
Residential and commercial buildings energy use, 1015 Btu 0
10
20
30
40
50
60
70
80
15.89
1950 1960 1970 1980 1990
16.09
0
5
10
15
FIGURE 1.3 Residential and commercial building energy use and commercial building floor space increase between
1992 and 1950 in the United States.
percent. Energy efficiency will be a challenge to everyone in the HVAC&R industry in this generation
as well as in many, many generations to come.
Environmental Problems—CFCs and Global Warming
The surface of the earth is surrounded by a layer of air, called the atmosphere. The lower atmosphere
is called the homosphere, and the upper atmosphere is called the stratosphere.
In the mid-1980s, chlorofluorocarbons (CFCs) were widely used as refrigerants in mechanical
refrigeration systems, to produce thermal insulation foam, and to produce aerosol propellants for
many household consumer products. CFC-11 (CCl3F) and CFC-12 (CCl2F2) are commonly used
CFCs. They are very stable. Halons are also halogenated hydrocarbons. If CFCs and halons leak or
are discharged from a refrigeration system during operation or repair to the lower atmosphere, they
will migrate to the upper stratosphere and decompose under the action of ultraviolet rays throughout
their decades or centuries of life. The free chlorine atoms will react with oxygen atoms of the
ozone layer in the upper stratosphere and will cause a depletion of this layer. The theory of the depletion
of the ozone layer was proposed in the early and middle-1970s. The ozone layer filters out
harmful ultraviolet rays, which may cause skin cancer and are a serious threat to human beings.
Furthermore, changes in the ozone layer may significantly influence weather patterns. Since 1996,
actions have been taken to ban the production of CFCs and halons, before it is too late.
A cloudless homosphere is mainly transparent to short-wave solar radiation but is quite opaque
to long-wave infrared rays emitted from the surface of the earth. Carbon dioxide (CO2) has the
greatest blocking effect of all; water vapor and synthetic CFCs also play important roles in blocking
the direct escape of infrared energy. The phenomenon of transparency to incoming solar radiation
and blanketing of outgoing infrared rays is called the greenhouse effect. The increase of the
CO2, water vapor, CFCs, and other gases, often called greenhouse gases (GHGs), eventually will
result in a rise in air temperature near the earth’s surface. This is known as the global warming
effect.
Over the past 100 years, global warming has caused an increase of 1°F. For the same period,
there was a 25 percent increase of CO2. During the 1980s, the release of CO2 to the atmosphere
was responsible for about 50 percent of the increase in global warming that was attributable to
human activity, and the release of CFCs had a 20 percent share. Some scientists have predicted an
accelerated global warming in the coming 50 years because of the increase in the world’s annual
energy use. Further increases in global temperatures may lead to lower rainfalls, drop in soil moisture,
more extensive forest fires, more flood, etc.
CFC production in developed countries has been banned since January 1, 1996. Carbon dioxide
is the product of many combustion processes. Alternative refrigerants to replace CFCs must also
have a low global warming potential. Designers and operators of the HVAC&R systems can reduce
the production of CO2 through energy efficiency and the replacement of coal, petroleum, and natural
gas power by hydroelectric, solar, and nuclear energy, etc. More studies and research are
needed to clarify the theory and actual effect of global warming.
Air Conditioning or HVAC&R Industry
The air conditioning or HVAC&R industry in the United States is an expanding and progressing industry.
In 1995, the installed value of nonresidential air conditioning hit $20 billion. According to
American Refrigeration Institute (ARI) and Heating/Piping/Air Conditioning data, from 1985 to
1995, the annual rate of increase of installed value of air conditioning systems is 8.7 percent.
Because of the replacement of the old chillers using CFCs, in 1994, the installed value of retrofit,
remodeling, and replacement accounted for up to two-thirds of all HVAC&R expenditures. This
trend may continue in the beginning of the new century.
Based on data from ARI, in 1995, shipment of packaged units reached a record figure of 5
million products, of which heat pumps comprised a one-fifth share. Centrifugal and screw chiller
INTRODUCTION 1.15
1.16 CHAPTER ONE
shipments were 7500 units, absorption chillers 502 units, and reciprocating chillers, at a
significantly smaller capacity, were 14,000 units.
1.8 AIR CONDITIONING PROJECT DEVELOPMENT
Basic Steps in Development
The basic steps in the development and use of a large air conditioning system are the design, construction,
commissioning, operation, energy efficiency upgrading, and maintenance. Figure 1.4 is a
diagram which outlines the relationship between these steps and the parties involved. The owner
sets the criteria and the requirements. Design professionals in mechanical engineering consulting
FIGURE 1.4 Steps in the development and use of air conditioning systems in buildings.
firms design the air conditioning system and prepare the design documents. Manufacturers supply
the equipment, instruments, and materials. Contractors install and construct the air conditioning
system. After construction, the air conditioning system is commissioned by a team, and then it is
handed over to the operation and maintenance group of the property management for daily operation.
Following a certain period of operation, an energy service company (ESCO) may often be
required to upgrade the energy efficiency of the HVAC&R system (energy retrofit).
Design-Bid and Design-Build
There are two types of project development: design-bid and design-build. A design-bid project
separates the design and installation responsibilities, whereas in a design-build project, engineering
is done by the installing contractor. Some reasons for a design-build are that the project is too small
to retain an engineering consultant, or that there is insufficient time to go through the normal
design-bid procedures.
According to Bengard (1999), the main advantages of design-build include established firm
price early, single-source responsibility, accelerated project delivery, and performance guarantees.
The market has experienced nearly 300 percent domestic growth since 1986.
The Goal—An Environmentally Friendlier, Energy-Efficient, and
Cost-Effective HVAC&R System
The goal is to provide an HVAC&R system which is environmentally friendlier, energy-efficient,
and cost-effective as follows:
 Effectively control indoor environmental parameters, usually to keep temperature and humidity
within required limits.
 Provide an adequate amount of outdoor ventilation air and an acceptable indoor air quality.
 Use energy-efficient equipment and HVAC&R systems.
 Minimize ozone depletion and the global warming effect.
 Select cost-effective components and systems.
 Ensure proper maintenance, easy after-hour access, and necessary fire protection and smoke control
systems.
Major HVAC&R Problems
In Coad’s paper (1985), a study by Wagner-Hohn-Ingles Inc. revealed that many large new buildings
constructed in the early 1980s in the United States suffer from complaints and major defects.
High on the list of these problems is the HVAC&R system. The major problems are these:
1. Poor indoor air quality (IAQ)—sick building syndrome. Poor indoor air quality causes the
sick building syndrome. The National Institute of Occupational Safety and Health (NIOSH) in 1990
reported that between 1971 and 1988, 529 field investigations found that lack of outdoor air, improper
use, and poor operation and maintenance of HVAC&R systems were responsible for more
than one-half of sick building syndrome incidents. Field investigations found that 20 to 30 percent
of the buildings had poor-air-quality problems. Sick building syndrome is covered in detail in
Chap. 4.
2. Updated technology. In recent years, there has been a rapid change in the technology of air
conditioning. Various types of VAV systems, air and water economizer, heat recovery, thermal storage,
desiccant dehumidification, variable-speed drives, and DDC devices have become more effective
and more advanced for energy efficiency. Many HVAC&R designers and operators are not
INTRODUCTION 1.17
1.18 CHAPTER ONE
properly equipped to apply and use these systems. Unfortunately, these sophisticated systems are
managed, constructed, and operated under the same budget and schedule constraints as the less
sophisticated systems. After years of operation, most HVAC&R equipment and controls need to be
upgraded for energy efficiency.
3. Insufficient communication between design professionals, construction groups, and operators.
Effective operation requires a knowledgeable operator to make adjustments if necessary. The operator
will operate the system at his or her level of understanding. If adequate operating and maintenance
documents are not provided by the designer and the contractor for the operator, the
HVAC&R system may not be operated according to the designer’s intentions.
4. Overlooked commissioning. Commissioning means testing and balancing all systems, functional
testing and adjusting of components and the integrated system, and adjusting and tuning the
direct digital controls. An air conditioning system is different from the manufacturing products having
models and prototypes. All the defects and errors of the prototypes can be checked and corrected
during their individual tests, but the more complicated HVAC&R system, as constructed and
installed, is the end product. Therefore, proper commissioning, which permits the system to perform
as specified in the design documents, is extremely important. Unfortunately, the specifications
seldom clearly designate competent technicians for the responsibility of commissioning the entire
integrated system. Commissioning is covered in detail in Chap. 32.
5. Reluctant to try innovative approaches. In addition to Coad’s paper, a survey was conducted
by the American Consulting Engineers Council (ACEC) in 1995. For the 985 engineering firms that
responded out of about 5500 firms total, there were 522 legal claims filed against them, 49 percent
of suits resolved without payment. Among the 522 legal claims, 9 percent were filed against mechanical
engineering firms. Fifty-two percent of 1995 legal claims were filed by owners, and
13 percent were filed by contractors or subcontractors.
Because of the legal claims, insurance companies discourage innovation, and engineering
consulting firms are reluctant to try innovative techniques. Twenty-one percent of these firms said
they were very much reluctant, 61 percent said they were somewhat affected, and 25 percent said
they were a little affected. Support and encouragement must be given to engineering firms for carefully
developed innovative approaches to projects. In addition to the ASHRAE Technology Awards,
more HVAC&R innovative awards should be established in large cities in the United States to
promote innovative approaches.
1.9 DESIGN FOR AIR CONDITIONING SYSTEM
System design determines the basic characteristics. After an air conditioning system is constructed
according to the design, it is difficult and expensive to change the design concept.
Engineering Responsibilities
The normal procedure in a design-bid project includes the following steps and requirements:
1. Initiation of a construction project by owner or developer
2. Selection of design team
3. Setting of the design criteria and indoor environmental parameters
4. Selection of conceptual alternatives for systems and subsystems; preparation of schematic layouts
of HVAC&R
5. Preparation of contract documents, working drawings, specifications, materials and construction
methods, commissioning guidelines
6. Competitive bidding by contractors
7. Evaluation of bids; negotiations and modification of contract documents
8. Advice on awarding of contract
9. Review of shop drawings and commissioning schedule, operating and maintenance manuals
10. Monitoring, supervision, and inspection of construction
11. Supervision of commissioning: testing and balancing; functional performance tests
12. Modification of drawings to the as-built condition and the finalization of the operation and
maintenance manual
13. Acceptance
Construction work starts at contract award following the bidding and negotiating and ends at the
acceptance of the project after commissioning.
It is necessary for the designer to select among the available alternatives for optimum comfort,
economics, energy conservation, noise, safety, flexibility, reliability, convenience, and maintainability.
Experience, education, and judgment all enter into the selection process. If both a complicated
system and a simple system yield the same performance, the simple system is preferred for its reliability,
operator convenience, and lower cost.
Coordination between Air Conditioning and Other Trades, Teamwork
Air conditioning, plumbing, and fire protection systems are mechanical systems in a building. Both
mechanical and electrical systems provide building services for the occupants and goods inside the
building. Coordination between the mechanical engineer for HVAC&R and the architect, as well as
between mechanical and structural or electrical engineers, or teamwork, becomes important.
Factors requiring input from both the architect and the mechanical engineer include the following:
1. Shape and the orientation of the building
2. Thermal characteristics of the building envelope, especially the type and size of the windows
and the construction of external walls and roofs
3. Location of the ductwork and piping to avoid interference with each other, or with other trades
4. Layout of the diffusers and supply and return grilles
5. Minimum clearance provided between the structural members and the suspended ceiling for the
installation of ductwork and piping
6. Location and size of the rooms for central plant, fan rooms, duct and pipe shafts
If the architect makes a decision that is thermally unsound, the HVAC&R engineer must offset
the additional loads by increasing the HVAC&R system capacity. Lack of such coordination results
in greater energy consumption.
HVAC&R and other building services must coordinate the following:
1. Utilization of daylight and the type of artificial light to be installed
2. The layout of diffusers, grilles, return inlets, and light troffers in the suspended ceiling
3. Integration with fire alarm and smoke control systems
4. Electric power and plumbing requirements for the HVAC&R equipment and lighting for equipment
rooms
5. Coordination of the layout of the ductwork, piping, electric cables, etc.
Retrofit, Remodeling, and Replacement
Retrofit of an HVAC&R system must be tailored to the existing building and integrated with existing
systems. Each retrofit project has a motivation which may be related to environment, safety and
health, indoor environment, energy conservation, change of use, etc. Because of ozone depletion,
INTRODUCTION 1.19
the production of CFCs ceased in the United States on January 1, 1996. All refrigeration systems
using CFC refrigerants will be replaced by or converted to non-CFC refrigerant systems. For the
sake of safety, smoke control systems and stairwell pressurization are items to be considered. For
the sake of the occupants’ health, an increase in the amount of outdoor air may be the right choice.
If energy conservation is the first priority, the following items are among the many that may be considered:
efficiency of lighting; energy consumption of the fans, pumps, chillers, and boilers; insulation
of the building envelope; and the remodeling or replacement of fenestration. For improving the
thermal comfort of the occupant, an increase of the supply volume flow rate, an increase of cooling
and heating capacity, and the installation of appropriate controls may be considered.
Engineer’s Quality Control
In 1988, the American Consulting Engineers Council chose “People, Professionalism, Profits: A
Focus on Quality” as the theme of its annual convention in New York City. Quality does not mean
perfection. In HVAC&R system design, quality means functionally effective, for health, comfort,
safety, energy conservation, and cost. Quality also implies the best job, using accepted professional
practices and talent. Better quality always means fewer complaints and less litigation. Of course, it
also requires additional time and higher design cost. This fact must be recognized by the owner and
developer.
The use of safety factors allows for the unpredictable in design, installation, and operation. For
example, a safety factor of 1.1 for the calculated cooling load and of 1.1 to 1.15 for the calculated
total pressure drop in ductwork is often used to take into account unexpected inferiority in fabrication
and installation. An HVAC&R system should not be overdesigned by using a greater safety factor
than is actually required. The initial cost of an overdesigned HVAC&R system is always higher,
and it is energy-inefficient.
When an HVAC&R system design is under pressure to reduce the initial cost, some avenues to
be considered are as follows:
1. Select an optimum safety factor.
2. Minimize redundancy, such as standby units.
3. Conduct a detailed economic analysis for the selection of a better alternative.
4. Calculate the space load, the capacity of the system, and the equipment requirements more
precisely.
5. Adopt optimum diversity factors based on actual experience data observed from similar buildings.
A diversity factor is defined as the ratio of the simultaneous maximum load of a system to the
sum of the individual maximum loads of the subdivisions of a system. It is also called the simultaneous-
use factor. For example, in load calculations for a coil, block load is used rather than the sum
of zone maxima for sizing coils for AHUs installed with modulating control for the chilled water
flow rate.
The need for a higher-quality design requires that engineers have a better understanding of the
basic principles, practical aspects, and updated technology of HVAC&R systems in order to avoid
overdesign or underdesign and to produce a satisfactory product.
Design of the Control System
Controlling and maintaining the indoor environmental parameters within predetermined limits
depends mainly on adequate equipment capacity and the quality of the control system. Energy can
be saved when the systems are operated at part load with the equipment’s capacity following the
system load accurately by means of capacity control.
1.20 CHAPTER ONE
Because of the recent rapid change of HVAC&R controls from conventional systems to energy
management systems, to DDC with microprocessor intelligence, and then to open-protocol BACnet,
many designers have not kept pace. In 1982, Haines did a survey on HVAC&R control system
design and found that many designers preferred to prepare a conceptual design and a sequence of
operations and then to ask the representative of the control manufacturer to design the control system.
Only one-third of the designers designed the control system themselves and asked the representative
of the control manufacturer to comment on it.
HVAC&R system control is a decisive factor in system performance. Many of the troubles with
HVAC&R arise from inadequate controls and/or their improper use. The designer should keep pace
with the development of new control technology. He or she should be able to prepare the sequence
of operations and select the best-fit control sequences for the controllers from a variety of the manufacturers
that offer equipment in the HVAC&R field. The designer may not be a specialist in the
details of construction or of wiring diagrams of controllers or DDC modules, but he or she should
be quite clear about the function and sequence of the desired operation, as well as the criteria for
the sensors, controllers, DDC modules, and controlled devices.
If the HVAC&R system designer does not perform these duties personally, preparation of a systems
operation and maintenance manual with clear instructions would be difficult. It would also be
difficult for the operator to understand the designer’s intention and to operate the HVAC&R system
satisfactorily.
Field Experience
It is helpful for the designer to visit similar projects that have been operated for more than 2 years
and talk with the operator before initiating the design process. Such practice has many advantages.
1. The designer can investigate the actual performance and effectiveness of the air conditioning
and control systems that he or she intends to design.
2. According to the actual operating records, the designer can judge whether the system is overdesigned,
underdesigned, or the exact right choice.
3. Any complaints or problems that can be corrected may be identified.
4. The results of energy conservation measures can be evaluated from actual performance instead
of theoretical calculations.
5. The designer can accumulate valuable practical experience from the visit, even from the defi-
ciencies.
New Design Technologies
Computer-aided design and drafting (CADD) and knowledge-based system (KBS) or expert system
assisted design are prospective new design technologies that have been used in HVAC&R design.
CADD is covered in detail in Sec. 1.12, and KBS is covered in Chap. 5.
1.10 DESIGN DOCUMENTS
In the United States, the construction (installation) of air conditioning or HVAC&R systems is
usually performed according to a contract between the owner or developer and the installer, the
contractor. The owner specifies the work or job that is committed to be accomplished within a time
period. The contractor shall furnish and install equipment, ductwork, piping, instruments, and the
related material of the HVAC&R system for a given compensation. The construction of an air conditioning
or HVAC&R system is usually a part of the construction of a building.
INTRODUCTION 1.21
1.22 CHAPTER ONE
Both drawings and specifications are legal documents. They are legal because the designer
conveys the requirements of the owner or developer to the contractor through these documents.
Drawings and specifications define the work to be done by the contractor. They should be clear and
should precisely and completely show the work to be accomplished. Drawings and specifications
are complements to each other. Things to be defined should be shown or specified only once, either
in drawing or in specification. Repetition often causes ambiguity and error.
Drawings
The layout of an HVAC&R system and the locations and dimensions of its equipment, instruments,
ducts, pipes, etc., are best shown and illustrated by drawings. HVAC&R drawings consist of mainly
the following:
 Floor plans. System layout including plant room, fan rooms, mechanical room, ductwork, and
pipelines is always illustrated on floor plans. Each floor has at least one floor plan. HVAC&R
floor plans are always drawn over the same floor plan of the architectural drawing.
 Detail drawings. These drawings show the details of a certain section of an HVAC&R system, or
the detail of the installation of certain equipment, or the connection between equipment and ductwork
or pipeline. Standard details are often used to save time.
 Sections and elevations. Sectional drawings are helpful to show the inner part of a section of a
system, a piece of equipment, or a device. They are especially useful for places such as the plant
room, fan room, and mechanical room where lots of equipment, ductwork, and pipelines are
found. Elevations often show clearly the relationship between the HVAC&R components and the
building structure.
 Piping diagram. This diagram shows the piping layout of the water system(s) and the flow of
water from the central plant to the HVAC&R equipment on each floor.
 Air duct diagram. This diagram illustrates the air duct layout as well as the airflow from the airhandling
unit or packaged unit to the conditioned spaces on each floor through space diffusion
devices.
 Control diagrams. These diagrams show the zone level control systems, each type of functional
control system for air-handling units or packaged units, water systems, outdoor air ventilation
systems, sequencing of compressors, network communication, etc.
 Equipment schedule. This schedule provides the quatity and performance characteristics of the
equipment or device in tabulated form as drawings. Drawings are always available to the installer
at any time but the specifications are not. This is why the equipment schedule should appear on
drawings instead of inside the specification.
 Legends. Symbols and abbreviations are often defined in a legend. ASHRAE has proposed a set
of symbols in the Fundamentals handbook.
Sometimes, three-dimensional isometric drawings are necessay for piping and air duct diagrams.
For a floor plan, a scale of 1/8 in. 	 1 ft (1:100) is often used. The size of the drawings should be
selected according to the size of the project. Drawing sheet sizes of 24  36 in., 30  42 in., and
36  48 in. (610  915 mm, 762  1067 mm, and 915  1219 mm) are widely adopted for large
projects. After completion, every drawing should be carefully checked for errors and omissions.
Specifications
Detailed descriptions of equipment, instruments, ductwork, and pipelines, as well as performances,
operating characteristics, and control sequences are better defined in specifications. Specifications
usually consist of the legal contract between the owner and the contractor, installer, or vendor, and
the technical specifications that specify in detail the equipment and material to be used and how
they are installed.
The Construction Specifications Institute (CSI) has developed a format, called the
Masterformat, for specifications. This Masterformat is widely adopted by most HVAC&R construction
projects. Masterformat promotes standardization and thereby facilitates the retrieval of information
and improves construction communication. The 1988 edition of Masterformat consists of 16
divisions: 01000 General Requirements, 02000 Site Work, 03000 Concrete, . . ., 14000 Conveying
Systems, 15000 Mechanical, and 16000 Electrical. In mechanical, it is subdivided into the following
major sections:
15050 Basic Mechanical Materials and Methods
15250 Mechanical Insulation
15300 Fire Protection
15400 Plumbing
15500 Heating, Ventilating, and Air Conditioning
15550 Heat Generation
15650 Refrigeration
15750 Heat Transfer
15850 Air Handling
15880 Air Distribution
15950 Controls
15990 Testing, Adjusting, and Balancing
Each of the above-mentioned sections may contain: general considerations, equipment and material,
and field installation.
According to whether the wanted vendor is specified or not, specifications can be classified into
two categories: (1) performance specification, which depends only on the performance criteria, and
(2) or-equal specification, which specifies the wanted vendor.
Here are some recommendations for writing an HVAC&R specification:
 The required indoor environmental parameters to be maintained in the conditioned space during
summer and winter outdoor design conditions, such as temperature, humidity, outdoor ventilation
air rates, air cleanliness, sound level, and space pressure, shall be clearly specified in the general
consideration of section 15500.
 Use simple, direct, and clear language without repetition.
 All the terms must be well defined and written in a consistent manner.
 Don’t write specifications or refer to other works without having personal knowledge of the content
or even understanding its meaning.
 All the specifications must be tailored to fit the designed HVAC&R system. Never list an item that
is not listed in the specified HVAC&R system, such as the return ducts in a fan-coil system.
1.11 CODES AND STANDARDS
Codes are generally mandatory state or city laws or regulations that force the designer to create the design
without violating human safety and welfare. State and city codes concerning structural integrity,
electrical safety, fire protection, and prevention of explosion of pressure vessels must be followed.
Standards describe consistent methods of testing, specify confirmed design guidelines, and recommend
standard practices. Conformance to standards is usually voluntary. However, if design
INTRODUCTION 1.23
criteria or system performance is not covered in local codes, ASHRAE Standards become the vital
document to assess the indoor air quality in a lawsuit against designer or contractor.
Since the energy crisis of 1973, the U. S. federal government and Congress have published legislation
which profoundly affects the design and operation of HVAC&R systems. As mentioned
before, in 1975, the Energy Policy and Conservation Act and in 1978 the National Energy Conservation
Policy Act were enacted for energy conservation. Also in 1992, the Clean Air Act Amendments
phase out the use of CFCs and recycling of refrigerants, to protect the ozone layer as well as
to provide an acceptable indoor air quality.
Among the HVAC&R related standards published by various institutions, the following by
ASHRAE and some other institutions have greatly influenced the design and operation of air conditioning
systems:
 ASHRAE/IES Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, which is often called the Energy Standard. This standard is under continuous maintenance
by a standing standard committee (SSPC) for regular publication of addenda or revisions.
 ASHRAE Standard 62-1999, Ventilation for Acceptable Indoor Air Quality, which is often called
the Indoor Air Quality Standard. This standard is under continuous maintenance.
 ANSI/ASHRAE Standard 55-1992, Thermal Environmental Conditions for Human Occupancy
 ASHRAE Standard 15-1992, Safety Code for Mechanical Refrigeration.
 ANSI/ASHRAE Standard 34-1997, Number Designation and Safety Classification of Refrigerants.
This standard is under continuous maintenance.
 ANSI/ASHRAE Standard 135-1995, BACnet: A Data Communication Protocol for Building Automation
and Control Networks.
 Air Conditioning and Refrigeration Institute, Standards for Unitary Equipment.
 National Fire Protection Association NFPA 90A, Standard for the Installation of Air Conditioning
and Ventilating Systems.
 Sheet Metal and Air Conditioning Contractors’ National Association, HVAC Duct Construction
Standards—Metal and Flexible.
These standards are covered in detail in later chapters.
Recently, many HVAC&R consulting firms seek compliance with the International Organization
for Standardization (ISO) 9000 for quality control. ISO is a specialized international agency for
standardization, at present comprising the national standards institutions of about 90 countries. The
American National Standards Institue (ANSI) is a member that represents the United States.
The ISO 9000 series has now become the quality mangement standard worldwide. It set the guidelines
for management participation, design control, review of specifications, supplier monitoring, and
services provided. The goal of ISO 9000 is to guarantee that products or the services provided by a
manufacturer or an engineering consulting firm are appropriate to the specifications and are within
the tolerances agreed upon. The ISO 9000 series consists of five quality system standards:
ISO 9000 Guidelines of the Selection and the Application of Quality Management and Quality
Assurance Standards
ISO 9001 Modeling of Quality Assurance in Design Development, Manufacturing, Installation
and Servicing
ISO 9002 Modeling of Quality Assurance in Production and Installation
ISO 9003 Modeling of Quality Assurance in Final Inspection and Testing
ISO 9004 Guidelines for the Implementation of Quality Management and to Provide Quality
Systems
There are no legal requirements to be registered to ISO 9000 standards. Meeting ISO 9000 standards
does not mean that the firm provides better service or products than those of nonregistered
companies.
1.24 CHAPTER ONE
Many companies are assembling quality assurance programs which show that they are accepting
compliance to ISO 9000 standards instead of registration. Such an arrangement will identify the
ISO 9000 standard that can be best applied to your operation and also support your customer’s ISO
9000 quality audit.
1.12 COMPUTER-AIDED DESIGN AND DRAFTING (CADD)
Because personal computers (PCs) provide a low-cost tool for computations and graphics and
owing to the availability of lots of design computer software during the 1980s, computer-aided
design and drafting (CADD) for air conditioning systems has grown quickly in recent years. Today,
computer software is often used to develop design documents, drawings, and specifications in common
use in engineering consulting firms. According to the 1994 CADD Application and User
Survey of design firms in Engineering Systems (June 1994), “. . . firms with high productivity
reported that they perform 95 percent of projects on CADD.”
In addition to the CADD, there is software available for computer-aided facilities management
(CAFM) for operation and maintenance.
Features of CADD
Compared with conventional design and drafting, CADD of air conditioning systems has the following
features:
 Different trade designers such as those in architecture, HVAC&R, plumbing, fire protection, and
electrical engineers can merge their work in plot output files to produce composite drawings in
different layers. Mechanical system design can integrate with equipment selection programs.
 Graphical model construction of air conditioning systems and subsystems in two-dimensional
(2D) or three-dimensional (3D) presentation uses architectural and structural models as backgrounds.
 CADD links the engineering design models based on calculation and the graphical model based
on drafting.
 It provides the ability to develop and compare the alternate design schemes quickly as well as the
capability to redesign or to match the changes during construction promptly.
 CADD establishes database libraries for design and graphical models.
 It develops design documents including as-built drawings and equipment schedules.
A saving of design time up to 40 percent has been claimed by some engineering consulting firms.
Computer-Aided Design
Current CADD for HVAC&R systems can be classified into two categories: (1) engineering design
and simulation, and (2) drafting graphical model construction. The software for engineering design
can again be subdivided into the following groups:
 Load calculations, energy simulation, and psychrometric analysis. Software for space load calculations
could be a part of the energy simulation software. However, load calculations are often
mainly used to determine the peak or block load at design conditions. These loads are the inputs
for equipment selection. However, the software for energy simulation is used mainly to determine
the optimum selection from different alternatives. Psychrometric analysis is sometimes a useful
tool for load calculation and energy simulation.
INTRODUCTION 1.25
1.26 CHAPTER ONE
 Equipment selection. All large HVAC&R equipment manufacturers offer software, often called an
electronic catalog, to help customers select their product according to the capacity, size, configuration,
and performance.
 Equipment schedule and specifications. Software generates equipment schedules or CSI masterformat
for specifications.
 Price list. Software reports the price of base unit, accessories, and materials.
Computer-Aided Drafting (CAD)
 Software to reproduce architectural drawings is the foundation of CADD. Graphical models of
items such as ducts, pipes, equipment, and system components in HVAC&R systems are developed
against backgrounds of architecture and structural models. Automated computer-aided drafting
(AutoCAD) is the leading PC-based drafting software used by most design firms to assist the
designer during design with automated drawings, databases, and layering schemes.
 Software applications to develop duct and piping work layouts are the two primary CADs used in
air conditioning system design. They can interface with architectural, electrical, and plumbing
drawings through AutoCAD. CAD for duct and piping work can also have hydraulic modeling capacities.
Tags and an HVAC&R equipment schedule including components and accessories can be
produced as well. CAD for ducts and piping work is covered in detail in corresponding chapters.
 Software is also available to produce details of the refrigerating plant, heating plant, and fan room
with accessories, duct, and piping layout. Many manufacturers also supply libraries of files for
AutoCAD users to input the details of their products into CAD drawings.
 HVAC&R CAD for ducts and piping work layout and details often uses graphical model construction.
The elements of the graphical model are linked to its associated information and stored
in the databases. Model construction starts by locating an end device such as a diffuser or terminal.
Add the supply and return pipes from different layers. Software for graphical model construction
has the features of copying, repeating, moving, rotating, and mirror-imaging which enable
the designer to construct the graphical models quickly. After supply and return main ducts, equipment,
and components are added, this completes the supply and return duct systems of this floor.
Duct and piping layouts can also be accomplished by drawing a single-line layout on the CAD,
and the software will convert to two-dimensional or three-dimensional drawings.
Software Requirements
 A CADD software tailored to a specific HVAC&R design or drafting purpose is needed.
 Built-in error checking and error finding are necessary features.
 Transferability of drawings and data among different CADD systems is important. ASHRAE
Handbook 1995, HVAC Applications, recommended that software interface to the most prominent
formats: the Initial Graphics Exchange Specification (IGES), the Standard Exchange Format
(SEF), and the Data Exchange Format (DXF).
 While using the many available computer software programs for loads, energy, and hydraulic
property calculations, one should verify and calibrate the results against field-measured values in
order to improve the effectiveness of the computer software when the conditions and affecting
factors may be varied.
REFERENCES
Amistadi, H., Design and Drawing: Software Review, Engineering Systems, June 1993, pp. 18–29.
Amistadi, H., HVAC Product Software, Engineering Systems, January 1994, pp. 56–62.
INTRODUCTION 1.27
Amistadi, H., Software Review: HVACR Product Software, Engineering Systems, January 1995, pp. 60–69.
Arnold, D., The Evolution of Modern Office Buildings and Air Conditioning, ASHRAE Journal, no. 6, 1999,
pp. 40–54.
ASHRAE, ASHRAE Handbook 1995, HVAC Applications, ASHRAE Inc., Atlanta, GA, 1995.
Bengard, M., The Future of Design-Build, Engineered Systems, no. 12, 1999, pp. 60–64.
Census Bureau, Snapshot Portrays Profile of Heating and Cooling Units 1991, AC, Heating & Refrigeration
News, Aug. 28, 1995, pp. 23–24.
Coad,W. J., Courses and Cures for Building System Defects, Heating/Piping/Air Conditioning, February 1985,
pp. 98–100.
Coad,W. J., Safety Factors in HVAC Design, Heating/Piping/Air Conditioning, January 1985, pp. 199–203.
Crandell, M. S., NIOSH Indoor Air Quality Investigations 1971 through 1987, Proceedings of the Air Pollution
Control Association Annual Meeting, Dallas, 1987.
Denny, R. J., The CFC Footprint, ASHRAE Journal, November 1987, pp. 24–28.
DiIorio, E., and Jennett, Jr., E. J., High Rise Opts for High Tech, Heating/Piping/Air Conditioning, January
1989, pp. 83–87.
Department of Energy/Energy Information Administration (DOE/EIA), Household Energy Consumption and
Expenditures 1993, Part I: National Data, DOE/EIA-0321 Washington, DC, 1995.
Department of Energy/Energy Information Administration (DOE/EIA), Nonresidential Building Energy Consumption
Survey: Characteristics of Commercial Buildings, 1995, DOE/EIA-0246(95),Washington, DC, 1997.
Department of Energy/Energy Information Administration (DOE/EIA), Nonresidential Building Energy Consumption
Survey: Commercial Buildings Consumption and Expenditures 1995, DOE/EIA-0318(95),Washington,
DC, 1998.
Energy Information Administration, Commercial Buildings Characteristics 1992 Commercial Building Energy
Consumption Survey,Washington, D.C., 1994.
Faust, F. H., The Early Development of Self-Contained and Packaged Air Conditioner, ASHRAE Transactions,
vol. 92, 1986, Part II B, pp. 353–360.
Grant,W. A., From ’36 to ’56—Air Conditioning Comes of Age, ASHVE Transactions, vol. 63, 1957,
pp. 69–110.
Guedes, P., Encyclopedia of Architectural Technology, section 4, 1st ed., McGraw-Hill, New York, 1979.
Haines, R.W., How Are Control Systems Designed? Heating/Piping/Air Conditioning, February 1982, p. 94.
Haines, R.W., Operating HVAC Systems, Heating/Piping/Air Conditioning, July 1984, p. 106.
Haines, R.W., and Wilson, C. L., HVAC Systems Design Handbook, McGraw-Hill, New York, 1994.
Houghten, F. C., Blackshaw, J. L., Pugh, E. M., and McDermott, P., Heat Transmission as Influenced by Heat
Capacity and Solar Radiation, ASHVE Transactions, vol. 38, 1932, pp. 231–284.
Kohloss, F. H., The Engineer’s Liability in Avoiding Air Conditioning System Overdesign, ASHRAE
Transactions, vol. 87, 1983, Part I B, pp. 155–162.
Korte, B., Existing Buildings: Vast HVAC Resource, Heating/ Piping/ Air Conditioning, March 1989,
pp. 57–63.
Korte, B., The Health of the Industry, Heating/Piping/and Air Conditioning, January 1996, pp. 69–70.
Lewis, L. L., and Stacey, Jr., A. E., Air Conditioning the Hall of Congress, ASHVE Transactions, vol. 36, 1930,
pp. 333–346.
MacCraken, C. D., The Greenhouse Effect on ASHRAE, ASHRAE Journal, June 1989, pp. 52–54.
Maynich, P., and Bettano, M., CADD Aids Fast-Track Hospital Expansion, Engineering Systems, June 1994,
pp. 38–40.
McClive, J. R., Early Development in Air Conditioning and Heat Transfer, ASHRAE Transactions, vol. 92,
1986, Part II B, pp. 361–365.
Miller, A., Thompson, J. C., Peterson, R. E., and Haragan, D.R., Elements of Meteorology, 4th ed., Bell &
Howell Co., Columbus, OH, 1983.
Nagengast, B., The First Centrifugal Chiller: The German Connection, ASHRAE Journal, no. 1, 1998,
pp. 18–19.
Penny, T., and Althof, J., Trends in Commercial Buildings, Heating/Piping/Air Conditioning, September 1992,
pp. 59–66.
Rowland, F. S., The CFC Controversy: Issues and Answers, ASHRAE Journal, December 1992, pp. 20–27.
Simens, J., A Case for Unitary Systems, Heating/Piping/Air Conditioning, May 1982, pp. 60, 68–73.
Thomas, V. C., Mechanical Engineering Design Computer Programs for Buildings, ASHRAE Transactions,
1991, Part II, pp. 701–710.
Turner, F., Industrial News: Study Shows 49% of Suits Resolved without Payment, ASHRAE Journal, April
1996, pp. 10–11.
Whalen, J. M., An Organized Approach to Energy Management, Heating/Piping/Air Conditioning, September
1985, pp. 95–102.
Wilson, L., A Case for Central Systems, Heating/Piping/Air Conditioning, May 1982, pp. 61–67.
1.28 CHAPTER ONE
CHAPTER 2
PSYCHROMETRICS
2.1
2.1 PSYCHROMETRICS 2.1
Moist Air 2.1
Equation of State of an Ideal Gas 2.2
Equation of State of a Real Gas 2.2
Calculation of the Properties of
Moist Air 2.3
2.2 DALTON’S LAW AND THE GIBBSDALTON
LAW 2.3
2.3 AIR TEMPERATURE 2.4
Temperature and Temperature
Scales 2.4
Thermodynamic Temperature Scale 2.5
Temperature Measurements 2.6
2.4 HUMIDITY 2.7
Humidity Ratio 2.7
Relative Humidity 2.7
Degree of Saturation 2.8
2.5 PROPERTIES OF MOIST AIR 2.8
Enthalpy 2.8
Moist Volume 2.9
Density 2.10
Sensible Heat and Latent Heat 2.10
Specific Heat of Moist Air at
Constant Pressure 2.10
Dew-Point Temperature 2.11
2.6 THERMODYNAMIC WET-BULB
TEMPERATURE AND THE WET-BULB
TEMPERATURE 2.11
Ideal Adiabatic Saturation Process 2.11
Thermodynamic Wet-Bulb
Temperature 2.12
Heat Balance of an Ideal Adiabatic
Saturation Process 2.12
Psychrometer 2.12
Wet-Bulb Temperature 2.12
Relationship between Wet-Bulb
Temperature and Thermodynamic
Wet-Bulb Temperature 2.14
2.7 SLING AND ASPIRATION
PSYCHROMETERS 2.14
2.8 HUMIDITY MEASUREMENTS 2.16
Mechanical Hygrometers 2.16
Electronic Hygrometers 2.16
Comparison of Various Methods 2.19
2.9 PSYCHROMETRIC CHARTS 2.19
2.10 DETERMINATION OF
THERMODYNAMIC PROPERTIES ON
PSYCHROMETRIC CHARTS 2.22
Computer-Aided Psychrometric
Calculation and Analysis 2.25
REFERENCES 2.25
2.1 PSYCHROMETRICS
Psychrometrics is the study of the thermodynamic properties of moist air. It is used extensively to
illustrate and analyze the characteristics of various air conditioning processes and cycles.
Moist Air
The surface of the earth is surrounded by a layer of air called the atmosphere, or atmospheric air.
From the point of view of psychrometrics, the lower atmosphere, or homosphere, is a mixture of
dry air (including various contaminants) and water vapor, often known as moist air.
The composition of dry air is comparatively stable. It varies slightly according to geographic
location and from time to time. The approximate composition of dry air by volume percent is the
following:
The amount of water vapor present in moist air at a temperature range of 0 to 100°F (17.8 to
37.8°C) varies from 0.05 to 3 percent by mass. It has a significant influence on the characteristics of
moist air.
Water vapor is lighter than air. A cloud in the sky is composed of microscopic beads of liquid
water that are surrounded by a thin layer of water vapor. These layers give the cloud the needed
buoyancy to float in the air.
Equation of State of an Ideal Gas
The equation of state of an ideal gas indicates the relationship between its thermodynamic properties,
or
pv  RTR (2.1)
where p  pressure of gas, psf (Pa)
v  specific volume of gas, ft3 /lb (m3 /kg)
R  gas constant, ft  lbf / lbm°R (J/kgK)
TR  absolute temperature of gas, °R (K)
Since v  V/m, then Eq. (2.1) becomes
pv  mRTR (2.2)
where V  total volume of gas, ft3 (m3)
m  mass of gas, lb (kg)
Using the relationship m  nM, and R  Ro /M, we can write Eq. (2.2) as
pv  nRoTR (2.3)
where n  number of moles, mol
M  molecular weight
Ro  universal gas constant, ft  lbf / lbm°R (J/mol K)
Equation of State of a Real Gas
A modified form of the equation of state for a real gas can be expressed as
(2.4)
where A, B, C,   virial coefficients and Z  compressibility factor. The compressibility factor
Z illustrates the degree of deviation of the behavior of the real gas, moist air, from the ideal gas
due to the following:
1. Effect of air dissolved in water
pv
RTR
 1  Ap  Bp2  Cp3      Z
Nitrogen 78.08
Oxygen 20.95
Argon 0.93
Carbon dioxide 0.03
Other gases such as neon, sulfur dioxide, etc. 0.01
2.2 CHAPTER TWO
2. Variation of the properties of water vapor attributable to the effect of pressure
3. Effect of intermolecular forces on the properties of water vapor itself
For an ideal gas, Z  1. According to the information published by the former National Bureau
of Standards of the United States, for dry air at standard atmospheric pressure (29.92 in. Hg, or 760
mm Hg) and a temperature of 32 to 100°F (0 to 37.8°C) the maximum deviation is about 0.12
percent. For water vapor in moist air under saturated conditions at a temperature of 32 to 100°F
(0 to 37.8°C), the maximum deviation is about 0.5 percent.
Calculation of the Properties of Moist Air
The most exact calculation of the thermodynamic properties of moist air is based on the formulations
developed by Hyland and Wexler of the U.S. National Bureau of Standards. The psychrometric
chart and tables of ASHRAE are constructed and calculated from these formulations.
Calculations based on the ideal gas equations are the simplest and can be easily formulated. According
to the analysis of Nelson and Pate, at a temperature between 0 and 100°F (17.8 and 37.8°C),
calculations of enthalpy and specific volume using ideal gas equations show a maximum deviation of
0.5 percent from the exact calculations by Hyland and Wexler. Therefore, ideal gas equations will be
used in this text for the formulation and calculation of the thermodynamic properties of moist air.
Although air contaminants may seriously affect the health of occupants of the air conditioned
space, they have little effect on the thermodynamic properties of moist air since their mass concentration
is low. For simplicity, moist air is always considered as a binary mixture of dry air and water
vapor during the analysis and calculation of its properties.
2.2 DALTON’S LAW AND THE GIBBS-DALTON LAW
Dalton’s law shows that for a mixture of gases occupying a given volume at a certain temperature,
the total pressure of the mixture is equal to the sum of the partial pressures of the constituents of the
mixture, i.e.,
pm  p1  p2    (2.5)
where pm  total pressure of mixture, psia (Pa)
p1, p2, . . .  partial pressure of constituents 1, 2, . . . , psia (Pa)
The partial pressure exerted by each constituent in the mixture is independent of the existence of
other gases in the mixture. Figure 2.1 shows the variation of mass and pressure of dry air and water
vapor, at an atmospheric pressure of 14.697 psia (101,325 Pa) and a temperature of 75°F (23.9°C).
The principle of conservation of mass for nonnuclear processes gives the following relationship:
mm  ma  mw (2.6)
where mm  mass of moist air, lb (kg)
ma  mass of dry air, lb (kg)
mw  mass of water vapor, lb (kg)
Applying Dalton’s law for moist air, we have
pat  pa  pw (2.7)
where pat  atmospheric pressure or pressure of the outdoor moist air, psia (Pa)
pa  partial pressure of dry air, psia (Pa)
pw  partial pressure of water vapor, psia (Pa)
PSYCHROMETRICS 2.3
Dalton’s law is based on experimental results. It is more accurate for gases at low pressures.
Dalton’s law can be further extended to state the relationship of the internal energy, enthalpy, and
entropy of the gases in a mixture as the Gibbs-Dalton law:
mmum  m1u1  m2u2    
mmhm  m1h1  m2h2     (2.8)
mmsm  m1s1  m2 s2    
where mm  mass of gaseous mixture, lb (kg)
m1, m2, . . .  mass of the constituents, lb (kg)
um  specific internal energy of gaseous mixture, Btu/ lb (kJ /kg)
u1, u2, . . .  specific internal energy of constituents, Btu/ lb (kJ /kg)
hm  specific enthalpy of gaseous mixture, Btu/ lb (kJ /kg)
h1, h2, . . .  specific enthalpy of constituents, Btu/ lb (kJ /kg)
sm  specific entropy of gaseous mixture, Btu/ lb °R (kJ/kgK)
s1, s2, . . .  specific entropy of constituents, Btu/ lb °R (kJ/kgK)
2.3 AIR TEMPERATURE
Temperature and Temperature Scales
The temperature of a substance is a measure of how hot or cold it is. Two systems are said to have
equal temperatures only if there is no change in any of their observable thermal characteristics
when they are brought into contact with each other. Various temperature scales commonly used to
measure the temperature of various substances are illustrated in Fig. 2.2.
In conventional inch-pound (I-P) units, at a standard atmospheric pressure of 14.697 psia
(101,325 Pa), the Fahrenheit scale has a freezing point of 32°F (0°C) at the ice point, and a boiling
point of 212°F (100°C). For the triple point with a pressure of 0.08864 psia (611.2 Pa), the magnitude
on the Fahrenheit scale is 32.018°F (0.01°C). There are 180 divisions, or degrees, between the
boiling and freezing points in the Fahrenheit scale. In the International System of Units (SI units),
the Celsius or Centigrade scale has a freezing point of 0°C and a boiling point of 100°C. There are
2.4 CHAPTER TWO
FIGURE 2.1 Mass and pressure of dry air, water vapor, and moist air.
100 divisions between these points. The triple point is at 0.01°C. The conversion from Celsius scale
to Fahrenheit scale is as follows:
°F  1.8(°C)  32 (2.9)
For an ideal gas, at TR  0, the gas would have a vanishing specific volume. Actually, a real gas
has a negligible molecular volume when TR approaches absolute zero. A temperature scale that includes
absolute zero is called an absolute temperature scale. The Kelvin absolute scale has the same
boiling-freezing point division as the Celsius scale. At the freezing point, the Kelvin scale is 273.15 K.
Absolute zero on the Celsius scale is 273.15°C. The Rankine absolute scale division is equal to
that of the Fahrenheit scale. The freezing point is 491.67°R. Similarly, absolute zero is 459.67°F
on the Fahrenheit scale.
Conversions between Rankine and Fahrenheit and between Kelvin and Celsius systems are
R  459.67  °F (2.10)
K  273.15  °C (2.11)
Thermodynamic Temperature Scale
On the basis of the second law of thermodynamics, one can establish a temperature scale that is
independent of the working substance and that provides an absolute zero of temperature; this is
called a thermodynamic temperature scale. The thermodynamic temperature T must satisfy the
following relationship:
(2.12)
where Q  heat absorbed by reversible engine, Btu/h (kW)
Qo  heat rejected by reversible engine, Btu/h (kW)
TR  temperature of heat source of reversible engine, °R (K)
TRo  temperature of heat sink of reversible engine, °R (K)
Two of the ASHRAE basic tables, “Thermodynamic Properties of Moist Air” and “Thermodynamic
Properties of Water at Saturation,” in ASHRAE Handbook 1993, Fundamentals, are based on
the thermodynamic temperature scale.
TR
TRo

Q
Qo
PSYCHROMETRICS 2.5
FIGURE 2.2 Commonly used temperature scales.
Temperature Measurements
During the measurement of air temperatures, it is important to recognize the meaning of the terms
accuracy, precision, and sensitivity.
1. Accuracy is the ability of an instrument to indicate or to record the true value of the measured
quantity. The error indicates the degree of accuracy.
2. Precision is the ability of an instrument to give the same reading repeatedly under the same conditions.
3. Sensitivity is the ability of an instrument to indicate change of the measured quantity.
Liquid-in-glass instruments, such as mercury or alcohol thermometers, were commonly used in
the early days for air temperature measurements. In recent years, many liquid-in-glass thermometers
have been replaced by remote temperature monitoring and indication systems, made possible
by sophisticated control systems. A typical air temperature indication system includes sensors, amplifiers,
and an indicator.
Sensors. Air temperature sensors needing higher accuracy are usually made from resistance temperature
detectors (RTDs) made of platinum, palladium, nickel, or copper wires. The electrical
resistance of these resistance thermometers characteristically increases when the sensed ambient
air temperature is raised; i.e., they have a positive temperature coefficient . In many engineering
applications, the relationship between the resistance and temperature can be given by
(2.13)
where R  electric resistance, 	
R32, R212  electric resistance, at 32 and 212°F (0 and 100°C), respectively, 	
T  temperature, °F (°C)
The mean temperature coefficient  for several types of metal wires often used as RTDs is shown
below:
Many air temperature sensors are made from thermistors of sintered metallic oxides, i.e.,
semiconductors. They are available in a large variety of types: beads, disks, washers, rods, etc. Thermistors
have a negative temperature coefficient. Their resistance decreases when the sensed air temperature
increases. The resistance of a thermistor may drop from approximately 3800 to 3250 	 when
the sensed air temperature increases from 68 to 77°F (20 to 25°C). Recently developed high-quality
thermistors are accurate, stable, and reliable. Within their operating range, commercially available
thermistors will match a resistance-temperature curve within approximately 0.1°F (0.056°C). Some
manufacturers of thermistors can supply them with a stability of 0.05°F (0.028°C) per year. For direct
digital control (DDC) systems, the same sensor is used for both temperature indication, or monitoring,
and temperature control. In DDC systems, RTDs with positive temperature coefficient are
widely used.
Measuring range,°F , 	/°F
Platinum 400 to 1350 0.00218
Palladium 400 to 1100 0.00209
Nickel 150 to 570 0.0038
Copper 150 to 400 0.0038


R212  R32
180 R32
R  R32(1 
T )
2.6 CHAPTER TWO
Amplifier(s). The measured electric signal from the temperature sensor is amplified at the solid
state amplifier to produce an output for indication. The number of amplifiers is matched with the
number of the sensors used in the temperature indication system.
Indicator. An analog-type indicator, one based on directly measurable quantities, is usually a
moving coil instrument. For a digital-type indicator, the signal from the amplifier is compared with
an internal reference voltage and converted for indication through an analog-digital transducer.
2.4 HUMIDITY
Humidity Ratio
The humidity ratio of moist air w is the ratio of the mass of water vapor mw to the mass of dry air ma
contained in the mixture of the moist air, in lb / lb (kg/kg). The humidity ratio can be calculated as
(2.14)
Since dry air and water vapor can occupy the same volume at the same temperature, we can apply the
ideal gas equation and Dalton’s law for dry air and water vapor. Equation (2.14) can be rewritten as
(2.15)
where Ra, Rw  gas constant for dry air and water vapor, respectively, ftlbf / lbm°R(J/kgK). Equation
(2.15) is expressed in the form of the ratio of pressures; therefore, pw and pat must have the
same units, either psia or psf (Pa).
For moist air at saturation, Eq. (2.15) becomes
(2.16)
where pws  pressure of water vapor of moist air at saturation, psia or psf (Pa).
Relative Humidity
The relative humidity  of moist air, or RH, is defined as the ratio of the mole fraction of water vapor
xw in a moist air sample to the mole fraction of the water vapor in a saturated moist air sample
xws at the same temperature and pressure. This relationship can be expressed as
(2.17)
And, by definition, the following expressions may be written:
(2.18)
(2.19) xws 
nws
na  nws
xw 
nw
na  nw


xw
xws
T,p
ws  0.62198
pws
pat  pws

53.352
85.778
pw
pat  pw
 0.62198
pw
pat  pw
w 
mw
ma

pwVRaTR
PaVRwTR

Ra
Rw
pw
pat  pw
w 
mw
ma
PSYCHROMETRICS 2.7
where na  number of moles of dry air, mol
nw  number of moles of water vapor in moist air sample, mol
nws  number of moles of water vapor in saturated moist air sample, mol
Moist air is a binary mixture of dry air and water vapor; therefore, we find that the sum of the mole
fractions of dry air xa and water vapor xw is equal to 1, that is,
xa  xw  1 (2.20)
If we apply ideal gas equations pwV  nwRoTR and paV  naRoTR, by substituting them into
Eq. (2.19), then the relative humidity can also be expressed as
(2.21)
The water vapor pressure of saturated moist air pws is a function of temperature T and pressure p,
which is slightly different from the saturation pressure of water vapor ps. Here ps is a function of
temperature T only. Since the difference between pws and ps is small, it is usually ignored.
Degree of Saturation
The degree of saturation 	 is defined as the ratio of the humidity ratio of moist air w to the humidity
ratio of the saturated moist air ws at the same temperature and pressure. This relationship can be
expressed as
(2.22)
Since from Eqs. (2.15), (2.20), and (2.21) w  0.62198 xw/xa and ws  0.62198 xws /xa, Eqs. (2.20),
(2.21), and (2.22) can be combined, so that
(2.23)
In Eq. (2.23), pws pat ; therefore, the difference between and 	is small. Usually, the maximum
difference is less than 2 percent.
2.5 PROPERTIES OF MOIST AIR
Enthalpy
The difference in specific enthalpy 
h for an ideal gas, in Btu/ lb (kJ / kg), at a constant pressure can
be defined as

h  cp (T2  T1) (2.24)
where cp  specific heat at constant pressure, Btu/ lb °F (kJ/kgK)
T1, T2  temperature of ideal gas at points 1 and 2, °F (°C)
As moist air is approximately a binary mixture of dry air and water vapor, the enthalpy of the moist
air can be evaluated as
h  ha  Hw (2.25)


	
1  (1 
	)xws

	
1  (1 
	)( pws /pat )
	

w
ws
T,p


pw
pws
T,p
2.8 CHAPTER TWO
where ha and Hw are, respectively, enthalpy of dry air and total enthalpy of water vapor, in Btu/ lb
(kJ/ kg). The following assumptions are made for the enthalpy calculations of moist air:
1. The ideal gas equation and the Gibbs-Dalton law are valid.
2. The enthalpy of dry air is equal to zero at 0°F (17.8°C).
3. All water vapor contained in the moist air is vaporized at 0°F (17.8°C).
4. The enthalpy of saturated water vapor at 0°F (17.8°C) is 1061 Btu/lb (2468 kJ/ kg).
5. For convenience in calculation, the enthalpy of moist air is taken to be equal to the enthalpy of a mixture
of dry air and water vapor in which the amount of dry air is exactly equal to 1 lb (0.454 kg).
Based on the preceeding assumptions, the enthalpy h of moist air can be calculated as
h  ha  whw (2.26)
where hw  specific enthalpy of water vapor, Btu / lb (kJ /kg). In a temperature range of 0 to 100°F
(17.8 to 37.8°C), the mean value for the specific heat of dry air can be taken as 0.240 Btu/ lb °F
(1.005 kJ/kg K). Then the specific enthalpy of dry air ha is given by
ha  cpd T  0.240 T (2.27)
where cpd  specific heat of dry air at constant pressure, Btu/ lb °F (kJ/kgK)
T  temperature of dry air, °F (°C)
The specific enthalpy of water vapor hw at constant pressure can be approximated as
hw  hg0  cpsT (2.28)
where hg0  specific enthalpy of saturated water vapor at 0°F (17.8°C)—its value can be taken
as 1061 Btu/lb (2468 kJ/kg)
cps  specific heat of water vapor at constant pressure, Btu/ lb °F (kJ/kgK)
In a temperature range of 0 to 100°F (17.8 to 37.8°C), its value can be taken as 0.444 Btu/ lb °F
(1.859 kJ/kg K). Then the enthalpy of moist air can be evaluated as
h  cpdT  w(hg0  cpsT)  0.240 T  w(1061  0.444 T) (2.29)
Here, the unit of h is Btu/ lb of dry air (kJ / kg of dry air). For simplicity, it is often expressed as
Btu/ lb (kJ / kg).
Moist Volume
The moist volume of moist air v, ft3 /lb (m3 / kg), is defined as the volume of the mixture of the dry
air and water vapor when the mass of the dry air is exactly equal to 1 lb (1 kg), that is,
(2.30)
where V  total volume of mixture, ft3 (m3)
ma  mass of dry air, lb (kg)
In a moist air sample, the dry air, water vapor, and moist air occupy the same volume. If we apply
the ideal gas equation, then
(2.31) v 
V
ma

RaTR
pat  pw
v 
V
ma
PSYCHROMETRICS 2.9
where pat and pw are both in psf (Pa). From Eq. (2.15), pw  patw/(w  0.62198). Substituting this
expression into Eq. (2.31) gives
(2.32)
According to Eq. (2.32), the volume of 1 lb (1 kg) of dry air is always smaller than the volume of
the moist air when both are at the same temperature and the same atmospheric pressure.
Density
Since the enthalpy and humidity ratio are always related to a unit mass of dry air, for the sake of
consistency, air density a, in lb / ft3 (kg/m3), should be defined as the ratio of the mass of dry air to
the total volume of the mixture, i.e., the reciprocal of moist volume, or
(2.33)
Sensible Heat and Latent Heat
Sensible heat is that heat energy associated with the change of air temperature between two state
points. In Eq. (2.29), the enthalpy of moist air calculated at a datum state 0°F (17.8°C) can be
divided into two parts:
h  (cpd  wcps)T  whg0 (2.34)
The first term on the right-hand side of Eq. (2.34) indicates the sensible heat. It depends on the temperature
T above the datum 0°F (17.8°C).
Latent heat hfg (sometimes called hig) is the heat energy associated with the change of the state
of water vapor. The latent heat of vaporization denotes the latent heat required to vaporize liquid
water into water vapor. Also, the latent heat of condensation indicates the latent heat to be removed
in the condensation of water vapor into liquid water. When moisture is added to or removed from a
process or a space, a corresponding amount of latent heat is always involved, to vaporize the water
or to condense it.
In Eq. (2.34), the second term on the right-hand side, whg0 , denotes latent heat. Both sensible
and latent heat are expressed in Btu/ lb (kJ / kg) of dry air.
Specific Heat of Moist Air at Constant Pressure
The specific heat of moist air at constant pressure cpa is defined as the heat required to raise its temperature
1°F (0.56°C) at constant pressure. In (inch-pound) I-P units, it is expressed as Btu/lb°F (in SI
units, as J/kgK). In Eq. (2.34), the sensible heat of moist air qsen, Btu/h (W), is represented by
(2.35)
where mass flow rate of moist air, lb/h (kg/ s). Apparently
cpa  cpd  wcps (2.36)
Since cpd and cps are both a function of temperature, cpa is also a function of temperature and, in addition,
a function of the humidity ratio.
For a temperature range of 0 to 100°F (17.8 to 37.8°C), cpd can be taken as 0.240 Btu/ lb °F
(1005 J/kgK) and cps as 0.444 Btu/ lb °F (1859 J/kg K). Most of the calculations of cpa(T2  T1)
m?a 
qsen  m?a(cpd  wcps)T  m?acpaT
a 
ma
V

1
v
v 
RaTR(1  1.6078 w)
Pat
2.10 CHAPTER TWO
have a range of w between 0.005 and 0.010 lb/lb (kg/ kg). Taking a mean value of w  0.0075
lb / lb (kg / kg), we find that
cpa  0.240  0.0075 0.444  0.243 Btu/ lb°F (1020 J/kgK)
Dew-Point Temperature
The dew-point temperature Tdew is the temperature of saturated moist air of the same moist air sample,
having the same humidity ratio, and at the same atmospheric pressure of the mixture pat. Two
moist air samples at the same Tdew will have the same humidity ratio w and the same partial pressure
of water vapor pw. The dew-point temperature is related to the humidity ratio by
ws ( pat, Tdew)  w (2.37)
where ws  humidity ratio of saturated moist air, lb/ lb (kg / kg). At a specific atmospheric pressure,
the dew-point temperature determines the humidity ratio w and the water vapor pressure pw of the
moist air.
2.6 THERMODYNAMIC WET-BULB TEMPERATURE AND
THE WET-BULB TEMPERATURE
Ideal Adiabatic Saturation Process
If moist air at an initial temperature T1, humidity ratio w1, enthalpy h1, and pressure p flows over a
water surface of infinite length in a well-insulated chamber, as shown in Fig. 2.3, liquid water will
evaporate into water vapor and will disperse in the air. The humidity ratio of the moist air will gradually
increase until the air can absorb no more moisture.
As there is no heat transfer between this insulated chamber and the surroundings, the latent heat
required for the evaporation of water will come from the sensible heat released by the moist air.
This process results in a drop in temperature of the moist air. At the end of this evaporation process,
the moist air is always saturated. Such a process is called an ideal adiabatic saturation process,
where an adiabatic process is defined as a process without heat transfer to or from the process.
PSYCHROMETRICS 2.11
FIGURE 2.3 Ideal adiabatic saturation process.
Thermodynamic Wet-Bulb Temperature
For any state of moist air, there exists a thermodynamic wet-bulb temperature T* that exactly
equals the saturated temperature of the moist air at the end of the ideal adiabatic saturation process
at constant pressure. Applying a steady flow energy equation, we have
(2.38)
where enthalpy of moist air at initial state and enthalpy of saturated air at end of ideal adiabatic
saturation process, Btu/ lb (kJ /kg)
 humidity ratio of moist air at initial state and humidity ratio of saturated air at end
of ideal adiabatic saturation process, lb / lb (kg /kg)
 enthalpy of water as it is added to chamber at a temperature T*, Btu/ lb (kJ /kg)
The thermodynamic wet-bulb temperature T*, °F (°C), is a unique property of a given moist air
sample that depends only on the initial properties of the moist air—w1, h1 and p. It is also a fictitious
property that only hypothetically exists at the end of an ideal adiabatic saturation process.
Heat Balance of an Ideal Adiabatic Saturation Process
When water is supplied to the insulation chamber at a temperature T* in an ideal adiabatic saturation
process, then the decrease in sensible heat due to the drop in temperature of the moist air is just
equal to the latent heat required for the evaporation of water added to the moist air. This relationship
is given by
(2.39)
where T1  temperature of moist air at initial state of ideal adiabatic saturation process, °F (°C)
hfg *  latent heat of vaporization at thermodynamic wet-bulb temperature, Btu/ lb (J /kg)
Since cpa  cpd  w1cps, we find, by rearranging the terms in Eq. (2.39),
(2.40)
Also
(2.41)
Psychrometer
A psychrometer is an instrument that permits one to determine the relative humidity of a moist air
sample by measuring its dry-bulb and wet-bulb temperatures. Figure 2.4 shows a psychrometer,
which consists of two thermometers. The sensing bulb of one of the thermometers is always kept
dry. The temperature reading of the dry bulb is called the dry-bulb temperature. The sensing bulb of
the other thermometer is wrapped with a piece of cotton wick, one end of which dips into a cup of
distilled water. The surface of this bulb is always wet; therefore, the temperature that this bulb
measures is called the wet-bulb temperature. The dry bulb is separated from the wet bulb by a radiation-
shielding plate. Both dry and wet bulbs are cylindrical.
Wet-Bulb Temperature
When unsaturated moist air flows over the wet bulb of the psychrometer, liquid water on the surface
of the cotton wick evaporates, and as a result, the temperature of the cotton wick and the wet bulb
drops. This depressed wet-bulb reading is called the wet-bulb temperature T, and the difference between
the dry-bulb and wet-bulb temperatures is called the wet-bulb depression.
Let us neglect the conduction along the thermometer stems to the dry and wet bulbs and also
assume that the temperature of the water on the cotton wick is equal to the wet-bulb temperature of
the moist air. Since the heat transfer from the moist air to the cotton wick is exactly equal to the
latent heat required for vaporization, then, at steady state, the heat and mass transfer per unit area of
the wet-bulb surface can be calculated as
(2.42)
where hc, hr  mean convective and radiative heat transfer coefficients, Btu/h ft2 °F (W/m2 K)
hd  mean convective mass-transfer coefficient, lb /h ft2 (kg/ s m2)
T  temperature of undisturbed moist air at a distance from wet bulb, °F (°C)
T  wet-bulb temperature, °F (°C)
Tra  mean radiant temperature, °F (°C)
w1, ws  humidity ratio of moist air and saturated air film at surface of cotton wick, lb / lb
(kg/kg)
latent heat of vaporization at wet-bulb temperature, Btu/ lb (J /kg)
Based on the correlation of cross-flow forced convective heat transfer for a cylinder, NuD  C Ren
Pr0.333, and based on the analogy between convective heat transfer and convective mass transfer, the
following relationship holds: hd  hc/(cpaLe0.6667). Here, Nu is the Nusselt number, Re the
Reynolds number, and Le the Lewis number. Also C is a constant.
Substituting this relationship into Eq. (2.42), we have
(2.43)

The term T  T indicates the wet-bulb depression.
Combining Eqs. (2.43) and (2.44) then gives
(2.45)
Relationship between Wet-Bulb Temperature and Thermodynamic
Wet-Bulb Temperature
Wet-bulb temperature is a function not only of the initial state of moist air, but also of the rate of
heat and mass transfer at the wet bulb. Comparing Eq. (2.40) with Eq. (2.45), we find that the wetbulb
temperature measured by using a psychrometer is equal to thermodynamic wet-bulb temperature
only when the following relationship holds:
(2.46)
2.7 SLING AND ASPIRATION PSYCHROMETERS
Sling and aspiration psychrometers determine the relative humidity through the measuring of the
dry- and wet-bulb temperatures.
A sling psychrometer with two bulbs, one dry and the other wet, is shown in Fig. 2.5a. Both dry
and wet bulbs can be rotated around a spindle to produce an airflow over the surfaces of the dry and
wet bulbs at an air velocity of 400 to 600 fpm (2 to 3 m/ s). Also a shield plate separates the dry
and wet bulbs and partly protects the wet bulb against surrounding radiation.
An aspiration psychrometer that uses a small motor-driven fan to produce an air current flowing
over the dry and wet bulbs is illustrated in Fig. 2.5b. The air velocity over the bulbs is usually kept
at 400 to 800 fpm (2 to 4 m/ s). The dry and wet bulbs are located in separate compartments and are
entirely shielded from the surrounding radiation. When the space dry-bulb temperature is within a
range of 75 to 80°F (24 to 27°C) and the space wet-bulb temperature is between 65 and 70°F (18
and 21°C), the following wet-bulb constants K can be used to calculate the humidity ratio of the
moist air:
Aspiration psychrometer K  0.000206 1/°F
Sling psychrometer K  0.000218 1/°F
After the psychrometer has measured the dry- and wet-bulb temperatures of the moist air, the
humidity ratio w can be calculated by Eq. (2.43). Since the saturated water vapor pressure can be
found from the psychrometric table, the relative humidity of moist air can be evaluated through
Eq. (2.15).
According to the analysis of Threlkeld (1970), for a wet-bulb diameter of 0.1 in. (2.5 mm) and
an air velocity flowing over the wet bulb of 400 fpm (2 m/ s), if the dry-bulb temperature is 90°F
(32.2°C) and the wet-bulb temperature is 70°F (21.1°C ), then (T  T*)/(T  T) is about 2.5
percent. Under the same conditions, if a sling psychrometer is used, then the deviation may be
reduced to about 1 percent. If the air velocity flowing over the wet bulb exceeds 400 fpm (2 m/ s),
there is no significant reduction in the deviation.
Distilled water must be used to wet the cotton wick for both sling and aspiration psychrometers.
Because dusts contaminate them, cotton wicks should be replaced regularly to provide a clean surface
for evaporation.

2.8 HUMIDITY MEASUREMENTS
Humidity sensors used in HVAC&R for direct humidity indication or operating controls are separated
into the following categories: mechanical hygrometers and electronic hygrometers.
Mechanical Hygrometers
Mechanical hygrometers operate on the principle that hygroscopic materials expand when they
absorb water vapor or moisture from the ambient air. They contract when they release moisture
to the surrounding air. Such hygroscopic materials include human and animal hairs, plastic polymers
like nylon ribbon, natural fibers, wood, etc. When these materials are linked to mechanical
linkages or electric transducers that sense the change in size and convert it into electric signals, the
results in these devices can be calibrated to yield direct relative-humidity measurements of the
ambient air.
Electronic Hygrometers
There are three types of electronic hygrometers: Dunmore resistance hygrometers, ion-exchange
resistance hygrometers, and capacitance hygrometers.
Dunmore Resistance Hygrometer. In 1938, Dunmore of the National Bureau of Standards developed
the first lithium chloride resistance electric hygrometer in the United States. This instrument
depends on the change in resistance between two electrodes mounted on a hygroscopic material.
Figure. 2.6a shows a Dunmore resistance sensor. The electrodes could be, e.g., a double-threaded
winding of noble-metal wire mounted on a plastic cylinder coated with hygroscopic material. The
wires can also be in a grid-type arrangement with a thin film of hygroscopic material bridging the
gap between the electrodes.
At a specific temperature, electric resistance decreases with increasing humidity. Because the
response is significantly influenced by temperature, the results are often indicated by a series of
isothermal curves. Relative humidity is generally used as the humidity parameter, for it must
be controlled in the indoor environment. Also the electrical response is more nearly a function of
relative humidity than of the humidity ratio.
The time response to accomplish a 50 percent change in relative humidity varies directly according
to the air velocity flowing over the sensor and also is inversely proportional to the saturated
vapor pressure. If a sensor has a response time of 10 s at 70°F (21°C), it might need a response time
of 100 s at 10°F (12°C).
Because of the steep variation of resistance corresponding to a change in relative humidity, each
of the Dunmore sensors only covers a certain range of relative-humidity measurements. A set of
several Dunmore sensors is usually needed to measure relative humidity between 1 percent and 100
percent.
A curve for output, in direct-current (dc) volts, versus relative humidity is shown in Fig. 2.6b
for a typical Dunmore sensor. It covers a measuring range of 10 to 80 percent, which is usually
sufficient for the indication of relative humidity for a comfort air conditioned space. This typical
Dunmore sensor has an accuracy of 3 percent when the relative humidity varies between 10 and
60 percent at a temperature of 70°F (21°C, see Fig. 2.6b). Its accuracy reduces to 4 percent when
the relative humidity is in a range between 60 and 80 percent at the same temperature.
In addition to lithium chloride, lithium bromide is sometimes used as the sensor.
Ion-Exchange Resistance Hygrometer. The sensor of a ion-exchange resistance electric hygrometer
is composed of electrodes mounted on a baseplate and a high-polymer resin film, used as a
humidity-sensing material, cross-linking the electrodes as shown in Fig. 2.7a and b. Humidity is
measured by the change in resistance between the electrodes. When the salt contained in the humidity-
sensitive material bridging the electrodes becomes ion-conductive because of the presence of
water vapor in the ambient air, mobile ions in the polymer film are formed. The higher the relative
humidity of the ambient air, the greater the ionization becomes, and therefore, the greater the
concentration of mobile ions. On the other hand, lower relative humidity reduces the ionization and
results in a lower concentration of mobile ions. The resistance of the humidity-sensing material
reflects the change of the relative humidity of the ambient air.
In Fig. 2.7b, the characteristic curves of an ion-exchange resistance electric hygrometer show
that there is a nonlinear relationship between resistance R and relative humidity I/>. These sensors
cover a wider range than Dunmore sensors, from 20 to 90 percent relative humidity.
Capacitance Hygrometer. The commonly used capacitance sensor consists of a thin-film plastic
foil. A very thin gold coating covers both sides of the film as electrodes, and the film is mounted
2.18 CHAPTER TWO
inside a capsule. The golden electrodes and the dividing plastic foil form a capacitor. Water vapor
penetrates the gold layer, which is affected by the vapor pressure of the ambient air and, therefore,
the ambient relative humidity. The number of water molecules absorbed on the plastic foil determines
the capacitance and the resistance between the electrodes.
FIGURE 2.7 Ion-exchange resistance-type electric hygrometer. (a)
Front and side view of sensor; (b) characteristic curve of R versus .
(Reprinted by permission of General Eastern Instruments.)
Comparison of Various Methods
The following table summarizes the sensor characteristics of various methods to be used within a
temperature range of 32 to 120°F (0 to 50°C) and a range of 10 to 95 percent RH:
Psychrometers are simple and comparatively low in cost. They suffer no irreversible damage at 100
percent RH, as do the sensors of electric hygrometers. Unfortunately, complete wet-bulb depression
readings of psychrometers become difficult when relative humidity drops below 20 percent or when
the temperature is below the freezing point. For remote monitoring, it is difficult to keep sufficient
water in the water reservoir. Therefore, psychrometers are sometimes used to check the temperature
and relative humidity in the air conditioned space manually.
Mechanical hygrometers directly indicate the relative humidity of the moist air. They are also
simple and relatively inexpensive. Their main drawbacks are their lack of precision over an extensive
period and their lack of accuracy at extreme high and low relative humidities. Electronic
hygrometers, especially the polymer film resistance and the capacitance types, are commonly used
for remote monitoring and for controls in many air conditioning systems. Both the electronic and
mechanical hygrometers need regular calibration. Initial calibrations are usually performed either
with precision humidity generators using two-pressure, two-temperature, and divided-flow systems
or with secondary standards during manufacturing (refer to ASHRAE Standard 41.6-1982, Standard
Method for Measurement of Moist Air Properties). Regular calibrations can be done with a
precision aspiration psychrometer or with chilled mirror dew-point devices.
Air contamination has significant influence on the performance of the sensor of electronic and
mechanical hygrometers. This is one of the reasons why they need regular calibration.
2.9 PSYCHROMETRIC CHARTS
Psychrometric charts provide a graphical representation of the thermodynamic properties of moist
air, various air conditioning processes, and air conditioning cycles. The charts are very helpful during
the calculation, analysis, and solution of the complicated problems encountered in air conditioning
processes and cycles.
Basic Coordinates. The currently used psychrometric charts have two types of coordinates:
 h-w chart. Enthalpy h and humidity ratio w are basic coordinates. The psychrometric charts published
by ASHRAE and the Chartered Institution of Building Services Engineering (CIBSE) are
h-w charts.
 T-w chart. Temperature T and humidity ratio w are basic coordinates. Most of the psychrometric
charts published by the large manufacturers in the United States are T-w charts.
For an atmospheric pressure of 29.92 in. Hg (760 mm Hg), an air temperature of 84°F (28.9°C),
and a relative humidity of 100 percent, the humidity ratios and enthalpies found from the psychrometric
charts published by ASHRAE and Carrier International Corporation are shown below:
Operating method Accuracy, % RH
Psychrometer Wet-bulb depression 3
Mechanical Dimensional change 3 to 5
Dunmore Electric resistance 1.5
Ion-exchange Electric resistance 2 to 5
Capacitance Electric capacitance and resistance 3 to 5
PSYCHROMETRICS 2.19
The last digit for humidity ratios and for enthalpies read from ASHRAE’s chart is an approximation.
Nevertheless, the differences between the two charts are less than 1 percent, and these are considered
negligible.
In this handbook, for manual psychrometric calculations and analyses, ASHRAE’s chart will be
used.
Temperature Range and Barometric Pressure. ASHRAE’s psychrometric charts are constructed
for various temperature ranges and altitudes. In Appendix B only the one for normal temperature,
that is, 32 to 120°F (0 to 50°C), and a standard barometric pressure at sea level, 29.92 in. Hg
(760 mm Hg), is shown. The skeleton of ASHRAE’s chart is shown in Fig. 2.8.
Enthalpy Lines. For ASHRAE’s chart, the molar enthalpy of moist air is calculated from the
formulation recommended by Hyland and Wexler (1983) in their paper “Formulations for the Thermodynamic
Properties of Dry Air from 173.15 K to 473.15 K, and of Saturated Moist Air from
173.15 K to 473.15 K, at Pressures to 5 MPa.” For ASHRAE’s chart, the enthalpy h lines incline at an
angle of 25° to the horizontal lines. The scale factor for the enthalpy lines Ch, Btu/lb ft (kJ/kgm), is
(2.47)
where Lh  shortest distance between enthalpy lines h2 and h1, ft (m).

Humidity Ratio Lines. In ASHRAE’s chart, the humidity ratio w lines are horizontal. They form
the ordinate of the psychrometric chart. The scale factor Cw, lb/lb ft (kg /kgm), for w lines in
ASHRAE’s chart is
(2.48)
where Lw  vertical distance between w2 and w1, ft (m). For ASHRAE’s chart, the humidity ratio w
can be calculated by Eq. (2.15).
Constant-Temperature Lines. For ASHRAE’s chart, since enthalpy is one of the coordinates,
only the 120°F constant-temperature T line is a true vertical. All the other constant-temperature
lines incline slightly to the left at the top. From Eq. (2.27), T  ha /cpd; therefore, one end of the T
line in ASHRAE’S chart can be determined from the enthalpy scale at w  0. The other end can be
determined by locating the saturated humidity ratio ws on the saturation curve.
Saturation Curve. A saturation curve is a locus representing a series of state points of saturated
moist air. For ASHRAE’s chart, the enthalpy of saturation vapor over liquid water or over ice at a
certain temperature is calculated by the formula recommended in the Hyland and Wexler paper. The
humidity ratios of the saturated moist air ws between 0 and 100°F (17.8 and 37.8°C) in the psychrometric
chart can also be calculated by the following simpler polynomial:
(2.49)
where Ts  saturated temperature of moist air, °F (°C)
a1  0.00080264
a2  0.000024525
a3  2.5420 106
a4   2.5855 108
a5  4.038 1010
If we use Eq. (2.49) to calculate ws, the error is most probably less than 0.000043 lb/lb (kg / kg). It
is far smaller than the value that can be identified on the psychrometric chart, and therefore the calculated
ws is acceptable.
Relative-Humidity and Moist Volume Lines. For ASHRAE’s chart, relative-humidity  lines,
thermodynamic wet-bulb T* lines, and moist volume v lines all are calculated and determined
based on the formulations in Hyland and Wexler’s paper.
Thermodynamic Wet-Bulb Lines. For ASHRAE’s chart, only thermodynamic wet-bulb T* lines
are shown. Since the dry-bulb and the thermodynamic wet-bulb temperatures coincide with each
other on the saturation curve, one end of the T* line is determined. The other end of the T* line can
be plotted on the w line where w  0. Let the state point of the other end of the T* line be represented
by 1. 
Cooling and Dehumidifying Curves. The two cooling and dehumidifying curves plotted
on ASHRAE’s chart are based on data on coil performance published in the catalogs of U.S.
manufacturers. These curves are very helpful in describing the actual locus of a cooling and dehumidifying
process as well as determining the state points of air leaving the cooling coil.
2.10 DETERMINATION OF THERMODYNAMIC PROPERTIES ON
PSYCHROMETRIC CHARTS
There are seven thermodynamic properties or property groups of moist air shown on a psychrometric
chart:
1. Enthalpy h
2. Relative humidity 
3. Thermodynamic wet-bulb temperature T*
4. Barometric or atmospheric pressure pat
5. Temperature T and saturation water vapor pressure pws
6. Density  and moist volume v
7. Humidity ratio w, water vapor pressure pw, and dew-point temperature Tdew
The fifth, sixth, and seventh are thermodynamic property groups. These properties or properties
groups are independent of each other except that the difference in slope between the enthalpy h line
and thermodynamic wet-bulb temperature T* line is small, and it is hard to determine their intersection.
Usually, atmospheric pressure pat is a known value based on the altitude of the location. Then, in
the fifth property group, pws is a function of temperature T only. In the sixth property group, according
to Eq. (2.33), a  1/v; that is, air density and moist volume are dependent on each other. In the
seventh property group, for a given value of pat, properties w, pw, and Tdew are all dependent on each
other.
When pat is a known value, and if the moist air is not saturated, then any two known independent
thermodynamic properties can determine the magnitudes of the remaining unknown properties. If
the moist air is saturated, then any independent property will determine the remaining magnitudes.
Example 2.1. The design indoor air temperature and relative humidity of an air conditioned space
at sea level are 75°F (23.9°C) and 50 percent. Find the humidity ratio, the enthalpy, and the density
of the indoor moist air
1. By using the ASHRAE chart
2. By calculation
Determine also the dew-point and thermodynamic wet-bulb temperatures of the moist air. The
following information is required for the calculations:

1. Plot the space point r on ASHRAE’s chart by first finding the space temperature Tr  75°F on
the abscissa and then following along the 75°F constant-temperature line up to a relative humidity
 50 percent, as shown in Fig. 2.9.
Draw a horizontal line from the space point r. This line meets the ordinate, humidity ratio w, at
a value of wr  0.00927 lb/lb (kg/ kg). This is the humidity ratio of the indoor space air. Draw a
line parallel to the enthalpy line from the space point r. This line meets the enthalpy scale line at
a value of hr  28.1 Btu/ lb. This is the enthalpy of the indoor space air. Draw a horizontal line
from the space point r to the left. This line meets the saturation curve at a dew-point temperature
of 55°F (12.8°C). Draw a line parallel to the thermodynamic wet-bulb temperature lines through
the space point r. The perpendicular scale to this line shows a thermodynamic wet-bulb temperature
 62.5°F (16.9°C). Draw a line parallel to the moist volume lines through the space point r.
The perpendicular scale to this line shows a moist volume vr  13.68 ft3/lb (0.853 m3 / kg).
2. The calculations of the humidity ratio, enthalpy, and moist volume are as follows:
From Eq. (2.49), the humidity ratio of the saturated air at the dry-bulb temperature is
ws  0.00080264  0.000024525T  2.542e-06T 2  2.5855e-08T 3  4.038e-10T 4
 0.00080264  0.0018394  0.014299  0.010908  0.012776
 0.018809 lb/ lb
According to Eq. (2.16), the saturated water vapor pressure of the indoor air is
T*r
PSYCHROMETRICS 2.23
FIGURE 2.9 Thermodynamic properties determined from ASHRAE’s psychrometric chart.
From Eq. (2.21), the water vapor pressure of indoor air is
pw  pws  0.5 0.4314  0.2157 psia
Then, from Eq. (2.15), the humidity ratio of the indoor air is
From Eq. (2.29), the enthalpy of the indoor moist air is
hr  cpdTF  w (hg  cpsT)
hr  0.240 75  0.009264(1061  0.444 75)
 28.14 Btu/ lb
From Eq. (2.32), the moist volume of the indoor air is
Then, from Eq. (2.33), the density of the indoor moist air is
Comparison of the thermodynamic properties read directly from ASHRAE’s chart and the calculated
values is as follows:
Apparently, the differences between the readings from ASHRAE’s chart and the calculated values
are very small.
Example 2.2. An HVAC&R operator measured the dry- and wet-bulb temperatures in an air
conditioned space as 75°F (23.9°C) and 63°F (17.2°C), respectively. Find the relative humidity
of this air conditioned space by using ASHRAE’s chart and by calculation. The humidity ratios
of the saturated air at temperatures of 75 and 63°F are 0.018809 and 0.012355 lb/lb (kg/ kg),
respectively.
Solution. It is assumed that the difference between the wet-bulb temperature as measured by
sling or aspiration psychrometer and the thermodynamic wet-bulb temperature is negligible. At a
measured dry-bulb temperature of 75°F (23.9°C) and a wet-bulb temperature of 63°F (17.2°C), the
relative humidity read directly from ASHRAE’s chart is about 51.8 percent. From Sec. 2.6, the wetbulb
constant K for a sling psychrometer is 0.000218 1/°F. Then, from Eq. (2.43), the humidity
ratio of the space air can be calculated as

From Eq. (2.15), the vapor pressure of the space air is
And from Eq. (2.16), the saturated vapor pressure at a space temperature of 75°F is
Hence, from Eq. (2.21), the calculated relative humidity of the space air is
The difference between the value read directly from ASHRAE’s chart and the calculated one is
52.53 percent  51.8 percent  0.7 percent.
Computer-Aided Psychrometric Calculation and Analysis
There are two kinds of psychrometric computer-aided software available on the market: psychrometric
calculations and psychrometric graphics. Most of the psychrometric software is Windowsbased
computer programs.
Software for psychrometric calculations can determine any one of the thermodynamic properties
of the moist air if two of the independent properties are known. Psychrometric calculation software
usually also finds the thermodynamic property of the mixture of airstreams and provides altitude
effect adjustments. Software programs for psychrometric graphics are far more powerful tools than
psychrometric calculations. On the computer screen, the following is shown: a psychrometic chart,
any number of labeled air conditioned state points, the corresponding air conditioned processes and
the air conditioning cycle, and the spreadsheet that lists the thermodynamic properties, airflow in
cubic feet per minute, and the heat transfer during the air conditioning processes. The thermodynamic
properties of any of the state points and therefore the characteristics of the air conditioning
process and cycle can be varied. As a result, the operation of the air system at either full load or part
load, under cooling or heating modes, can be investigated and analyzed.

Aslam, S., Charmchi, M., and Gaggioli, R. A., Psychrometric Analysis for Arbitrary Dry-Gas Mixtures and
Pressures Using Microcomputers, ASHRAE Transactions, 1986, Part I B, pp. 448–460.
Hedlin, C. P., Humidity Measurement with Dunmore Type Sensors, Symposium at ASHRAE Semiannual
Meeting, ASHRAE Inc., New York, February 1968.
Hyland, R. W., and Wexler, A., Formulations for the Thermodynamic Properties of Dry Air from 173.15 K to
473.15 K, and of Saturated Moist Air from 173.15 K to 372.15 K, at Pressures to 5 MPa, ASHRAE Transactions,
1983, Part II A, pp. 520–535.
Hyland, R. W., and Wexler, A., Formulations for the Thermodynamic Properties of the Saturated Phases of H2O
from 173.15 K to 473.15 K, ASHRAE Transactions, 1983, Part II A, pp. 500–519.
Kamm, V., New Psychrometric Software Offer Free on the Internet, The Air Conditioning, Heating and Refrigeraion
News , July 22, 1996, pp. 18–19.
McGee, T. D., Principles and Methods of Temperature Measurements, 1st ed.,Wiley, New York, 1988.
Nelson, R. M., and Pate, M., A Comparison of Three Moist Air Property Formulations for Computer Applications,
ASHRAE Transactions, 1986, Part I B, pp. 435–447.
Stewart, R. B., Jacobsen, R. T., and Becker, J. H., Formulations for Thermodynamic Properties of Moist Air at
Low Pressures as Used for Construction of New ASHRAE SI Unit Psychrometric Charts, ASHRAE Transactions,
1983, Part II A, pp. 536–548.
Threlkeld, J. L., Thermal Environmental Engineering, 2d ed., Prentice-Hall, Englewood Cliffs, NJ, 1970.
The Trane Company, Psychrometry, La Crosse, WI, 1979.
Wang, S. K., Air Conditioning, vol. 1, 1st ed., Hong Kong Polytechnic, Hong Kong, 1987.
2.26 CHAPTER TWO
CHAPTER 3
HEAT AND MOISTURE
TRANSFER THROUGH
BUILDING ENVELOPE
3.1
3.1 BUILDING ENVELOPE 3.2
3.2 HEAT-TRANSFER
FUNDAMENTALS 3.2
Conductive Heat Transfer 3.3
Convective Heat Transfer 3.4
Radiant Heat Transfer 3.5
Overall Heat Transfer 3.6
Heat Capacity 3.8
3.3 HEAT-TRANSFER COEFFICIENTS 3.8
Coefficients for Radiant Heat
Transfer 3.8
Coefficients for Forced Convection 3.9
Coefficients for Natural Convection 3.10
Surface Heat-Transfer Coefficients 3.10
3.4 MOISTURE TRANSFER 3.11
Sorption Isotherm 3.11
Moisture-Solid Relationship 3.12
Moisture Migration in Building
Materials 3.13
Moisture Transfer from the Surface of
the Building Envelope 3.14
Convective Mass-Transfer
Coefficients 3.15
Moisture Transfer in Building
Envelopes 3.16
3.5 CONDENSATION IN BUILDINGS 3.17
Visible Surface Condensation 3.17
Concealed Condensation within the
Building Envelope 3.18
3.6 THERMAL INSULATION 3.18
Basic Materials and Thermal
Properties 3.19
Moisture Content of Insulation
Material 3.19
Economic Thickness 3.21
Thermal Resistance of Airspaces 3.21
3.7 SOLAR ANGLES 3.22
Basic Solar Angles 3.22
Hour Angle and Apparent Solar
Time 3.24
Solar Angle Relationships 3.24
Angle of Incidence and Solar
Intensity 3.24
3.8 SOLAR RADIATION 3.25
Solar Radiation for a Clear Sky 3.26
Solar Radiation for a Cloudy Sky 3.28
3.9 FENESTRATION 3.29
Types of Window Glass (Glazing) 3.29
Optical Properties of Sunlit Glazing 3.30
3.10 HEAT ADMITTED THROUGH
WINDOWS 3.32
Heat Gain for Single Glazing 3.32
Heat Gain for Double Glazing 3.34
Shading Coefficients 3.36
Solar Heat Gain Factors and Total
Shortwave Irradiance 3.37
Selection of Glazing 3.39
3.11 SHADING OF GLASS 3.40
Indoor Shading Devices 3.40
External Shading Devices 3.42
Shading from Adjacent Buildings 3.43
3.12 HEAT EXCHANGE BETWEEN THE
OUTER BUILDING SURFACE AND ITS
SURROUNDINGS 3.46
Sol-Air Temperature 3.47
3.13 COMPLIANCE WITH ASHRAE/IESNA
STANDARD 90.1-1999 FOR BUILDING
ENVELOPE 3.48
General Requirements 3.48
Mandatory Provisions 3.49
Prescriptive Building Envelope
Option 3.49
Building Envelope Trade-Off
Option 3.50
3.14 ENERGY-EFFICIENT AND COSTEFFECTIVE
MEASURES FOR BUILDING
ENVELOPE 3.50
Exterior Walls 3.50
Windows 3.50
Infiltration 3.51
Energy-Efficient Measures for
Commercial Buildings in
the United States 3.51
REFERENCES 3.51
3.1 BUILDING ENVELOPE
Building envelope consists of the building components that enclose conditioned spaces. Heat,
moisture, and contaminants may be transferred to or from the outdoors or unconditioned spaces and
therefore affect the indoor environment of the conditioned space.
Building envelope used for air conditioned space in buildings consists of mainly walls, roofs,
windows, ceilings, and floors. There are two types of partitions: exterior partitions and demising
partitions. An exterior partition is an opaque, translucent, or transparent solid barrier that separates
conditioned space from outdoors or space which is not enclosed. A demising partition is a solid
barrier that separates conditioned space from enclosed unconditioned space.
 An exterior wall is a solid exterior partition which separates conditioned space from the outdoors.
Exterior walls are usually made from composite layers including any of the following: stucco,
bricks, concrete, concrete blocks, wood, thermal insulation, vapor barrier, airspace, and interior finish.
The gross exterior wall area is the sum of the window area, door area, and exterior wall area.
 A partition wall is an interior solid barrier which separates a conditioned space from others. Partition
walls are usually made from interior surface finishes on two sides, wooden studs and boards,
concrete blocks, concrete, bricks, and thermal insulation. A demising partition wall separates a
conditioned space from an enclosed unconditioned space.
 Wall below grade is a solid barrier below ground level which often separates a basement or a
crawl space from soil.
 A roof is an exterior partition that has a slope less than 60° from horizontal and has a conditioned
space below directly or through a ceiling indirectly. Roofs are usually made from clay tile, waterproof
membrane, concrete and lightweight concrete, wood, and thermal insulation.
 A ceiling is an interior partition that separates the conditioned space from a ceiling plenum. The
ceiling plenum may or may not be air conditioned. Ceilings are usually made from acoustic tile or
boards, thermal insulation, and interior surface finishes.
 An exterior floor is a horizontal exterior partition under conditioned space. A floor placed over a
ventilated basement or a parking space is an exterior floor. Exterior floors are usually made from
wood, concrete, thermal insulation, and face tiles.
 Slab on grade is a concrete floor slab on the ground. There is usually a vapor barrier, thermal
insulation, and gravel and sand fill between the concrete slab and the ground.
 A window is glazing of any transparent or translucent material plus sash, frame, mullions, and
dividers in the building envelope. Glazing is usually made from glass and transparent plastics.
Frames are often made from wood, aluminium, and steel. The window area is the area of the
surface of glazing plus the area of frame, sash, and mullions. A fenestration is any area on the
exterior building envelope which admits light indoors. Fenestrations include windows, glass
doors, and skylights.
 A skylight is glazing having a slope of less than 60° from the horizontal. There is often a conditioned
space below skylight(s).
3.2 HEAT-TRANSFER FUNDAMENTALS
Heat transfer between two bodies, two materials, or two regions is the result of temperature
difference. The science of heat transfer has provided calculations and analyses to predict rates of
heat transfer. The design of an air conditioning system must include estimates of heat transfer
between the conditioned space, its contents, and its surroundings, to determine cooling and heating
loads. Heat-transfer analysis can be described in three modes: conduction, convection, and
radiation.
3.2 CHAPTER THREE
Conductive Heat Transfer
Conduction is the mechanism of heat transfer in opaque solid media, such as through walls and
roofs. For one-dimensional steady-state heat conduction qk, Btu /h (W), Fourier’s law gives the following
relationship:
(3.1)
where k  thermal conductivity, Btu /h ft °F (W/m°C)
A  cross-sectional area normal to heat flow, ft2 (m2)
T  temperature, °F (°C)
x  coordinate dimension along heat flow, ft (m)
Equation (3.1) shows that the rate of heat transfer is directly proportional to the temperature gradient
dT/dx, the thermal conductivity k, and the cross-sectional area A. The minus sign indicates that
the heat must flow in the direction of decreasing temperature; i.e., if the temperature decreases as x
increases, the gradient dT/dx is negative, so that heat conduction is a positive quantity.
For steady-state heat conduction through a plane composite wall with perfect thermal contact
between each layer, as shown in Fig. 3.1, the rate of heat transfer through each section of the
composite wall must be the same. From Fourier’s law of conduction,
(3.2)
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.3
FIGURE 3.1 Steady-state one-dimensional heat conduction through a
composite wall.
kA, kB, kC  thermal conductivity of layers A, B, and C, respectively, of composite wall,
Btu/h ft °F (W/m°C)
Eliminating T2 and T3, we have
(3.3)
For a multilayer composite wall of n layers in perfect thermal contact, the rate of conduction heat
transfer is given as
(3.4)
Subscript n indicates the nth layer of the composite wall.
In Eq. (3.2), conduction heat transfer of any of the layers can be written as
(3.5)
where R*  thermal resistance, h°F/Btu (°C/W). In Eq. (3.5), an analogy can be seen between
heat flow and Ohm’s law for an electric circuit. Here the temperature difference T  T1  T2 indicates
thermal potential, analogous to electric potential. Thermal resistance R* is analogous to electric
resistance, and heat flow qk is analogous to electric current.
The total conductive thermal resistance of a composite wall of n layers RT, h°F/Btu (°C/W),
can be calculated as
(3.6)
where  thermal resistances of layers 1, 2,   , n layer of the composite wall,
h°F/Btu (°C/W). The thermal circuit of a composite wall of three layers is shown in the lower
part of Fig. 3.1.
Convective Heat Transfer
Convective heat transfer occurs when a fluid comes in contact with a surface at a different temperature,
such as the heat transfer taking place between the airstream in a duct and the duct wall.
Convective heat transfer can be divided into two types: forced convection and natural or free
convection. When a fluid is forced to move along the surface by an outside motive force, heat is
transferred by forced convection. When the motion of the fluid is caused by the density difference
of the two streams in the fluid as a product of contacting a surface at a different temperature, the
result is called natural or free convection.
The rate of convective heat transfer qc, Btu /h (W), can be expressed in the form of Newton’s
law of cooling as
qc  hcA(Ts  T) (3.7)
where hc  average convective heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
Ts  surface temperature, °F (°C)
T  temperature of fluid away from surface, °F (°C)
In Eq. (3.7), the convective heat-transfer coefficient hc is usually determined empirically. It is
related to a dimensionless group of fluid properties, such as the correlation of flat-plate forced
R*1 , R*2 ,    R*n
R*T  R*1 	 R*2 	    	 R*n
R* 
L
k A
qk 
k A
L
T 
T
R*
qk 
T1  Tn	1
L1 /(k1 A) 	 L2 /(k2 A) 	    	 Ln /(kn A)
qk 
T1  T4
LA /(kA A) 	 LB /(kB A) 	 LC /(kC A)
3.4 CHAPTER THREE
convection, as shown in the following equation:
(3.8)
where C  constant and k  thermal conductivity, Btu /h ft °F (W/m°C). In Eq. (3.8), NuL is the
Nusselt number, which is based on a characteristic length L. Characteristic length can be the length
of the plate, the diameter of the tube, or the distance between two plates, in feet (meters). The
Reynolds number ReL is represented by the following dimensionless group:
(3.9)
where  density of fluid, lb/ ft3 (kg/m3)
v  velocity of fluid, ft/s (m/s)
	 absolute viscosity of the fluid, lb/ ft  s (kg /m s)
 kinematic viscosity of fluid, ft2/s (m2/s)
The Prandtl number Pr is represented by
(3.10)
where cp  specific heat at constant pressure of the fluid, Btu / lb°F (J/kg°C). Fluid properties
used to calculate dimensionless groups are usually related to the fluid temperature Tf . That is,
(3.11)
Convective heat transfer can also be considered analogous to an electric circuit. From Eq. (3.7),
the convective thermal resistance Rc*, h°F/Btu (°C/W), is given as
(3.12)
Radiant Heat Transfer
In radiant heat transfer, heat is transported in the form of electromagnetic waves traveling at the
speed of light. The net rate of radiant transfer qr, Btu /h (W), between a gray body at absolute
temperature TR1 and a black surrounding enclosure at absolute temperature TR2 (for example, the
approximate radiation exchange between occupant and surroundings in a conditioned space) can be
calculated as
(3.13)
where 
 Stefan-Boltzmann constant  0.1714  108 Btu/h ft2 °R4 (W/m2 K4 )
A1  area of gray body, ft2 (m2 )
1  emissivity of surface of gray body
TR1, TR2  absolute temperature of surfaces 1 and 2, °R (K)
If we multiply the right-hand side of Eq. (3.13) by (TR1  TR2)/(TR1  TR2), and let the thermal
resistance
(3.14) R*r 
TR1  TR2

A1 1(T R1
4  T R2
4 )

1
hr A1
qr 

A1 1(T R1
4  T R2
4 )
Rc* 
1
hc A
Tf 
Ts  T
2
Pr 
3600 	cp
k
ReL 
vL
	

vL

NuL 
hc L
k
 C ReL n Prm
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.5
where hr  radiant heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C), we find that as a
consequence,
(3.15)
where T1, T2  temperature of surfaces 1 and 2, °F (°C).
If either of the two surfaces is a black surface or can be approximated as a black surface, as was
previously assumed for the conditioned space surroundings, the net rate of radiant heat transfer
between surfaces 1 and 2 can be evaluated as
(3.16)
where F1–2  shape factor for a diffuse emitting area A1 and a receiving area A2. The thermal resistance
can be similarly calculated.
Overall Heat Transfer
In actual practice, many calculations of heat-transfer rates are combinations of conduction, convection,
and radiation.
Consider the composite wall shown in Fig. 3.1; in addition to the conduction through the
wall, convection and radiation occur at inside and outside surfaces 1 and 4 of the composite wall.
At the inside surface of the composite wall, the rate of heat transfer qi, Btu /h (W), consists of convec-
tive heat transfer between fluid, the air, and solid surface qc and the radiant heat transfer qr, as
follows:
(3.17)
where Ti  indoor temperature, °F (°C). From Eq. (3.17), the inside surface heat-transfer coefficient
hi at the liquid-to-solid interface, Btu /h ft2 °F (W/m2 °C) is
and the thermal resistance Ri* of the inner surface due to convection and radiation, h°F/Btu
(°C/W), is
(3.18)
Similarly, at the outside surface of the composite wall, the rate of heat-transfer qo, Btu/h ft2 °F
(W/m2 °C), is
(3.19)
where ho  outside surface heat-transfer coefficient at fluid-to-solid interface, Btu /h ft2 °F
(W/m2 °C)
A4  area of surface 4, ft2 (m2 )
T4  temperature of surface 4, °F (°C)
To  outdoor temperature, °F (°C)
The outer thermal resistance Ro*, h°F/Btu (°C/W), is
For one-dimensional steady-state heat transfer, the overall heat-transfer rate of the composite
R*o 
1
ho A4
qo  ho A4(T4  To)
Ri* 
1
hi A1
hi  hc 	 hr
 hi A1(Ti  T1)
qi  qc 	 qr  hc A1(Ti  T1) 	 hr A1(Ti  T1)
qr 

A1F1– 2(T R1
4  T R2
4 )
qr  hr A1(T1  T2)
3.6 CHAPTER THREE
wall q, Btu /h (W), can be calculated as
(3.20)
where U  overall heat-transfer coefficient, often called the U value, Btu/h ft2 °F (W/m2 °C)
A  surface area perpendicular to heat flow, ft2 (m2 )
RT*  overall thermal resistance of composite wall, h°F/Btu (°C/W)
and
(3.21)
Also, the thermal resistances can be written as
Therefore, the overall heat transfer coefficient U is given as
(3.22)
For plane surfaces, area A  A1  A2  A3  A4. For cylindrical surfaces, because the inside
and outside surface areas are different, and because UA  1/RT*, it must be clarified whether the
area is based on the inside surface area Ai, outside surface area Ao, or any chosen surface area.
For convenient HVAC&R heat-transfer calculations, the reciprocal of the overall heat-transfer
coefficient, often called the overall R value RT, h ft2 °F/Btu (m2 °C/W), is used. So RT can be
expressed as
(3.23a)
where Ri, Ro  R values of inside and outside surfaces of composite wall, h ft2 °F/Btu
(m2 °C/W)
R1, R2,   , Rn  R values of components 1, 2, . . ., n, h ft2 °F/Btu (m2 °C/W)
and
(3.23b)
Sometimes, for convenience, the unit of R is often omitted; for example, R-10 means the R value
equals 10 h ft2 °F/Btu (m2 °C/W).
A building envelope assembly, or a building shell assembly, includes the exterior wall assembly
(i.e., walls, windows, and doors), the roof and ceiling assembly, and the floor assembly. The areaweighted
average overall heat-transfer coefficient of an envelope assembly Uav, Btu/h ft2 °F
(W/m2 °C), can be calculated as
(3.24)
where A1, A2 . . ., An  area of individual elements 1, 2, . . ., n of envelope assembly, ft2 (m2 )
Ao  gross area of envelope assembly, ft2 (m2)
U1, U2, . . ., Un  overall heat-transfer coefficient of individual paths 1, 2, . . ., n of the
envelope assembly, such as paths through windows, paths through walls,
and paths through roof, Btu /h ft2 °F (W/m2 °C)
Uav 
U1 A1 	 U2 A2 	    	 Un An
Ao
Ri 
1
hi
R1 
L1
k1
R2 
L2
k2
   Rn 
Ln
kn
Ro 
1
ho
RT 
1
U
 Ri 	 R1 	 R2 	    	 Rn 	 Ro
U 
1
1/hi 	 LA / kA 	 LB / kB 	 LC / kC 	 1/ho
R*A 
LA
kA A R*B 
LB
kB B R*C 
LC
kCC
R*T  R*i 	 R*A 	 R*B 	 R*C 	 R*o 
1
UA
q  qi  qk  qo  UA(Ti  To) 
Ti  To
R*T
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.7
Heat Capacity
The heat capacity (HC) per square foot (meter) of an element or component of a building envelope
or other structure depends on its mass and specific heat. Heat capacity HC, Btu/ ft2 °F (kJ/m2 °C),
can be calculated as
(3.25)
where m  mass of building material, lb (kg)
c  specific heat of building material, Btu/ lb°F (kJ/kg°C)
A  area of building material, ft2 (m2)
 density of building material, lb / ft3 (kg/m3 )
L  thickness or height of building material, ft (m)
3.3 HEAT-TRANSFER COEFFICIENTS
Determination of heat-transfer coefficients to be used for load calculations or year-round energy
estimates is complicated by the following types of variables:
 Building envelopes, exterior wall, roof, glass, partition wall, ceiling, or floor
 Fluid flow, turbulent flow, or laminar flow, forced or free convection
 Heat flow, horizontal heat flow in a vertical surface, or an upward or downward heat flow in a
horizontal surface
 Space air diffusion, ceiling or sidewall inlet, or others
 Time of operation—summer, winter, or other seasons
Among the three modes of heat transfer, convection processes and their related coefficients are the
least understood, making analysis difficult.
Coefficients for Radiant Heat Transfer
For a radiant exchange between the inner surface of an exterior wall and the surrounding surfaces
(such as the surfaces of partition walls, ceilings, and floors) in an air conditioned room, the sum of
the shape factors F1– n can be considered as unity. If all surfaces are assumed to be black, then the
radiative heat-transfer coefficient hr, Btu/h ft2 °F (W/m2 °C), can be calculated as
(3.26)
where TRis  absolute temperature of inner surface of exterior wall, roof, or external window glass,
°R (K)
TRrad  absolute mean radiant temperature of surrounding surfaces, °R (K)
Often TRrad is approximately equal to the air temperature of the conditioned space TRr when both are
expressed in degrees Rankine (kelvins).
Radiant heat transfer coefficients calculated according Eq. (3.26) for various surface temperatures
and temperature differences between the inner surface of any building envelope and the surrounded
surfaces are presented in Table 3.1. From Table 3.1, hr depends on the absolute temperature of inner
surface TRis and the temperature difference Tis  Trad.
hr 

 (T Ris
4  T Rrad
4 )
TRis  TRrad
HC
A

Lc
3.8 CHAPTER THREE
Coefficients for Forced Convection
Before one can select the mean convective heat-transfer coefficient hc in order to calculate the rate
of convective heat transfer or to determine the overall heat-transfer coefficient U value during cooling
load calculations, the type of convection (forced or natural) must be clarified. Forced convection
between the surfaces of the building envelope and the conditioned space air occurs when the airhandling
system is operating or there is a wind over the outside surface. In an indoor space, free
convection is always assumed when the air-handling system is shut off, or there is no forced-air
motion over the surface involved.
The convective heat-transfer coefficient hc also depends on the air velocity v flowing over the
surface, the configuration of the surface, the type of space air diffusion, and the properties of the
fluid. According to Kays and Crawford (1980), a linear or nearly linear relationship between hc and
v holds. Even though the air velocity v in the occupied zone may be only 30 fpm (0.15 m/ s) or even
lower when the air-handling system is operating, however, the mode of heat transfer is still considered
as forced convection and can be expressed as
(3.27)
where n  exponential index, usually between 0.8 and 1
v  bulky air velocity of fluid 0.5 to 1 ft (0.15 to 0.30 m) from surface, fpm (m/min)
A, B  constants
On the basis of test data from Wong (1990), Sato et al. (1972), and Spitler et al. (1991), as well as
many widely used energy estimation computer programs, the convective heat-transfer coefficient
for forced convection hc, Btu/h ft2 °F (W/m2 °C), for indoor surfaces can be determined as
(3.28)
For outside surfaces, the surface heat-transfer coefficient ho  hc 	 hr, Btu/h ft2 °F (W/m2 °C),
can be calculated as
(3.29)
where vwind  wind speed, fpm.
ho  1.8 	 0.004vwind
hc  1.0 	 0.008v
hc  A 	 Bv
n
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.9
TABLE 3.1 Radiative Heat Transfer Coefficients hr
Temperature
difference
Inner surface temperature Tis, °F
Tis–Trad, 60 70 75 80 85
°F (520°R) (530°R) (535°R) (540°R) (545°R)
Btu/h ft2 °F
1 0.871 0.909 0.935 0.968 0.990
2 0.866 0.910 0.936 0.966 0.990
3 0.862 0.908 0.935 0.963 0.989
5 0.856 0.904 0.930 0.958 0.984
10 0.844 0.892 0.918 0.945 0.971
20 0.819 0.868 0.893 0.919 0.945
Note: Calculations made by assuming that Trad  Tr and the emissivity of inner
surface i  1.
Coefficients for Natural Convection
For natural convection, the empirical relationship between the dimensionless groups containing the
convective heat-transfer coefficient hc can be expressed as follows:
(3.30)
In Eq. (3.30), GrL, called the Grashof number, is based on the characteristic length L.
(3.31)
where  coefficient of volume expansion of fluid, 1/ °R (1/K)
g  acceleration of gravity, ft / s2 (m/s2)
 density of fluid, lb/ ft3 (kg/m3)
C  constant
And Ra is called the Rayleigh number, and Ra  GrPr.
Natural convective heat transfer is difficult to evaluate because of the complexity of the recirculating
convective stream of room air that is the result of the temperature distribution of the surfaces
and the variation of temperature profile of the stream. Computer programs using numerical techniques
have been developed recently. In actual practice, simplified calculations are often adopted.
Altmayer et al. (1983) in their recent experiments and analyses found that “The ASHRAE free
convection correlations provide a fair prediction of the heat flux to the room air from cold and hot
surfaces.” The errors are mainly caused by the variation of temperature of the convective airstreams
after contact with cold or hot surfaces. In simplified calculations, this variation can only be included
in the calculation of the mean space air temperature T.
Many rooms have a vertical wall height of about 9 ft (2.7 m). If the temperature difference
Ts T  1°F (0.56°C), the natural convection flow is often turbulent. If Ts  T  1°F (0.56°C),
as in many cooling load calculations between a partition wall and space air, the flow is laminar.
Based on data published in ASHRAE Handbook 1997, Fundamentals, the natural convection
coefficients hc, Btu/h ft2 °F (W/m2 °C), are given as follows:
Vertical plates:
Large plates, turbulent flow hc  0.19(Tsa)0.33 (3.32)
Small plates, laminar flow (3.33)
Horizontal plates, facing upward when air is heated or facing downward when air is cooled:
Large plates, turbulent flow hc  0.22(Tsa)0.33 (3.34)
Small plates, laminar flow (3.35)
Horizontal plates, facing upward when air is cooled or facing downward when air is heated:
Small plates (3.36)
Here, Tsa indicates the temperature difference between the surface and the air.
Surface Heat-Transfer Coefficients
The surface heat-transfer coefficient h, sometimes called the surface conductance, Btu/h ft2 °F
(W/m2 °C), is the combination of convective and radiant heat-transfer coefficients; that is,
hc  0.12Tsa
L 0.25
hc  0.27Tsa
L 0.25
hc  0.29Tsa
L 0.25
GrL  g 2(Ts  T )
	2
NuL  C(GrLPr)n  C RaL n
3.10 CHAPTER THREE
h  hc 	 hr. Table 3.2 lists the h values for various surface types at Tsa  10°F (5.6°C) and
Tsa  1°F (0.56°C) during summer and winter design conditions. The values are based upon the
following:
 Trad  Ta, where Ta is the air temperature, °F (°C).
 Tsa indicates temperature difference between surface and air, °F (°C).
 Emissivity of the surface  0.9 and  0.2.
3.4 MOISTURE TRANSFER
Moisture is water in the vapor, liquid, and solid states. Building materials exposed to excessive
moisture may degrade or deteriorate as a result of physical changes, chemical changes, and biological
processes. Moisture accumulated inside the insulating layer also increases the rate of heat
transfer through the building envelope. Moisture transfer between the building envelope and the
conditioned space air has a significant influence on the cooling load calculations in areas with hot
and humid climates.
Sorption Isotherm
Moisture content X, which is dimensionless or else in percentage, is defined as the ratio of the mass
of moisture contained in a solid to the mass of the bone-dry solid. An absorption isotherm is a
constant-temperature curve for a material in which moisture content is plotted against an increased
ambient relative humidity during an equilibrium state; i.e., the rate of condensation of water vapor
on the surface of the material is equal to the rate of evaporation of water vapor from the material. A
desorption isotherm is also a constant-temperature curve for a material. It is a plot of moisture content
versus a decreased ambient relative humidity during equilibrium state.
Many building materials show different absorption and desorption isotherms. The difference in
moisture content at a specific relative humidity between the absorption and desorption isotherms is
called hysteresis. Figure 3.2 shows absorption and desorption isotherms of a building material.
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.11
TABLE 3.2 Surface Heat-Transfer Coefficients h, Btu/h ft2 °F
Surface emissivity 
0.90 0.20
Indoor surface Indoor surface
Tsa  10°F Tsa  10°F
Direction of Outdoor
Description heat flow Summer Winter Tsa  1°F surface Summer Winter Tsa  1°F
Forced convection
30 fpm 2.21 2.11 2.16 1.46 1.43 1.44
50 fpm 2.37 2.27 2.32 1.62 1.59 1.60
660 fpm (7.5 mph) 4.44
1320 fpm (15 mph) 7.08
Free convection
Horizontal surface Upward 1.36 1.27 1.17 0.62 0.60 0.44
Vertical surface 1.42 1.33 1.15 0.68 0.66 0.30
Horizontal surface Downward 1.03
Note: Assume space temperature Tr  74°F year-round and Tr  Trad; here Trad indicates the mean radiant temperature of the surroundings.
Temperature also has an influence on the moisture content of many building materials. At a constant
relative humidity in ambient air, the moisture content of a building material will be lower at a
higher temperature.
When a building material absorbs moisture, heat as heat of sorption is evolved. If liquid water is
absorbed by the material, an amount of heat ql, Btu / lb (kJ /kg) of water absorbed, similar to the
heat of solution, is released. This heat results from the attractive forces between the water molecules
and the molecules of the building material. If water vapor is absorbed, then the heat released
qv, Btu / lb (kJ / kg), is given by
(3.37)
where hfg  latent heat of condensation, Btu/ lb (kJ /kg). Heat of sorption of liquid water ql varies
with equilibrium moisture content for a given material. The lower the X and the  of ambient air,
the higher will be the value of ql. For pine, ql may vary from 450 Btu/lb (1047 kJ/kg) for nearly
bone-dry wood to 20 Btu/ lb (46.5 kJ /kg) at a moisture content of 20 percent. Many building materials
have very low ql values compared with hfg, such as a ql of 40 Btu/ lb (93 kJ / kg) for sand.
Moisture-Solid Relationship
Many building materials have numerous interstices and microcapillaries of radius less than 4 
106 in. (0.1 	m), which may or may not be interconnected. These interstices and microcapillaries
provide large surface areas to absorb water molecules. Moisture can be bound to the solid surfaces
by retention in the capillaries and interstices, or by dissolution into the cellular structures of fibrous
materials.
When the relative humidity  of ambient air is less than 20 percent, moisture is tightly bound to
individual sites in the monomolecular layer (region A, in Fig. 3.2). Moisture moves by vapor diffusion.
The binding energy is affected by the characteristics of the surface, the chemical structure of
the material, and the properties of water.
qv  ql 	 hfg
3.12 CHAPTER THREE
FIGURE 3.2 Sorption isotherms.
When relative humidity  is 20 to 60 percent (region B, Fig. 3.2), moisture is bound to the
surface of the material in multimolecular layers and is held in microcapillaries. Moisture begins to
migrate in liquid phase, and its total pressure is reduced by the presence of moisture in microcapillaries.
The binding energy involved is mainly the latent heat of condensation.
When   60 percent, i.e., in region C of Fig. 3.2, moisture is retained in large capillaries. It is
relatively free for removal and chemical reactions. The vapor pressure of the moisture is influenced
only moderately because of the moisture absorbed in regions A and B. Moisture moves mainly in a
liquid phase.
Because the density of the liquid is much greater than that of water vapor, the moisture content
in building materials is mostly liquid. Unbound moisture can be trapped in interstices having a
radius greater than 4  105 in. (1 	m), without significantly lowering the vapor pressure. Free
moisture is the moisture in excess of the equilibrium moisture content in a solid. For insulating
materials in which interstices are interconnected, air penetrates through these open-cell structures,
and, moisture can be accumulated in the form of condensate and be retained in the large capillaries
and pores.
Moisture Migration in Building Materials
Building envelopes are not constructed only with open-cell materials. The airstream and its associated
water vapor cannot penetrate building envelopes. Air leakage can only squeeze through the
cracks and gaps around windows and joints. However, all building materials are moisturepermeable;
in other words, moisture can migrate across a building envelope because of differences
in moisture content or other driving potentials.
According to Wong and Wang (1990), many theories have been proposed by scientists to predict
the migration of moisture in solids. The currently accepted model of moisture flow in solids is
based upon the following assumptions:
 Moisture migrates in solids in both liquid and vapor states. Liquid flow is induced by capillary
flow and concentration gradients; vapor diffusion is induced by vapor pressure gradients.
 During the transport process, the moisture content, the vapor pressure, and the temperature are
always in equilibrium at any location within the building material.
 Heat transfer within the building material is in the conduction mode. It is also affected by latent
heat from phase changes.
 Vapor pressure gradients can be determined from moisture contents by means of sorption
isotherms.
 Fick’s law is applicable.
 Only one-dimensional flow across the building envelope is considered. Building materials are
homogeneous.
If the temperature gradient is small, for simplicity, the mass flux for one-dimensional
flow, lb/h ft2 (kg/hm2), can be expressed as
(3.38)
where Dlv  mass diffusivity of liquid and vapor, ft2 /h (m2/h)
X  moisture content, lb / lb (kg / kg) dry solid
w  density of water, lb/ ft3 (kg/m3 )
x  coordinate dimension along moisture flow, ft (m)
A  area of building envelope perpendicular to moisture flow, ft2 (m2 )
Mass diffusivities of some building materials as a function of moisture content are shown in
Fig. 3.3.
m?w /A  wDlv
dX
dx
m? w /A
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.13
Moisture Transfer from the Surface of the Building Envelope
At a certain time instant, moisture migrating from any part in the building envelope to its surface
must be balanced by convective moisture transfer from the surface of the building envelope to the
ambient air and the change of the moisture content as well as the corresponding mass concentration
at the surface of the building envelope. Such a convective moisture transfer is often a part of the
space latent cooling load. Analogous to the rate of convective heat transfer [Eq. (3.7)], the rate of
convective moisture transfer , lb/h (kg/ h), can be calculated as
(3.39)
where hm  convective mass-transfer coefficient, ft /h (m/h)
Aw  contact area between moisture and ambient air, ft2 (m2 )
Cws, Cwr  mass concentration of moisture at surface of building envelope and of space air, lb/ ft3
(kg/m3)
It is more convenient to express the mass concentration difference Cws  Cwr in terms of a
humidity ratio difference ws  wr. Here ws and wr represent the humidity ratio at the surface of the
building envelope and of the space air, respectively, lb/lb (kg / kg). In terms of mass concentration
we can write
(3.40)
(3.41)
Then the rate of convective moisture transfer can be rewritten as
(3.42)
In Eq. (3.42), the surface humidity ratio ws depends on the moisture content Xs, the temperature Ts,
and therefore the vapor pressure pws in the interstices of the surface of the building envelope. From
the known Xs, Ts, and the sorption isotherm, the corresponding relative humidity s at the surface
can be determined. If the difference between relative humidity and the degree of saturation 	is
m?w 
ahmAw(ws  wr)
Cwr 
awr
Cws 
aws
m? w  hm Aw (Cws  Cwr)
m? w
3.14 CHAPTER THREE
FIGURE 3.3 Mass diffusivity Dlv of some building materials. (Abridged
with permission from Bruin et al. Advances in Drying, Vol. 1, 1979.)
ignored, then
(3.43)
where 	s  degree of saturation at surface. In Eq. (3.43), wss represents the humidity ratio of the
saturated air. It can be determined from Eq. (2.49) since Ts is a known value.
In Eq. (3.42), the contact area Aw between the liquid water at the surface of the building envelope
and the space air is a function of moisture content Xs. A precise calculation of Aw is very complicated,
but if the surface area of the building material is As, then a rough estimate can be made
from
(3.44)
Convective Mass-Transfer Coefficients
The Chilton-Colburn analogy relates the heat and mass transfer in these forms:
(3.45)
where v  air velocity remote from surface, ft / s (m/ s)
cpa  specific heat of moist air, Btu / lb°F (J /kg°C)
hm  convective mass-transfer coefficient, ft /h (m/ s)
 kinematic viscosity, ft2/s (m2/s)
 thermal diffusivity of air, ft2/s (m2/s)
Daw  mass diffusivity for water vapor diffusing through air, ft2/s (m2/s)
Sc  Schmidt number
Subscript a indicates dry air and w, water vapor.
For air at 77°F (25°C), cpa  0.243 Btu/ lb°F (1020 kJ/kg°C), va  1.74  104 ft2/s
(1.62  105 m2 / s),  2.44  104 ft2 / s (2.27  105 m2 / s), Daw  2.83 ft2 / s (0.263 m2 / s),
and a  0.0719 lb/ ft3 (1.15 kg/m3 ), we can show that
and that
At a space temperature of 77°F (25°C), therefore,

hc (0.713)2/3
0.0719  0.243  (0.614)2/3  63.2hc
hm 
hc Pr 2/3
a cpa Sc2/3
Sc 
a
Daw

1.74  104
2.83  104  0.614
Pr 
a


1.74  104
2.44  104
 0.713
hc
acpa
a
2/3
 hm  vw
Daw2/3
hc
a cpa
Pr2/3 
hm
v
Sc2/3
jH  jD
Aw  Xs As
ws 
	swss 
swss
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.15
Moisture Transfer in Building Envelopes
Moisture transfer in building envelopes of a typical residential building can proceed along two
paths, as shown in Fig. 3.4:
1. Moisture migrates inside the building material mainly in the form of liquid if the relative humidity
of the ambient air is more than 50 percent. It will be transported to the indoor or outdoor air
by convective mass transfer. The driving potentials are the moisture content of the building material,
the vapor pressure gradient inside the building material, and the humidity ratio gradient between
the surface and the ambient air.
2. Air leakage and the associated water vapor infiltrates or exfiltrates through the cracks and
gaps around the windows, doors, fixtures, outlets, and between the joints. Air and moisture enter the
cavities and the airspace in the building envelope. If the sheathing is not airtight, air leakage and its
water vapor penetrate the perforated insulating board and discharge to the atmosphere, as shown in
Fig. 3.4. If the sheathing is a closed-cell, airtight insulating board, then the airstream may infiltrate
through gaps between the joints of the insulating board or through cracks between the window sill
and the external wall and discharge to the atmosphere.
The driving potential of the air leakage and the associated water vapor is the total pressure
differential between indoor and outdoor air across the building envelope due to wind effects, stack
effect, mechanical ventilation, or a combination of these. Moisture is moving in the vapor form.
In leaky buildings, the moisture transfer by means of air leakage is often far greater than the
moisture migration through solids. For better-sealed commercial buildings, moisture migration
through the building envelope may be important.
3.16 CHAPTER THREE
FIGURE 3.4 Moisture transfer in building envelope along two paths.
Installation of vapor retarders is an effective means to block or to reduce the moisture transfer in
the building envelope. Vapor retarders are covered in greater detail later.
3.5 CONDENSATION IN BUILDINGS
When moist air contacts a solid surface whose temperature is lower than the dew point of the moist
air, condensation occurs on the surface in the form of liquid water, or sometimes frost. Condensation
can damage the surface finish, deteriorate the material and cause objectionable odors, stains,
corrosion, and mold growth; reduce the quality of the building envelope with dripping water; and
fog windows. Two types of condensation predominate in buildings:
1. Visible surface condensation on the interior surfaces of external window glass, below-grade
walls, floor slabs on grade, and cold surfaces of inside equipment and pipes
2. Concealed condensation within the building envelope
Visible Surface Condensation
To avoid visible surface condensation, either the dew point of the indoor air must be reduced to a
temperature below that of the surface, or the indoor surface temperature must be raised to a level
higher than the dew point of the indoor air. The indoor surface temperature of the building envelope,
such as T1 of the composite exterior wall in Fig. 3.1, or any cold surface where condensation
may occur can be calculated from Eqs. (3.17) and (3.20) as
(3.46)
where q  rate of heat transfer through building envelope, Btu /h (W)
hi  inside surface heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
Increasing the thermal insulation is always an effective and economical way to prevent condensation
because it saves energy and raises the surface temperature of the building envelope and other
cold surfaces.
The dew point temperature of the indoor air Tr  dew is a function of humidity ratio wr of the space
air. At a specific space temperature Tr, the lower the relative humidity, the lower the Tr  dew. There
are several ways to lower the dew point of the indoor space air:
 By blocking and controlling infiltration of hot, humid outside air
 By reducing indoor moisture generation
 By using a vapor retarder, such as polyethylene film or asphalt layer, to prevent or decrease the
migration of moisture from the soil and the outside wetted siding and sheathing
 By introducing dry outdoor air through mechanical or natural ventilation when doing so is not in
conflict with room humidity criteria
 By using dehumidifiers to reduce the humidity ratio of the indoor air
During winter, any interior surface temperature of the wall, roof, or glass is always lower than
the indoor space temperature. Better insulation and multiple glazing increase the interior surface
temperature and, at the same time, reduce the heat loss, providing a more comfortable indoor environment
than is possible by decreasing the relative humidity. To reduce excessive indoor humidity
in many industrial applications, a local exhausting booth that encloses the moisture-generating
source is usually the remedy of first choice.
T1  Ti 
U(Ti  To)
hi
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.17
During summer, the outer surface of chilled water pipes and refrigerant pipes (even the cold
supply duct) is at a lower temperature than the indoor air temperature. Sufficient pipe and duct
insulation is needed to prevent surface condensation. When an air-handling system is shut down,
the heavy construction mass often remains at a comparatively lower surface temperature as the indoor
humidity rises. To avoid indoor surface condensation, the enclosure must be tight enough to
prevent the infiltration of hot, humid air from outdoors.
Concealed Condensation within the Building Envelope
Usually, concealed condensation within building materials does not accumulate as a result of moisture
migration. Excessive free moisture, usually in the form of liquid or frost, at any location inside
the building envelope results in a higher moisture content than in surrounding areas. Therefore, it
produces a moisture migration outward from that location rather than into it.
During cold seasons, concealed condensation in building envelopes is mainly caused by the
warm indoor air, usually at a dew-point temperature of 32°F (0°C) and above. It leaks outward
through the cracks and openings and enters the cavities, the gaps between component layers, or
even the penetrable insulating material. It ultimately contacts a surface at a temperature lower than
the dew point of the indoor air. Donnelly et al. (1976) discovered that a residential stud wall panel
with a poor vapor retarder accumulated about 3 lb(14.6 kg) of moisture per square foot (meter) of
wall area during a period of 31 days in winter. This moisture accumulated in the mineral fiber insulating
board adjacent to the cold side of the sheathing and at the interface between the insulating
board and the sheathing.
The rate of condensation , lb/h (kg/ h), can be calculated as
(3.47a)
where  volume flow rate of air leakage, cfm (m3 /min)
r  density of indoor air, lb/ ft3 (kg/m3 )
ws con  saturated humidity ratio corresponding to temperature of surface upon which condensation
occurs, lb / lb (kg /kg)
Vapor retarders are effective for reducing moisture transfer through the building envelope. They
not only decrease the moisture migration in the building material significantly, but also block air
leakage effectively if their joints are properly sealed. Vapor retarders are classified as rigid, flexible,
and coating types. Rigid type includes reinforced plastic, aluminum, and other metal sheets. Flexible
type includes metal foils, coated films, and plastic films. Coatings are mastics, paints, or fusible
sheets, composed of asphalt, resin, or polymeric layers. The vapor retarder is generally located on
the warm side of the insulation layer during winter heating.
In an area where summer cooling is dominant, a vapor retarder to prevent condensation in the
external brick wall and insulation layer would often be located on the outside of the insulation. The
absorptive brick wall takes substantial amounts of water during rainfall. The subsequent sunny
period drives the water vapor farther inside the wall and into the insulation layer where the water
vapor condenses.
3.6 THERMAL INSULATION
Thermal insulation materials, usually in the form of boards, slabs, blocks, films, or blankets, retard
the rate of heat transfer in conductive, convective, and radiant transfer modes. They are used within
building envelopes or applied over the surfaces of equipment, piping, or ductwork to achieve the
following benefits:
 Savings of energy by reducing heat loss and heat gain from the surroundings
V?
lk
m? con  60V?lk
r (wr  ws con )
m? con
3.18 CHAPTER THREE
 Prevention of surface condensation by increasing the surface temperature above the dew point of
the ambient air
 Reduction of temperature difference between the inside surface and the space air for the thermal
comfort of the occupants, when radiant heating or cooling is not desired
 Protection of the occupant from injury due to contact with hot pipes and equipment
Basic Materials and Thermal Properties
Basic materials in the manufacture of thermal insulation for building envelopes or air conditioning
systems include
 Fibrous materials such as glass fiber, mineral wool, wood, cane, or other vegetable fibers
 Cellular materials such as cellular glass, foam rubber, polystyrene, and polyurethane
 Metallic reflective membranes
Most insulating materials consist of numerous airspaces, either closed cells (i.e., airtight cellular
materials) or open cells (i.e., fibrous materials, penetrable by air).
Thermal conductivity k, an important physical property of insulating material, indicates the rate of
heat transfer by means of a combination of gas and solid conduction, radiation, and convection
through an insulating material, expressed in Btu in./h ft2 °F or Btu/h ft°F (W/m°C). The thermal
conductivity of an insulating material depends on its physical structure (such as cell size and shape or
diameter of the fibrous materials), density, temperature, and type and amount of binders and additives.
Most of the currently used thermal insulating materials have thermal conductivities within a
range of 0.15 to 0.4 Btu in./h ft2 °F (0.021 to 0.058 W/m°C). Thermal properties of some building
and insulating materials, based on data published in the ASHRAE Handbook 1993, Fundamentals,
are listed in Table 3.3.
The thermal conductivity of many cellular insulating materials made from polymers like
polyurethane and extruded polystyrene is not significantly affected by change in density . Other
insulating materials have a density at which the thermal conductivity is minimum. Deviating from
this , conductivity k increases always, whether  is greater or smaller in magnitude. The thermal
conductivity of an insulating material apparently increases as its mean temperature rises. For
polystyrene, k increases from 0.14 to 0.32 Btu in./h ft2 °F (0.020 to 0.046 W/m°C) when its
mean temperature is raised from 300 to 570°F (150 to 300°C).
Moisture Content of Insulation Material
From the point of view of moisture transfer, penetrability is an important characteristic. Closed-cell
airtight board or block normally cannot have concealed condensation except at the gap between the
joints and at the interface of two layers. Only when air and its associated water vapor penetrate an
open-cell insulating material can concealed condensation form if they contact the surfaces of the
interstices and pores at a temperature lower than the dew point of the penetrating air. Concealed
condensation might also accumulate in open-cell insulating materials. If there is excess free moisture
in an insulating material, the thermal insulation may be degraded. The increase of moisture in
the thermal insulation layer is often due to the absorption from the ambient air, space air, or moisture
transfer from adjacent layer, ground, or wetted surface.
Smolenski (1996) brought forward the question of how much is too much. In roofing insulation,
many consider the answer to be the moisture content to produce a more than 20 percent loss in thermal
efficiency or a thermal resistance ratio (TRR) of less than 80 percent. The thermal resistance
ratio is defined as the ratio of wet to dry thermal resistance of the thermal insulation, in percent, or
(3.47b) TRR 
(wet thermal resistance)  100
dry thermal resistance
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.19
The equilibrium moisture content of most commomly used insulation material at 90 percent relative
humidity ambient air, as well as the moisture content of insulation material at 80 percent TRR by
weight (percent of dry weight) and by volume (percent of volume of insulating material) are listed
below:
Density,
Moisture content, %
Insulation material lb/ft3  90% 80% TRR, by weight 80% TRR, by volume
Cellular glass 8.4 0.2 23 3.1
Expanded polystyrene 1.0 2.0 383 6.1
Glass fiber 9.2 1.1 42 6.2
Urethane 2.1 6.0 262 8.8
Phenolic foam 2.6 23.4 25 1.0
For instance, for cellular glass when in contact with an ambient air of 90 percent relative humidity
at normal room temperaure that reaches an equilibrium state during moisture absorption, it has
3.20 CHAPTER THREE
TABLE 3.3 Thermal Properties of Selected Materials
Thermal Specific
Density, conductivity, heat,
lb / ft3 (Btu/h ft °F) (Btu/ lb °F) Emissivity
Aluminum (alloy 1100) 171 128 0.214 0.09
Asbestos: insulation 120 0.092 0.20 0.93
Asphalt 132 0.43 0.22
Brick, building 123 0.4 0.2 0.93
Brass (65% Cu, 35% Zn) 519 69 0.09 0.033 Highly polished
Concrete (stone) 144 0.54 0.156
Copper (electrolytic) 556 227 0.092 0.072 Shiny
Glass: crown (soda-lime) 154 0.59 0.18 0.94 Smooth
Glass wool 3.25 0.022 0.157
Gypsum 78 0.25 0.259 0.903 Smooth plate
Ice (32°F) 57.5 1.3 0.487 0.95
Iron: cast 450 27.6 0.12 0.435 Freshly turned
Mineral fiberboard:
acoustic tile, wet-molded 23 0.035 0.14
wet-felted 21 0.031 0.19
Paper 58 0.075 0.32 0.92
Polystyrene, expanded, molded beads 1.25 0.021 0.29
Polyurethane, cellular 1.5 0.013 0.38
Plaster, cement and sand 132 0.43 0.91 Rough
Platinum 1340 39.9 0.032 0.054 Polished
Rubber: vulcanized, soft 68.6 0.08 0.48 0.86 Rough
hard 74.3 0.092 0.95 Glossy
Sand 94.6 0.19 0.191
Steel (mild) 489 26.2 0.12
Tin 455 37.5 0.056 0.06 Bright
Wood: fir, white 27 0.068 0.33
oak, white 47 0.102 0.57 0.90 Planed
plywood, Douglas fir 34 0.07 0.29
Wool: fabric 20.6 0.037
Source: Adapted with permission from ASHRAE Handbook 1989, Fundamentals.
only a moisture content of 0.2 percent by dry weight, which is far lower than the moisture content
of 23 percent by weight at 80 percent TRR. Most above-listed insulating materials have a far lower
equilibrium moisture content of dry weight at an ambient air of 90 percent relative humidity than
moisture content at 80 percent TRR. Also, a closed-cell structure cellular glass requires months or
years to reach an equilibrium moisture content by volume corrresponding to 80 percent TRR
whereas mineral wool and calcium silica absorb moisture in only hours.
Economic Thickness
The economic thickness of insulation is the thickness with the minimum owning and operating
costs. Owning cost is the net investment cost of the installed insulation Cin, in dollars, less any capital
investment that can be made as a result of lower heat loss or gain Cpt, in dollars. Theoretically,
for a new plant, some small savings might be made because of the reduction of the size of a central
plant; but in actual practice, this is seldom considered. For an existing plant where add-on insulation
is being considered, Cpt is zero because the plant investment has already been made.
Operating cost Cen includes the annualized cost of energy over the life of a new plant, or the
remaining life of an existing plant, in n years. It can also be taken as the number of years over
which an owner wishes to have a total return of the net investment, considering both the interest and
the fuel escalation rate. Total cost of insulation Ct, in dollars, for any given thickness is
(3.48)
When the thickness of the insulating material increases, the quantity CinCpt also increases, as
shown in Fig. 3.5, and Cen decreases. As a result, Ct first decreases, drops to a minimum, and then
increases. The optimum economic thickness occurs when Ct drops to a minimum. The closest commercially
available thickness is the optimum thickness.
Thermal Resistance of Airspaces
Thermal resistance of an enclosed airspace Ra has a significant effect on the total thermal resistance
RT of the building envelope, especially when the value of RT is low. Thermal resistance Ra depends
on the characteristic of the surface (reflective or nonreflective), the mean temperature, the temperature
difference of the surfaces perpendicular to heat flow, the width across the airspace along the heat flow,
Ct  Cin  Cpt 	 Cen
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.21
FIGURE 3.5 Optimum thickness of insulation material.
and the direction of airflow. The R values of the enclosed airspaces Ra, h ft2 °F/Btu (m2 °C/W),
abridged from data published in ASHRAE Handbook 1989, Fundamentals, are presented in Table 3.4.
3.7 SOLAR ANGLES
Basic Solar Angles
The basic solar angles between the sun’s rays and a specific surface under consideration are shown
in Figs. 3.6 and 3.7.
3.22 CHAPTER THREE
FIGURE 3.6 Basic solar angles and position of sun’s rays at summer solstice.
TABLE 3.4 R Values of Enclosed Airspace Ra, h ft2 °F/Btu
Mean Temperature Type of Direction Width of
Emissivity E†
temperature, °F difference, °F surface of heat flow airspace, in. 0.05 0.2 0.82
Summer, 90 10 Horizontal Upward 0.5 2.03 1.51 0.73
3.5 2.66 1.83 0.80
Vertical Horizontal 0.5 2.34 1.67 0.77
3.5 3.40 2.15 0.85
Horizontal Downward 0.5 2.34 1.67 0.77
3.5 8.19 3.41 1.00
Winter, 50 10 Horizontal Upward 0.5 2.05 1.60 0.84
3.5 2.66 1.95 0.93
Vertical Horizontal 0.5 2.54 1.88 0.91
3.5 3.40 2.32 1.01
Horizontal Downward 0.5 2.55 1.89 0.92
3.5 9.27 4.09 1.24
†Emissity E  1/(1/e1 	 1/e2  1), where e1, e2 indicate the emittances on two sides of the airspace.
Source: Abridged with permission from ASHRAE Handbook 1989, Fundamentals.
3.23
FIGURE 3.7 Solar intensity and angle of incidence.
 Solar altitude angle (Fig. 3.7a and b) is the angle ROQ on a vertical plane between the sun’s ray
OR and its projection on a horizontal plane on the surface of the earth.
 Solar azimuth  (Fig. 3.7a) is the angle SOQ on a horizontal plane between the due-south direction
line OS and the horizontal projection of the sun’s ray OQ.
 Solar declination angle  (Fig. 3.6) is the angle between the earth-sun line and the equatorial
plane. Solar declination changes with the times of the year. It is shown in Fig. 3.6 on June 21.
 Surface-solar azimuth  (Fig. 3.7a and c) is the angle POQ on a horizontal plane between the
normal to a vertical surface OP and the horizontal projection of the sun’s rays OQ.
 Surface azimuth  (Fig. 3.7a) is the angle POS on a horizontal plane between OP and the direction
line SN.
 Latitude angle L (Fig. 3.6) is the angle SOO on the longitudinal plane between the equatorial
plane and the line OO that connects the point of incidence of the sun’s ray on the surface of earth
O and the center of the earth O.
Hour Angle and Apparent Solar Time
Hour angle H (Fig. 3.6) is the angle SOQ on a horizontal plane between the line OS indicating the
noon of local solar time tls and the horizontal projection of the sun’s ray OQ. The values of the hour
angle H before noon are taken to be positive. At 12 noon, H is equal to 0. After 12 noon, H is negative.
Hour angle H, in degrees, can be calculated as
H  0.25 (time in minutes from local solar noon) (3.49)
The relationship between apparent solar time tas, as determined by a sundial and expressed in apparent
solar time, and local standard time tst, both in minutes, is as follows:
(3.50)
where M  local standard time meridian and G  local longitude, both in degrees. In Eq. (3.50), teq,
in minutes, indicates the difference in time between the mean time indicated by a clock running at a
uniform rate and the solar time due to the variation of the earth’s orbital velocity throughout the year.
Solar Angle Relationships
The relationship among the solar angles is given as
sin  sin  sinL 	 cos  cosH cosL (3.51)
and
(3.52)
Angle of Incidence and Solar Intensity
The angle of incidence  (Fig. 3.7a) is the angle between the sun’s rays radiating on a surface and
the line normal to this surface. For a horizontal surface, the angle of incidence H is ROV; for a vertical
surface, the angle of incidence V is ROP; and for a tilted surface, the angle of incidence  is
ROU. Here,  is the angle between the tilted surface ABCD and the horizontal surface.
Let IDN be the intensity of direct normal radiation, or solar intensity, on a surface normal to the
sun’s ray, in Btu/h ft2 (W/m2). In Fig. 3.7b, IDN is resolved into the vertical component RQ  IDN
cos  
sin
 sin L  sin 
cos  cos L
tas  teq 	 tst  4(M  G)
3.24 CHAPTER THREE
sin  and the horizontal component OQ  IDN cos . In Fig. 3.7c, for the right triangle OPQ, angle
POQ  , and hence OP  IDN cos  cos . In Fig. 3.7d, OPL is a right triangle. From point P, a
line PT can be drawn perpendicular to the line normal OU, and hence, two right triangles are
formed: OTP and LTP. In OTP, because angle OPT  , the horizontal component of IDN along the
line normal of the tilting surface OU is OT  IDN cos  cos  sin. In Fig. 3.7e, draw line OU
parallel to line OU. Again, line QM can be drawn from point Q perpendicular to OU. Then the
right triangles PTL and QMR are similar. Angles PLT and QRM are both equal to . In the right
triangle QMR, the component of RQ that is parallel to the line normal to the tilting surface is MR 
IDN sin cos.
The intensity of solar rays normal to a tilted surface I, Btu/h ft2 (W/m2) is the vector sum of
the components of the line normal to the tilted surface, or
(3.53)
3.8 SOLAR RADIATION
Solar radiation provides most of the energy required for the earth’s occupants, either directly or
indirectly. It is the source of indoor daylight and helps to maintain a suitable indoor temperature
during the cold seasons. At the same time, its influence on the indoor environment must be reduced
and controlled during hot weather. The sun is located at a mean distance of 92,900,000 mi,
(149,500,000 km) from the earth, and it has a surface temperature of about 10,800°F (6000°C). It
emits electromagnetic waves at wavelengths of 0.29 to 3.5 	m (micrometers). Visible light has
wavelengths of 0.4 to 0.7 	m and is responsible for 38 percent of the total energy striking the earth.
The infrared region contains 53 percent. At the outer edge of the atmosphere at a mean earth-sun
distance, the solar intensity, called the solar constant Isc, is 434.6 Btu/h ft2 (1371 W/m2 ). The
extraterrestrial intensity Io, Btu/h ft2 (W/m2) varies as the earth-sun distance changes during the
earth’s orbit.
Based on the data from Miller et al. (1983), the breakdown of solar radiation reaching the
earth’s surface and absorbed by the earth is listed in Table 3.5. As listed in Table 3.5, only 50 percent
of the solar radiation that reaches the outer edge of the earth’s atmosphere is absorbed by the
clouds and the earth’s surface. At any specific location, the absorption, reflection, and scattering of
solar radiation depend on the composition of the atmosphere and the path length of the sun’s rays
through the atmosphere, expressed in terms of the air mass m. When the sun is directly overhead,
m  1.
 IDN (cos
 cos  sin  	 sin  cos 
I  IDN cos 
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.25
TABLE 3.5 Components of Solar Radiation That Traverse the Earth’s
Atmosphere
Components Breakdowns
Scattered by air 11% Reflected to space 6%
Scattered to earth 5
Absorbed by water vapor,
dust, etc. 16
Intercepted by clouds 45 Reflected to space 20
Absorbed by clouds 4
Diffused through clouds
and absorbed by earth 21
Traversed through air 28 Absorbed by earth 24
Reflected by earth 4
In Table 3.5, the part of solar radiation that gets through the atmosphere and reaches the earth’s
surface, in a direction that varies with solar angles over time, is called direct radiation. The part that
is diffused by air molecules and dust, arriving at the earth’s surface in all directions, is called diffuse
radiation. The magnitude of solar radiation depends on the composition of the atmosphere, especially
on the cloudiness of the sky. Therefore, different models are used to calculate the solar radiation
reaching the surface of building envelopes.
Solar Radiation for a Clear Sky
ASHRAE recommends use of the following relationships to calculate the solar radiation for a clear
sky. The solar intensity of direct normal radiation IDN, Btu/h ft2 (W/m2 ), can be calculated as
(3.54)
where A  apparent solar radiation when air mass m  0 (its magnitudes are listed in Table 3.6),
Btu/hft (W/m2 )
B  atmospheric extinction coefficient (Table 3.6), which depends mainly on the amount of
water vapor contained in the atmosphere
In Eq. (3.54), the term 1/sin  in exponential form denotes the length of the direct radiation path
through the atmosphere. The term Cn is the clearness number of the sky; Cn takes into account the
dryness of the atmosphere and the dust contained in the air at a geographic location. Estimated Cn
values for nonindustrial locations in the United States are shown in Fig. 3.8.
From Eq. (3.53), the direct radiation radiated onto a horizontal surface through a clear sky IDH,
Btu/h ft2 (W/m2), is the vertical component of IDN, that is,
(3.55)
and the direct radiation irradiated onto a vertical surface for clear sky IDV, Btu/h ft2 (W/m2 ), is the
horizontal component of IDN, or
(3.56) IDV  IDN cos V  IDN cos  cos 
IDH  IDN cos H  IDN sin 
IDN 
ACn
exp (B/sin )
3.26 CHAPTER THREE
TABLE 3.6 Extraterrestrial Solar Radiation and Related Data for 21st Day of Each Month,
Base Year 1964
A B C
Equation of Declination, (Dimensionless
I0, Btu/h ft2 time, min deg Btu /h ft2 ratios)
January 448.8 11.2 20.0 390 0.142 0.058
Febuary 444.2 13.9 10.8 385 0.144 0.060
March 437.7 7.5 0.0 376 0.156 0.071
April 429.9 	1.1 	11.6 360 0.180 0.097
May 423.6 	3.3 	20.0 350 0.196 0.121
June 420.2 1.4 	23.45 345 0.205 0.134
July 420.3 6.2 	20.6 344 0.207 0.136
August 424.1 2.4 	12.3 351 0.201 0.122
September 430.7 	7.5 0.0 365 0.177 0.092
October 437.3 	15.4 10.5 378 0.160 0.073
November 445.3 	13.8 19.8 387 0.149 0.063
December 449.1 	1.6 23.45 391 0.142 0.057
A: apparent solar radiation; B: atmospheric extinction coefficient; C: diffuse radiation factor.
Source: ASHRAE Handbook 1997, Fundamentals. Reprinted with permission.
From Eq. (3.53), for the direct radiation radiated onto a tilted surface at an angle  with the horizontal
plane through clear sky ID, Btu/h ft2 (W/m2 ), can be evaluated as
(3.57)
Most of the ultraviolet solar radiation is absorbed by the ozone layer in the upper atmosphere. Direct
solar radiation, through an air mass m  2, arrives on the earth’s surface at sea level during a
clear day with a spectrum of  3 percent in the ultraviolet, 44 percent in the visible, and 53 percent
in the infrared.
The diffuse radiation Id, Btu/h ft2 (W/m2 ), is proportional to IDN on cloudless days and can be
approximately calculated as
(3.58)
where C  diffuse radiation factor, as listed in Table 3.6, and Cn  clearness number of sky from
Fig. 3.8. In Eq. (3.58), Fss indicates the shape factor between the surface and the sky, or the fraction
of shortwave radiation transmitted through the sky that reaches the surface. For a vertical surface
Fss  0.5, for a horizontal surface Fss  1, and for any tilted surface with an angle 
(3.59)
The total or global radiation on a horizontal plane IG, Btu/h ft2 (W/m2), recorded by the U.S.
Fss 
1.0 	 cos 
2
Id 
CIDN Fss
Cn 2
ID  I  IDN cos   IDN (cos
 cos
 sin  	 sin
 cos)
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.27
FIGURE 3.8 Estimated atmospheric clearness numbers in the United States for nonindustrial localities.
National Climatic Data Center (NCDC), can be calculated as
(3.60)
The reflection of solar radiation from any surface Iref, Btu/h ft2 (W/m2 ), is given as
(3.61)
where s  reflectance of the surface and Fsr  shape factor between the receiving surface and the
reflecting surface. The ground-reflected diffuse radiation falling on any surface Isg, Btu/h ft2
(W/m2), can be expressed as
Isg  g Fsg IG (3.62)
where Fsg  shape factor between the surface and the ground and g  reflectance of the ground.
For concrete, g  0.23, and for bitumen and gravel, g  0.14. A mean reflectance g  0.2 is
usually used for ground.
The total intensity of solar radiation It, Btu/h ft2 (W/m2 ), falling on a surface at a direction
normal to the surface on clear days, is given by
(3.63)
where I  component of reflected solar intensity in direction normal to surface, Btu /h ft2 (W/m2 )
Iref.DN  component of reflected solar intensity in direction of sun ray, Btu/h ft2 (W/m2 )
Solar Radiation for a Cloudy Sky
For cloudy skies, the global horizontal irradiation IG*, Btu/h ft2 (W/m2), usually can be obtained
from the NCDC. If it is not available, then it can be predicted from the following relationship:
(3.64a)
Here Ccc indicates the cloud cover, on a scale of 0 to 10, and can be calculated by
(3.64b)
where CT  total cloud amount
Ccir. j  clouds covered by cirriforms, including cirrostratus, cirrocumulus, and cirrus, in j  1
to 4 layers
Both CT and Ccir. j values can be obtained from the major weather stations. The values of coefficients
P, Q, and R, according to Kimura and Stephenson (1969), are listed below:
P Q R
Spring 1.06 0.012  0.0084
Summer 0.96 0.033  0.0106
Autumn 0.95 0.030  0.0108
Winter 1.14 0.003  0.0082
Ccc  CT  0.5 4
j1
Ccir. j
I*G  1 	
Ccc Q
P
	
Ccc
2 R
P IG
 (IDN 	 Iref.DN) cos 
	 Id
It  ID 	 Id 	 Iref. S
Iref  sFsr (ID 	 Id)
IG  ID 	 Id  IDN sin 
	
C
Cn 2
3.28 CHAPTER THREE
The direct radiation for a cloudy sky ID*, Btu/h ft2 (W/m2), can be calculated as
(3.65)
The diffuse radiation for a cloudy sky Id*, Btu/h ft2 (W/m2 ), is calculated as
Id*  IG*  ID* (3.66)
3.9 FENESTRATION
Fenestration is the term used for assemblies containing glass or light-transmitting plastic, including
appurtenances such as framing, mullions, dividers, and internal, external, and between-glass shading
devices, as shown in Fig. 3.12a. The purposes of fenestrations are to (1) provide a view of the
outside world, (2) permit entry of daylight, (3) admit solar heat as a heating supplement in winter,
(4) act as an emergency exit for single-story buildings, and (5) add to aesthetics.
Solar radiation admitted through a glass or window pane can be an important heat gain for
commercial buildings, with greater energy impact in the sun belt. HVAC&R designers are asked to
control this solar load while providing the required visibility, daylight, and winter heating benefits
as well as fire protection and safety features.
Types of Window Glass (Glazing)
Most window glasses, or glazing, are vitreous silicate consisting of silicon dioxide, sodium oxide,
calcium oxide, and sodium carbonate. They can be classified as follows:
 Clear plate or sheet glass or plastic. Clear plate glass permits good visibility and transmits more
solar radiation than other types.
 Tinted heat-absorbing glass. Tinted heat-absorbing glass is fabricated by adding small amounts of
selenium, nickel, iron, or tin oxides. These produce colors from pink to green, including gray or
bluish green, all of which absorb infrared solar heat and release a portion of this to the outside
atmosphere through outer surface convection and radiation. Heat-absorbing glass also reduces
visible light transmission.
 Insulating glass. Insulating glass consists of two panes—an outer plate and a inner plate—or
three panes separated by metal, foam, or rubber spacers around the edges and hermetically sealed
in a stainless-steel or aluminum-alloy structure. The dehydrated space between the glass panes
usually has a thickness of 0.125 to 0.75 in. (3.2–19 mm) and is filled with air, argon, or other
inert gas. Air- or gas-filled space increases the thermal resistance of the fenestration.
 Reflective coated glass. Reflective glass has a microscopically thin layer of metallic or ceramic
coating on one surface of the glass, usually the inner surface of a single-pane glazing or the outer
surface of the inner plate for an insulating glass. For a single pane, the coating is often protected
by a layer of transparent polyester. The chromium and other metallic coatings give excellent re-
flectivity in the infrared regions but reduced transmission of visible light compared to clear plate
and heat-absorbing glass. Reflections from buildings with highly reflective glass may blind drivers,
or even kill grass in neighboring yards.
 Low-emissivity (low-E) glass coatings. Glazing coated with low-emissivity, or low-E, films has
been in use since 1978. It is widely used in retrofit applications. A low-emissivity film is usually
a vacuum-deposited metallic coating, usually aluminum, on a polyester film, at a thickness of
about 4  107 in. (0.01 	m). Because of the fragility of the metal coating, protection by another
polyester film against abrasion and chemical corrosion must be provided. Recently, copper and
I*D 
I*G sin  (1  Ccc /10)
sin 
	 C/Cn 2
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.29
silver coatings on polyethylene and polypropylene film for protection have been used for better
optical transmission. A low-E film coating reduces the U value about 25 to 30 percent for single
panes. When combined with other solar control devices, these films can reduce solar heat gain
further.
Optical Properties of Sunlit Glazing
When solar radiation inpinges on the outer surface of a plate of glass with an intensity of I at an
angle of incidence , as shown in Fig. 3.9, a portion of the solar radiation is transmitted, another
portion is reflected from both inner and outer surfaces, and the remaining portion is absorbed.
Let r indicate the portion reflected and a the portion absorbed. Also let  be the angle of incidence.
In Fig. 3.9, it can be seen that the portion of the solar radiation transmitted through the glass
is actually the sum of the transmittals after successive multiple reflections from the outer and inner
surfaces. The decimal portion of I transmitted through the glass, or , represents the transmittance
of the window glass. Similarly, the portion of solar radiation reflected from the window glass is the
sum of the successive reflections from the outer surfaces after multiple reflections and absorptions,
3.30 CHAPTER THREE
FIGURE 3.9 Simplified representation of multiple transmissions, reflections, and absorptions of solar radiation
at glass surfaces.
and it is identified as , the reflectance. The portion absorbed is the sum of the successive absorptions
within the glass , or its absorptance.
Figure 3.10 illustrates the spectral transmittance of several types of glazing. All are transparent
for shortwave solar radiation at a wavelength between 0.29 and 3	m and are opaque to longwave
radiation in the infrared range with a wavelength greater than 3	m. Most interior furnishings,
equipment, and appliances have an outer surface temperature lower than 250°F (120°C), emitting
almost all longwave radiation. At such temperatures, glass is opaque to longwave radiation emitted
from inside surfaces and lets only shortwave radiation through. This trapping of longwave radiation
indoors is called the greenhouse effect.
In Fig. 3.10 one can also see that clear plate glass has a high transmittance  0.87 for visible
light, and that  0.8 for infrared from 0.7 to 3	m. Heat-absorbing glass has a lower  and higher
absorptance  for both visible light and infrared radiation. Bluish-green heat-absorbing glass has a
higher  in the visible light range and a lower  in the infrared range than gray heat-absorbing
glass. Some reflective glazing has a high reflectance and a significantly lower  in the visible light
range and is opaque to radiation at wavelengths greater than 2	m. Such characteristics for heatabsorbing
and reflective glasses are effective for reducing the amount of solar heat entering the
conditioned space during cooling as well as heating seasons. This fact sometimes presents a
dilemma for the designer, who must finally compromise to get the optimum combination of con-
flicting properties.
Another important property of glass is that both  and  decrease and  increases sharply as the
incident angle  increases from 60 to 90°. At 90°,  and  are 0 and  is equal to 1. That is why the
solar radiation transmitted through vertical glass declines sharply at noon during summer with solar
altitudes   70°.
For all types of plate glass, the sum of these radiation components is always equal to 1, that is,
	 	  1
and
I  I 	 I 	 I (3.67)
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.31
FIGURE 3.10 Spectral transmittance for various types of window glasses.
(Source: ASHRAE Handbook 1989, Fundamentals. Reprinted by permission.)
3.10 HEAT ADMITTED THROUGH WINDOWS
For external glazing without shading, the heat gain admitted into the conditioned space through
each square foot of sunlit area As of window Qwi /As, Btu/h ft2 (W/m
2), can be calculated as follows:
Heat gain through each ft2

solar radiation transmitted
	
inward heat flow from glass
of sunlit window through window glass inner surface into conditioned space
That is,
(3.68)
where QRCi  inward heat flow from the inner surface of an unshaded sunlit window, Btu/h (W).
Heat Gain for Single Glazing
For an external, sunlit single-glazed window without shading, the inward heat flow from the inner
surface of the glass, as shown in Fig. 3.11, can be evaluated as
QRCi  inward absorbed radiation 	 conductive heat transfer
(3.69)  UAs It
ho
	 To  Ti
Qwi
As

It 	 QRCi
As
3.32 CHAPTER THREE
FIGURE 3.11 Heat admitted through a single-glazing window
glass.
where ho  heat-transfer coefficient for outdoor surface of window glass, Btu/h ft2 °F (W/m2 °C)
To  outdoor air temperature, °F (°C)
Ti  indoor air temperature,°F (°C)
Heat admitted through a unit area of the single-glazing window glass is
(3.70)
Solar heat gain coefficient (SHGC) is the ratio of solar heat gain entering the space through the
window glass to the incident solar radiation, total shortwave irradiance for a single-glazed window
is given as
(3.71)
In these last three equations, U indicates the overall heat-transfer coefficient of the window, in
Btu/h ft2 °F (W/m2 °C), and can be calculated as
(3.72)
In Eq. (3.72), A represents area, in square feet (square meters); subscript wg indicates glass, and eg
signifies the edge of the glass including the sealer and spacer of the insulating glass. The edge of
glass has a width of about 2.5 in. (64 mm). The subscript f means the frame of the window. Some U
values for various types of windows at winter design conditions are listed in Table 3.7. For summer
U 
Uwg Awg 	 Ueg Aeg 	 Uf Af
Awg 	 Aeg 	 Af


	
U
ho
SHGC  Qws / It As
 SHGCIt 	 U(To  Ti)
Qwi
As

It 	 U 
It
ho
	 To  Ti
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.33
TABLE 3.7 Overall Heat-Transfer Coefficient U Values for Windows at Winter Conditions* with Commercial Type of Frame,
Btu/h ft2 °F)
Aluminum
frame with Wood
Gas Space Emittance† Aluminum thermal or vinyl
between between of low-E frame of break, with frame of
Type glasses glasses, in. film Glass Edge Uf  1.9 Uf  1.0 Uf  0.41
Overall coefficient Btu/h ft2 °F
Single glass 1.11 1.23 1.10 0.98
Double glass Air 3/8 0.52 0.62 0.74 0.60 0.51
Double glass Air 3/8 0.40 0.43 0.55 0.67 0.54 0.45
Double glass Air 3/8 0.15 0.36 0.51 0.62 0.48 0.39
Double glass Argon 3/8 0.15 0.30 0.48 0.57 0.43 0.34
Triple glass Air 3/8 0.34 0.50 0.60 0.46 0.38
Triple glass Air 3/8 0.40 0.30 0.48 0.57 0.44 0.35
Triple or double
glass with polyester
film suspended
in between Argon 3/8 0.15 0.17 0.43 0.47 0.34 0.25
*Winter conditions means 70°F indoor, 0°F outdoor temperature and a wind speed of 15 mph.
†Low-E film can be applied to surface 2 for double glass (see Fig. 3.12b) or surface 2, 3, 4, or 5 for triple glass (any surface other than outer and
inner surfaces).
Source: Abridged with permission from ASHRAE Handbook, 1989, Fundamentals.
design conditions at 7.5 mph (3.3 m/s) wind speed, the listed U values should be multiplied by a
factor of 0.92. The U values of windows depend on the construction of the window, the emissivity
of the surfaces of glass or plastic sheets, and the air velocity flowing over the outdoor and indoor
surfaces. Qws indicates the solar heat gain entering the space, in Btu/h (W).
As shown in Fig. 3.11, solar radiation having a total intensity I, Btu/h ft2 (W/m2), when it is radiated
on the outer surface of a vertical pane with an angle of incidence v, the line normal indicating
total shortwave irradiance It, in Btu/h ft2 (W/m2) actually consists of
(3.73)
Heat Gain for Double Glazing
For a double-glazed window, the inward heat flow per square foot of the inner surface of the glass,
as shown in Fig. 3.12b, is
(3.74)  U 
Io
ho
	  1
ho
	
1
ha  Ii 	 To  Ti 	
QRCi
As
 Nio Io 	 Nii Ii 	 U(To  Ti)
It  I cos v
I  IDN 	 Id 	 Iref. DN
3.34 CHAPTER THREE
FIGURE 3.12 Heat flow through an insulating glass (double-pane) window. (a) Construction of a typical insulating glass; (b) heat flow
and temperature profiles.
where Nio  inward fraction of solar radiation absorbed by outdoor glass
Nii  inward fraction of solar radiation absorbed by indoor glass
Io, Ii  solar intensity irradiated on outdoor and indoor glass, Btu/h ft2 (W/m2)
The solar radiation absorbed by the outdoor and indoor panes, in Btu/h ft2 (W/m2), is
Io  o It Ii  i It (3.75)
Absorptance of the outdoor panes o is
(3.76)
Absorptance of the indoor pane is
(3.77)
Subscript o indicates outdoor; i, indoor; and 1, 2, 3, and 4, the surfaces of the panes as shown in
Fig. 3.12b.
The airspace heat-transfer coefficient ha, Btu/h ft2 °F (W/m2 °C), is the reciprocal of the R
value of the airspace Ra listed in Table 3.4, that is,
(3.78)
The transmittance of solar radiation through both outdoor and indoor panes oi can be calculated
as
(3.79)
Then the heat admitted per square foot (square meter) through a double-glass window Qwoi /As
Btu/h ft2 (W/m2), can be calculated as
(3.80)
The SHGCoi for a double-glass window can be calculated as
(3.81)
Because a glass plate is usually between 0.125 and 0.25 in. (3 and 6 mm) thick, with a thermal
conductivity k of about 0.5 Btu/h ft °F (W/m°C), there is only a small temperature difference
between the inner and outer surfaces of the plate when solar radiation is absorbed. For the sake
of simplicity, it is assumed that the temperature of the plate Tg is the same in the direction of heat
flow.
For a double-glazed window, the glass temperature of the outdoor pane Tgo, °F (°C), can be
calculated as
(3.82) Tgo  To 	 Io 	
Ii 
QRCi
As  1
ho
	
Rgo
2 
SHGCoi 
oi 	 U o
ho
	
i
ha
 SHGCoi It 	 U(To  Ti)
Qwoi
As

oi It 	 Uo
h

	
i
h

	
i
ha 	 It 	 U(To  Ti)
oi 
o i
1 
2 3
ha 
1
Ra
i 
3 o
1 
2 3
o 
1 	
2 oi 3
1 
2 3
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.35
The temperature of the indoor pane Tgi, °F (°C), is calculated as
(3.83)
where Rgo, Rgi  R values of outdoor and indoor panes, h ft2 °F/Btu (m2 °C/W)
hi  heat-transfer coefficient of inside surface 4 (see Fig. 3.12b) of inside plate,
Btu/h ft2 °F (W/m2 °C)
Shading Coefficients
The shading coefficient is defined as the ratio of solar heat gain of a glazing assembly of specific
construction and shading devices at a summer design solar intensity and outdoor and indoor
temperatures, to the solar heat gain of a reference glass at the same solar intensity and outdoor and
indoor temperatures. The reference glass is double-strength sheet glass (DSA) with transmittance
  0.86, reflectance  0.08, absorptance  0.06, and FDSA  0.87 under summer design
conditions. The shading coefficient SC is an indication of the characteristics of a glazing and the
associated shading devices, and it can be expressed as
(3.84)
where SHGCw  solar heat gain coefficient of specific type of window glass
SHGCDSA  solar heat gain coefficient of standard reference double-strength sheet glass
Shading coefficients of various types of glazing and shading devices are presented in Table 3.8.

SHGCw
SHGCDSA

SHGCw
0.87
 1.15 SHGCwi
SC 
solar heat gain of specific type of window glass
solar heat gain of double-strength sheet glass
Tgi  Tr 	 QRCi  1
hi
	
Rgi
2 
3.36 CHAPTER THREE
TABLE 3.8 Shading Coefficients for Window Glass with Indoor Shading Devices
Venetian blinds Roller shade Draperies
Thickness Solar
Type of glass of glass, in. transmittance Glass Med.* Light† Opaque, white Translucent Med.‡ Light§
Clear 0.87 to 0.79 0.74 0.67 0.39 0.44 0.62 0.52
Heat-absorbing or 0.46 0.57 0.53 0.30 0.36 0.46 0.44
0.34 0.54 0.52 0.28 0.32
0.24 0.42 0.40 0.28 0.31 0.38 0.36
Reflective-coated 0.30 0.25 0.23
0.40 0.33 0.29
0.50 0.42 0.38
Insulating glass Outer Inner
Clear out , or 0.87 0.87 0.62 0.58 0.35 0.40
Clear in
Heat-absorbing out 0.46 0.80 0.39 0.36 0.22 0.30
Clear in
Reflective glass 0.20 0.19 0.18
0.30 0.27 0.26
0.40 0.34 0.33
*Med. indicates medium color.
† Light indicates light color.
‡Draperies Med. represents draperies of medium color with a fabric openness of 0.10 to 0.25 and yarn reflectance of 0.25 to 0.50.
§ Draperies Light repesents draperies of light color with a fabric openness below 0.10 and yarn reflectance over 0.50.
Source: Adapated with permission from ASHRAE Handbook 1989, Fundamentals.
1
4
1
8
3
32
3
8
1
4
3
16
3
32
Solar Heat Gain Factors and Total Shortwave Irradiance
The solar heat gain factor (SHGF), Btu/h ft2 (W/m2), is designated as the average solar heat gain
during cloudless days through DSA glass. In the ASHRAE Handbook 1993, Fundamentals are tabulated
SHGF value for various latitudes, solar times, and orientations for load and energy calculations.
For calculating the summer cooling peak load, the concept of maximum SHGF has been
introduced. This is the maximum value of SHGF on the 21st of each month for a specific latitude,
as listed in Table 3.9. For high elevations and on very clear days, the actual SHGF may be 15
percent higher than the value, listed in Table 3.9. In dusty industrial areas or at very humid locations,
the actual SHGF may be lower.
According to ASHRAE Handbook 1997, Fundamentals, Gueymard (1995) provides a comprehensive
model for calculating the spectral and broadband total shortwave irradiance It, in Btu/h ft2
(W/m2) for cloudless sky conditions and allows the input of the concentrations of a variety atmospheric
constituents.
Example 6.1. A double-glass window of a commercial building facing west consists of an outdoor
clear plate glass of 0.125-in. and an indoor reflective glass of 0.25-in. thickness with a reflective
film on the outer surface of the indoor glass, as shown in Fig. 3.12b. This building is located at
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.37
TABLE 3.9 Maximum Solar Heat Gain Factors (Max SHGF)
Max SHGF, Btu /h ft2
N
(shade) NNE/NNW NE/NW ENE/WNW E/W ESE/WSW SE/SW SSE/SSW S HOR*
North latitude, 40°
Jan. 20 20 20 74 154 205 241 252 254 133
Feb. 24 24 50 129 186 234 246 244 241 180
Mar. 29 29 93 169 218 238 236 216 206 223
Apr. 34 71 140 190 224 223 203 170 154 252
May 37 102 165 202 220 208 175 133 113 265
June 48 113 172 205 216 199 161 116 95 267
July 38 102 163 198 216 203 170 129 109 262
Aug. 35 71 135 185 216 214 196 165 149 247
Sept. 30 30 87 160 203 227 226 209 200 215
Oct. 25 25 49 123 180 225 238 236 234 177
Nov. 20 20 20 73 151 201 237 248 250 132
Dec. 18 18 18 60 135 188 232 249 253 113
North latitude, 32°
Jan. 24 24 29 105 175 229 249 250 246 176
Feb. 27 27 65 149 205 242 248 232 221 217
Mar. 32 37 107 183 227 237 227 195 176 252
Apr. 36 80 146 200 227 219 187 141 115 271
May 38 111 170 208 220 199 155 99 74 277
June 44 122 176 208 214 189 139 83 60 276
July 40 111 167 204 215 194 150 96 72 273
Aug. 37 79 141 195 219 210 181 136 111 265
Sept. 33 35 103 173 215 227 218 189 171 244
Oct. 28 28 63 143 195 234 239 225 215 213
Nov. 24 24 29 103 173 225 245 246 243 175
Dec. 22 22 22 84 162 218 246 252 252 158
*Horizontal surface
Source: Abridged with permission from ASHRAE Handbook 1989, Fundamentals.
40° north latitude. The detailed optical properties of their surfaces are as follows:
The R value of the 0.25-in. (6-mm) thickness indoor glass is Rg  0.035 h ft2 °F/Btu (0.0063
m2 °C/W), and that of the enclosed airspace is Ra  1.75 h ft2 °F/Btu (0.32 m2 °C/W).
At 4 P.M. on July 21, the outdoor temperature at this location is 100°F (37.8°C), the indoor temperature
is 76°F (24.4°C), and the total solar intensity at a direction normal to this west window is
248 Btu/h ft2 (782 W/m2). The outdoor surface heat-transfer coefficient ho  4.44 Btu/h ft2 °F
(25.2 W/m2°C), and the heat-transfer coefficient of the inner surface of the indoor glass hi  2.21
Btu/h ft2 °F (12.5 W/m2 °C). Calculate the following:
1. The inward heat flow of this window
2. The temperatures of the outdoor and indoor glasses
3. The shading coefficient of this double-glass window
4. The total heat gain admitted through this window
Solution
1. From Eq. (3.76), the absorption coefficient of the outdoor glass can be calculated as
And from Eq. (3.77), the absorption coefficient of the indoor glass is
From Eq. (3.75), the heat absorbed by the outdoor glass is
Also, the heat absorbed by the indoor glass is calculated as
And then, from Eq. (3.22), the overall heat-transfer coefficient of this double-glass window is

1
1/2.21 	 0.035 	 1.75 	 0.035 	 1/4.44
 0.400 Btu / h ft2F
U 
1
1/hi 	 Rg 	 Ra 	 Rg 	 1/ho
Ii  aiIt  0.135  248  33.5 Btu / h ft2
Io 
o It  0.189  248  46.9 Btu / h ft2
i 
3 o
1 
2 3

0.16  0.8
1  0.08  0.68
 0.135
 0.12 	
0.12  0.8  0.68
1  0.08  0.68
 0.189
o 
1 	
2 o 3
1 
2 3
 0.80 1  0.08 1  0.12 e1  0.84
2  0.08 2  0.12 e2  0.84
i  0.16 3  0.68 3  0.16 e3  0.15
4  0.08 4  0.76 e4  0.84
3.38 CHAPTER THREE
The inward heat flow from the inner surface of the indoor glass, from Eq. (3.74), is given as
2. From Eq. (3.82), the temperature of the outdoor glass is
and the temperature of the indoor glass is
3. From Eq. (3.79), the transmittance for both panes is
and from Eq. (3.81), the solar heat gain coefficient for the double-glass window is
Then, from Eq. (3.84), the shading coefficient of this double-glass window is given by
4. From Eq. (3.80), the total heat gain admitted through this double-glass window per square
foot of sunlit area is
Selection of Glazing
During the selection of glazing, the following factors should be considered: visual communication,
use of daylight, thermal comfort, summer and winter solar heat gain, street noise attenuation, safety
 64.2 	 9.6  73.8 Btu / h ft2 (233 W/ m2)
Qwi
As
 SHGCoi It 	 U(To  Ti)  0.259  248 	 0.40(100  76)
SC 
SHGCoi
SHGCDSA

0.259
0.87
 0.297
 0.135 	
0.40  0.189
4.44
	 0.40
4.44
	 0.40  1.75 0.135  0.259
SHGCoi 
oi 	
Uo
ho
	 U
ho
	
U
ha i
oi 
o i
1 
2 3

0.8  0.16
1  0.08  0.68
 0.135
 76 	 40.29  1
2.21
	
0.035
2   94.9F (34.9C)
Tgi  Ti 	
QRCi
As
1
hi
	
Rg
2 
 100 	 (46.9 	 33.5  40.29) 1
4.44
	
0.035
2   109.7F
Tgo  To 	 o It 	
i It 
QRCi
As
1
ho
	
Rg
2 
 0.400  46.9
4.44
	 33.5
1
4.44
	 1.75 	 100  76	  40.29 Btu / h ft2 (127.1 W/ m2)
QRCi
As
 U 
o It
ho
	
i It
1
ho
	
1
ha
	 To  Tr	
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.39
and fire protection, and life-cycle cost analysis. Energy conservation considerations, including the
control of solar heat with the optimum combination of absorbing and reflective glass and various
shading devices, are covered in the next section.
To reduce the heat loss through glass during winter, one can install double or triple glazing,
storm windows, or low-emission film coating on the surface of the glass. Elmahdy (1996) and de
Abreu et al. (1996) tested the thermal performance of seven insulating glass units. If a clear,
double-glazed insulating glass unit with a silicone foam spacer of 0.5 in. (13 mm) between two
panes is taken as the base unit, the U values of these seven insulating glass units are as follows:
Low-e coating reduces the U value by about 30 percent, and the triple-glazing drops about onethird
compared with a clear, double-glazed insulating glass unit. If the width of the airspace is
decreased to 0.25 in. (6.5 mm), its U value will increase 15 percent. When the airspace is wider
than 0.5 in. (13 mm), regardless of whether a metal or silicone foam spacer is adopted, neither has
any significant effect on U value of the insulation glass unit.
3.11 SHADING OF GLASS
Shading projected over the surface of glass significantly reduces its sunlit area. Many shading
devices increase the reflectance of the incident radiation. There are two types of shading: deliberately
installed shading devices, which include indoor and external shading devices, and shading
from adjacent buildings.
Indoor Shading Devices
Indoor shading devices not only provide privacy but also are usually effective in reflecting part of
the solar radiation back to the outdoors. They also raise the air temperature of the space between
the shading device and the window glass, which in turn reduces the conductive heat gain in summer.
Indoor shading devices are easier to operate and to maintain and are more flexible in operation
than external shading devices. Three types of indoor shading devices are commonly used: venetian
blinds, draperies, and roller shades.
Venetian Blinds. Most horizontal venetian blinds are made of plastic or aluminum slats, spaced 1
to 2 in. (25 to 50 mm) apart, and some are made of rigid woven cloth. The ratio of slat width to slat
spacing is generally 1.15 to 1.25. For light-colored metallic or plastic slats at a 45° angle, typical
optical properties are  0.05,  0.55, and  0.40. Vertical venetian blinds with wider slats
are widely used in commercial buildings.
Consider a single-glazed window combined with indoor venetian blinds at a slat angle   45°,
as shown in Fig. 3.13. Let the subscripts g, v, and a represent the glass, the venetian blinds, and the
U value,
Unit Glazing Airspace, in. Spacer Btu/h ft2 °F
1 Clear, double-glazed 0.5 Foam 0.51
2 Clear, double-glazed 0.5 Aluminum 0.51
3 Clear, double-glazed 0.25 Foam 0.58
4 Clear, double-glazed 0.75 Foam 0.51
5 Low-e, double-glazed 0.5 Foam 0.36
6 Clear, Triple-glazed 0.5 Foam 0.32
7 Clear, Triple-glazed 0.25 Foam 0.39
3.40 CHAPTER THREE
air between the venetian blinds and the glass. Also let o indicate the outward direction and i the
inward direction. Of the solar radiation transmitted through the glass and radiating on the surface of
the slats g It,
 A fraction vo g It is reflected outward from the surface of each slat.
 Another fraction vi g It is reflected from the slat surface into the conditioned space.
 A third fraction absorbed by the slat is either convected away by the space air or reflected from
the surface to the indoor surroundings in the form of longwave radiation.
If the glass has a high transmittance, the slat temperature Tv1 will be higher, as shown by the lower
solid temperature curve in Fig. 3.13. If the glass has a high absorptance, its temperature Tg2 will be
higher. In Table 3.8 the shading coefficients of various combinations of venetian blinds and glazing
are listed.
According to results of a field survey by Inoue et al. in four office buildings in Japan in 1988,
60 percent of the venetian blinds were not operated during the daytime. The incident angle of the
direct solar radiation had greater influence on operation of the blinds than the intensity. Automatic
control of slat angle and of raising and lowering the blinds is sometimes used and may become
more popular in the future.
Draperies. These are fabrics made of cotton, regenerated cellulose (such as rayon), or synthetic
fibers. Usually they are loosely hung, wider than the window, and pleated; and they can be drawn
open or closed as required. Drapery-glass combinations reduce the solar heat gain in summer and
increase the thermal resistance in winter. Reflectance of the fabric is the dominant factor in the
reduction of heat gain, and visibility is a function of the openness of the weave.
Roller Shades. These are sheets made of treated fabric or plastic that can be pulled down to cover
the window or rolled up. Roller shades have a lower SC than do venetian blinds and draperies.
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.41
FIGURE 3.13 Heat transfer through a single-glazed window
combined with venetian blinds.
When glass is covered with shades, any outdoor visual communication is blocked, and the visible
light transmittance is less than that of other indoor shading devices.
External Shading Devices
External shading devices include overhangs, side fins, egg-crate louvers, and pattern grilles, as
shown in Fig. 3.14. They are effective in reducing the solar heat gain by decreasing the sunlit area.
However, the external shading devices do not always fit into the architectural requirements and are
less flexible and more difficult to maintain. Pattern grilles impair visibility significantly.
Figure 3.15 shows the shaded area of a glass pane constructed with both overhang and side fin.
The profile angle  is defined as the angle between a horizontal plane and a tilted plane that includes
the sun’s rays. We see that tan   UQ/OR  tan / cos . Let SW be the width of the
shadow and SH be the height of the shadow projected on the plane of the glass by direct solar radiation,
in feet (meters). Also, let W be the width of the glass and H be the height. Then the shadow
width on the plane of the glass is
SW  PV tan  (3.85)
and the shadow height is
(3.86)
where PV  projection of side fin plus mullion and reveal, ft (m)
PH  horizontal projection of overhang plus transom and reveal, ft (m)
The projection factor due to the external shading device Fpro can be calculated as
(3.87) Fpro 
PH
H
SH  PH tan   PH
tan 
cos 
3.42 CHAPTER THREE
FIGURE 3.14 Types of external shades. (a) Overhang; (b) egg-crate
louver; (c) pattern grilles.
The net sunlit area of the glass As, ft2 (m2), can then be calculated as
As  (W  SW)(H  SH) (3.88a)
and the shaded area of the glass Ash, ft2 (m2), is given by
(3.88b)
where Ag  area of the glass, ft2 (m2).
Shading from Adjacent Buildings
Shadows on glass cast by adjacent buildings significantly reduce the sunlit area of the glass.
For example, in Fig. 3.16 we see the area on building A shaded because of the presence of
building B. Let two sides of the shaded building A coincide with the X and Y axes on the plan
view shown in Fig. 3.16. In the elevation view of Fig. 3.16, the shadow height on the facade of
building A is
(3.89)
and the shadow width SW on the facade of building A is
(3.90) SW  WOB 	 WB  LAB tan 
SH  HB  LAB tan 
Ash  Ag  As  WH  (W  SW)(H  SH)
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.43
FIGURE 3.15 Shaded area of a window glass constructed with an overhang and side fin.
where HB  height of building B, ft (m)
WB  width of building B, ft (m)
LAB  distance between two buildings along X axis, ft (m)
WOB  distance between X axis and building B, ft (m)
Subscript A indicates the shaded building and B the shading building.
Because of the sun’s varying position, the solar altitude  and the surface solar azimuth  change
their values from time to time. Table 3.10 lists the solar altitude  and solar azimuth  at north
latitudes 32° and 40°. To evaluate the shaded area of the outer surface of a building, a computer program
can determine which of several hundred representative points on this outer surface are sunlit or
shaded at specific times of the day. Then a ratio of shaded area to total area can be calculated.
At a certain time instant, a window glass area can be shaded due to the effect of the overhang
and vertical fins, or the effect of the adjacent building, or both. For the calculation of the combined
shading effect of overhangs, vertical fins, and adjacent building, it is recommended that the number
of sunlit and shaded windows of an outer surface of a building under the effect of the adjacent
building be calculated first. According to Eq. (3.88a), the total sunlit area can be calculated as the
product of the net sunlit area for each window because of the overhang and vertical fins and the
number of windows in the sunlit area of that outer surface.
3.44 CHAPTER THREE
FIGURE 3.16 Shading from adjacent building.
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.45
TABLE 3.10 Solar Altitude (ALT) and Solar Azimuth (AZ)
North latitude
32° 40°
Solar time Solar position Solar time Solar time Solar position Solar time
Date AM ALT AZ PM AM ALT AZ PM
Dec. 8 10 54 4 8 5 53 4
9 20 44 3 9 14 42 3
10 28 31 2 10 21 29 2
11 33 16 1 11 25 15 1
12 35 0 12 12 27 0 12
Jan. 7 1 65 5 8 8 55 4
	 8 13 56 4 9 17 44 3
Nov. 9 22 46 3 10 24 31 2
10 31 33 2 11 28 16 1
11 36 18 1 12 30 0 12
12 38 0 12
Feb. 7 7 73 5 7 4 72 5
	 8 18 64 4 8 15 62 4
Oct. 9 29 53 3 9 24 50 3
10 38 39 2 10 32 35 2
11 45 21 1 11 37 19 1
12 47 0 12 12 39 0 12
Mar. 7 13 82 5 7 11 80 5
	 8 25 73 4 8 23 70 4
Sept. 9 37 62 3 9 33 57 3
10 47 47 2 10 42 42 2
11 55 27 1 11 48 23 1
12 58 0 12 12 50 0 12
Apr. 6 6 100 6 6 7 99 6
	 7 19 92 5 7 19 89 5
Aug. 8 31 84 4 8 30 79 4
9 44 74 3 9 41 67 3
10 56 60 2 10 51 51 2
11 65 37 1 11 59 29 1
12 70 0 12 12 62 0 12
May 6 10 107 6 5 2 115 7
	 7 23 100 5 6 13 106 6
July 8 35 93 4 7 24 97 5
9 48 85 3 8 35 87 4
10 61 73 2 9 47 76 3
11 72 52 1 10 57 61 2
12 78 0 12 11 66 37 1
12 70 0 12
June 5 1 118 7 5 4 117 7
6 12 110 6 6 15 108 6
7 24 103 5 7 26 100 5
8 37 97 4 8 37 91 4
9 50 89 3 9 49 80 3
10 62 80 2 10 60 66 2
11 74 61 1 11 69 42 1
12 81 0 12 12 73 0 12
Source: ASHRAE Handbook 1981, Fundamentals. Reprinted with permission.
3.12 HEAT EXCHANGE BETWEEN THE OUTER BUILDING
SURFACE AND ITS SURROUNDINGS
Because atmospheric temperature is lower at high altitudes, there is always a radiant heat loss from
the outer surface of the building to the sky vault without clouds. However, it may be offset or partly
offset by reflected solar radiation from the ground on a sunny day. Radiant heat loss from the building
needs to be calculated during nighttime and included in year-round energy estimation.
In commercial and institutional buildings using glass, concrete, or face brick on the outside surface
of the building envelope, the migration of moisture through the glass pane is rather small.
Because of the heavy mass of the concrete wall, the influence of the diurnal cyclic variation of the
relative humidity of outdoor air on moisture transfer through the building envelope is also small.
Therefore, for simplicity, the moisture transfer between the building envelope and the outside air
can be ignored.
The heat balance at the outer building surface, as shown in Fig. 3.17, can be expressed as
follows:
(3.91)
In Eq. (3.91), Qsol represents the solar radiation absorbed by the outer surface of the building envelope,
in Btu/ h (W). It can be calculated as
(3.92)
where os  absorptance of the outer surface of the building envelope. From Eq. (3.61), the reflection
of solar radiation from any reflecting surface to the outer surface of the building and absorbed
by it, or qref , Btu /h (W), is given by
(3.93)
where A  total area of the outer surface of the building envelope, . The
term qos indicates the convective heat transfer from the outer surface of the building outward, in
A  As 	 Ash, ft2 (m2)
Qref  A os Iref  A s os Fsr (ID 	 Id )
Qsol 
os [As (ID 	 Id ) 	 Ash Id]
Qsol 	 Qref  Qos 	 (Qrad  Qat) 	 Qoi
3.46 CHAPTER THREE
FIGURE 3.17 Heat balance at the outer surface of a building.
Btu/h (W). From Eq. (3.7), it can be calculated as
(3.94)
where Tos , To  outer surface temperature of the building and the outdoor air temperature, respectively,
°F (°C)
hc  convective heat-transfer coefficient of outer surface, Btu /h ft2 °F (W/m2 °C)
The term Qoi denotes the inward heat flow from the outer building surface, in Btu/h (W). The term
Qrad  Qat indicates the net heat emitted from the outer surface of the building because of the radiation
exchange between the surface and the atmosphere, in Btu/h (W). Here, Qrad represents
the longwave radiation emitted from the surface, and Qat indicates the atmospheric radiation to the
surface.
Kimura (1977) found that atmospheric radiation can be expressed as
(3.95)
where Ccc  cloud cover factor, which can be obtained from local climate records, dimensionless
Kcc  cloudy reduction factor
Usually, the smaller the value of Ccc, the higher the clouds and the smaller the Kcc value. For simplicity,
it can be calculated as
(3.96)
And Br is actually the emissivity of the atmosphere, and it can be expressed by an empirical formula
developed by Brunt:
(3.97)
where pw  water vapor pressure, psia. In Eq. (3.95), TRg represents the absolute ground temperature,
°R (K), and TRo the absolute outdoor temperature, °R (K). Then, radiant heat loss can be written as
(3.98)
where os  emissivity of outer surface of building
Fsat  shape factor between outer surface and atmosphere
TRos  absolute outer surface temperature of building, °R (K)
Sol-Air Temperature
For a sunlit outer surface of a building, if Qref is mainly from ground-reflected solar radiation, it is
offset by and Eq. (3.91) becomes
Let . Substituting for Qoi, then, gives
and
(3.99)
In Eq. (3.99), Tsol is called sol-air temperature, in °F (°C). It is a fictitious outdoor temperature that
combines the effect of the solar radiation radiated on the outer surface of the building and the inward
heat transfer due to the outdoor–indoor temperature difference.
Tsol  To 	
os It
ho
os It  ho(Tos  To) 	 ho(Tsol  Tos)
Qoi  ho A(Tsol  Tos)
Qsol  Qos 	 Qoi
Qrad  Qat,
Qrad  Qat 
os AFsat {T Ros
4  [(1  CccKcc )T Ro
4
Br 	 Ccc Kcc T Rg
4 ]}
Br  0.51 	 0.55?pw
Kcc  0.83  0.4Ccc
Qat  (1  CccKcc )  T Ro
4 Br 	 CccKcc
 T Rg
4
Qos  hc A(Tos  To)
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.47
Example 3.2. At midnight of July 21, the outdoor conditions of an office building are as follows:
Outdoor temperature 75°F (23.9°C)
Water vapor pressure of outdoor air 0.215 psia
Ground temperature 72°F (22.2°C)
Cloud cover factor Ccc 0.1
Emissivity of the outer surface 0.90
Shape factor between the surface and sky 0.5
If the outer surface temperature of this building is 76°F (24.4°C) and the Stefan-Boltzmann constant

  0.1714  108 Btu/h ft2 °R4 (5.67  108 W/m2 K4), find the radiant heat loss from
each square foot (square meter) of the vertical outer surface of this building.
Solution. From Eq. (3.97) and given data, we can see
And from Eq. (3.96), the cloudy reduction factor is
From the given, TRos  76 	 460  536°R, TRo  75 	 460  535°R, and TRg  72 	 460 
532°R; then, from Eq. (3.98), the radiant heat loss from the outer surface of the building due to
radiant exchange between the surface and the atmosphere is
3.13 COMPLIANCE WITH ASHRAE/IESNA STANDARD 90.1-1999
FOR BUILDING ENVELOPE
An energy-efficient and cost-effective building envelope should meet the requirements and design
criteria in ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise
Residential Buildings; the DOE Code of Federal Regulations, Title10, Part 435, Subpart A, Performance
Standard for New Commercial and Multi-Family High-Rise Residential Buildings; and local
energy codes. The DOE Code is very similar to Standard 90.1.
The design and selection of the building envelope are generally the responsibility of an architect
with the assistance of a mechanical engineer or contractor. Building envelopes are usually designed,
or even constructed, before the HVAC&R system is designed. A speculative building is built of
known use and type of occupancy, but the exact tenants are unknown. A shell building is built before
the use and occupancy are determined.
Compliance with the ASHRAE/IES Standard 90.1-1999 for building envelopes includes the following.
Refer to Standard 90.1-1999 for exceptions and details.
General Requirements
The requirements apply to the exterior building envelope which separates conditioned space from
the outdoors and the semiexterior building envelope which separates the conditioned space from
semiheated space or unconditioned space, or separated semiheated space from the unconditioned
space or outdoors.
Semiheated space is an enclosed space, but not a conditioned space, within a building that is
heated by a heating system whose output capacity is greater than or equal to 3.4 Btu/h ft (10.71
	 0.1  0.79  (5.32)4]}  14.27 Btu / h ft2 (45.02 W/ m2)
 0.90  0.5  0.174{(5.36)4  [(1  0.1  0.79)(5.35)4  0.765
Qrad  Qat 
os Fsat

{T Ros
4  [(1  CccKcc )T Ro
4 Br 	 CccKcc T Rg
4 ]}
Kcc  0.83  0.4Ccc  0.83  0.4(0.1)  0.79
Br  0.51 	 0.55?pw  0.51 	 0.55?0.215  0.765
3.48 CHAPTER THREE
W/m2). Compliance includes mandatory provisions, the prescriptive building envelope option, and
the building envelope trade-off option.
Mandatory Provisions
The standard mandates that insulation materials shall be installed in accordance with the manufacturer’s
recommendation to achieve rated R value of insulation. Rated R value of insulation is the
thermal resistance of the insulation alone in units of h  ft2 °F/Btu (m2 °C/W) at a mean temperature
of 75°F (23.9°C). Rated R value refers to the thermal resistance of the added insulation in
framing cavities or insulated sheathing only and does not include the thermal resistance of other
building materials or air films.
Insulation shall be installed in a permanent manner in substantial contact with the inside surface.
The roof insulation shall not be installed on a suspended ceiling with removable ceiling panels. Insulation
outdoors shall be covered with a protective material to prevent damage from sunlight,
moisture, landscaping operations, equipment maintenance, and wind. Insulation materials contacts
with the ground shall have a water absoption rate no greater than 0.3 percent.
Fenestration performance shall be determined from production line units or representative units purchased.
U factors shall be determined in accordance with National Fenestration Rating Council (NFRC)
100. Solar heat gain coefficient (SHGC) for the overall fenestration area shall be determined in accordance
with NFRC 200. Visible light transmittance shall be determined in accordance with NFRC 300.
The following areas of building envelope shall be sealed, caulked, gasketed, or weather-stripped
until air leakage is minimal:
 Joints around fenestration and door frames
 Junctions between walls and foundations, between wall corners, between walls and roofs, wall
and floors, or walls and panels
 Openings because of penetrations of utility services through roofs, walls, and floors
 Fenestrations and doors built at site
 Building assemblies used as ducts and plenums
 Joints, seams, and penetrations due to vapor retarders
 Other openings in the building envelope
Air leakage of fenestration and doors shall be determined in accordance with NFRC 400. Air
leakage shall not exceed 1.0 cfm/ft2 (5.0 L/s m2) for glazed swinging entrance doors and revolving
doors and also 0.4 cfm/ft2 (2.0 L/sm2) for all other products. A door that separates conditioned
space from the exterior shall be protected with an enclosed vestibule with all doors into and out of
the vestibule installed with self-closing devices.
All mandatory provisions in Standard 90.1-1999 will be presented in the form “standard mandates
that”, whereas for nonmandatory provisions, only “standard specifics that” will be used.
Prescriptive Building Envelope Option
The exterior building envelope shall comply with the requirements for the appropriate conditioned
space in Table 5.3 for various climate (located in Normative Appendix B of Standard 90.1-1999). In
Table B1 to B26 of I-P edition:
 There are 26 tables, each of them has a number of heating degree days of 65°F (HDD65), and a
number of cooling degree days of 50°F (CDD 50). Select a table with number of HDD 65 and
CDD 50 equal to or most nearly equal to the values where the proposed building locates.
 There are three kinds of conditioned space: nonresidential, residential, semiheated.
 Total vertical fenestration area, including both fixed and operable fenestration shall be less than
50 percent of the gross wall area. The total skylight area, including glass and plastic skylights
with or without a curb, shall be less than 5 percent of the roof area.
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.49
Fenestration, including fixed and operable vertical fenestration, shall have a U factor not greater
than that specified in Table 5.3, 90.1-1999. Vertical fenestration shall have an SHGC not greater
than that specified for all orientations in Table 5.3, 90.1-1999. There are only visible light transmittance
criteria in the Building Envelope Trade-Off Option.
 All roofs shall have a rated R value of insulation not less than that specified in Table 5.3,90.1-
1999. Skylight curbs shall be insulated to the level of the roofs with the insulation entirely above
the deck or R-5, whichever is less.
 All above-grade walls shall have a rated value of insulation not less than that specified in
Table 5.3, 90.1-1999. Mass wall heating capacity shall be determined from Table A-6 or A-7 in
Standard 90.1-1999. Below-grade walls shall have a rated R value of insulation not less than that
specified in Table 5.3, 90.1-1999.
 All floors and heated or unheated slab-on-grade floors shall have a rated R value of insulation not
less than that specified in Table 5.3, 90.1-1999. Slab-on-grade floor insulation shall be installed
around the perimeter of the slab-on-grade floor to the distance specified.
 All opaque doors shall have a U factor not greater than that specified in Table 5.3, 90.1-1999.
Building Envelope Trade-Off Option
The building envelope complies with the standard if the proposed building satisfied the provisions
of 5.1 and 5.2 of Standard 90.1-1999 and the envelope performance factor of the proposed building
is less than or equal to the envelope performance factor of the budget building. The envelope performance
factor considers only the building envelope components. Schedules of operation, lighting
power input, equipment power input, occupant density, and mechanical systems shall be the same
for both the proposed building and budget building. Envelope performance factors shall be calculated
using the procedures of Normative Appendix C, Standard 90.1-1999. Refer to Standard 90.1-
1999 for details.
3.14 ENERGY-EFFICIENT AND COST-EFFECTIVE MEASURES FOR
BUILDING ENVELOPE
According to the economic parametric analysis of the thermal design in Johnson et al. (1989), the
following are energy-efficient and cost-effective measures for the design of building envelopes for
office buildings.
Exterior Walls
An increase in the mass of the exterior wall, i.e., its thermal capacitance, reduces only the peak
heating and cooling loads. The increase in electrical usage and capital investment often offsets the
benefit of the decrease of peak loads. The variation of life-cycle costing is often negligible.
Increasing the insulation to more than 2-in. (50-mm) thick decreases only the natural gas usage
for heating. This is cost-effective for areas with very cold winters. In some cases, the increase in the
annual cost due to the increase in the capital investments of the insulation may balance the reduction
of gas usage. A careful analysis is required.
Windows
The vertical fenestration area ratio (VFR) is the single parameter that most influences the building
life-cycle costing among building envelopes for high-rise buildings. For many office buildings, VFR
lies between 0.2 and 0.3.
3.50 CHAPTER THREE
Heat-absorbing and -reflective glasses produce a significant building cost savings in areas where
solar heat control is important in summer. Double-panes, triple-panes, and low-emission films are
effective in reducing the U value of the window assembly and therefore, the heating and cooling
loads. Indoor shading devices with window management systems are cost-effective. For example,
an indoor shading device can be turned on when solar heat gain exceeds 20 Btu/h ft2 (63 W/m2).
Although overhangs reduce the cooling loads, they may increase the need for electric lighting for
daylit buildings. The effect of overhang usage on building life-cycle costing is not significant in
many instances.
Infiltration
Infiltration has a significant influence on heating and cooling loads. Indoor air quality must be guaranteed
by mechanical ventilation systems and sufficient outdoor air intake through these systems. It
is desirable that windows and cracks in joints be well sealed. In multiple-story buildings, infiltration
through elevator shafts, pipe shafts, and duct shafts should be reduced.
Energy-Efficient Measures for Commercial Buildings in the United States
According to the EIA’s Commercial Buildings Characteristics, in 1992, the breakdown of
energy conservation features for commercial buildings for a total area of 67,876 million ft2 (6308
million m2) in the United States is as follows:
Roof or ceiling insulation 74%
Wall insulation 49%
Storm or multiple glazing 44%
Tinted, reflective, or shading glass 37%
Exterior or interior shading devices 50%
Windows that are openable 43%
REFERENCES
de Abreu, P. F., Fraser, R. A., Sullivan, H. F., and Wright, J. L., A Study of Insulated Glazing Unit Surface
Temperature Profiles Using Two-Dimensional Computer Simulation, ASHRAE Transactions, 1996, Part II,
pp. 497–507.
Altmayer, E. F., Gadgil, A. J., Bauman, F. S., and Kammerud, R. C., Correlations for Convective Heat Transfer
from Room Surfaces, ASHRAE Transactions, 1983 Part II A, pp. 61–77.
ASHRAE, ASHRAE Handbook 1989, Fundamentals, Atlanta, GA, 1989.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, ASHRAE Inc., Atlanta, GA, 1997.
ASHRAE, ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, Atlanta, GA, 1999.
ASHRAE, Procedure for Determining Heating and Cooling Loads for Computerizing Energy Calculations, Algorithms
for Building Heat Transfer Subroutines, Atlanta, GA, 1976.
Bauman, F., Gadgil, A., Kammerud, R., Altmayer, E., and Nansteel, M., Convective Heat Transfer in Buildings:
Recent Research Results, ASHRAE Transactions, 1983, Part I A, pp. 215–233.
Chandra, S., and Kerestecioglu, A. A., Heat Transfer in Naturally Ventilated Rooms: Data from Full Scale
Measurements, ASHRAE Transactions, 1984, Part I B, pp. 211–225.
Dahlen, R. R., Low-E Films for Window Energy Control, ASHRAE Transactions, 1987, Part I, pp. 1517–1524.
Deringer, J. J., An Overview of Standard 90.1: Building Envelope, ASHRAE Journal, no. 2, 1990, pp. 30–34.
HEAT AND MOISTURE TRANSFER THROUGH BUILDING ENVELOPE 3.51
Donnelly, R. G., Tennery, V. J., McElroy, D. L., Godfrey, T. G., and Kolb, J. O., Industrial Thermal Insulation,
An Assessment, Oak Ridge National Laboratory Report TM-5283, TM-5515, and TID-27120, 1976.
Energy Information Administration, Commercial Buildings Characteristics 1992, Commercial Buildings Energy
Consumption Survey April 1994,Washington, 1994.
Elmahdy, H., Surface Temperature Measurement of Insulating Glass Units Using Infrared Thermography,
ASHRAE Transactions, 1996, Part II, pp. 489–496.
Fairey, P. W., and Kerestecioglu, A. A., Dynamic Modeling of Combined Thermal and Moisture Transport in
Buildings: Effect on Cooling Loads and Space Conditions, ASHRAE Transactions, 1985 Part II A,
pp. 461–472.
Galanis, N., and Chatigny, R., A Critical Review of the ASHRAE Solar Radiation Model, ASHRAE Transactions,
1986, Part I A, pp. 410–419.
Glicksman, L. R., and Katsennelenbogen, S., A Study of Water Vapor Transmission Through Insulation under
Steady State and Transient Conditions, ASHRAE Transactions, 1983, Part II A, pp. 483–499.
Gueymard, C. A., A Simple Model of the Atmospheric Radiative Transfer of Sunshine: Algorithms and Performance
Assessment, Report FSEC-PF-270-95, Florida Solar Energy Center, Cocoa, Fla., 1995.
Hagentoft, C-E., Moisture Conditions in a North-Facing Wall with Cellulose Loose-Fill Insulation: Construction
with and without a Vapor Retarder and Air Leakage, ASHRAE Transactions, 1995, Part I, pp. 639–646.
Handegrod, G. O. P., Prediction of the Moisture Performance of Walls, ASHRAE Transactions, 1985, Part II B,
pp. 1501–1509.
Inoue, T., Kawase, T., Ibamoto, T., Takakusa, S., and Matsuo,Y., The Development of an Optimal Control System
for Window Shading Devices Based on Investigations in Office Buildings, ASHRAE Transactions, 1988,
Part II, pp. 1034–1049.
Johnson, C. A., Besent, R. W., and Schoenau, G. J., An Economic Parametric Analysis of the Thermal Design of a
Large Office Building under Different Climatic Zones and Different Billing Schedules, ASHRAE Transactions,
1989, Part I, pp. 355–369.
Kays, M. M., and Crawford, M. E., Convective Heat and Mass Transfer, 2d ed., McGraw-Hill, New York, 1980.
Kimura, K., Scientific Basis of Air Conditioning, Applied Science Publishers, London, 1977.
Kimura, K., and Stephenson, D. G., Solar Radiation on Cloudy Days, ASHRAE Transaction, 75(1), 1969, pp.
1–8.
Miller, A., Thompson, J. C., Peterson, R. E., and Haragan, D. R., Elements of Meteorology, 4th ed., Bell and
Howell Co., Columbus, Ohio, 1983.
Robertson, D. K., and Christian, J. E., Comparison of Four Computer Models with Experimental Data from
Test Buildings in Northern New Mexico, ASHRAE Transactions, 1985, Part II B, pp. 591–607.
Sato, A., Eto, N., Kimura, K., and Oka, J., Research on the Wind Variation in the Urban Area and Its Effects in
Environmental Engineering No. 7 and No. 8—Study on Convective Heat Transfer on Exterior Surface of
Buildings, Transactions of Architectural Institute of Japan, no. 191, January 1972.
Smolenski, C. P., Absorption in Thermal Insulation: How Much Is Too Much? HPAC no. 11, 1996, pp. 49–58.
Spitler, J. D., Pedersen, C. O., and Fisher, D. E., Interior Convective Heat Transfer in Buildings with Large Ventilative
Flow Rates, ASHRAE Transactions, 1991, Part I, pp. 505–514.
Stewart,W. E., Effect of Air Pressure Differential on Vapor Flow through Sample Building Walls, ASHRAE
Transactions, 1998, Part II, pp. 17–24.
Verschoor, J. D., Measurement of Water Vapor Migration and Storage in Composite Building Construction,
ASHRAE Transactions, 1985, Part II A, pp. 390–403.
Wang, S. K., Air Conditioning, vol. 1, Hong Kong Polytechnic, Hong Kong, 1987.
Wong, S. P. W., Simulation of Simultaneous Heat and Moisture Transfer by Using the Finite Difference Method
and Verified Tests in a Test Chamber, ASHRAE Transactions, 1990, Part I, pp. 472–485.
Wong, S. P .W., and Wang, S. K., Fundamentals of Simultaneous Heat and Moisture Transfer between the
Building Envelope and the Conditioned Space Air, ASHRAE Transactions, 1990, Part II, pp. 73–83.
3.52 CHAPTER THREE
4.1
CHAPTER 4
INDOOR AND OUTDOOR
DESIGN CONDITIONS
4.1 INDOOR DESIGN CONDITIONS 4.1
4.2 HEAT EXCHANGE BETWEEN
HUMAN BODY AND INDOOR
ENVIRONMENT 4.2
Two-Node Model of Thermal
Interaction 4.2
Steady-State Thermal Equilibrium 4.3
Transient Energy Balance 4.3
4.3 METABOLIC RATE AND SENSIBLE
HEAT LOSSES FROM HUMAN BODY 4.4
Metabolic Rate 4.4
Mechanical Work 4.4
Sensible Heat Exchange 4.5
Clothing Insulation 4.7
4.4 EVAPORATIVE HEAT LOSSES 4.7
Respiration Losses 4.7
Evaporative Heat Loss from Skin
Surface 4.7
Maximum Evaporative Heat Loss due to
Regulatory Sweating 4.7
Diffusion Evaporative Heat Loss and
Total Skin Wetness 4.8
4.5 MEAN RADIANT TEMPERATURE
AND EFFECTIVE TEMPERATURE 4.9
Mean Radiant Temperature 4.9
Effective Temperature 4.14
4.6 FACTORS AFFECTING THERMAL
COMFORT 4.14
4.7 THERMAL COMFORT 4.15
Fanger’s Comfort Equation 4.15
ASHRAE Comfort Zones 4.17
Comfort-Discomfort Diagrams 4.17
4.8 INDOOR AIR TEMPERATURE AND AIR
MOVEMENTS 4.20
Comfort Air Conditioning Systems 4.20
Design Considerations 4.21
Indoor Design Temperatures for Comfort
Air Conditioning 4.21
Process Air Conditioning Systems 4.23
4.9 HUMIDITY 4.23
Comfort Air Conditioning Systems 4.23
Process Air Conditioning Systems 4.24
4.10 SICK BUILDING SYNDROME AND
INDOOR AIR QUALITY 4.27
Indoor Air Contaminants 4.27
Basic Strategies to Improve Indoor Air
Quality 4.29
Outdoor Air Requirements for
Occupants 4.30
4.11 AIR CLEANLINESS 4.31
4.12 SOUND LEVEL 4.32
Sound and Sound Level 4.32
Sound Power Level and Sound Pressure
Level 4.32
Octave Bands 4.33
Addition of Sound Levels 4.33
Human Response and Design
Criteria 4.34
4.13 SPACE PRESSURE
DIFFERENTIAL 4.37
4.14 OUTDOOR DESIGN
CONDITIONS 4.38
Summer and Winter Outdoor Design
Conditions 4.39
The Use of Outdoor Weather Data in
Design 4.39
Outdoor Weather Characteristics and
Their Influence 4.42
REFERENCES 4.42
4.1 INDOOR DESIGN CONDITIONS
Indoor design parameters are those that the air conditioning system influences directly to produce a
required conditioned indoor environment in buildings. They are shown below and grouped as follows:
1. Basic design parameters
 Indoor air temperature and air movements
 Indoor relative humidity
2. Indoor air quality
 Air contaminants
 Outdoor ventilation rate provided
 Air cleanliness for processing
3. Specific design parameters
 Sound level
 Pressure differential between the space and surroundings
The indoor design parameters to be maintained in an air conditioned space are specified in the
design document and become the targets to be achieved during operation. In specifying the indoor
design parameters, the following points need to be considered:
1. Not all the parameters already mentioned need to be specified in every design project. Except
for the indoor air temperature which is always an indoor design parameter in comfort air conditioning,
it is necessary to specify only the parameters which are essential to the particular situation
concerned.
2. Even for process air conditioning systems, the thermal comfort of the workers should also be
considered. Therefore, the indoor design parameters regarding health and thermal comfort for the
occupants form the basis of design criteria.
3. When one is specifying indoor design parameters, economic strategies of initial investment
and energy consumption of the HVAC&R systems must be carefully investigated. Design criteria
should not be set too high or too low. If the design criteria are too high, the result will be an excessively
high investment and energy cost. Design criteria that are too low may produce a poor indoor
air quality, resulting in complaints from the occupants, causing low-quality products, and possibly
leading to expensive system alternations.
4. Each specified indoor design parameter is usually associated with a tolerance indicated
as a  sign, or as an upper or lower limit. Sometimes there is a traditional tolerance understood
by both the designers and the owners of the building. For instance, although the summer
indoor design temperature of a comfort air conditioning system is specified at 75 or 78°F
(23.9 or 25.6°C), in practice a tolerance of 2–3°F (1.1 – 1.7°C) is often considered
acceptable.
5. In process air conditioning systems, sometimes a stable indoor environment is more important
than the absolute value of the indoor parameter to be maintained. For example, it may not be necessary
to maintain 68°F (20°C) for all areas in precision machinery manufacturing. More often, a
72°F (22.2°C) or even a still higher indoor temperature with appropriate tolerance will be more
suitable and economical.
4.2 HEAT EXCHANGE BETWEEN HUMAN BODY AND
INDOOR ENVIRONMENT
Two-Node Model of Thermal Interaction
In 1971, Gagge et al. recommended a two-node model of human thermal interaction. In this model,
the human body is composed of two compartments: an inner body core, including skeleton, muscle
and internal organs; and an outer shell of skin surface. The temperatures of the body core and the
surface skin are each assumed to be uniform and independent. Metabolic heat production, external
mechanical work, and respiratory losses occur only in the body core. Heat exchange between the
body core and the skin surface depends on heat conduction from direct contact and the peripheral
blood flow of the thermoregulatory mechanism of the human body.
4.2 CHAPTER FOUR
Steady-State Thermal Equilibrium
When the human body is maintained at a steady-state thermal equilibrium, i.e., the heat storage
at the body core and skin surface is approximately equal to zero, then the heat exchange between
the human body and the indoor environment can be expressed by the following heat balance
equation:
M  W  C  R  Esk  Eres (4.1)
where M  metabolic rate, Btu /h ft2 (W/m2)
W  mechanical work performed, Btu/h ft2 (W/m2)
C  R  convective and radiative, or sensible heat loss from skin surface,
Btu/h ft2 (W/m2)
Esk  evaporative heat loss from skin surface, Btu /h ft2 (W/m2)
Eres  evaporative heat loss from respiration, Btu/h ft2 (W/m2)
In Eq. (4.1), the ft2 in the unit Btu /h ft2 applies to the skin surface area. The skin surface area of a
naked human body can be approximated by an empirical formula proposed by Dubois in 1916
AD  0.657mb
0.425 Hb
0.725 (4.2)
where AD  Dubois surface area of naked body, ft2 (m2)
mb  mass of human body, lb (kg)
Hb  height of human body, ft (m)
In an air conditioned space, a steady-state thermal equilibrium is usually maintained between the
human body and the indoor environment.
Transient Energy Balance
When there is a transient energy balance between the human body and the indoor environment, the
thermal interaction of the body core, skin surface, and indoor environment forms a rate of positive
or negative heat storage both in the body core and on the skin surface.
The human body needs energy for physical and mental activity. This energy comes from the oxidation
of the food taken into the human body. The heat released from this oxidation process is
called metabolic heat. It dissipates from the skin surface of the human body into the surroundings.
In a cold environment, the thermoregulatory mechanism reduces the rate of peripheral blood circulation,
lowering the temperature of the skin and preventing any greater heat loss from the human
body. However, if the heat loss and the mechanical work performed are greater than the rate of
metabolic heat produced, then the temperatures of both the body core and the skin surface fall, and
shivering or other spontaneous activities occur to increase the production of heat energy within the
human body.
On the other hand, in a hot environment, if a large amount of heat energy needs to be dissipated
from the human body, the physiological control mechanism increases the blood flow to the skin
surface. This raises the skin temperature. If the heat produced is still greater than the heat actually
dissipated and the temperature of the body core increased from its normal temperature of about
97.6 to about 98.6°F (36.4 to about 37.0°C), then liquid water is released from the sweat glands for
evaporative cooling.
For a transient state of energy balance between the human body and the indoor environment, the
rate of heat storage in the body core Scr and the skin surface Ssk, both in Btu/h ft2 (W/m2), can be
calculated as
Scr  Ssk  M  W  (C  R)  Esk  Eres (4.3)
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.3
4.3 METABOLIC RATE AND SENSIBLE HEAT LOSSES FROM
HUMAN BODY
Metabolic Rate
The metabolic rate M is the rate of energy release per unit area of skin surface as a result of the
oxidative processes in the living cells. Metabolic rate depends mainly on the intensity of the physical
activities performed by the human body. The unit of metabolic rate is called the met. One met is
defined as 18.46 Btu/h ft2 (58.24 W/m2) of metabolic heat produced in the body core. In Table 4.1
are listed the metabolic rates of various activities.
Mechanical Work
Some of the energy released from the oxidative processes within the body core can be partly transformed
to external mechanical work through the action of the muscles. Mechanical work W is
usually expressed as a fraction of the metabolic rate and can be calculated as
W  	M (4.4)
where 	 mechanical efficiency. For most office work, mechanical efficiency is 	 less than 0.05.
Only when there is a large amount of physical activity such as bicycling, lifting and carrying, or
walking on a slope may increase 	 to a value of 0.2 to 0.24.
4.4 CHAPTER FOUR
TABLE 4.1 Metabolic Rate for Various Activities
Metabolic rate
Activity level Met Btu/h ft2
Resting
Sleeping 0.7 13
Seated, quiet 1.0 18
Office work
Reading, seated 1.0 18
Typing 1.1 20
Teaching 1.6 30
Domestic work
Cooking 1.6–2.0 29–37
House cleaning 2.0–3.4 37–63
Walking
Speed 2 mph 2.0 37
4 mph 3.8 70
Machine work
Light 2.0–2.4 37–44
Heavy 4.0 74
Dancing, social 2.4–4.4 44–81
Sports
Tennis, singles 3.6–4.0 66–74
Basketball 5.0–7.6 90–140
Wrestling 7.0–8.7 130–160
Source: Adapted with permission from ASHRAE
Handbook 1989, Fundamentals.
Sensible Heat Exchange
Sensible heat loss or, occasionally, sensible heat gain R  C represents the heat exchange between
the human body and the indoor environment through convective and radiative heat transfer.
Figure 4.1 shows the sensible heat exchange between the human body and the environment. The
combined convective and radiative heat transfer can be calculated as
(4.5)
where Tcl  mean surface temperature of clothing, °F (°C). The operative temperature To is defined
as the weighted average of the mean radiant temperature Trad and indoor air temperature Ta, both in
°F (°C), that is,
(4.6)
The surface heat-transfer coefficient is defined as
h  hc  hr (4.7)
and the ratio of the clothed surface area to the naked surface area is
(4.8)
where hc, hr  convective and radiative heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
Acl  surface area of clothed body, ft2 (m2)
The mean radiant temperature Trad is discussed in greater detail in later sections.
According to Seppenan et al. (1972), the convective heat-transfer coefficient hc for a person
standing in moving air, when the air velocity is 30 v 300 fpm (0.15 v 1.5 m/ s), is
hc  0.0681v0.69 (4.9)
fcl 
Acl
AD
To 
hrTrad  hcTa
hr  hc
 fclh(Tcl  To)
C  R  fclhc (Tcl  Ta)  fclhr(Tcl  Trad)
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.5
FIGURE 4.1 Sensible heat exchange between the human body and the
indoor environment.
When the air velocity v	30 fpm (0.15 m/ s), hc  0.7 Btu/h ft2 °F (4 W/m2 °C). For typical
indoor temperatures and a clothing emissivity nearly equal to unity, the linearized radiative heattransfer
coefficient hr  0.83 Btu/h ft2 °F (4.7 W/m2 °C). Let Rcl be the R value of clothing, in
h ft2 °F/Btu (m2 °C/W). Then
(4.10)
where Tsk  mean skin surface temperature, °F (°C). If the human body is able to maintain a thermal
equilibrium with very little evaporative loss from the skin surface, then the skin temperature
Tsk.n will be around 93°F (33.9°C).
Combining Eqs. (4.5) and (4.10), and eliminating Tcl, we find
(4.11)
In this equation, the dimensionless clothing efficiency Fcl is defined as
(4.12)
In Eq. (4.11), if To Tsk, then C  R could be negative, i.e., there could be a sensible heat gain.
Fcl 
1
Rcl fcl h  1

Tcl  To
Tsk  To
 Fcl fcl h(Tsk  To)
C  R 
fcl h(Tsk  To)
Rcl fcl h  1
C  R 
Tsk  Tcl
Rcl
4.6 CHAPTER FOUR
TABLE 4.2 Insulation Values for Clothing Ensembles*
Rcl,
h ft2 °F
Ensemble description† Rcl, clo Btu fcl
Walking shorts, short-sleeve shirt 0.41 0.36 1.11
Fitted trousers, short-sleeve shirt 0.50 0.44 1.14
Fitted trousers, long-sleeve shirt 0.62 0.55 1.19
Same as above, plus suit jacket 0.96 0.85 1.23
Loose trousers, long-sleeve shirt,
long-sleeve sweater, T-shirt 1.01 0.89 1.28
Sweat pants, sweatshirt 0.77 0.68 1.19
Knee-length skirt, short-sleeve shirt,
pantyhose (no socks), sandals 0.54 0.48 1.26
Knee-length skirt, long-sleeve shirt,
full slip, pantyhose (no socks) 0.67 0.59 1.29
Long-sleeve coveralls, T-shirt 0.72 0.63 1.23
Overalls, long-sleeve shirts, long
underwear tops and bottoms,
flannel long-sleeve shirt 1.00 0.88 1.28
*For mean radiant temperature equal to an air temperature and air velocity less
than 40 fpm.
†Unless otherwise noted, all ensembles included briefs or panties, shoes,
and socks.
Source: Adapted from McCullough and Jones (1984). Reprinted with permission.
Clothing Insulation
Clothing insulation Rcl can be determined through measurements on a heated manikin, a model of
the human body for laboratory experiments. After C  R is measured from the thermal manikin in
a controlled indoor environment, Rcl can be calculated from Eq. (4.11) since fcl, Tsk, To, and h are
also known values. Clothing insulation Rcl can be expressed either in h  ft2 °F/Btu (m2 °C/W) or
in a new unit called clo, where 1 clo  0.88 h ft2 °F/Btu (0.16 m2 °C/W).
Clothing insulation Rcl values and area ratios fcl for typical clothing ensembles, taken from
McCullough and Jones (1984), are listed in Table 4.2.
4.4 EVAPORATIVE HEAT LOSSES
Evaporative heat loss E is heat loss due to the evaporation of sweat from the skin surface Esk and
respiration losses Eres. Actually, metabolic heat is mainly dissipated to the indoor air through the
evaporation of sweat when the indoor air temperature is nearly equal to the skin temperature.
Respiration Losses
During respiration, there is convective heat loss Cres that results from the temperature of the
inhaled air being increased to the exhaled air temperature, or about 93°F (33.9°C). There is also a
latent heat loss Lres due to evaporation of liquid water into water vapor inside the body core.
The amount of respiration loss depends mainly on the metabolic rate. In summer, at an indoor
temperature of 75°F (23.9°C) and a relative humidity of 50 percent, respiratory losses Eres  Cres
 Lres are approximately equal to 0.09M. In winter, Eres is slightly greater. For simplicity, let
Eres  0.1M.
Evaporative Heat Loss from Skin Surface
Evaporative heat loss from the skin surface Esk consists of (1) the evaporation of sweat as a result of
thermoregulatory mechanisms of the human body Ersw and (2) the direct diffusion of liquid water
from the skin surface Edif. Evaporative heat loss due to regulatory sweating Ersw, Btu/h ft2 (W/m2),
is directly proportional to the mass of the sweat produced, i.e.,
(4.13)
where mass flow rate of sweat produced, lb/h ft2 (kg/ s m2)
hfg  latent heat of vaporization at 93°F (33.9°C), Btu/ lb (J /kg)
Maximum Evaporative Heat Loss due to Regulatory Sweating
The wetted portion of the human body needed for the evaporation of a given quantity of sweat wrsw
is
(4.14)
In Eq. (4.14), Emax represents the maximum evaporative heat loss due to regulatory sweating when
the skin surface of the human body is entirely wet. Its magnitude is directly proportional to the
vapor pressure difference between the wetted skin surface and the indoor ambient air, and it can be
wrsw 
Ersw
Emax
m? rsw 
Ersw
m? rswhfg
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.7
calculated as
Emax  he,c(psk, s  pa) (4.15)
where psk, s  saturated water vapor pressure at skin surface temperature, psia (kPa)
pa  water vapor pressure of ambient air, psia (kPa )
he,c  overall evaporative heat-transfer coefficient of clothed body, in Btu/h ft2psi
(W/m2 kPa)
Woodcock (1962) proposed the following relationship between he, c and hs, the overall sensible heat
transfer coefficient, in Btu/h ft2 psi (W/m2 kPa):
(4.16)
The moisture permeability index im denotes the moisture permeability of the clothing and is dimensionless.
Clothing ensembles worn indoors usually have an im  0.3 to 0.5. The moisture permeability
indexes im of some clothing ensembles are presented in Table 4.3.
The Lewis relation LR in Eq. (4.16) relates the evaporative heat-transfer coefficient he and the
convective heat-transfer coefficient hc, both in Btu /h ft2 °F(W/m2°C). LR  f (he / hc) has a magnitude
of 205°F/psi (16.5°C/kPa). In Eq. (4.16), the overall sensible heat-transfer coefficient hs can
be calculated as
(4.17)
where Rt  total resistance to sensible heat transfer between the skin and the indoor environment,
h ft2 °F/Btu (m2°C/W).
Diffusion Evaporative Heat Loss and Total Skin Wetness
The minimum level of evaporative heat loss from the skin surface occurs when there is no regulatory
sweating and the skin wetness due to direct diffusion Edf, min, Btu/h ft2 (W/m2), is approximately
equal to 0.06Emax under normal conditions, or
Edf, min  0.06Emax (4.18)
When there is a heat loss from regulatory sweating Ersw, the diffusion evaporative heat loss
Edif, Btu/h ft2 (W/m2), for the portion of skin surface that is not covered with sweat can be
hs 
1
Rt

fcl h
fcl hRcl  1
im LR 
he,c
hs
4.8 CHAPTER FOUR
TABLE 4.3 Moisture Permeability of Clothing Ensembles
Ensemble description im
Cotton/ polyester long-sleeve shirt, long
trousers, street shoes, socks, briefs 0.385
Cotton short-sleeve shirt, long trousers, work
boots, socks, briefs, cotton gloves 0.41
Cotton/nylon long-sleeve shirt, cotton/nylon
trousers, combat boots, socks, helmet liner
(army battle dress uniform) 0.36
*Measured with Trad  Ta, and air velocity  40 fpm.
Source: Adapted with permission from ASHRAE Handbook
1989, Fundamentals.
calculated as
Edif  (1  wrsw)0.06Emax (4.19)
Therefore, the total evaporative heat loss from the skin surface is
Esk  Ersw  Edif
 wrswEmax  (1  wrsw)0.06Emax  (0.06  0.94wrsw)Emax
 wskEmax (4.20)
In Eq. (4.20), wsk is called the total skin wetness; it is dimensionless, and it can be calculated as
(4.21)
4.5 MEAN RADIANT TEMPERATURE AND EFFECTIVE
TEMPERATURE
Mean Radiant Temperature
Mean radiant temperature TRad is defined as the temperature of a uniform black enclosure in which
an occupant would have the same amount of radiative heat exchange as in an actual indoor environment.
Mean radiant temperature TRad, °R(K), can be calculated by the expression
T 4
Rad  T 4R1F0 – 1  T 4
R2 F0 –2  T 4
RnF0 – n (4.22)
where TR1, TR2,   , TRn  absolute temperature of surrounding surfaces of indoor
environment, °R(K)
F0 – 1  shape factor denoting fraction of total radiant energy leaving surface
of occupant’s clothing 0 and arriving on the surface 1
F0 – 2  fraction of total radiant energy leaving surface 0 and arriving on
surface 2, etc.
Shape factors F0 – n depend on the position and orientation of the occupant as well as the dimensions
of the enclosure. One can use Figs. 4.2 and 4.3 to estimate the mean value of the shape factor between
a seated person and rectangular surfaces. The sum of the shape factors of all the surfaces
with respect to the seated occupant in an enclosure is unity.
The temperature measured by a globe thermometer, called the globe temperature, is often used
to estimate the mean radiant temperature. The globe thermometer consists of a copper hollow
sphere of 6-in. (152-mm) diameter that is coated with black paint on the outer surface. A precision
thermometer or thermocouple is inserted inside the globe with the sensing bulb or the thermojunction
located at the center of the sphere. Because the net radiant heat received at the globe surface is
balanced by the convective heat transfer from the globe surface in reaching a thermal equilibrium,
according to Bedford and Warmer (1935), such a relationship gives
T 4
Rad  T 4
Rg  0.247  109 v0.5(TRg  TRa) (4.22a)
where TRad  absolute mean radiant temperature, °R (K)
TRg  absolute globe temperature, °R (K)
v  ambient air velocity, fpm (m/s)
TRa  absolute air temperature, °R (K)
After TRg, v, and TRa are measured, the mean radiant temperature TRad can be calculated from Eq.
(4.22). The mean radiant temperature indicates the effect, due to the radiant energy from the
wsk 
Esk
Emax
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.9
4.10 CHAPTER FOUR
FIGURE 4.2 Mean value of shape factor between a sedentary person and a horizontal plane. (Source: P. O. Fanger,
Thermal Comfort Analysis and Applications in Environmental Engineering, 1972. Reprinted with permission.)
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.11
FIGURE 4.3 Mean value of shape factor between a seated person and a vertical plane. (Source: P. O. Fanger,
Thermal Comfort Analysis and Applications in Environmental Engineering, 1970. Reprinted with permission.)
surroundings, on radiant exchange between an occupant or any substance and the enclosure. Such
an influence may be significant if the mean radiant temperature is several degrees higher than the
temperature of the indoor air.
Example 5.1. The dimensions of a private office and the location of a person seated within it are
shown in Fig. 4.4. The surface temperatures of the enclosure are as follows:
West window 88°F (31.1°C)
West wall 80°F (26.7°C)
North partition wall 75°F (23.9°C)
East partition wall 75°F (23.9°C)
South partition wall 75°F (23.9°C)
Floor 78°F (25.6°C)
Ceiling 77°F (25°C)
Calculate the mean radiant temperature of the enclosure that surrounds this office. The orientation
of the seated occupant is unknown.
Solution
1. Regarding the north partition wall, the shape factor denotes the fraction of the total radiant energy
that leaves the outer surface of the clothing of the occupant (surface 0) and arrives directly on
the north partition wall (surfaces 1, 2, 3, and 4) and is given by
F0 – 1,2,3,4  F0 – 1  F0 – 2  F0 – 3  F0 – 4
Here F0 – 1 is the shape factor for the radiation from surface 0 to surface 1 of the north partition wall.
Based on the curves in Fig. 4.3, for a ratio of b/L  1.8/3  0.6 and a ratio of a/L  4.5/3  1.5,
4.12 CHAPTER FOUR
FIGURE 4.4 Dimension, in feet, of a private office.
the shape factor F0 –1  0.04. Here L is the horizontal distance from the occupant to the north partition
wall. The shape factors F0 – 2, F0 – 3, and F0 – 4 can be calculated in the same manner.
2. To determine the shape factor F0 – 5 for the radiation from the occupant (surface 0) to the floor
(surface 5), we note that the vertical distance L from the center of the seated occupant to the floor is
1.8 ft. From the curves in Fig. 4.2, the ratio b/L  3/1.8  1.67, and the ratio a/L  4.5/1.8 
2.5; thus the shape factor F0 – 5  0.068. All the remaining shape factors can be determined in the
same manner as listed in Table 4.4.
3. For the north partition wall (surface 1),
T4
R1F0 – 1  (75  460)4  0.04  32.77  108
Other products T 4
Rn F0 – n can be similarly calculated, as listed in Table 4.4. The sum of the products
T4
Rn F0– n  832.56  108. Therefore
T 4
Rad  832.56  108
That is,
TRad  537.2°R or Trad  77.2°F(25.1°C)
4. The sum of the shape factors F0 – n  0.994 is nearly equal to 1.
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.13
TABLE 4.4 Values of F0 – n and T4
RnF0 – n in Example 4.1
Surface Shape
Surface temperature, °R factor b/L a/L F0– n T 4
RnF0– n  108
North partition wall 535 F0 – 1 0.6 1.5 0.04 32.77
F0 – 2 0.6 3.5 0.045 36.87
F0 – 3 2.2 1.5 0.07 57.35
F0 – 4 2.2 3.5 0.087 71.27
East partition wall 535 F0 – 17 0.63 0.29 0.014 11.47
F0 – 18 0.63 0.86 0.03 24.58
F0 – 19 0.17 0.29 0.004 3.28
F0 – 20 0.17 0.86 0.009 7.37
South partition wall 535 F0 – 9 0.73 0.5 0.023 18.84
F0 – 10 0.73 1.17 0.038 31.13
F0 – 11 0.2 0.5 0.008 6.55
F0 – 12 0.2 1.17 0.013 10.65
West wall 540 F0 – 13 0.4 0.67 0.018 15.31
540 F0 – 14 0.4 2 0.03 25.51
West window 548 F0 – 15 1.6 0.67 0.04 36.07
548 F0 – 16 1.6 2 0.07 63.13
Floor 538 F0 – 5 1.67 2.5 0.068 56.97
F0 – 6 1.67 5.8 0.073 61.16
F0 – 7 5 2.5 0.087 72.89
F0 – 8 5 5.8 0.102 85.45
Ceiling 537 F0 – 21 1.36 0.68 0.033 27.44
F0 – 22 1.36 1.59 0.052 43.24
F0 – 23 0.45 0.68 0.015 12.47
F0 – 24 0.45 1.59 0.025 20.79
 0.994 8.32.56  108
Effective Temperature
The effective temperature ET* is the temperature of an environment that causes the same total heat
loss from the skin surface as in an actual environment of an operative temperature equal to ET* and
at a relative humidity of 50 percent. And ET* can be calculated as
ET*  To  wsk im LR (0.5pET, s) (4.23)
where pET, s  saturated water vapor pressure at ET*, psia (kPa abs.). The right-hand side of
Eq. (4.23) describes the conditions of the indoor air regarding the total heat loss from the human
body. The same value of the combination To  wsk imLR(0.5pET, s ) results in the same amount of total
heat loss from the skin surface, if other parameters remain the same. Theoretically, total skin
wetness wsk and clothing permeability index im are constants for a specific ET* line.
Because the effective temperature is based on the operative temperature To, it is a combined index
of Ta, Trad, and pa. In an indoor air temperature below 77°F (25°C), the constant-ET* lines are
nearly parallel to the skin temperature lines for sedentary occupants with a clothing insulation of
0.6 clo; therefore, ET* values are reliable indexes to indicate thermal sensations at normal indoor
air temperature during low activity levels. The term effective temperature was originally proposed
by Houghton and Yaglou in 1923. A new definition of ET* and its mathematical expression were
developed by Gagge et al. in 1971. It is the environmental index commonly used in specifying and
assessing thermal comfort requirements.
4.6 FACTORS AFFECTING THERMAL COMFORT
Daily experience and many laboratory experiments all show that thermal comfort occurs only under
these conditions:
1. There is a steady-state thermal equilibrium between the human body and the environment; i.e.,
heat storage of the body core Scr and the skin surface Ssk are both equal to zero.
2. Regulatory sweating is maintained at a low level.
From the heat balance equation at steady-state thermal equilibrium Eq. (4.1) we have
M  W  C  R  Esk  Eres
Let Eres  0.1M and the mechanical efficiency 	 0.05M. From Eq. (4.11), C  R can be determined.
Also, from Eqs. (4.15), (4.16), and (4.20), Esk is a known value. If we substitute into
Eq. (4.1), the heat balance equation at steady-state thermal equilibrium can be expressed as
M(1  0.05  0.1)  Fcl fclh(Tsk  To)  wsk im LR hs (psk, s  pa)
or
0.85M  Fcl fcl h(Tsk  To)  wsk imLRhs ( psk, s  pa) (4.24)
In Eq. (4.24), the physiological and environmental factors that affect the balance—the metabolic
rate and the heat losses on the two sides of the equation—are as follows:
1. Metabolic rate M determines the magnitude of the heat energy that must be released from the
human body, i.e., the left-hand side of the equation.
2. Indoor air temperature Ta is a weighted component of the operating temperature To. It also
affects the sensible heat loss and the vapor pressure of indoor air pa in the calculation of the
evaporative loss from the skin surface.
4.14 CHAPTER FOUR
3. Mean radiant temperature Trad is another weighted component of the operating temperature To. It
affects the sensible heat loss from the human body.
4. Relative humidity of the ambient air a is the dominating factor that determines the difference
psk, s  pa in the evaporative loss from the skin surface. Air relative humidity becomes important
when the evaporative heat loss due to regulatory sweating is the dominating heat loss from the
human body.
5. Air velocity va influences the heat transfer coefficient h and the clothing efficiency Fcl in the
term in Eq. (4.24) for the sensible heat loss from the human body. It also affects the overall sensible
heat transfer coefficient hs in the evaporative heat loss term and the clothing permeability
im term in Eq. (4.24).
6. Clothing insulation Rcl affects the clothing efficiency Fcl, the area ratio fcl, the heat transfer coef-
ficient h, the clothing permeability index im, and the overall sensible heat-transfer coefficient hs.
4.7 THERMAL COMFORT
Thermal comfort is defined as the state of mind in which one acknowledges satisfaction with regard
to the thermal environment. In terms of sensations, thermal comfort is described as a thermal sensation
of being neither too warm nor too cold, defined by the following seven-point thermal sensation
scale proposed by ASHRAE:
3  cold
2  cool
1  slightly cool
0  neutral
1  slightly warm
2  warm
3  hot
Fanger’s Comfort Equation
A steady-state energy balance is a necessary condition for thermal comfort, but is not sufficient by
itself to establish thermal comfort. Fanger (1970) calculated the heat losses for a comfortable person,
experiencing a neutral sensation, with corresponding skin temperature Tsk and regulatory
sweating Ersw. The calculated heat losses L are then compared with the metabolic rate M. If L  M,
the occupant feels comfortable. If L M, then this person feels cool; and if L 	 M, then this person
feels warm.
Using the responses of 1396 persons during laboratory experiments at Kansas State University
of the United States and Technical University of Denmark, Fanger developed the following equation
to calculate the predicted mean vote (PMV) in the seven-point thermal sensation scale:
PMV  (0.303e0.036M  0.276)(M  L) (4.25)
In Eq. (4.25), the metabolic rate M and heat losses L are both in Btu/h ft2 (W/m2). According to
Fanger’s analysis, the predicted percentage of dissatisfied (PPD) vote for thermal comfort at a
PMV  0 is 5 percent, and at a PMV  1 is about 27 percent.
Tables of PMV and comfort charts including various combinations of operating temperature To,
air velocity v, metabolic rate M, and clothing insulation Rcl have been prepared to determine comfortable
conditions conveniently. Fanger’s comfort charts also include relative humidity. Six of his
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.15
4.16
FIGURE 4.5 Fanger’s comfort charts. (Abridged with permission from ASHRAE Handbook 1981, Fundamentals.)
comfort charts at various activity levels, wet-bulb temperatures, relative humidities, and air velocities
are shown in Fig. 4.5. In the first four charts, the air temperatures are equal to the mean radiant
temperatures. Three of these four have a clothing insulation of 0.5 clo. The other is for 1-met
activity level and 1 clo, because one rarely finds an occupant with such heavy outwear at an activity
level of 2 or 3 met. In the fifth and sixth charts, the air temperature could be different from the
mean radiant temperature with a constant relative humidity of 50 percent. These comfort variations
clearly show that all six factors—air temperature Ta , mean radiant temperature Trad, relative humidity
, air velocity v, metabolic rate M, and clothing insulation Rcl—seriously affect the thermal
comfort.
For example, from Fanger’s comfort chart, a sedentary occupant at an activity level of 1.0 met,
with a clothing insulation of 0.5 clo, in an air conditioned space at a relative humidity of 50 percent
and an air velocity less than 20 fpm (0.1 m/ s), feels comfortable with an air temperature equal to
the mean radiant temperature of 78°F (25.6°C). If all values were identical except for a 2-met activity
level, the temperature would need to be 67°F (19.4°C) for the same level of comfort.
Another factor, the duration of exposure to the indoor thermal environment, should be discussed
here. If an indoor environment can provide thermal comfort for the occupant, the duration of the
exposure has no significant influence upon the physiological responses of the person’s thermal regulatory
mechanism. If the indoor environment is uncomfortable, subjecting the subject to a certain
degree of heat or cold stress, the time exposure will influence the person’s physiological response.
ASHRAE Comfort Zones
Based on results of research conducted at Kansas State University and at other institutions,
ANSI/ASHRAE Standard 55-1992 specified winter and summer comfort zones to provide for the
selection of the indoor parameters for thermal comfort (see Fig. 4.6). This chart is based upon an
occupant activity level of 1.2 met (69.8 W/m2). For summer, typical clothing insulation is 0.5 clo,
that is, light slacks and short-sleeve shirt or comparable ensemble; there is no minimum air speed
that is necessary for thermal comfort. Standard 55-1992 recommended a summer comfort zone with
an effective boundary temperature ET*  73 to 79°F (22.5 to 26°C) at 68°F (20°C) wet-bulb as its
upper-slanting boundary and dew-point temperature 36°F (2.2°C) as its bottom flat boundary. If the
clothing insulation is 0.1 clo higher, the boundary temperatures both should be shifted 1°F (0.6°C)
lower. Rohles et al. (1974) and Spain (1986) suggested that the upper boundary of the summer
comfort zone can be extended to 85 or 86°F (29.4 or 30°C) ET* if the air velocity of the indoor air
can be increased to 200 fpm (1 m/ s) by a ceiling fan or other means.
The winter comfort zone is based upon a 0.9-clo insulation including heavy slacks, long-sleeve
shirt, and sweater or jacket at an air velocity of less than 30 fpm (0.15 m/s). Standard 55-1992 recommended
a winter comfort zone with an effective boundary temperature ET*  68 to 74°F (20 to
23.3°C) at 64°F (17.8°C) wet-bulb as its slanting upper boundary and at dew-point 36°F (2.2°C) as
its bottom flat boundary.
Indoor air parameters should be fairly uniform in order to avoid local discomfort. According to
Holzle et al., 75 to 89 percent of the subjects tested found the environment within this summer
comfort zone to be thermally acceptable.
ASHRAE comfort zones recommend only the optimal and boundary ET* for the determination
of the winter and summer indoor parameters. For clothing insulation, activity levels, and indoor air
velocities close to the values specified in Standard 55-1992, a wide range of indoor design conditions
are available.
Comfort-Discomfort Diagrams
A comfort diagram provides a graphical presentation of the total heat loss from the human body at
various operative or air temperatures and indoor relative humidities when the activity level, level of
clothing insulation, and air velocity are specified. The abscissa of the comfort diagram is the opera-
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.17
tive temperature To or ambient air temperature Ta, whereas the ordinate can be either water vapor
pressure p or humidity ratio w. Figure 4.7 shows a comfort diagram with a sedentary activity level,
a clothing insulation value of 0.6 clo, and still-air conditions, i.e., an air velocity v 	 20 fpm
(0.1 m/ s). The curved lines represent relative humidity, and the straight lines represent effective
temperature ET*. The short dash curves diverging from the ET* lines are total skin wetness wsk
lines. The figure is based upon To  Ta.
4.18 CHAPTER FOUR
Winter
Summer
w, lb/lb
Dew-point temperature Tdew, F
68 F wet bulb
64 F wet bulb
Operative temperature T0, F
70
60
50
40
30
ET* 68 F
73
74
79
20
10
60 70 80 90
0
0.005
0.010
0.015
100%
50%
30%
70%
60%
FIGURE 4.6 ASHRAE comfort zones. (Adapted with permission from ANSI/ASHRAE Standard, 55–1992.)
Effective temperature ET* lines are calculated according to Eq. (4.23). Because the total skin
wetness wsk is a constant for a specific ET* below 79°F ET*, the ET* lines and wsk lines coincide
with one other. At higher ET* values, wsk lines curve to the left at high relative humidities. For low
ambient air temperatures, evaporative heat loss from the skin surface Esk is small; therefore, ET*
and wsk lines are nearly vertical. As Esk becomes greater and greater, the slopes of the ET* and wsk
lines decrease accordingly.
The comfort diagram is divided into five zones by the ET* and wsk lines:
1. Body cooling zone. For the condition given in Fig. 4.7, if the effective temperature ET* 	
73°F (22.8°C), the occupant will feel cold in this zone. Because the heat losses exceed the net metabolic
rate, the skin and body core temperatures tend to drop gradually.
2. Comfort zone. This is the zone between the lower boundary ET*  73°F (22.8°C), and wsk 
0.06, and the higher boundary ET*  86°F (30°C), and wsk  0.25. Steady-state thermal equilibrium
is maintained between the occupant and the environment, and regulatory sweating is at a
rather low level. The occupant will feel comfortable in this zone, and the heart rate (HR) is between
76 and 87 beats per minute.
ASHRAE’s winter and summer comfort zones are a part of this zone. The lower boundary of the
ASHRAE winter comfort zone forms the lower boundary of this comfort zone. The reason that the
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.19
FIGURE 4.7 Comfort-discomfort diagram.
lower boundary in this diagram is ET*  73°F (22.8°C) whereas ET*  68°F (20°C) in
ASHRAE’s winter comfort zone is that a lower clothing insulation of 0.6 clo is used here.
3. Uncomfortable zone. In this zone, 86°F 	 ET*  95°F (30°C  ET*  35°C) and 0.25 	
wsk  0.45. Thermal equilibrium also exists between the occupant and the environment, and the
evaporative heat loss due to regulatory sweating dominates. Heart rate shows a range between 87
and 100. The occupant feels uncomfortable, i.e., warm or hot, when his or her physiological parameters
are in these ranges.
4. Very uncomfortable zone. In this zone, 95°F 	 ET* 	 106.5°F (35°C 	 ET* 	 41.4°C) and
0.45 	 wsk 	 1. Although thermal equilibrium is still maintained with zero heat storage at the skin
and the body core, there is a danger of a heat stroke when ET* 95°F (35°C). Toward the upper
boundary of this zone, the skin surface is nearly entirely wet, and the heart rate exceeds 120. Under
these conditions, the occupant will feel very hot and very uncomfortable.
5. Body heating zone. When ET*  106.5°F (41.4°C) and wsk  1, thermal regulation by evaporation
fails. At a higher ET* or wsk, the environment is intolerable, and the temperatures of the
body core and skin tend to rise gradually.
For an air conditioned space with an occupant at low activity levels (M 	 2 met), the indoor
environment is usually maintained within the comfort zone, and the physiological and thermal responses
of the occupant are also in the comfort zone. Only at higher activity levels do the thermal
responses occasionally fall into the discomfort zone.
4.8 INDOOR AIR TEMPERATURE AND AIR MOVEMENTS
Comfort Air Conditioning Systems
For comfort air conditioning systems, most occupants have a metabolic rate of 1.0 to 1.5 met. The
indoor clothing insulation in summer is usually 0.35 to 0.6 clo, and in winter it is 0.8 to 1.2 clo.
Relative humidity has a lesser influence on thermal comfort, and will be discussed in the next section,
but indoor air temperature and air velocity are discussed here.
Many researchers have conducted tests to determine the effects of airspeed on the preferred indoor
air temperature and the thermal comfort of occupants. The relationship between the preferred
indoor air temperatures and various airspeeds is presented in Fig. 4.8. Most of the data were taken
under these conditions: metabolic rate M  400 Btu/h (117 W), clothing insulation Rcl  0.6 clo,
Trad  Ta, and relative humidity of the indoor air   50 percent. The one exception is the students
in Holzle’s experiments, who had 0.54 clo for summer and 0.95 clo for winter. Examination of
Fig. 4.8 shows the following:
 Higher indoor air temperature requires greater indoor air velocity to provide thermal comfort.
 Variation of airspeed has a greater influence on preferred indoor air temperature at lower air
temperatures.
ANSI/ASHRAE Standard 55-1992 recommended that within the thermally acceptable temperature
ranges in the ASHRAE summer and winter comfort zones discussed in Sec. 4.7, there be no minimum
airspeed (nondirectional) that is necessary for thermal comfort.
If temperature is increased above the level allowed for the comfort zone, means must be provided
to elevate the airspeed. For instance, when indoor air temperature Ta  Trad (mean radiant
temperature), given that the airspeed for a summer comfort zone of 79°F (26°C) is 40 fpm (0.2
m/ s), if Ta has an increase of 2°F (1.1°C) from 79 to 81°F (26 to 27.2°C), according
to ANSI/ASHRAE Standard 55-1992, there must be a relevant increase of airspeed of about
70 fpm (0.35 m/ s).
Before specifying Ta for summer conditions, one needs to determine whether occupants are
likely to wear suit jackets, such as members of a church congregation or guests in a multipurpose
4.20 CHAPTER FOUR
hall. In such cases, a reduction of 4°F of summer optimal ET* may be necessary because of the increase
in clothing insulation of about 0.4 clo.
Design Considerations
When one is specifying indoor design conditions, thermal comfort must be provided at optimum
cost while using energy efficiently. These principles should be considered:
1. To determine the optimum summer and winter indoor design temperatures, consider the local
clothing habits and the upper and lower acceptable limits on clothing insulation at various operative
temperatures To, as shown in Fig. 4.9.
2. It is always more energy-efficient to use different indoor design temperatures for summer and
winter than a year-round constant value. An unoccupied-period setback during winter always
saves energy. There are also buildings in which a constant indoor temperature is required for the
health and comfort of the occupants, such as in many health care facilities.
3. For short-term occupancies, or when the metabolic rate is higher than 1.2 met, a strategy of using
a lower energy-use ceiling fan or a wall-mounted fan to provide higher air velocity may be
considered. Thus a higher indoor design temperature within the extended summer comfort zone
may be acceptable for occupants, especially in industrial settings.
Indoor Design Temperatures for Comfort Air Conditioning
According to ANSI/ASHRAE Standard 55-1992, Thermal Environmental Conditions for Human
Occupancy, and ASHRAE/IES 90.1-1999, Energy-Standard for Buildings Except Low-Rise
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.21
FIGURE 4.8 Preferred indoor air temperatures at various air velocities.
Residential Buildings, the following indoor design temperature and air speed apply for comfort air
conditioning systems when the activity level is 1.2 met, there is a relative humidity of 50 percent in
summer, mean airspeed  30 fpm (0.15 m/ s), and Ta  Trad:
According to ANSI/ASHRAE Standard 55-1992, “within the thermally acceptable temperature
ranges, there is no minimum air speed that is necessary for thermal comfort.” If the summer indoor
temperature is 79°F (26°C), an airspeed of 40 fpm (0.2 m/s) is recommended. If Ta 79°F (26°C),
a relevant increase of airspeed in the indoor occupied zone should be considered. Refer to
ANSI/ASHRAE Standard 55-1992.
If the space relative humidity can be lowered to 35 to 40 percent in summer, then the higher
limit 78°F (25.5°C) is often specified.
Indoor badminton and table tennis tournament arenas should have air velocities below 30 fpm
(0.15 m/ s). To avoid nonuniformity and to prevent local discomfort, the air temperature difference
between 4 in. from the floor and 67 in. above the floor should not exceed 5°F (3°C). The radiant
temperature asymmetry in the vertical direction should be less than 9°F (5°C), and in the horizontal
direction less than 18°F (10°C).
To determine whether the specified air temperature and airspeed are met, they should be measured
at 4-, 24-, and 43-in. (0.1-, 0.6-, and 1.1-m) levels for sedentary occupants, and 4-, 43-, and
67-in. (0.1-, 1.1-, and 1.7-m) levels for standing activity. The duration for determining the mean
value of the air movement should be 3 min. or 30 times the 90 percent response time of the measuring
instrument, whichever is greater.
Typical clothing Optimum operative Indoor design
insulation, clo temperature temperature range
Winter 0.9 71°F (22°C) 69–74°F (20.5–23.5°C)
Summer 0.5 76°F (24.5°C) 74–79°F (23.5–26°C)
4.22 CHAPTER FOUR
FIGURE 4.9 Relationship between clothing insulation Rcl and operating
temperature To. (Source: ASHRAE Transactions 1983, Part I B. Reptinted
with permission.)
Process Air Conditioning Systems
For process air conditioning systems, indoor design temperature is usually based on previous experiences.
For precision manufacturing projects, a basic temperature plus a tolerance, such as 72 
2°F (22.2  1.1°C) for a precision machinery assembling workshop, is often specified. First, the
tolerance should be neither too tight nor too loose. Second, either the temperature fluctuation at
various times within the working period or the temperature variation within the working space,
or both, should be included in this tolerance.
In unidirectional-flow clean rooms, the velocity of the airstream in the working area is often
specified as 90  20 fpm (0.45  0.1 m/ s) to prevent contamination of the products.
4.9 HUMIDITY
Comfort Air Conditioning Systems
According to ANSI/ASHRAE Standard 55-1992, for the zone occupied by people engaged in light,
primarily sedentary activity (1.2 met), the relative humidity should conform with the limits of
ASHRAE winter and summer comfort zones, as shown in Fig. 4.6. These limits are intended to
maintain acceptable thermal conditions for the occupants’ comfort.
ASHRAE/IESNA Standard 90.1-1999 mandates that where a zone is served by system(s) with
both humidification or dehumidification capacity, means shall be provided to prevent simultaneous
operation of humidification and dehumidification equipment. Standard 90.1-1999 also specifies that
where humidistatic controls are provided, such controls shall be capable of preventing reheating,
mixing of hot and cold air streams, and simultaneous heating and cooling. Refer to Section 29.12
for more details.
The following results should be considered during the design and evaluation of the performance
of comfort air conditioning systems:
 Maintaining an indoor space relative humidity r between 20 and 30 percent in winter prevents or
reduces the condensation at the inner side of the window glass.
 In high-occupancy applications, it may be economical and is still comfortable to specify the summer
indoor relative humidity at 55 to 60 percent if the indoor temperature is within the summer
comfort zone.
 The indoor relative humidity at part load may be considerably higher than at full load in some air
conditioning systems in summer.
 When the indoor relative humidity is below 25 percent, the incidence of respiratory infections increases
significantly. If, simultaneously, indoor temperatures are low, such as below 70°F (21°C),
the induced static electricity in carpeted rooms may cause uncomfortable shocks to occupants
contacting metal furniture or decorations.
 Because the increase in the outdoor ventilation rates in ASHRAE Standard 62-1999 for the air
conditioning system serves the building located in areas where the humidity ratio of outdoor air is
very low during winter, the system has greater difficulty maintaining an indoor relative humidty
of 20 to 30 percent without winter humidification.
 If a humidifier is installed, its humidifying capacity should not exceed the actual humidifying
requirements so that wet surfaces do not occur inside the air-handling unit, packaged unit, and
supply ducts. Wet surface and dirt cause the growth of microorganisms and poor indoor air quality.
Therefore, for comfort air conditioning systems, the recommended indoor relative humidity
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.23
levels are as follows:
Process Air Conditioning Systems
Humidity affects the physical properties of many materials and, therefore, their manufacturing
processes.
Moisture Content. Relative humidity has a marked influence on the moisture content of hygroscopic
materials such as natural textile fibers, paper, wood, leather, and foodstuffs. Moisture content
affects the weight of the products and sometimes their strength, appearance, and quality.
Dimensional Variation. Hygroscopic materials often extend at higher relative humidity and contract
at lower humidity. A 2 percent increase in moisture content may result in a 0.2 percent increase in
dimension of paper. That is why lithographic printing requires a relative humidity of 45  2 percent.
Corrosion and Rust. Corrosion is an electrochemical process. Moisture encourages the formation
of electrolytes and therefore the corrosion process. A relative humidity greater than 50 percent may
affect the smooth operation of bearings in precision instruments. When indoor relative humidity
exceeds 70 percent, rust may be visible on the surface of the machinery and on parts made of steel
and iron.
Static Electricity. Static electricity may cause minute particles to repel or attract one another,
which is detrimental to many manufacturing processes. Static electricity charges minute dust particles,
in the air, causing them to cling to equipment and work surfaces. Static electricity exists in an
indoor environment at normal air temperatures when relative humidity is less than about 40 percent.
Loss of Water. Vegetables and fruits lose water vapor through evaporation from their surfaces
during storage. Low temperatures and high relative humidities, such as  90 to 98 percent, may
reduce water loss and delay desiccation.
It is important to specify the exact relative humidity required for product quality and cost
control. For process air conditioning systems, the specified relative humidity is either a year-round
single value or a range. A strict relative-humidity requirement always includes a basic value and a
tolerance, such as the relative humidity for lithographic printing mentioned before. When temperature
and relative humidity controls are both required, they should be specified as a combination.
Consider this example:
Case Study 4.1. A factory workshop has the following environmental parameters during summer:
Indoor air temperature 79°F (26.1°C)
Indoor air relative humidity 50 percent
Temperature, °F (°C) Relative humidity, %
Clean room 72  2 (22.2  1.1) 45  5
Relative humidity, %
Summer 30–65
Winter
Commercial and public buildings 20–60
Health care buildings 30–60
4.24 CHAPTER FOUR
Space air velocity 20 fpm (0.1 m/s)
Clothing insulation of the workers 0.5 clo
Activity level 3 met
As a result of comfort complaints by personnnel, you are asked to recommend effective and economical
corrective measures to improve the indoor environment. The following table includes the
information required during analysis:
Area ratio of the clothed body fcl 1.2
Permeability index of clothing im 0.4
Skin surface temperature,Tsk 92.7°F (33.7°C)
Saturation water vapor pressure at 92.7°F 0.764 psia (5.27 kPa abs.)
at 79°F 0.491 psia (3.39 kPa abs.)
Solution
1. When the space air velocity v  20 fpm, the convective heat-transfer coefficient can be calculated,
from Eq. (4.9), as
hc  0.0681v0.69  0.0681  200.69  0.538 Btu/h ft2 °F
Because at normal indoor conditions the radiative heat-transfer coefficient hr  0.83 Btu/h ft2 °F,
the surface heat-transfer coefficient h is
h  hc  hr  0.538  0.83  1.368 Btu/h ft2 °F
The clothing insulation Rcl  0.5  0.88  0.44 h ft2 °F/Btu. From Eq. (4.12), the clothing effi-
ciency is found to be
Therefore, from Eq. (4.11), we find the sensible heat loss from the skin surface of the worker
C  R  Fcl fclh(Tsk  To)  0.5806  1.2  1.368(92.7  79)
 13.06 Btu/h ft2
From Eq. (4.17), the overall sensible heat-transfer coefficient hs is
From Eq. (4.16), the overall evaporative heat-transfer coefficient he,c is
he,c  hs im LR  0.9531  0.4  205  78.16 Btu/h ft2  psi
The maximum evaporative heat loss Emax due to regulatory sweating can be calculated from
Eq. (4.15) as
Emax  he,c(psk, s  pa)  78.16(0.764  0.5  0.491)
 40.24 Btu/h ft2

1.2  1.368
1.2  1.368  0.44  1
 0.9531 Btu/h ft2F
hs 
fclh
fclhRcl  1
Fcl 
1
Rcl fclh  1

1
0.44  1.2  1.368  1
 0.5806
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.25
In Eq. (4.24),
0.85M  0.85  3  18.46  47.07 Btu/h ft2
When 	 0.05 and Eres  0.1M, the total evaporative loss from the skin surface of the worker,
from Eq. (4.1), is
Esk  0.85M  (C  R)  47.07  13.06  34.01 Btu/h ft2 (107.3 W/m2)
Therefore, the total skin wetness is
In Fig. 4.7, when wsk  0.845, a person is very uncomfortable.
2. From Fanger’s comfort chart, shown in Fig. 4.5, for a person with an activity level of 3 met
and a clothing insulation of 0.5 clo, at a relative humidity of 50 percent and an air velocity of 20
fpm and with Ta  Trad for neutral thermal sensation, the indoor air temperature should be 56°F
(13.3°C). Obviously, this is not economical because too much refrigeration is required.
3. Let us analyze the results if ceiling fans or wall fans are used to increase the space air velocity
v to 300 fpm (1.5 m/ s). Then the convective heat-transfer coefficient hc is
hc  0.681  3000.69  3.486 Btu/h ft2 °F
The surface heat-transfer coefficient is
h  3.486  0.83  4.316 Btu/h ft2 °F
Also the clothing efficiency is calculated as
Then the sensible heat loss is equal to
C  R  0.305  1.2  4.316(92.7  79)  21.64 Btu/h ft2
The overall sensible heat-transfer coefficient is
The overall evaporative heat-transfer coefficient he,c can be shown to be
he,c  1.58  0.4  205  129.5 Btu/h ft2  psi
Then the maximum evaporative heat loss due to regulatory sweating is
Emax  129.5(0.7604  0.5  0.491)  66.68 Btu/h ft2 (210.3 W/m2)
The total skin wetness is
wsk  47.07  21.64/66.68  0.381
That is, wsk has been greatly reduced compared with the value at v  20 fpm (0.1 m/ s). In Fig. 4.7,
wsk  0.381 is in the uncomfortable zone. Workers will feel warm, but the indoor environment has
been considerably improved. This may be the most cost-effective solution.
hs 
1.2  4.316
1.2  4.136  0.44  1
 1.58 Btu /h ft2F
Fcl 
1
0.44  1.2  4.316  1
 0.305
wsk 
Esk
Emax

34.01
40.24
 0.845
4.26 CHAPTER FOUR
4.10 SICK BUILDING SYNDROME AND INDOOR AIR QUALITY
Sick building syndrome is a kind of building-related illness that has received public attention since
the 1970s. ASHRAE (1987) defined the sick building as “. . . a building in which a significant
number (more than 20 percent) of building occupants report illness perceived as being building related.
This phenomenon, also known as ‘sick building syndrome’ is characterized by a range of
symptoms including, but not limited to, eye, nose, and throat irritation, dryness of mucous membranes
and skin, nose bleeds, skin rash, mental fatigue, headache, cough, hoarseness, wheezing,
nausea, and dizziness. Within a given building, there usually will be some commonality among the
symptoms manifested as well as temporal association between occupancy in the building and appearance
of symptoms.” If there are signs of actual illnesses, these illness are classified as buildingrelated
illnesses.
Poor indoor air quality (IAQ) is the dominant factor that causes sick building syndrome. Indoor
air quality is defined as an indication of harmful concentrations of the indoor air contaminants that
affect the health of the occupants or the degree of satisfaction of a substantial majority (80 percent
or more) of occupants exposed to such an indoor environment. Poor control of the indoor air temperature
and relative humidity are causes of discomfort symptoms. They may also increase the
indoor air contaminants. However, unsatisfactory indoor temperature and indoor relative humidity
are only indirect causes of poor indoor air quality.
National Institute for Occupational Safety and Health (NIOSH) of the United States (1989),
according to the results of 529 building investigations between 1971 and 1988, and Health and
Welfare Canada (HWC), according to the results of 1362 building investigations between 1984
and 1989, classified the reasons for sick building syndrome as follows:
Inadequate ventilation includes lack of outdoor air, poor air distribution, poor thermal control, and
inadequate maintenance; and it is the primary cause of indoor air quality. The survey found 70 to 80
percent of the investigated buildings had no known problems. Effective operation and control of the
HVAC&R system will be discussed in later chapters.
In the United States, most people spend about 90 percent of their time indoors. The purpose of
specifying the indoor design conditions in the design documents is to provide the occupants with a
satisfactory indoor environment at optimum cost.
After the energy crisis in 1973, a lower outdoor ventilation rate, a tighter building shell, and the
use of variable-air-volume (VAV) systems at part-load operation may reduce the amount of outdoor
air intake significantly. Indoor air quality therefore has become one of the critical HVAC&R problems
especially in commercial buildings since the 1980s.
Indoor Air Contaminants
Based on the results of the field investigations of three office buildings by Bayer and Black in 1988,
the indoor air contaminants that relate to indoor air quality and the symptoms of the sick building
syndrome are mainly the following:
NIOSH, 529 Buildings HWC, 1362 Buildings
No. of buildings Percent No. of buildings Percent
Inadequate ventilation 280 53 710 52
Indoor contaminants 80 15 165 12
Outdoor contaminants 53 10 125 9
Biological contaminant 27 5 6 0.4
Building fabric contamination 21 4 27 2
Unknown sources 68 13 329 24
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.27
1. Total particulate concentration. This parameter includes particulates from building materials,
combustion products, and mineral and synthetic fibers. In February 1989, the U.S. Environmental
Protection Agency (EPA) specified the allowable indoor concentration level of particulates of
10 	m and less in diameter (which can penetrate deeply into the lungs, becoming hazardous to
health) as follows:
50 	g/m3 (0.000022 gr/ ft3): 1 year
150 	g/m3 (0.000066 gr/ ft3): 24 h
According to ASHRAE Handbook 1997, Fundamentals, particles less than 2 	m in diameter are
most likely retained in the lungs, and particles less than 0.1 to 0.5 	m in diameter may leave the
lungs with the exhaled breath. Particles larger than 8 to 10 	m in diameter are separated and retained
in the upper respiratory tract. Particles between 2 and 8 	m in diameter are deposited mainly
in the conducting airways of the lungs and are swallowed or coughed out quickly.
2. Combustion products. Carbon monoxide (CO) is a colorless, odorless gas, a product of incomplete
combustion. CO interferes with the delivery of oxygen throughout the body. NO2 is a
combustion product from gas stoves and other sources. There is growing evidence that NO2 may
cause respiratory disease. Indoor concentrations for CO and NO2 are the same as specified in the
National Primary Ambient-Air Quality Standard by the EPA later in this section.
3. Volatile organic compounds (VOCs). These include formaldehyde and a variety of aliphatic,
aromatic, oxygenated, and chlorinated compounds. Mucous membrane irritation caused by
formaldehyde is well established. U.S. Department of Housing and Urban Development (HUD)
specifies a target level of indoor concentration of formaldehyde for manufacturing homes of
0.4 ppm.
4. Nicotine. Environmental tobacco smoke is clearly a discomfort factor to many adults who do
not smoke. Nicotine and other components of tobacco smoke are also a health risk for human beings.
5. Radon. Radon is a colorless, odorless, inert radiative gas widely found in soil, rocks, and water,
created by the decay of the radium and uranium. It travels through the pores of rock and soil and in-
filtrates into a house along cracks and other openings in the basement slab or walls; pressure-driven
radon containing soil gas is caused by thermal stack, wind, and the mechanical ventilation system.
The annual average concentration of radon in residential buildings in the United States is about
1.25 pCi/L. Only about 6 percent of U.S. homes have an annual average radon concentration exceeding
the EPA recommended annual average indoor concentration of 4 pCi/L for residential and
school occupancies. Pressure-driven flow of radon containing soil gas is the primary source for elevated
concentrations. At various locations in the United States, the indoor radon concentrations may
vary hourly, daily, and seasonally, sometimes by as much as a factor of 10 to 20 on a daily basis.
The radiative decay of radon produces a series of radioactive isotopes called progeny. These progeny
are chemically active. They can deposit directly into the lung, or attach to airborne particles and
then deposit into the lung. Some of the progeny are alpha particle emitters and may lead to cellular
changes and initiate lung cancer.
6. Occupant-generated contaminants and odors. These include odors and emissions (bioeffluents)
from the human body, particulates, and other contaminants.
7. Bioaerosols. These contaminants include bacteria, mold and midew, viruses, and pollens.
Bacteria and viruses are airborne, carried by dust or transmitted by people and animals; standing
water (wet surface) and dirt (nutrients) can become the breeding ground for mold, mildew, and
other biological contaminants. Pollens originate from plants.
In addition to the preceding indoor air contaminants, others such as sulfur dioxide and ozone
can have a significant effect on occupants. Carbon dioxide (CO2) is a kind of gas released from
human beings and is not an indoor contaminant at the concentrations found in most buildings.
ASHRAE Standard 62-1999 specifies guidelines for indoor concentration for ozone during continuous
exposure time as 100 	g/m3.
4.28 CHAPTER FOUR
Basic Strategies to Improve Indoor Air Quality
There are three basic strategies to improve indoor air quality: control the contaminated source,
remove air contaminants from the indoor air by air cleaner, and use outdoor ventilation air to dilute
the concentrations of indoor air contaminants.
To eliminate or to reduce the emissions of air contaminants from the contaminated source is often
the most effective way to improve the IAQ as well as periodically cleaning the duct’s interior
surface and coil’s condensate pan and using building materials and carpets that do not release or release
only negligible volatile organic compounds and dust. However, emissions and odor released
from occupants are difficult to eliminate or to reduce, and smoking is now prohibited in many public
places and limited to specified areas in many commercial buildings.
The volatile organic compounds and combustion products contain numerous minute particles of
size between 0.003 and 1 	m. Only high-efficiency air filters and activated carbon filters can remove
these minute particles and odors from the airstream effectively. High efficiency air filters and
carbon filters are expensive to install, operate, and maintain.
Adequate outdoor ventilation air to dilute the air contaminations in practice has been proved an
essential, practical, and cost-effective means to improve the indoor air quality. ASHRAE Standard
62-1999, Ventilation for Acceptable Indoor Air Quality, specifies two alternative procedures to obtain
acceptable IAQ: the ventilation rate procedure and indoor air quality procedure.
 In the ventilation rate procedure, acceptable indoor air quality is achieved by providing ventilation
air of specified quality and quantity to the space.
 In the IAQ procedure, acceptable air quality is achieved within the space by controlling known
and specifiable contaminants.
The ventilation rate and IAQ procedurs are discused again in Sec. 23.2.
ASHRAE Standard 62-1999 defined acceptable indoor air quality as air in which there are no
known contaminants at harmful concentrations as determined by cognizant authorities and with
which a substantial majority (80 percent or more) of the people exposed do not express dissatisfaction.
ASHRAE Standard 62-1999 defines ventilation air as that portion of supply air that is outdoor
air plus any recirculated air that has been treated for the purpose of maintaining acceptable indoor
air quality. Ventilation is the process of supplying and removing ventilation by natural and mechanical
means, and the ventilation rate means the rate of ventilation air supplied to the conditioned
space through the air system.
If outdoor air is used to dilute the concentration of indoor contaminants, its quality must meet
the National Primary Ambient-Air Quality Standard provided by the EPA. Part of the time average
concentrations are shown below:
Long-term concentration Short-term concentration
Average period Average period
Pollutants 	g/m3 ppm of exposure 	g/m3 ppm of exposure
Particulate matter 50 1 year 150 24 hours
Sulfur oxides 80 0.03 1 year 365 0.14 24 hours
Carbon monoxide 40,000 35 1 hour
10,000 9 8 hours
Nitrogen dioxide 100 0.055 1 year
Oxidants (ozone) 235 0.12 1 hour
Lead 1.5 3 months
Only particulate matter is expressed in annual geometric means; the other two are annual arithmetic
means. For carbon monoxide and ozone, both values are not to be exceeded more than once a year.
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.29
Outdoor Air Requirements for Occupants
For both comfort and process air conditioning systems, outdoor air is required to do the following:
 To meet metabolic requirements of the occupants
 To dilute the indoor air contaminants, odors, and pollutants to maintain an acceptable indoor air
quality
 To support any combustion process or replace the amount of exhaust air required in laboratories,
manufacturing processes, or restrooms
 To provide makeup for the amount of exfiltrated air required when a positive pressure is to be
maintained in a conditioned space
The amount of outdoor air required for metabolic oxidation processes for occupants is actually
rather small. ASHRAE Standard 62-1999 noted that where only dilution ventilation is used
to control indoor air quality, a CO2 indoor-to-outdoor differential concentration is
not greater than about 700 ppm, and the CO2 production of a sedentary occupant who is eating
a normal diet is 0.0106 cfm (0.3 L / min). The amount of outdoor air required for each indoor
occupant can be calculated as
(4.26)
Thayer (1982), using data from different authors, developed a dilution index that indicates
15 cfm (7 L/s) of outdoor air per person will satisfy more than 80 percent of the occupants in
the space.
Usually, for comfort air conditioning systems, the same outdoor air used for the dilution of the
concentration of air contaminants including human bioeffluents is sufficient for the metabolic oxygen
requirement, for exhausting air from restrooms, and for replacing exfiltrated air lost from the
conditioned space as a result of positive pressure. If the outdoor air supply is used to dilute the concentration
of a specific indoor air contaminant, the rate of outdoor air supply in cfm (L/min),
can be calculated as
(4.27)
where  rate of generation of contaminants in space, mg/ s
Ci, Co  concentrations of air contaminants indoors and outdoors, respectively, mg/m3
Values of Co can be found from the EPA National Primary Ambient-Air Quality Standards. The indoor
concentration of CO2 and other contaminants should meet the specified value as stated before.
Some of the outdoor air requirements for ventilation, often called the ventilation rate, specified in
ASHRAE Standard 62-1999 are indicated in Table 4.5. For clean rooms, Federal Standard 209B
specifies the rate of outdoor air, or makeup air, to be 5 to 20 percent of the supply air.
When only dilution ventilation is used to control indoor air quality and CO2 is used as an indicator
of human bioeffluents, ASHRAE Standard 62-1999 noted that an indoor-outdoor differential
concentration not greater than about 700 ppm of CO2 indicates that comfort (odor) criteria related
to bioeffluents are likely to be satisfied. The CO2 concentrations in outdoor air typically are between
300 to 350 ppm. Using CO2 as an indicator of bioeffluents does not elimnate the need for
consideration of other contaminants.
The refrigeration capacity required to cool and dehumidify the outdoor air can be a major porm
? par
V?
o 
2118m? par
Ci  Co
V?
o,
 15 cfm (7 L / s)
V?
o,oc 
V?
CO2
Ci,CO2  Co

0.0106
0.0007
V?
o, oc
V?
CO2
(Ci,CO2  Co)
4.30 CHAPTER FOUR
tion of the total refrigeration requirement during summer, depending upon the occupant density and
the amount of exhaust air. The amount of infiltration depends on the wind speed and direction, as
well as the outdoor and indoor temperatures and pressure differences, which are variable. Therefore,
infiltration is not a reliable source of outdoor air supply. Infiltration cannot replace specified
outdoor air ventilation requirements.
4.11 AIR CLEANLINESS
The manufacturing process of semiconductors, pharmaceutical, aerospace, and operating rooms in
health care facilities need clean indoor environment, clean spaces, and clean rooms. A clean room is
a constructed enclosed area in which air cleanliness is expressed in terms of particle count of air
contaminants and in which the associated temperature, humidity, air pressure, and lighting are controlled
within specific limits. A clean space is a defined area in which air cleanliness and environmental
conditions are controlled within specific limits.The quality of their manufactured products is
closely related to the size and number of particulates contained in the space air. An indoor air quality
of allowable total particulate annual-average concentration of 50 	g/m3 cannot meet the air
cleanliness requirements. Therefore, Federal Standard (FS) 209E specifies the following classes for
clean spaces and clean rooms:
Class 1. Particle count not to exceed 1 particle/ ft3 (35 particles /m3) of a size of 0.5 	m and
larger, with no particle exceeding 5 	m.
Class 10. Particle count not to exceed 10 particles/ ft3 (353 particles /m3) of a size of 0.5 	m and
larger, with no particle exceeding 5 	m.
Class 100. Particle count not to exceed 100 particles/ ft3 (3531 particles /m3) of a size of 0.5 	m
and larger.
Class 1000. Particle count not to exceed 1000 particles/ ft3 (35,315 particles/m3) of a size of
0.5 	m and larger.
Class 10,000. Particle count not to exceed 10,000 particles/ ft3 (353,150 particles/m3) of a size
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.31
TABLE 4.5 Outdoor Air Requirements
Recommended by ASHRAE Standard 62–1999
Application cfm/person
Dining room 20
Bar and cocktail lounges 30
Conference rooms 20
Office spaces 20
Office conference rooms 20
Retail stores 0.2–0.3*
Beauty shops 25
Ballrooms and discos 25
Spectator areas 15
Theater auditoriums 15
Transportation waiting rooms 15
Classrooms 15
Hospital patient rooms 25
Residences 0.35†
Smoking lounges 60
*cfm/ft2 floor area
†Air changes/h
Source: Abridged with permission from ASHRAE
Standard 62-1999.
of 0.5 	m and larger or 65 particles / ft3 (2295 particles / ft3) of a size 5.0 	m and larger.
Class 100,000. Particle count not to exceed 100,000 particles/ ft3 (3,531,500 particles/m3) of
a size of 0.5 	m and larger or 700 particles / ft3 (24,720 particles / ft3) of a size 5.0 	m and larger.
Refer to Federal Standard 209E for more details.
Since workers in these clean rooms wear protective gowns and hats, in order to provide these air
cleanliness classes, a year-round constant temperature and associated relative humidity and a speci-
fied unidirectional airflow should be maintained. High-efficiency particulate air (HEPA) filters and
ultra-low-penetration air (ULPA) filters should be installed in the air conditioning systems. Special
building materials with hard and clean surfaces should be used as the building envelope. More details
are covered in Chap. 30.
4.12 SOUND LEVEL
Sound and Sound Level
Sound can be defined as a variation in pressure due to vibration in an elastic medium such as air. A
vibrating body generates pressure waves, which spread by alternate compression and rarefaction of
the molecules within the transmitting medium. Airborne sound is a variation of air pressure, with
atmospheric pressure as the mean value. Because sound is transmitted by compression and expansion
of molecules, it cannot travel in a vacuum. The denser the material, the faster the traveling
speed of a sound wave. The velocity of a sound wave in air is approximately 1130 fps. In water, it is
about 4500 fps and in steel 15,000 fps.
Noise is any unwanted sound. In air systems, noise should be compensated for, either by attenuation
(the process of reducing the amount of sound that reaches the space) or by masking it with
other, less objectionable sounds.
Sound Power Level and Sound Pressure Level
Sound power is the ability to radiate power from a sound source excited by an energy input. The intensity
of sound power is the power output from a sound source expressed in watts (W). Because of
the wide variation of sound output—from the threshold hearing level of 1012 W to a level of
108 W, generated by the launching of a Saturn rocket, a ratio of 1020 to 1—it is more appropriate
and convenient to use a logarithmic expression to define sound power level, i.e.,
(4.28)
where Lw  sound power level, dB
w  sound source power output,W
Here 1012W, or 1 pW (picowatt), is the international reference base, and re indicates the reference
base.
The human ear and microphones are pressure-sensitive. Analogous to the sound power level, the
sound pressure level is defined as
(4.29)
where Lp  sound pressure level, dB
p  sound pressure, Pa
Lp  20 log
p
2  105 Pa
re 20 	Pa
Lw  10 log
w
1012 W
re 1 pW
4.32 CHAPTER FOUR
Here the reference sound pressure level is 2  10–5 Pa (pascal), or 20 	Pa, corresponding to the
hearing threshold. Because sound power is proportional to the square of the sound pressure, 10 log
p2  20 log p. Sound pressure levels of various sources are listed in Table 4.6.
The sound power level of a specific source is a fixed output. It cannot be measured directly and
can only be calculated through the measurement of the sound pressure level. On the other hand,
sound pressure level is the sound level measured at any one point and is a function of distance from
the source and characteristics of the surroundings.
Octave Bands
Sound waves, like other waves, are characterized by the relationship between wavelength, speed,
and frequency:
(4.30)
People can hear frequencies from 20 Hz to 20 kHz. To study and analyze sound, we must break it down
into components. A convenient way is to subdivide the audible range into eight octave bands or sometimes
- octave bands. An octave is a frequency band in which the frequency of the upper band limit
is double the frequency of the lower band limit. The center frequency of an octave or a octave band is
the geometric mean of its upper and lower band limits. An octave or octave band is represented by its
center frequency. The eight octave bands and their center frequencies are listed in Table 4.7.
Addition of Sound Levels
Because sound levels, in dB, are in logarithmic units, two sound levels cannot be added arithmetically.
If sound levels A, B, C, . . ., in dB, are added, the combined overall sound level L can be
calculated as
L  10 log (100.1A  100.1B  100.1C  · · ·) (4.31)
1
3-
1
3-
24
13
Wavelength 
speed
frequency
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.33
TABLE 4.6 Typical Sound Pressure Levels
Sound Sound pressure Subjective
Source pressure, Pa level dB re 20 	Pa reaction
Military jet takeoff at 100 ft 200 140 Extreme
Passenger’s ramp at jet Threshold
airliner (peak) 20 120 of pain
Platform of subway station
(steel wheels) 2 100
Computer printout room* 0.2 80
Conversational speech at 3 ft 0.02 60
Window air conditioner* 0.006 50 Moderate
Quiet residential area* 0.002 40
Whispered conversation at 6 ft 0.0006 30
Buzzing insect at 3 ft 0.0002 20
Threshold of good hearing 0.00006 10 Faint
Threshold of excellent youthful Threshold
hearing 0.00002 0 of hearing
*Ambient.
Source: Abridged from ASHRAE Handbook 1989, Fundamentals. Reprinted by permission.
Human Response and Design Criteria
The human brain does not respond in the same way to lower frequencies as to higher frequencies.
At lower sound pressure levels, it judges a 20-dB sound at 1000 Hz to have the same loudness as a
52-dB sound at 50 Hz. However, at high sound pressure levels, a 100-dB sound at 1000 Hz seems
as loud as 110 dB at 50 Hz.
The purpose of noise control in an air conditioned space is to provide a background sound low
enough to avoid interference with the acoustical requirements of the occupants. The distribution of
the background sound should be balanced over a wide range of frequencies, without whistle, hum,
rumble, or audible beats. Three types of criteria for sound control are currently used in indoor system
design:
1. A-weighted sound level dBA. The A-weighted sound level dBA tries to simulate the response
of the human ear to sound at low sound pressure levels. An electronic weighting network automatically
simulates the lower sensitivity of the human ear to lower-frequency sounds by subtracting a
certain number of decibels at various octave bands, such as approximately 27 dB in the first octave
band, 16 dB in the second, 8 dB in the third, and 4 dB in the fourth. The A-weighted sound level
gives a single value. It is simple and also takes into consideration the human judgment of relative
loudness at low sound pressure levels. Its main drawback is its failure to consider the frequency
spectrum or the subjective quality of sound.
2. Noise criteria, or NC, curves. NC curves represent actual human reactions during tests. The
shape of NC curves is similar to the equal loudness contour representing the response of the human
ear, as shown in Fig. 4.10. NC curves are intended to indicate the permissible sound pressure level
of a broadband noise at various octave bands by a single NC curve sound level rating. NC curves
are practical. They also consider the frequency spectrum of the broadband noise. The main problem
with NC curves is that the shape of the curve does not approach a balanced, bland-sounding spectrum
that is neither rumbly nor hissy.
3. Room criteria, or RC, curves. RC curves, as shown in Fig. 4.11, are similar to NC curves
except that the shape of an RC curve is a close approximation of a balanced, bland-sounding
spectrum.
ASHRAE recommends the indoor design RC or NC ranges presented in Table 4.8. For sounds
containing significant pure tones or impulsive sounds, a 5- to 10-dB lower value should be specified.
Noise is always an annoying element and source of complaints in indoor environments. Attenuation
4.34 CHAPTER FOUR
TABLE 4.7 Octave Bands and Their Center
Frequencies
Band
Band frequency, H2
number Lower Center Upper
22.4 31.5 45
1 45 63 90
2 90 125 180
3 180 250 355
4 355 500 710
5 710 1,000 1,400
6 1,400 2,000 2,800
7 2,800 4,000 5,600
8 5,600 8,000 11,200
Source: Abridged with permission from ASHRAE
Handbook 1989, Fundamentals with permission.
to achieve an NC or RC goal less than 30 dB for a central air conditioning system or a packaged system
is very expensive. In actual practice, the NC or RC criteria range for private residences and
apartments varies greatly because of personal requirements and economic considerations.
To meet the listed design criteria range, the measured sound pressure levels of at least three of the
four octave bands between 250 and 2000 Hz should be within the listed NC or RC range. In
industrial workshops with machinery and equipment, Occupational Safety and Health Administration
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.35
FIGURE 4.10 Noise criteria (NC) curves. (Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with
permission.)
(OSHA) Standard Part 1910.95, published by the U.S. Department of Labor, specifies: “Feasible
administrative or engineering controls shall be utilized when employees are subjected to sound
levels exceeding those in Table G-16. If such controls fail to reduce sound levels to below those in
Table G-16, personal protective equipment shall be provided and used. Exposure to impulsive and
impact noise should not exceed 140 dB peak sound pressure level.” Table G-16 is reproduced in this
text as Table 4.9.
4.36 CHAPTER FOUR
FIGURE 4.11 Room criteria (RC) curves. (Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with
permission.)
4.13 SPACE PRESSURE DIFFERENTIAL
Most air conditioning systems strive to maintain a slightly higher indoor pressure than that of the
surroundings. This positive pressure tends to eliminate or reduce infiltration, the entry of untreated
air to the space. Negative space pressure may cause space high humidity levels, mold and mildew
growth, combustion equipment backdraft, and entry of sewer gas. For rooms where toxic,
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.37
TABLE 4.8 Recommended Criteria for Indoor Design RC
or NC Range
Recommended criteria
Type of area for RC or NC range
Private residence 25–30
Apartments 25–30
Hotels/motels
Individual rooms or suites 30–35
Meeting/banquet rooms 25–30
Halls, corridors, lobbies 35–40
Service/support areas 40–45
Offices
Executive 25–30
Conference 25–30
Private 30–35
Open-plan 30–35
Computer equipment rooms 40–45
Public circulation 40–45
Hospitals and clinics
Private rooms 25–30
Wards 30–35
Operating rooms 35–40
Corridors 35–40
Public areas 35–40
Churches 25–30
Schools
Lecture and classrooms 25–30
Open-plan classrooms 30–35
Libraries 35–40
Concert halls †
Legitimate theaters †
Recording studios †
Movie theaters 30–35
Note: These are for unoccupied spaces, with all systems
operating.
*Design goals can be increased by 5 dB when dictated by budget
constraints or when intrusion from other sources represents a limiting
condition.
†An acoustical expert should be consulted for guidance on these
critical spaces.
Source: ASHRAE Handbook 1987, HVAC Systems and Applications.
Reprinted with permission.
TABLE 4.9 Occupational Noise Exposure
(Occupational Health and Safety Administration
Table G-16)
Duration Sound level
per day, h dBA slow response
8 90
6 92
4 95
3 97
2 100
1 102
1 105
110
(or less) 115
Note: If the variations in noise level involve
maxima at intervals of 1 s or less, it is to be considered
continuous.
In all cases where the sound levels exceed the
values shown herein, a continuing, effective hearing
conservation program shall be administered.
Source: Occupational Safety and Health Standard,
Part 1910.95. Reprinted with permission.
1
4
1
2
1
2
hazardous, contaminated, or objectionable gases or substances are produced, a slightly lower
pressure, or an appropriate negative room pressure, should be maintained to prevent the diffusion of
these substances to the surroundings and at the same time prevent and reduce the damage of the
uncontrolled outdoor airflow.
The magnitude of the positive or negative pressure to be maintained in the space must be carefully
determined. A higher positive pressure always means a greater amount of exfiltrated air,
which creates requirements for greater volumes of outdoor air intake. A higher negative pressure
means a greater infiltration. Both result in higher initial and operating costs. Space pressurization
is closely related to the amount of effective leakage area, or tightness of the building. Cummings
et al. (1996) reported that the space pressure differentials between indoor and outdoor air and
building tightness vary widely for seven field-surveyed restaurants. For comfort systems in lowrise
buildings, the recommended pressure differential is 0.005 to 0.03 in. WC (1.25 to
7.5 Pa), which is often adopted for average building tightness. For a 44-story high-rise
building, the pressure differential between the entrance lobby outdoors and rooftop floor because
of the stack effect during winter may exceed 0.3 in. WC (75 Pa). The phrase in. WC represents
the pressure at the bottom of a water column that has a height of the specified number of
inches.
Another type of space pressurization control for the health and comfort of passengers in an aircraft
is cabin pressurization control. If an airplane is flying at an altitude of 28,000 ft (8540 m) with
an ambient pressure of 4.8 psia (33.1 kPa abs.), the minimum cabin pressure required is 10.8 psia
(75 kPa abs.); and a cabin pressurization control system must be installed to maintain a pressure
differential of 10.8  4.8  6 psi (41 kPa) between the cabin and the ambient air.
Process air conditioning systems, such as for clean rooms, need properly specified space
pressurization to prevent contaminated air from entering the clean, uncontaminated area from surrounding
contaminated or semicontaminated areas. According to Federal Standard 209B, Clean
Rooms and Work Station Requirements, Controlled Environment, published by the U.S. government
in 1973, the minimum positive pressure differential between the clean room and any adjacent area
of less clean requirements should be 0.05 in. WC (12.5 Pa) with all entryways closed. When the entryways
are open, an outward flow of air is to be maintained to minimize migration of contaminants
into the room.
Space pressurization is discussed in Chaps. 20 and 23.
4.14 OUTDOOR DESIGN CONDITIONS
In principle, the capacity of air conditioning equipment is selected so that indoor design conditions
can be maintained when the outdoor weather does not exceed the design values. The outdoor
weather affects the space cooling load and the capacity of the air system to condition the required
amount of outdoor air. It is not economical to choose the annual maximum or annual minimum values
as the design data. Outdoor design conditions are usually determined according to statistical
analysis of the weather data of the previous 30 years so that either 99.6 or 99 percent of the winter
indoor design conditions can then be attained annually, or only 0.4, 1.0, or 2.0 percent of the time
annually the summer indoor design conditions are exceeded. The greater the need for strict control
of indoor environmental parameters, the greater the percentage coverage of the winter outdoor design
hours or the smaller the percentage occurrence of the total hours that exceed the outdoor
design values.
The recommended design values are based on data from the National Climatic Data Center
(NCDC) from 1961 to 1991, or from 1982 to 1993, and Canadian Weather Energy and Engineering
Data Sets (CWEEDS) from 1953 through 1993. Per ASHRAE Handbook 1997, Fundamentals, and
ASHRAE/IESNA Standard 90.1-1999, the design conditions recommended for large cities in the
United States are listed in Table 4.10.
4.38 CHAPTER FOUR
Summer and Winter Outdoor Design Conditions
In Table 4.10, note the following:
 Summer outdoor design dry-bulb temperature for a specific locality To,s, in °F (°C), is the rounded
higher integral number of the statistically determined summer design temperature To,ss such that
the average number of hours of occurrence of outdoor temperatures To higher than To,ss annually
is on average 0.4 percent (35 h), 1 percent (88 h), or 2 percent (175 h).
 The summer outdoor mean coincident design wet-bulb temperature To, s, in °F (°C), is the mean
of all wet-bulb temperatures occurring at the specific summer outdoor design dry-bulb temperature
To,s during the summer.
 The 1.0 percent summer design wet-bulb temperature To1 is the design value having an average
annual occurrence of To  To1 of 88 h. Variable To1 is used for evaporative cooling systems,
cooling towers, and evaporative condensers, which are covered in later chapters.
 The mean daily range, in °F (°C), indicates the mean of the difference between the daily maximum
and minimum temperatures for the warmest month.
 The winter outdoor design dry-bulb temperature To,w, in °F, is the rounded lower integral value of
the statistically determined winter outdoor design temperature To,ws, such that the annual average
number of hours of occurrence of outdoor temperature at values To  To,ws should be equaled or
exceeded 99.6 or 99 percent of the total number of annual hours. The annual average number of
hours in winter when To 	 0.99To,ws is 35 h, and To 	 0.99To,ws is 88 h.
 The number of degree-days is the difference between a base temperature and the mean daily outdoor
air temperature for any one day Tbase  To,m, both in °F. Annually, the total number of heating
degree-days with a base temperature of 65°F, or HDD65, is
(4.32)
where n  number of days whose To,m 	 65°F per annum. The total number of cooling degreedays
with a base temperature of 50°F, or CDD50, is
(4.33)
where m  number of days whose To,m 50°F per annum. Heating and cooling degree-days with
different base temperatures have been used as climatic parameters to calculate the energy flux
through building envelope, or to determine the U value and the configuration of the building
envelope.
 In the last two columns, MWS/MWD to DB 99.6% indicates mean coincident wind speed
(MWS)/mean coincident wind direction (MWD), i.e., most frequently occurring with the heating
dry-bulb 99.6 percent. In MWD, wind direction is expressed in degrees; 270° represents west, and
180° south.
The Use of Outdoor Weather Data in Design
During the design of air conditioning systems, the following parameters are often adopted:
1. Indoor and outdoor design conditions are used to calculate the space cooling and heating loads.
2. Summer outdoor dry-bulb and coincident wet-bulb temperatures are necessary to evaluate the
coil load. The summer outdoor wet-bulb temperature is used to determine the capacity of the
evaporative coolers, cooling towers, and evaporative condensers.
CDD50  
m1
(To,m  50)
HDD65  
n1
(65  To,m)
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.39
4.40 TABLE 4.10 Climatic Conditions for the United States and Canada
Winter Summer
Annual Annual Annual
average daily cooling heating
Design incident degree- degree-
Mean wet- solar days days
daily bulb, energy on base base
Lat, Elevation, range, 1% east or west, 50°F, 65°F,
City State degree ft °F °F °F °F °F °F °F Btu / ft2  day CDD50 HDD65 MWS MWD
Albuquerque New Mexico 35.05 5,315 13 18 96/60 93/60 91/60 25.4 64 1,105 3,908 4,425 8 360
Anchorage Alaska 61.17 131 14 9 71/59 68/57 65/56 12.6 58 538 10,570 8 290
Atlanta Georgia 33.65 1,033 18 23 93/75 91/74 88/73 17.3 76 807 5,038 2,991 12 320
Atlantic
City CO New Jersey 39.45 66 8 13 91/74 88/73 86/72 18.1 76 9 310
Baltimore Maryland 39.18 154 11 15 93/75 91/74 88/73 18.8 76 739 3,709 4,707 10 290
Billings Montana 45.80 3,570 13 7 93/63 90/62 87/61 25.8 64 814 2,466 7,164 10 230
Birmingham Alabama 33.57 630 18 23 94/75 92/75 90/74 18.7 77 789 5,206 2,918 7 340
Bismarck North Dakota 46.77 1,660 21 16 93/68 90/67 86/66 26.5 70 766 2,144 8,968 7 290
Boise Idaho 43.57 2,867 2 9 96/63 94/63 91/62 30.3 64 916 2,807 5,861 6 130
Boston Massachusetts 42.37 30 7 12 91/73 87/71 84/70 15.3 74 659 2,897 5,641 17 320
Bridgeport Connecticut 41.17 16 8 12 86/73 84/72 82/71 14.1 74 2,997 5,537 14 320
Buffalo New York 42.93 705 2 5 86/70 84/69 81/68 17.7 72 609 2,468 6,747 12 270
Burlington Vermont 44.47 341 11 6 87/71 84/69 82/68 20.4 72 698 2,228 7,771 6 70
Caribou Maine 46.87 623 14 10 85/69 82/67 79/66 19.5 70 649 1,470 9,651 10 270
Casper Wyoming 42.92 5,289 13 5 92/59 89/58 86/58 30.4 61 961 2,082 7,682 9 260
Charleston South Carolina 32.90 49 25 28 94/78 92/77 90/77 16.2 79 796 6,188 2,013 7 20
Charleston West Virginia 38.37 981 6 11 91/73 88/73 86/71 19.1 75 667 3,655 4,646 7 250
Charlotte North Carolina 35.22 768 18 23 94/74 91/74 89/73 17.8 76 809 4,704 6 50
Chicago, Illinois 41.98 673 6 1 91/74 88/73 86/71 19.6 75 729 2,941 6,536 10 270
Cincinnati CO Ohio 39.10 482 5 12 93/74 90/75 88/73 20.0 76 3,733 4,988 9 260
Cleveland Ohio 41.42 804 1 6 89/73 86/72 84/71 18.6 74 2,755 6,201 12 230
Concord New Hampshire 43.20 344 8 2 90/71 87/70 84/68 24.1 73 630 2,087 7,554 4 320
Dallas/
Fort Worth Texas 32.90 597 17 24 100/74 98/74 96/74 20.3 77 875 6,587 2,259 13 350
Denver Colorado 39.75 5,331 3 3 93/60 90/59 87/59 26.9 63 971 2,732 6,020 6 180
Des Moines Iowa 41.53 965 9 4 93/76 90/74 87/73 18.5 76 788 3,371 6,497 11 320
Detroit Michigan 42.23 663 0 5 90/73 87/72 84/70 20.4 74 676 3,046 6,167 11 240
Honolulu Hawaii 21.35 16 61 63 89/73 88/73 87/73 12.2 75 953 9,949 0 5 320
Houston Texas 29.97 108 27 31 96/77 94/77 92/77 18.2 79 805 6,876 1,599 8 340
Indianapolis Indiana 39.73 807 3 3 91/75 88/74 86/73 18.9 77 692 3,453 5,615 8 230
Jackson Mississippi 32.32 331 21 25 95/77 93/76 92/76 19.2 79 833 5,900 2,467 7 340
Kansas City Missouri 39.32 1,024 1 4 96/75 93/75 90/74 18.8 77 3,852 5,393 10 320
Lansing Michigan 42.77 873 3 2 89/73 86/72 84/70 21.7 74 2,449 7,101 8 290
Las Vegas Nevada 36.08 2,178 27 30 108/66 106/66 103/65 24.8 70 1,136 6,745 2,407 7 250
Design
dry-bulb/
Design mean coincident
dry-bulb wet-bulb
99.6% 99% 0.4% 1% 2%
MWS/MWD
to DB
99.6%
4.41
Lincoln CO Nebraska 40.85 1,188 7 2 97/74 94/74 91/73 22.3 76 3,455 6,278 9 350
Little Rock Arkansas 34.92 312 16 21 97/77 95/77 92/76 19.5 79 831 5,299 3,155 9 360
Los Angeles CO California 33.93 105 43 45 85/64 81/64 78/64 10.9 69 962 4,777 1,458 6 70
Louisville Kentucky 38.18 489 6 12 93/76 90/75 88/74 18.2 77 727 4,000 4,514 10 290
Memphis Tennessee 35.05 285 16 21 96/78 94/77 92/77 16.8 79 806 5,467 3,082 10 20
Miami Florida 25.82 13 46 50 91/77 90/77 89/77 11.4 79 874 9,474 200 10 340
Milwaukee Wisconsin 42.95 692 7 2 89/74 86/72 83/70 16.6 74 724 2,388 7,324 13 290
Minneapolis/
St. Paul Minnesota 44.88 837 16 11 91/73 88/71 85/70 19.1 74 709 2,680 7,981
New Orleans Louisiana 29.98 30 30 34 93/79 92/78 90/78 15.5 80 838 6,910 1,513 7 340
New York New York 40.65 23 11 15 91/74 88/72 85/71 13.9 75 650 3,342 5,027 17 320
Norfolk Virginia 36.90 30 20 24 93/77 91/76 88/75 15.3 77 792 1,586 3,609 12 340
Phoenix Arizona 33.43 1,106 34 37 110/70 108/70 106/70 23 75 1,116 8,425 1,350 5 90
Pittsburgh Pennsylvania 40.50 1,224 2 7 89/72 86/70 84/69 19.5 73 642 2,836 5,968 10 260
Portland CO Oregon 45.60 39 22 27 90/67 86/66 83/64 21.6 67 647 2,517 4,522 13 120
Providence Rhode Island 41.73 62 5 10 89/73 86/71 83/70 17.4 74 677 2,743 5,884 12 340
Rapid City South Dakota 44.05 3,169 11 5 95/65 91/65 88/64 25.3 68 819 2,412 7,301 9 350
St. Louis Missouri 38.75 564 2 8 95/76 93/75 90/74 18.3 78 797 4,283 4,758 12 290
Salt Lake City Utah 40.78 4,226 6 11 96/62 94/62 92/61 27.7 65 975 3,276 5,765 7 160
San Antonio Texas 29.53 794 26 30 98/73 96/73 94/74 19.1 77 878 7,142 1,644 10 350
San Diego California 32.73 30 44 46 85/67 81/67 79/67 8.9 71 950 5,223 1,256 3 70
San Francisco California 37.62 16 37 39 83/63 78/62 74/61 16.7 63 941 2,883 3,016 5 160
CO
Seattle CO Washington 47.45 449 23 28 85/65 81/64 78/62 18.3 65 621 2,120 4,611 10 10
Shreveport Louisiana 32.47 259 22 26 97/77 95/77 93/76 19.1 79 843 6,166 2,264 9 360
Spokane Washington 47.62 2,461 1 7 92/62 89/61 85/60 26.1 63 758 2,032 6,842 7 50
Syracuse New York 43.12 407 3 2 88/72 85/71 83/70 20.3 73 611 2,399 6,834 7 90
Tucson Arizona 32.12 2,556 31 34 104/65 102/65 100/65 29.4 71 1,112 6,921 1,678 7 140
Tulsa Oklahoma 36.20 676 9 14 100/76 97/76 94/75 19.5 78 820 5,150 3,691 11 360
Washington DC 38.85 66 15 20 95/76 92/76 89/74 16.6 78 724 11 340
Wichita Kansas 37.65 1,339 2 8 100/73 97/73 94/73 22.2 76 4,351 4,791 13 360
Wilmington Delaware 39.68 79 10 14 91/75 89/74 86/73 17.0 76 726 3,557 4,937 11 290
Canada
Calgary Alberta 51.12 3,556 22 17 83/60 80/59 77/57 22.0 61 7 0
Montreal Quebec 45.47 118 12 7 85/71 83/70 80/68 17.6 72 7 250
Regina Saskatchewan 50.43 1,893 29 24 89/64 85/64 82/62 23.6 66 9 270
Toronto Ontario 43.67 568 4 1 87/71 84/70 81/68 20.2 72 9 340
Vancouver British Columbia 49.18 7 18 24 76/65 74/64 71/62 14.0 64 6 90
Winnipeg Manitoba 49.90 784 27 23 87/68 84/67 81/66 20.5 70 7 320
Note: CO designates locations within an urban area. Most of the data are taken from city airport temperature observations. Some semirural data are comparable to airport data.
Source: Abridged with permission from ASHRAE Handbook 1997, Fundamentals and ASHRAE/IES Standard 90.1-1999.
3. Outdoor weather data presented consecutively for a whole year of 8760 h, or other simplified
form, are sometimes used for year-round energy estimations.
4. Outdoor climate has a significant influence on the selection of an air conditioning system and its
components.
Outdoor Weather Characteristics and Their Influence
The following are characteristics of outdoor climate in the United States:
1. For clear days during summer, the daily maximum temperature occurs between 2 and 4 P.M. In
winter, the daily minimum temperature usually occurs before sunrise, between 6 and 8A.M.
2. The maximum combined influence of outdoor temperature and solar radiation on load calculations
usually occurs in July or August. January is often the coldest month in many locations.
3. The mean daily temperature range of the warmest month varies widely between different locations.
The smallest mean daily range, 10°F (5.6°C), occurs at Galveston, Texas. The greatest,
45°F (25°C), occurs at Reno, Nevada. Many cities have a mean daily range of 15 to 25°F (8.3 to
13.9°C). Usually, coastal areas have smaller mean daily ranges, and continental areas and areas
of high elevation have greater values.
4. Extremes in the difference between 1 percent design dry-bulb temperature in summer and the 99
percent design temperature in winter in the United States are
 A maximum of 119°F (66°C) at Bettles, Alaska
 A minimum of 18°F (10°C) at Kaneohe, MCAS, Hawaii
5. Snelling (1985) studied outdoor design temperatures for different locations all over the country.
He found that extremely cold temperatures may last for 3 to 5 days. Extremely hot temperatures
seldom last more than 24 h.
6. Among the 538 locations listed in Table 4.10, 138 locations have a summer mean coincident
wet-bulb temperature corresponding to 1 percent dry-bulb, Tos  68°F (20°C). There it is advantageous
to use evaporative cooling systems to replace part of or all the refrigeration capacity.
7. When commercial buildings are only occupied after 10 A.M. or even after 12 noon, a winter outdoor
design temperature higher than the 99.0 percent design dry-bulb temperature should be
considered.
REFERENCES
ANSI/ASHRAE, ANSI/ASHRAE Standard 55-1992, Thermal Environmental Conditions for Human Occupancy,
Atlanta, GA, 1992.
ASHRAE, ASHRAE Standard 62-1999, Ventilation for Acceptable Indoor Air Quality, Atlanta, GA, 1999.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
ASHRAE Evironmental Health Committee 1987, Indoor Air Quality Position Paper, Atlanta, GA, 1987.
Bayer, C. W., and Black, M. S., IAQ Evaluations of Three Office Buildings, ASHRAE Journal, July 1988,
pp. 48–52.
Bedford, T., and Warmer, C. G., The Globe Thermometer in Studies of Heating and Ventilation, Journal of the
Institution of Heating and Ventilating Engineers, vol. 2, 1935, p. 544.
Berglund, L., Mathematic Models for Predicting the Thermal Comfort Response of Building Occupants,
ASHRAE Transactions, 1978, Part I, pp. 735–749.
Cena, K., Spotila, J. R., and Avery, H. W., Thermal Comfort of the Elderly Is Affected by Clothing, Activity, and
Psychological Adjustment, ASHRAE Transactions, 1986, Part II A, pp. 329–342.
4.42 CHAPTER FOUR
Cummings, J. B.,Withers, Jr., C. R., Moyer, N. A., Fairey, P. W., and McKendry, B. B., Field Measurement of
Uncontrolled Airflow and Depressurization in Restaurants, ASHRAE Transactions, 1996, Part I, pp. 859–869.
DeBat, R. J., Humidity: The Great Equalizer, HPAC, no. 10, 1996, pp. 66–71.
Eade, R., Humidification Looming Larger in IAQ, Engineered Systems, no 1, 1996, pp. 49–58.
Fanger, P. O., Thermal Comfort Analysis and Applications in Environmental Engineering, McGraw-Hill, New
York, 1970.
Federal Supply Service, Federal Standard No. 209B, Clean Rooms and Work Station Requirements, Controlled
Environment, General Services Administration,Washington, 1973.
Gagge, A. P., Stolwijk, J. A. J., and Nishi,Y., An Effective Temperature Scale Based on a Simple Model of
Human Physiological Regulatory Response, ASHRAE Transactions, 1971, Part I, p. 247.
George, A. C., Measurement of Sources and Air Concentrations of Radon and Radon Daughters in Residential
Buildings, ASHRAE Transactions, 1985, Part II B, pp. 1945–1953.
Holzle, A. M., Munson, D. M., and McCullough, E. A., A Validation Study of the ASHRAE Summer Comfort
Envelope, ASHRAE Transactions, 1983, Part I B, pp. 126–138.
Janssen, J. E., Ventilation for Acceptable Indoor Air Quality, ASHRAE Journal, October 1989, pp. 40–46.
Kirkbride, J., Lee, H. K., and Moore, C., Health and Welfare Canada’s Experience in Indoor Air Quality Investigation,
Indoor Air ’90, vol. 5, D. S. Walkinshaw, ed., Ottawa: International Conference on Indoor Air and
Climate, 1990, pp. 99–106.
McCullough, E. A., and Jones, B. W., A Comprehensive Data Base for Estimating Clothing Insulation, IER
Technical Report 84–01, Kansas State University, 1984.
National Institute for Occupational Safety and Health, Indoor Air Quality: Selected Reference, Division of
Standards Development and Technology Transfer, Cincinnati, Ohio, 1989.
O’Sullivan, P., Energy and IAQ Can Be Complementary, Heating/Piping/Air Conditioning, February 1989,
pp. 37–42.
Persily, A., Ventilation Rates in Office Buildings, ASHRAE Journal, July 1989, pp. 52–54.
Rohles, Jr., F. H.,Woods, J. E., and Nevins, R. G., The Effect of Air Movement and Temperature on the Thermal
Sensations of Sedentary Man, ASHRAE Transactions, 1974, Part I, pp. 101–119.
Seppanen, O., McNall, P. E., Munson, D. M., and Sprague, C. H., Thermal Insulating Values for Typical Clothing
Ensembles, ASHRAE Transactions, 1972, Part I, pp. 120–130.
Snelling, H. J., Duration Study for Heating and Air Conditoning Design Temperature, ASHRAE Transactions,
1985, Part II, pp. 242–249.
Spain, S., The Upper Limit of Human Comfort from Measured and Calculated PMV Values in a National
Bureau of Standards Test House, ASHRAE Transactions, 1986, Part I B, pp. 27–37.
Sterling, E. M., Arundel, A., and Sterling, T. D., Criteria for Human Exposure to Humidity in Occupied Buildings,
ASHRAE Transactions, 1985, Part I B, pp. 611–622.
Thayer,W. W., Tobacco Smoke Dilution Recommendations for Comfort Ventilation, ASHRAE Transactions,
1982, Part II, pp. 291–304.
Woodcock, A. H., Moisture Transfer in Textile Systems, Textile Research Journal, vol. 8, 1962, pp. 628–633.
INDOOR AND OUTDOOR DESIGN CONDITIONS 4.43
CHAPTER 5
ENERGY MANAGEMENT AND
CONTROL SYSTEMS
5.1
5.1 AUTOMATIC CONTROL SYSTEM AND
HISTORICAL DEVELOPMENT 5.2
Air Conditioning Automatic Control
System 5.2
Historical Development 5.2
Distribution of Energy Management and
Control Systems 5.3
5.2 CONTROL LOOP AND CONTROL
METHODS 5.5
Control Loop 5.5
Sequence of Operations 5.5
Control Methods 5.7
Comparison of Control Methods
5.3 CONTROL MODES 5.9
Two-Position Control 5.9
Step Control and Modulating
Control 5.10
Floating Control 5.11
Proportional Control 5.11
Proportional plus Integral (PI)
Control 5.13
Proportional-Integral-Derivative (PID)
Control 5.14
Compensation Control, or Reset 5.15
Applications of Various Control
Modes 5.15
5.4 SENSORS AND TRANSDUCERS 5.16
Temperature Sensors 5.18
Humidity Sensors 5.18
Pressure Sensors 5.19
Flow Sensors 5.19
Carbon Dioxide and Air Quality
Sensors 5.20
Occupancy Sensors 5.20
Wireless Zone Sensors and Intelligent
Network Sensors 5.21
Transducers or Transmitters 5.21
5.5 CONTROLLERS 5.21
Direct-Acting or Reverse-Acting 5.21
Normally Closed or Normally Open 5.22
Pneumatic Controllers 5.22
Electric and Electronic Controllers 5.23
Direct Digital Controllers 5.23
5.6 WATER CONTROL VALVES AND
VALVE ACTUATORS 5.26
Valve Actuators 5.26
Types of Control Valves 5.27
Valve Characteristics and Ratings 5.27
Valve Selection 5.29
Valve Sizing 5.30
5.7 DAMPERS AND DAMPER
ACTUATORS 5.32
Types of Volume Control Dampers 5.32
Damper Actuators (Motors) 5.33
Volume Flow Control between Various
Airflow Paths 5.33
Flow Characteristics of Opposed- and
Parallel-Blade Dampers 5.35
Damper Selection 5.37
Damper Sizing 5.37
5.8 SYSTEM ARCHITECTURE 5.38
Architecture of a Typical EMCS with
DDC 5.38
System Characteristics 5.40
Future Development 5.41
5.9 INTEROPERABILIITY AND OPEN
PROTOCOL BACnet 5.41
Interoperability 5.41
BACnet—Open Data Communication
Protocol 5.41
Application Layer 5.42
Network Layer 5.43
Data Link/Physical Layer—Network
Technology 5.43
Connection between BACnet and
Proprietary Network 5.44
LonTalk Protocol 5.44
5.10 CONTROL LOGIC AND ARTIFICIAL
INTELLIGENCE 5.45
Fuzzy Logic 5.45
Knowledge-Based Systems and
Expert Systems 5.47
Artificial Neural Networks 5.51
5.11 PROGRAMMING FOR DDC
SYSTEMS 5.53
Evolution of DDC Programming 5.53
Graphical Programming 5.53
Templates 5.54
Graphical Programming for Mechanical
Cooling Control 5.55
5.12 TUNING DDC UNITS 5.55
Tuning PI Controllers 5.55
Self-Tuning PI and PID Controllers 5.55
Adaptive Control 5.56
5.13 FACTORS AFFECTING CONTROL
PROCESSES 5.56
Load 5.56
Climate Change 5.56
System Capacity 5.57
Performance of Control Processes 5.57
Thermal Capacitance 5.58
5.14 FUNCTIONAL CONTROLS 5.58
Generic Controls 5.59
5.1 AUTOMATIC CONTROL SYSTEM AND HISTORICAL
DEVELOPMENT
Air Conditioning Automatic Control System
An air conditioning automatic control system or simply a control system, primarily modulates
the capacity of the air conditioning equipment to maintain predetermined parameter(s) within an
enclosure or for the fluid entering or leaving the equipment to meet the load and climate changes at
optimum energy consumption and safe operation. The predetermined parameter or variable to
be controlled is called the controlled variable. In air conditioning or in HVAC&R, the controlled
variable can be temperature, relative humidity, pressure, enthalpy, fluid flow, contaminant concentration,
etc.
Because of the variation of the space load and the outdoor climate, a control system is one of the
decisive factors for an air conditioning, or an HVAC&R, system to achieve its goal: to effectively
control the indoor environmental parameters, provide an adequate amount of outdoor air, be energy-
efficient, and provide better security and safety. Today nearly all HVAC&R systems are installed
with a control system to provide effective operation and energy conservation.
Historical Development
Early controls for comfort air conditioning systems used in the Capitol building since 1928 were
pneumatic controls that included high- and low-limit thermostats in the supply air discharge of the
air-handling unit. These controls were used to maintain the desired supply air temperature. Dewpoint
control of the supply air maintained room humidity within a desirable range. A thermostat in
the return-air passage was used to control the air volume flow rate passing through the cooling coil,
and a 16-point recorder connected with resistance thermometers was used to record the room temperature
at various locations.
Shavit (1995) stated that the use of thermocouples offers the possibility of remote monitoring of
the space temperature. In coordination with pneumatic local control systems, the first centralized
monitoring system with a central panel and remote set point change was installed in the White
House in 1950. The concept of a building automation system (BAS) which centralizes the monitoring
management of building services was devised and developed in 1950s. The primary improvement
of the pneumatic control system occurred in 1963. The vapor and bulb-based sensors were
replaced by a remote pneumatic sensor using a bimetal strip and a nozzle flapper to regulate the
compressed air pressure between 3 and 15 psi (20.7 and 103.4 kPa).
Shavit also indicated that the first on-line computer was introduced in 1967 in One Main Place
in Dallas, Texas. This system had the first set energy conservation software including enthalpy control,
demand limit, optimum start/ stop, reset according to the zone of highest demand, chiller optimization,
and night purge. In 1970, the introduction of all-electronic, solid-state centralized control
system was another significant milestone. Solid-state components improved the scanning process
and serial transmission as well as reduced many wires in the trunk wiring to a single pair.
5.2 CHAPTER FIVE
Specific Controls 5.60
Commissioning and Maintenance 5.61
5.15 FAULT DETECTION AND
DIAGNOSTICS 5.61
Basics 5.61
Expert System Rule-Based
Diagnostics 5.62
ARX and ANN Model-Based
Diagnosis 5.63
5.16 CONTROLS IN ASHRAE/IESNA
STANDARD 90.1-1999 5.66
General 5.66
Off-Hour Controls 5.66
REFERENCES 5.67
The 1973 energy crisis greatly boosted the control industry. Many energy management systems
(EMSs) were installed for the purpose of saving energy. The microprocessor-based direct digital
control (DDC) was first introduced in 1981 and the DDC unit controller in 1986–1987. By means
of the incorporated hardware and software, it tremendously increased the control functions, developed
the control logic, and accelerated the data processing and analysis.
Distribution of Energy Management and Control Systems
A building automation system is a centralized monitoring, operation, and management system of the
building services in a building including air conditioning (HVAC&R), lighting, fire protection, and security.
Only air conditioning control systems and smoke control systems are covered in this handbook.
In the United States, a single-zone, two-position electric control system, as shown in Fig. 5.1, is
now widely used in residential and light commercial buildings that adopt DX packaged systems.
Single-zone means that the load characteristics of the whole conditioned space are similar, it is
controlled by one controller, and two-position means on-off control. This kind of control system is
simple, easy to manage, and low in cost. For a large high-rise building constructed in the 1990s, an
energy management and control system (EMCS), as shown in Fig. 5.2, is often the choice. Variable
air volume indicates that airflow is varied to match the variation of the system load. Today, an
EMCS is a microprocessor-based system with direct digital control which optimizes the operation
and the indoor environmental parameters of an HVAC&R system in order to maintain a healthy and
comfortable environment at optimum energy use. Today, an EMCS is an advanced-generation energy
management and control system and is also a part of a building automation system.
According to the surveyed results in Commercial Buildings Characteristics 1992 by EIA, the
percentages of the floor area served by the installed EMCS among the total conditioned area in
commercial buildings are as follows:
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.3
DX-coil
Compressor
Thermostat
Furnace
Condensing
unit
Refrigerant
pipes
Supply
fan
FIGURE 5.1 A single-zone, two-position electric control system.
Year built Building with EMCS, million ft2 Percentage of floor space
Before 1970 5431 18
1970–1979 3313 29
1980–1989 4343 36
1990–1992 1236 58
5.4 CHAPTER FIVE
3
1
1 1 5 6 4 2
3
P3
P1
M
SD
P
T
SD T2
T1
T4
2 1
T3
P2 DM4
DM2
DM3
MPS
DM1
Outdoor
air damper
Air-handler
Smoke detector
Pressure sensor
Temperature sensor
M
DM
MPS
Multicompensator
Damper actuator
Manual positioning switch
Relief
air damper
Relief
fan
14.0
70
60
50%
20%
13.5
50
40
40
0
S
S ru r
r ru
sf
sf cc
m
50 60 70
T, F
w, lb/lb
80 90 100
0.020
0
0.004
0.008
0.012
0.016
0
Cooling
coil
Outdoor air
reference
Conditioned
space
r
Recirculating air damper
Inlet
vanes
sf
S
ru
C
cc
FIGURE 5.2 Control diagram and psychrometric analysis of an energy management and control system for
a single-zone, VAV cooling packaged system.
The later the building is constructed and the larger the conditioned area of the commercial building,
the greater is the chance to install an energy management and control system (EMS).
5.2 CONTROL LOOP AND CONTROL METHODS
Control Loop
The basic element of a control system is a control loop. A control loop often consists of a sensor
(such as T2 in Fig. 5.2), which senses and measures the controlled variable of recirculating air; a
DDC unit controller which compares the sensed input signal with the predetermined condition (the
set point) and sends an output signal to actuate the third element, a controlled device or control element
(such as a damper or a valve in Fig. 5.2). Modulation, minute adjustments, or on-off control of
dampers and valves will change the controlled devices position or operating status, which affects
the controlled variable by changing the airflow and water flow, electric power supply, and so on.
The controlled variable is thus varied toward the predetermined value, the set point. Air and water
are control media or control agents. Airflow and water flow are the manipulated variables. The
equipment that varies the output capacity by changing the opening position of the dampers and
valves is called the process plant.
There are two types of control loops: open and closed. An open-loop system assumes a fixed
relationship between the controlled variable and the input signal being received. The sensed variable
is not the controlled variable, so there is no feedback. An example of an open-loop system
would be a ventilating fan that turns on when the outdoor temperature exceeds a specified set point.
The sensed variable is the outdoor air temperature, and the controlled variable is the state of the fan
(on or off).
A closed-loop system depends on sensing the controlled variable to vary the controller output
and modulate the controlled device. In Fig. 5.2, temperature sensor T2 senses the controlled
variable, the recirculating air temperature entering the air handler Tru; the DDC unit controller receives
this sensed signal input and produces an output according to the software stored to modulate
the fan speed, or the position of the inlet vanes of the supply fan. As the supply fan speed changes,
or inlet vanes open and close, the supply air volume flow rate varies and the space temperature Tr
and recirculating temperature Tru change accordingly. This change in Tru is sensed again by T2 and
fed back to the controller for further modulation of fan speed or inlet vanes to maintain values of
Tru that approach the set points. These components form a closed-loop system.
Figure 5.3 shows a block diagram of this closed-loop system. It shows a secondary input to the
controller, such as the outdoor air temperature To, which may reset the set point in the controller to
provide better and more economical control. The disturbances that affect the controlled variable
are load variations and changes in the outdoor weather. After the controller senses the signal feedback,
it sends a corrective signal to the controlled device based on the difference between the
sensed controlled variable and the set point. Thus, Tru is under continuous comparison and correction.
A control system or its component, control subsystem, used to control the controlled variable(s)
in a conditioned space, or within a mechanical device or equipment, may contain only one control
loop; or it may contain two or more loops.
Sequence of Operations
The sequence of operations is a description of the sequential order of the functional operations that
a control system is supposed to perform, which plans and guides the operation and control of an air
conditioning system. For a single-zone variable-air-volume (VAV) cooling packaged system shown
in Fig. 5.2, when cold air supply is required during full occupancy, the sequence of operations of
cooling mode is as follows:
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.5
 The supply fan is started and stopped by the scheduling software stored in the DDC unit
controller. Manual override is possible. When the time schedule puts this system in cooling
occupy mode, the microprocessor-based controller goes through a short initiation period, such as
a 2-min period. During this period, dampers are driven to fully open, minimum open, and fully
closed positions. These determine the effective range of the economizer potentiometer range. The
smoke detector in the return air or the low-temperature limit sensor of the mixed air will stop the
supply fan if necessary. The supply fan status (on or off) is determined by the pressure differential
switch across the fan. When the initiation period is completed, the supply fan is turned on.
 The control system tends to maintain the recirculating temperature Tru at around the cold set
point, and it uses the 100 percent all outdoor air free cooling economizer cycle as the first-stage
cooling. If the outdoor temperature To  Tru, the outdoor air damper is fully opened and the recirculating
air damper closed. If the outdoor air temperature To  Tru, the outdoor air damper is
closed to a minimum position to provide required outdoor ventilation air, and the recirculating
damper is fully opened.
 When the initiation period is completed, the supply fan is turned on from its zero speed. If Tru,
sensed by the temperature sensor T2, is at a value above the set point, Tru  Tc, set, also To 
Tru, the speed of the supply fan is gradually increased by the variable-speed drive inverter, resulting
in a higher supply volume flow rate. When the fan speed is raised to its upper limit, the
supply volume flow rate is then at its maximum value. A still higher space load further raises
Tru, and it exceeds the cold set point Tc, set, and if To  Tru, chilled water starts to flow to the
cooling coil to cool the air simultaneously with the outdoor air free cooling, in order to make
Tru  Tc, set.
 Only when the outdoor air damper is fully open and the static pressure difference between the
space air and outdoor air is greater than a preset value, such as ps – o  0.03 in. WC (7.5 Pa), will
the relief fan be energized. Its speed is modulated to maintain ps – o  0.03 in. WC (7.5 Pa).
 When Tru drops, the chilled water flow to the cooling coil is reduced first, and then the outdoor air
free cooling, outdoor air volume flow rate.
Refer to Sec. 21.2 for the details of the sequence of operations of single-zone VAV cooling
systems. In the design and operation of an EMCS system, the necessary documentation includes the
sequence of operation, control diagrams, specifications, operation, and maintenance manual.
5.6 CHAPTER FIVE
FIGURE 5.3 Block diagram for a closed-loop feedback control system.
Control Methods
According to the types of control signal and the different kinds of energy used to transmit the
signals, as well as whether a software is used during control operation, control methods can be
classified as direct digital, pneumatic, electric, and electronic.
Analog and Digital. There are two types of control signals: analog and digital. An analog signal
is in the form of a continuous variable. It often uses the magnitude of electric voltage or pneumatic
pressure to represent the air temperature. A digital signal is a series of on and off pulses used to
transmit information.
A conventional analog controller receives a continuous analog signal, such as a voltage or a
pneumatic signal, that is proportional to the magnitude of the sensed variable. The controller compares
the signal received from the sensor to the desired value (i.e., the set point) and sends a signal
to the actuator in proportion to the difference between the sensed value and the set point.
A digital controller, or microprocessor-based controller, receives an electric signal from
sensor(s). It converts the electric signal to digital pulses of different time intervals to represent the
signals values. The microprocessor of the digital controller performs the mathematical operations
and knowledge processing on these values. The output from the microprocessor can be either in
digital form to actuate relays or converted to an analog signal (say, a voltage or a pneumatic pressure)
to operate the actuator(s).
Direct Digital Control (DDC). A control system using DDC involves adopting a microprocessorbased
digital controller to perform mathematical operations and knowledge processing according to
the predetermined control algorithms or computer programs. The key element of DDC compared to
analog control is the software and hardware contained in the direct digital controller which expands
the control functions tremendously and adopts recently developed control logic. ADDC unit usually
has more precise sensors and uses the same type of controlled devices as other control methods.
Figure 1.2 shows an energy management and control system using DDC for an air-handling unit
in a typical floor of the NBC Tower, and Fig. 5.2 shows an EMS with DDC for a single-zone VAV
system.
Pneumatic Control. In a control system using pneumatic control, compressed air is used to operate
the sensors, controllers, and actuators and to transmit the signals. It consists of: a compressed
air supply and distribution system, sensors, controllers, and actuators. Figure 5.4 shows a typical
pneumatic control system. In Fig. 5.4, a filter is used to remove the dust particles, including submicrometer-
size particles, contained in the air. The function of the pressure-reducing valve is to reduce
the pressure of compressed air discharged from the air compressor to the required value in the
main supply line. The discharged compressed air is usually at a gauge pressure of 18 to 25 psig
(124 to 172 kPag), and the pressure signal required to actuate the valve or damper actuator is 3 to
13 psig (20.7 to 89.6 kPa g).
The advantages of pneumatic control are as follows:
 The compressed air itself is inherently a proportional control signal.
 The cost of modulating actuators is low, especially for large valves and dampers.
 Pneumatic controls require less maintenance and have fewer problems.
 Pneumatic controls are explosionproof.
The disadvantages stem mainly from comparatively fewer control functions, the high cost of
sophisticated pneumatic controllers, and the comparatively higher first cost of a clean and dry,
compressed air supply for small projects.
Electric or Electronic Control. Both types of control use electric energy as the energy source for
the sensors and controllers. A control system using electric control often offers two-position on-off
control, as shown in Fig. 5.1. Switches, relays, contactors, and electromechanical devices are
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.7
system components for electric control systems. They are generally used for low-cost, small, and
simpler control systems. In addition to the switches and relays, a control system using electronic
control has transistors, diodes, capacitors, and printed-circuit boards as system components. Electronic
control systems always have more accurate sensors and solid-state controllers with sophisticated
functions and can be easily interfaced with the building automation system. Electronic control
has a faster response and more accurate processing of data than electric control systems. Electronic
control systems cost more and need skilled personnel for maintenance and troubleshooting.
Comparison of Control Methods
Because of the increasing demand for more complicated controllers to satisfy the needs of better
indoor environmental control, satisfactory indoor air quality, improved energy saving, lower cost,
and greater reliability, the recent trend is to use EMCS with direct digital control for more demanding
and large projects. A modern DDC system consists of electronic sensors, microprocessor-based
controllers incorporated with electronic components, and electronic or pneumatic actuators. It is often
more effective and cheaper to operate a large pneumatic actuator than to use a large electronic
actuator in a DDC system.
An EMCS using DDC offers the following advantages:
1. Flexibility of providing required control functions and the ability to coordinate multiple functions
directly from complicated software programs (it can even mimic a human expert within a
certain knowledge domain and offer artificial intelligence)
2. More precise and faster-response control actions provided by the microprocessor-based controller
3. The possibility of using high-level self-checking and self-tuning system components, which
increases the system reliability
5.8 CHAPTER FIVE
FIGURE 5.4 A typical pneumatic control system.
One of the main disadvantages of DDC is a still-higher first cost. However, the cost per control
action is comparatively lower. As more and more DDC systems are installed, DDC will become
familiar to us. The cost of the microprocessor-based controllers and other DDC system components
will drop further in the future. In the 1990s, the trend in HVAC&R control for new and retrofit
projects is to use EMCS with DDC systems except for small projects.
5.3 CONTROL MODES
Control modes describe how the corrective action of the controller takes place as well as its effect
on the controlled variable. For applications in HVAC&R, control modes can be classified as twoposition,
step, floating, proportional, proportional-integral, and proportional-integral-derivative.
Before we discuss the control modes in detail, the term lag, or time lag, should be introduced.
According to ASHRAE terminology, lag is (1) the time delay required for the sensing element to
reach equilibrium with the controlled variable; or (2) any retardation of an output with respect to
the causal input, including the delay, because of the transport of material or the propagation of a
signal.
Two-Position Control
In two-position control, the controller controls the final control element at one of two positions:
maximum or minimum (except during the short period when it changes position). Examples of twoposition
control include starting and stopping the motor of a fan, pump, or compressor and turning
on or off an electric heater. Sometimes it is called on-off control.
Figure 5.5 shows a two-position control mode for an electric heater installed in a branch duct. In
the middle diagram, the ordinate indicates the controlled variable, the space temperature Tr; in the
lower diagram, the ordinate denotes the output capacity of the final control element, the electric
heater. If the controller turns on the electric heater when the sensor senses a space temperature
69.5°F (20.8°C), and turns off at 70.5°F (21.4°C), the result is a cyclic operation of the electric
heater and the rise and fall of Tr toward the two positions of 69.5°F (20.8°C) and 70.5°F (21.4°C).
The thermal storage effect of the electric heater, the branch duct, the building envelope that surrounds
the space, and the sensor itself will have a time lag effect on Tr. The rise of Tr is a convex
curve with an overshoot higher than 70.5°F (21.4°C), and the drop of Tr is a concave curve with an
undershoot lower than 69.5°F (20.8°C). The difference between the two points, on and off, is called
the differential.
If the heating capacity of the electric heater that results from an on-and-off cyclic operation is
represented by Qe,t, and if the actual space heating load , the slope of the rising Tr curve
will then be greater than the slope of the falling Tr curve. As a consequence, the “on” period will be
shorter than the “off” period. Such a cyclic Tr curve is shown in the upper left corner of Fig. 5.5. If
, the condition will be reversed (see the upper right corner of Fig. 5.5). The two-position
control varies the ratio of on and off periods to meet any variation in the space heating load.
To reduce the overshoot and undershoot of the controlled variable in a two-position control
mode, a modification called timed two-position control has been developed. A small heating
element attached to the temperature sensor is energized during on periods. This additional heating
effect on the sensor shortens the on timing. During off periods, the heating element is deenergized.
The differential of two-position control, the overshoot, and the undershoot all result in a greater
fluctuation of the controlled variable. A suitable differential is always desirable in two-position control
in order to prevent very short cycling, which causes hunting, a phenomenon of short cycling of
the controlled variable. Two-position control is not suitable for precise control of the controlled
variable, but it is often used for status control, such as opening or closing a damper, turning a small
single piece of equipment on or off for capacity control, etc., and for lower-cost control systems.
Qrh  1
2Qe,t
Qrh 	 1
2Qe,t
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.9
Step Control and Modulating Control
In step control, the controller operates the relays or switches in sequence to vary the output capacity
of the process plant in steps or stages. The greater the deviation of the controlled variable from the
set point and the faster the rate of change of the controlled variable, the higher will be the output
capacity of the process plant. In modulating control, the controller activates the control device continuously
and thus the change of the output capacity gradually.
During the earlier stage of step control, there must be a differential of controlled variable between
two on or two off points of a particular piece of equipment of the process plant. The result is a
greater fluctuation of the controlled variable. In the DDC controller-activated new generation of step
control, the software of the DDC unit determines how fast each capacity step is added or subtracted
according to the deviation from the set point and the rate of change of the controlled variable.
Figure 5.6 shows step control of the cooling capacity of refrigeration compressors using DDC to
maintain the discharge air temperature Tdis within predetermined limits in a packaged unit. When
the air economizer alone can no longer balance the DX coils load, and Tdis floats to the upper limit
of the control band, point 1, the first stage of cooling is energized. If the cycling and the continuously
energized first-stage cooling capacity still cannot balance the coils load, then Tdis continues to
rise until it reaches point 2, which is 1°F (0.56°C) higher than the upper limit of the control band.
The first-stage cooling is then locked on, and the second stage is energized to cycling. When Tdis
rises to point 2, all the present cooling stages will be locked on, and an additional cooling stage is
added for cycling. When Tdis drops to point 7, the current cycling cooling stage will be deenergized,
and the one cooling stage next to the deenergized stage will be cycling. It always needs a time
delay, say at least 4 min, to turn on or off again in order to prevent hunting and other damage.
5.10 CHAPTER FIVE
FIGURE 5.5 A typical two-position control mode.
Step control is another kind of on-and-off control having smaller varying capacity and thus a
fluctuating controlled variable. It has been widely used for the capacity control of refrigeration
compressors and the electric heaters.
Floating Control
In a floating control mode, the controller moves the control device by means of the actuator toward
the set point only when the control point is out of the differential, or dead band, as shown in
Fig. 5.7. A control point is the actual value of the controlled variable at a certain time instant.
Figure 5.7 represents a duct static pressure control system using a floating control mode by opening
and closing the inlet vanes of a supply fan. The control device—inlet vanes—can be moved in either
an opening or a closing direction depending on whether the control point is over the upper limit
of the differential or under the lower limit. In Fig. 5.7, when the controlled variable, the duct static
pressure ps, is above the upper limit of the differential, the controller then closes the inlet vanes. If
ps is below the lower limit of the differential, the controller opens the inlet vanes.
A floating control mode is more suitable for control systems with a minimal lag between the
sensor and the control medium. A control medium is the medium in which the controlled variable
exists.
Proportional Control
In a proportional control mode, the controller moves the controlled device to a position such that
the change in its output capacity is proportional to the deviation of the controlled variable from the
set point. The position of the controlled device is linearly proportional to the magnitude of the controlled
variable. Figure 5.8 represents the control of the space temperature through the modulation
of a two-way valve of a cooling coil using proportional control. In Fig. 5.8, the controlled variable
is the space temperature Tr, and the controlled device is the valve.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.11
FIGURE 5.6 DDC-activated discharge air temperature control for a VAV rooftop packaged system.
In a proportional control mode, the throttling range is the change in the controlled variable when
the controller moves the controlled device from the position of maximum output to the position of
minimum output. The controlled variable’s range of values that will move the proportional controller
through its operating range is called the proportional band. In a proportional control system,
the throttling range is equal to the proportional band.
5.12 CHAPTER FIVE
FIGURE 5.7 Floating control mode.
FIGURE 5.8 Proportional control mode.
The set point is the desired value of the controlled variable, or the desired control point that the
controller seeks to achieve. The difference between the control point and the set point is called
the offset, or deviation. In proportional control, when the controlled variable is at the bottom line of
the throttling range, the controller will position the actuator at the closed position. At the set
point, the actuator will be at 50 percent of the open position. When the controlled variable is at the
top of the throttling range, the actuator will be at 100 percent of the open position.
In a proportional control mode, since the output signal V of the controller is proportional to the
deviation of the control point from the set point, their relationship can be expressed as:
(5.1)
where Kp  proportional gain, proportional to 1 / (throttling range)
e  error signal, i.e., deviation or offset
M  output value when deviation is zero (usually, output value at middle of output range of
controller)
For space and discharge air temperature control using proportional control, the offset is directly
proportional to the space load and the coil load. The space or discharge temperature T, °F (°C), can
be calculated as:
(5.2)
where Tt,r  throttling range, expressed in terms of space or discharge temperature, °F (°C)
Tmin  space and discharge temperature when space load or coil load is zero, °F (°C)
In Eq. (5.2), Rload represents the load ratio of the space load or coil load, which is dimensionless and
can be evaluated as
(5.3)
The actual load and design load must be in the same units, such as Btu/h (W).
For a proportional control mode, a certain degree of offset, or deviation, is inherent. The position
of the controlled device is a function of the offset. Only when offset exists will the controller
position the actuator and the valve or damper at a position greater or smaller than 50 percent open.
The throttling range is primarily determined by the HVAC&R system characteristics and cannot be
changed after the system is designed. A proportional control mode is suitable for an HVAC&R system
that has a large thermal capacitance, resulting in a slow response and stable system and allowing
comparatively narrower throttling range and thus a smaller offset. Fast-reacting systems need a
large throttling range to avoid instability and short cycling, or hunting.
Proportional plus Integral (PI) Control
In a proportional plus integral control mode, a second component, the integral term, is added to the
proportional action to eliminate the offset. The output of the controller can thus be expressed as
(5.4)
where Ki  integral gain and t  time.
In Eq. (5.4), the second term on the right-hand side (the integral term) indicates that (1) the error,
or offset, is measured at regular time intervals and (2) the product of the sum of these measurements
and Ki is added to the output of the controller to eliminate the offset. The longer the offset
exists, the greater the response of the controller. Such a control action is equivalent to resetting the
set point in order to increase the controller output to eliminate the offset. As such, PI control is
sometimes called proportional plus reset control. Figure 5.9a shows the variations of a controlled
V  Kpe 	 Kie dt 	 M
Rload 
actual load
design load
T  RloadTt,r 	 Tmin
V  Kpe 	 M
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.13
variable for a proportional-integral control mode. In PI control, proportional-integral control signals
are additive.
For PI control, the controlled variable does not have any offset once it has achieved a stable condition,
except due to any inaccuracy of the instrumentation and measurement. Proper selection of
the proportional gain Kp and integral gain Ki as well as proper tuning is important for system stability
and control accuracy. PI control may be applied to fast-acting control systems with a greater
throttling range setting at the controller for better system stability, e.g., discharge air temperature,
discharge chilled water temperature, and duct static pressure control systems.
Proportional-Integral-Derivative (PID) Control
A PID control mode has additional control action added to the PI controller: a derivative function
that opposes to any change and is proportional to the rate of change. The output of such a controller
can be described by the following equation:
(5.5) V  Kpe 	 Kie dt 	 K
de
dt
	 M
5.14 CHAPTER FIVE
FIGURE 5.9 Proportional-integral and proportional-integral-derivative
control modes. (a) Proportional-integral; (b) proportional-integral-derivative.
where K is the derivative gain. The effect of adding the derivative function K de/dt is that the
quicker the control point changes, the greater the corrective action provided by the derivative function.
Figure 5.9b shows the variation of the controlled variable for a PID control mode. As with a PI
control mode, PID control mode also has no offset once the controlled variable has reached a stable
condition, except due to instrument inaccuracy.
Compared to the PI control mode, the PID control mode, which is a combination of proportional,
integral, and derivative actions, exhibits faster corrective action and a smaller overshoot and
undershoot following an offset and a change of the controlled variable. The controlled variable is
brought to the required set point in a shorter time. However, it is more difficult to determine properly
three constants, or gains (Kp, Ki, and K).
Compensation Control, or Reset
Compensation control, or reset, is a type of control mode in which a compensation sensor is generally
used to reset a main sensor to compensate for a variable change sensed by the compensation
sensor. The purpose is to achieve operation that is more effective, energy-efficient, or both.
In the design of a reset mode, the first things to decide are the control point at which the main
sensor will be reset and the variable to be sensed by the compensation sensor. The main sensor,
which senses the mixed air temperature in the air-handling unit, is usually reset by a compensation
sensor that senses the outdoor temperature, as shown in Fig. 5.10.
Another task to decide on is the relationship between the variables sensed by the main and compensation
sensors—the reset schedule. Many reset schedules have a different relationship between
these two sensors at various stages. For example, in Fig. 5.10, when the outdoor temperature To
is less than 30°F (1.1°C), it is within the range of stage I. No matter what the magnitude of To is,
the set point of the mixed temperature Tm is 65°F (18.3°C). When 30°F  To  95°F (1.1°C 
To  35°C), it is in stage II. The linear relationship between the temperatures sensed by the main
and compensation sensors in Fig. 5.10 can be expressed by the following equation:
(5.6)
When To  95°F (35°C), the set point of the mixed temperature is always 55°F (12.8°C).
Compensation control modes have been widely adopted to reset space temperature, discharged
air temperatures from air-handling units or packaged units, or water discharge temperatures from
central plants.
Applications of Various Control Modes
Selection of a suitable control mode depends on
 Operating characteristics
 Process or system characteristics, such as whether the thermal capacitance should be taken into
consideration
 Characteristics of load changes
If a simpler control mode can meet the requirements (say, two-position control versus PID control),
the simpler control mode is always the first choice. Except for two-position and step control, all the
other control modes are modulation controls. Modulation control is a control mode that is capable
of increasing or decreasing a variable according to the deviation from the required value in small
increments continuously. Also note that there is a significant difference between the two-position
control, step control, floating control, and proportional control modes with an offset and the offset
free PI and PID controls.
Tm  65 
65 55
95 30
(To 30)
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.15
For HVAC&R processes, PI control can satisfy most of the requirements, and it is most widely
used. Microprocessor-based PID control is a very powerful tool. PID control mode is more appropriate
for fast-acting duct static pressure control and airflow control, and it is recommended to set
the controller with a large proportional band for control system stability, a slow reset to eliminate
deviation, and a derivative action to provide a quick response.
5.4 SENSORS AND TRANSDUCERS
A sensor is a device that acts as a component in a control system to detect and measure the
controlled variable and to send a signal to the controller. A sensor consists of a sensing element and
accessories, as shown in Fig. 5.11. The term sensing element often refers to that part of the sensor
that actually senses the controlled variable. In HVAC&R systems, the most widely used sensors are
temperature sensors, humidity sensors, pressure sensors, and flow sensors. Recently, CO2 sensors,
air quality sensors, and occupancy sensors are being used in many new and retrofit projects. Electronic
sensors that send electric signals to electronic controllers, can also be used for the DDC
units, and the current trend is to use solid-state miniaturizing sensing elements. Electric output from
the sensors is usually expressed in 0 to 10 V dc or 4 to 20 mA. In the selection of sensors, accuracy,
sensitivity of response, reliability, long-term stability or drift, required calibration intervals, maintainability,
and especially the possibility of contamination by dust particles due to contact with the
control medium should be considered. Drift is a kind of offset. ASHRAE’s Terminology defines
drift as “change in an output-input relationship over time with the change unrelated to input, environment,
or load.” A calibration period of 1 year and longer is often considered acceptable.
Location of the sensor affects the sensor output directly. A sensor should be located in the critical
area where the controlled variable needs to be maintained within required limits. The sensing
element of an air sensor should be well exposed to the air, so that air can flow through the sensor
without obstruction. Air sensors should not be affected directly by the supply airstream nor should
space air sensors be affected by the outdoor airstream. Sensors should be shielded from radiation
5.16 CHAPTER FIVE
FIGURE 5.10 Reset control and schedule.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.17
FIGURE 5.11 (a) A temperature sensor and (b) a typical pitot tube flow-measuring station.
and mounted firmly on a structural member or duct wall, free from vibration. If a space air sensor
must be located on a concrete column, thermal insulation should be provided between the sensor
and the column.
Temperature Sensors
Temperature sensors fall into two categories: those that produce mechanical signals and those that
emit electric signals. Bimetal and rod-and-tube sensors that use sensing elements to produce a mechanical
displacement during a sensed temperature change either open or close an electric circuit in
an electric control system, as shown in Fig. 5.1, or adjust the throttling pressure by means of a
bleeding nozzle in a pneumatic control system. For a sealed bellow sensor, a change in temperature
causes a change in the pressure of a liquid in a remote bulb. The expansion and contraction of vapor
then move the mechanism of the controller.
Temperature sensors that produce electric signals, as shown in Fig. 5.11a, are the same as the
sensors for temperature measurement and indication mentioned in Sec. 2.3.
In addition to resistance temperature detectors (RTDs) and thermistors, sensors sometimes use
thermocouples. A thermocouple uses wires of two dissimilar metals, such as copper and constantan,
or iron and nickel, connected at two junctions, to generate an electromotive force between the junctions
that is directly proportional to the temperature difference between them. One of the junctions
is kept at constant temperature and is called the cold junction. Various systems have been developed
to maintain the cold junction at a constant temperature and to provide compensation if the cold
junction is not at 32°F (0°C). This task makes the use of thermocouples more complicated and
expensive. The electromotive force produced between the two junctions can be used as the signal
input to a controller.
Bimetal and rod-and-tube temperature sensors are simple and low in cost. However, they cannot
provide temperature indication and electric signals for DDC and are often used in electric control
systems. Platinum and nickel RTDs are stable, reliable, and accurate. They are very expensive compared
to thermistors and need calibration to compensate for the effects of having external leading
wires. RTDs are widely adopted in DDC for commercial applications. For a project that needs
precise temperature control, RTD is often the choice. High-quality thermistors can provide stable,
reliable, and interchangeable temperature sensors, and are also widely used in many commercial
applications. Petze (1986) reported that some high-quality thermistors exhibited better than

0.002°F (
0.001°C) stability for a 2-year period.
A typical space air temperature sensor has the following characteristics:
Sensing element Platinum film element, 3000 
Positive temperature coefficient, 4.8 /°F
(8.64 /°C)
Operating range 60 to 90°F (15.6 to 32.2°C)
Humidity Sensors
As mentioned in Sec. 2.8, humidity sensors fall into two categories: mechanical and electronic.
When the same humidity sensor is used for both monitoring and DDC, the ion-exchange resistancetype,
also called the bulk polymer resistance-type, and capacitance-type humidity sensors are often
adopted. Humidity sensors usually have different accuracy at very low, 20 to 80 percent mid-range,
and high relative humidities. The bulk polymer resistance humidity sensor is not accurate at low
relative humidities and generally provides stable and accurate readings within a range of 30 to 90
percent relative humidity. Its performance is affected by the air contamination. Capacitance humidity
sensors are accurate at 10 to 80 percent relative humidity. However, they become unstable at
high relative humidities.
5.18 CHAPTER FIVE
In an EMCS, an electronic humidity sensing element measures the relative humidity and sends
an electric signal to the DDC unit. A typical space air humidity sensor has the following specifications:
Power source 24 V ac
Humidity range (active) 10 to 80 percent
Nominal output range 0 to 5 V dc
Accuracy at 70°F (21.1°C) 
3 percent at 10 to 60 percent range

4 percent at 60 to 80 percent range
Speed of response 8 min (90 percent of response time)
Mounting Wall mount
Pressure Sensors
A pressure sensor usually senses the difference in pressure between the controlled medium (air and
water) and a reference pressure; or the pressure differential across two points, such as the pressure
differential across a filter. The reference pressure may be an absolute vacuum, atmospheric pressure,
or the pressure at any adjacent point. Output signals from the pressure sensors may be electric
or pneumatic, and analog or binary.
Pressure sensors used for HVAC&R systems can be divided into high-pressure and low-pressure
sensors. High-pressure sensors measure in pound per square inch or feet WC (kilo pascals), and
low-pressure sensors measure in inches WC (pascals). The sensing elements for high-pressure sensors
are usually Bourdon tube, bellows, and sometimes diaphragms. For low-pressure air sensors,
large diaphragms or flexible metal bellows are usually used.
To measure the duct static pressure, if a long section of straight duct of a length greater than
10 ft (3 m) is available, a single-point, pitot-tube type of duct static pressure sensor can be used.
Otherwise, a multipoint pitot-tube array or flow-measuring station with airflow straighteners should
be used. The small hole used to measure static pressure should never be directly opposite to an
airstream with a velocity pressure that can affect the reading. A reference pressure should be picked
up at a point with low air velocity outside the duct, at a point served by the same air system, or in
the ceiling plenum.
In a DDC system, low air pressure is often sensed by measuring the capacitance of two
diaphragms; one is allowed to move toward or away from the fixed one, depending on the pressure
on two sides of the diaphragm. An electric circuit is used to convert the capacitance to a voltage or
milliampere signal. The other method used to sense the low air pressure in DDC systems works because
the air pressure differential compresses or stretches a diaphragm as well as a strain gauge.
The change of the resistance of the gauge is detected and amplified, and this electric signal in the
form of voltage or milliamperes is sent to the controller.
A space air pressure sensor having an operating range of 	0.1 to 0.1 in WC (	25 Pa to 25
Pa) detects the resistance of a silicon diaphragm, and thus the space air pressure is then measured.
Stainless steel, rubber, etc., can also be used as the material of the diaphragms. A space air pressure
sensor should be located in an open area of the conditioned space where the air velocity is less than
40 fpm (0.2 m/s) and where its reading is not affected by the opening of the doors. The reference
pressure pickup is best located at the rooftop, at a level 10 ft (3 m) above the building to avoid the
influence of wind.
Flow Sensors
Flow sensors usually sense the rate of air flow and water flow in cfm (L/s) for air and gpm (L/ s)
for water. For airflow sensors, the average velocity pressure pv, in. WC (Pa), is often sensed and
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.19
measured. The average air velocity va, fpm, is then calculated as
(5.7)
where K is the flow coefficient, which depends on the type of pitot-tube array used and the dimension
of the round or rectangular duct. After that, the volume flow rate can easily be determined as
the product vaA. Here, A represents the cross-sectional area of the air passage, perpendicular to the
airflow, in ft2 (m2). Various forms of pitot-tube array have been developed and tested to determine
the average velocity pressure of a rectangular or circular duct section by measuring the difference
between the total and static pressures of the airstream. A typical pitot-tube flow-measuring station
with flow straighteners to provide more even airflow is shown in Fig. 5.11.
Electronic air velocity sensors such as hot-wire anemometers and thermistors have also been
widely adopted to measure airflow, especially for variable-air-volume (VAV) boxes. Thermistors are
cheaper than hot-wire anemometers. Heated thermistors need periodic calibration.
Water flow sensors of the differential-pressure type, such as orifice plates, flowing nozzles, and
pitot tubes, have only a limited measurement range. Turbine or magnet-type flowmeters can apply
to a wider range, but they are more expensive and also need periodic calibration.
Carbon Dioxide and Air Quality Sensors
A CO2 sensor detects and indicates the amount of carbon dioxide (CO2 ) contained inside the air,
which is a reliable indication of the body odor released by the occupants of the conditioned space.
Because of a certain relationship that exists between the CO2 contained in the outdoor air and
the CO2 released by the occupants, the concentration of CO2 in space air may sometimes be used as
a rough indication of amount of outdoor ventilation air supply to the conditioned space under the
specific conditions.
The sensing process used in CO2 sensors includes potentiometric and amperometric electrochemical
cells, an infrared detector, etc. A CO2 sensor usually has an operating range of 0 to 3000
ppm. For measurements within an accuracy of 
100 ppm, recalibration may be required on the
order of once per year.
An air quality sensor, also called a volatile organic compound (VOC) or a mixed-gas sensor, is
used to monitor and detect the relative concentrations of VOC or mixed gas, or total concentration
including acetone, ammonia, CO, CO2, SO2 , chlorine, formaldehyde, CFC-11, CFC-12, etc. The
concentration of the contaminant is often expressed in tested and commissioned units, such as 0 to
5 or 0 to 10 units.
The sensing element has a tin dioxide surface which is heated to a temperature above 130°F
(54°C). The change of the conductivity and thus its resistance are then amplified and fed to the controller
in terms of 0 to 10 V dc in proportion to the contamination. Air quality sensors are less
expensive than CO2 sensors and need less maintenance. The drift of an air quality sensor is unpredictable.
Both CO2 sensors and air quality sensors are now used for demand-based ventilation control to
provide the required amount of outdoor air for acceptable indoor air quality.
Occupancy Sensors
An occupancy sensor detects whether a room is occupied by occupant(s). As a result, the HVAC&R
and lighting in this room can be turned off when the room is not occupied, to save energy. There are
two types of occupancy sensors: ultrasonic and infrared. An ultrasonic occupancy sensor sends out
a rather lower level of ultrasonic wave and detects movement of the occupant when there are
changes in receiving patterns. Ultrasonic sensors sweep over the area that surrounds the sensor.
False signals may be received due to the motion of the papers or air movement from a diffuser. The
sensitivity of the ultrasonic sensor should be adjusted to avoid these problems.
va  4005K ?pv
5.20 CHAPTER FIVE
An infrared occupancy sensor senses the movement of the occupant(s) as the sensor receives
heat from the occupant when that person is moving. Infrared sensors need to see the occupant. As
long as there is no obstruction between the sensor and the occupant in the room, the sensing process
of an infrared sensor is effective. Occupancy sensors are often mounted on the ceiling or at a high
level on a wall. They are used in the guest rooms of the hotels and motels and other commercial
buildings. Ultrasonic occupancy sensors show a wider acceptance than infrared occupancy sensors.
Wireless Zone Sensors and Intelligent Network Sensors
In an HVAC&R system, a wireless zone sensor is an indoor spread spectrum radio-frequency transmitter
which sends room temperature and other status information to a local receiver located no
more than 1000 ft (300 m) away. A translator converts these data and sends them to a variable-airvolume
DDC unit controller through a wired communications link. This type of new technology
was developed recently for the sake of providing flexibility for the rearrangements of the office
layout.
Kovacs (1996) noted that intelligent network sensors “. . . can linearize the sensor signal, accept
an offset adjustment through the network, may have alarm- or decision-making algorithms on
board, and include self-test diagnostics to continuously validate performance.” An intelligent network
sensor may have a sensing element, an analog/ digital transducer, a neutron chip—an onboard
microprocessor, and a transceiver. Dew-point temperature, enthalpy, wet-bulb temperature, etc. can
be provided as a combined variable. An intelligent network sensor will help to drive the system
structure towards becoming a more cost-effective distributed control system in the future, a system
that has more locally processed signal and data.
Transducers or Transmitters
A transducer is a device that converts energy from one form to another or amplifies an input or output
signal. In HVAC&R control systems, a transducer may be used to convert an electric signal to a
pneumatic signal (E/P transducer), e.g., a pneumatic proportional relay that varies its branch air
pressure from 3 to 15 psig (20 to 103 kPa g) in direct proportion to changes in the electrical input
from 2 to 10 V. Also E/P transducers are used between microprocessor-based or electronic controllers
and pneumatic actuators. However, a pneumatic signal can be converted to an electric signal
in a P /E transducer. For example, a P/ E relay closes a contact when the air pressure falls and opens
the contact when the air pressure rises above a predetermined value. A transmitter is used to transmit
a signal, pneumatic or electric; through air, water, or other fluids. A sensor is also a transmitter.
However, the difference between them is that a sensor only senses the signal of the controlled variable
and transmits it.
5.5 CONTROLLERS
A controller receives input from the sensor, compares with the set point or implemention based on its
stored computer software, sends an output to or modulates the control device for maintaining a desirable
indoor environment. A thermostat is a combination of a temperature sensor and a temperature
controller, whereas a humidistat is a combination of a humidity sensor and a humidity controller.
Direct-Acting or Reverse-Acting
A direct-acting controller increases its output signal upon an increase in the sensed controlled variable,
and it decreases its output signal upon a decrease in the sensed controlled variable. Conversely,
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.21
a reverse-acting controller decreases its output signal upon an increase in the sensed controlled
variable and increases its output signal upon a decrease in the controlled variable.
Normally Closed or Normally Open
A controlled device, a valve or a damper, that is said to be in the normally closed (NC) position indicates
that when the input signal to the controller falls to zero or below a critical value, the control
device will be in the closed position. The closing of the valve or damper is most likely due to the
action of a spring or a supplementary power supply. A valve or a damper that is said to be normally
open (NO) indicates that when the input signal to the controller falls to zero or below a critical
value, then the valve, damper, or associated process plant will be in the open position.
Pneumatic Controllers
In a pneumatic controller, the basic mechanism used to control the air pressure in the branch line
supplied to the actuators is a nozzle-flapper assembly plus a restrictor. Figure 5.12 shows a pneumatic
controller with such a mechanism. Compressed air supplied from the main line flows through
the restrictor and discharges at the opening between the nozzle and the flapper, which has a spring
pulling it downward. The nozzle and restrictor are sized in such a manner that when the flapper
moves away from the nozzle, all the air escapes from the nozzle and the branch line pressure is
zero. When the flapper covers the nozzle or if there is no airflow in the branch line, the pressure of
the branch line is equal to that in the main line.
If a sensor, such as a bimetal sensor, moves the flapper upward according to the magnitude of
the controlled variable sensed by the sensor, then the input signal from the sensor determines
the opening between the nozzle and flapper and hence the compressed air pressure in the branch
line.
5.22 CHAPTER FIVE
FIGURE 5.12 A pneumatic controller with a nozzle-flapper assembly
and a restrictor.
For a direct-acting pneumatic temperature controller used to control the space temperature
during summer, the branch line pressure may change from 9 to 13 psig (62 to 89 kPag) when the
space temperature increases from 73 to 77°F (22.8 to 25°C). This nozzle-flapper assembly pneumatic
controller operates in proportional control mode. Many other more complicated pneumatic
controllers have been developed to perform other control modes and additional functions.
Electric and Electronic Controllers
An electric controller uses switches, relays, and a bridge circuit formed by potentiometers to position
the actuators in on-off, floating, and proportional control modes according to the input signal
from the sensor and the predetermined set point. An electronic controller can provide far more
functions than electric controllers can. It may receive input signals from both the main sensor and
the compensation sensor with amplification and combination.
In the control circuit, an electronic controller basically provides proportional or proportionalintegral
control modes. The output signal from the controller can be used to position an actuator or
to provide the sequencing of actuators, or to change to two-position, floating, or even PID control
modes in conjunction with additional circuits.
Direct Digital Controllers
A direct digital controller has a microprocessor to implement computer programs to provide various
control functions. In DDC units, there are analog-to-digital (A/D) and digital-to-analog (D/A)
converters to convert analog input to digital signals for processing, or to convert digital signals to
analog for actuators, if necessary. DDC units are stand-alone and microprocessor-based controllers.
Stand-alone means that the controller has sufficient capacity to execute the assigned control
functions alone. Today, there are mainly two types of DDC units: system controllers and unit controllers.
System Controllers. A system controller, also called a stand-alone panel (SAP), as shown in
Fig. 5.13, has the ability to coordinate communications between system controllers, between the
system controller and the personal computer (PC) in the workstation, and between the system controller
and the supported unit controllers. It also has the ability to provide and execute the control
programs for functional control, and to store user databases and trend log values. A system controller
can support 50 to 200 unit controllers on separate unit controller trunk(s).
Unit Controllers. A unit controller, also called a terminal controller, is shown in Fig. 5.14. It usually
has limited capacity to execute factory-loaded computer programs and to provide functional
control for a terminal or a piece of HVAC&R equipment. Unit controllers are often connected in a
separate network and supported by a system controller. The new-generation unit controllers have
greater memory to handle complicated control programs, and they provide time and calendar scheduling,
data storage, and other functions, such as limited programming.
Hardware. Many system controllers have only a single printed-circuit board. Single-board configuration
often offers lower first cost. Its disadvantage is that any component failure requires the
replacement of the complete board. Another approach is that the controller is made from various
modules. The modular approach isolates the component failures and plugs on the single-module
replacement quickly and inexpensively.
Memory. The types of memory included in DDC units are as follows:
 Read-only memory (ROM), which stores the software provided by the manufacturer and should
not be modified by the user.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.23
 Random-access memory (RAM), which stores custom control software developed during installation
or prepared by the user. This type of memory is volatile (i.e., it can be read from and written to)
and requires battery backup.
 Electric erasable programmable read-only memory (EEPROM), which stores custom control software
and is volatile. The advantage of EEPROM over RAM is that EEPROM does not need
5.24 CHAPTER FIVE
Universal analog
inputs;
Independent power
supplies for each
electronic module;
Tool-less installation;
Onboard diagnostic
displays and indications.
Context-sensitive
touchpad;
user interface.
Wireway space
Fully integrated
output relays,
overrides and
transducers;
FIGURE 5.13 A typical system controller. (Source: Johnson Controls. Reprinted by permission.)
battery backup. However, EEPROM cannot be used for as many writing, erasing, and rewriting
cycles as RAM.
 Flash erasable programmable read-only memory (flash EPROM), which is a kind of new memory
technology and allows the stored control software to remain untouchable indefinitely without
power.
System controllers often use a 16-bit microprocessor. A typical system controller has the following
memories:
RAM: 256 Kbytes
ROM: 128 Kbytes EEPROM
512 Kbytes flash EPROM
Input/Output (I/O). There are four kinds of I/O: analog input (AI); digital or binary input (BI);
analog output (AO); and pulsed or binary output (BO). Conventionally, a sensor input, a controller
output, or a control value, is referred to as a point or object. Current system controllers allow their
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.25
FIGURE 5.14 A typical unit controller. (Source: Honeywell Inc. Reprinted by
permission.)
input and output connections to be configured with great flexibility. Each input or output point can
be either analog or digital. Typical analog inputs (electric signal) are 0 to 10 V dc or 4 to 20 mA. A
system controller often has a total of 20 to 50 I /O points. Some system controllers can be extended
to 100 points if necessary. A typical system controller has the following I/O points capacity:
Analog/ digital inputs 18
Universal analog / digital input /outputs 6
Digital outputs 12
Totalizer inputs, i.e., pulsed inputs 4
A unit controller or a terminal controller usually has 4 to 20 I /O points. The I/O points in a typical
unit controller may take the following forms:
Analog inputs 0 to 10 V dc
Digital inputs Switch, relay, and transistor
Microbridge sensor 0 to 3 in. WC pressure differential (0 to 750 Pa)
Analog outputs 0 to 10 V dc, 4 to 20 mA
Digital outputs 30 V ac
Pneumatic 3 to 15 psig (20.6 to 103 kPag)
Triac On/off output for electric heater, fan motor, etc.
5.6 WATER CONTROL VALVES AND VALVE ACTUATORS
Water valves are used to regulate or stop water flow in a pipe either manually or by means of automatic
control systems. Water control valves adopted in water systems can modulate water flow rates
by means of automatic control systems.
Valve Actuators
An actuator, sometimes called an operator, is a device which receives an electric or pneumatic analog
control signal from the controller, either directly or through a digital-to-analog converter. It then
closes or opens a valve or damper, modulating the associated process plant, and causes the controlled
variable to change toward its set point. Valve actuators are used to position control valves.
They are mainly of the following types:
Solenoid Actuators. These use a magnetic coil to move a movable plunger connected with the
valve stem. Most solenoid valve actuators operate at two positions (on and off). They are used
mainly for small valves.
Electric Actuators. These move the valve stem by means of a gear train and linkage. Different
electric motor valve actuators can be classified according to the control mode they use:
1. On/off mode. For this type of actuator, the motor moves the valve in one direction, and when the
electric circuit breaks, the spring returns the valve stem to the top position (either open or closed
position depending on whether it is a normally open or closed valve).
2. Modulating mode. The motor can rotate in both directions, with spring return when the electric
circuit breaks.
3. Modulating mode with supplementary power supply. The motor rotates in two directions and
without a spring-return arrangement. When the power is cut off, a bypass signal is usually sent
5.26 CHAPTER FIVE
to the electric motor to drive the valve to its open or closed position, depending on whether it is
a normally open or closed valve.
It may take minutes to fully open a large valve using an electric motor valve actuator. Modern
electronic actuators use solid-state control boards to determine the speed, the action, and other
functions to meet more demanding requirements.
Pneumatic Actuators. A pneumatic valve actuator consists of an actuator chamber whose bottom
is made of a flexible diaphragm or bellows connected with the valve stem. When the air pressure in
the actuator chamber increases, the downward force overcomes the spring compression and pushes
the diaphragm downward, closing the valve. As the air pressure in the actuator chamber decreases,
the spring compresses the diaphragm, moving the valve stem and valve upward. A pneumatic valve
actuator is powerful, simple, and fast to respond.
Because of the increasing popularity of the DDC systems, there is an increasing demand for
electric actuators that can be interfaced with a DDC system.
Types of Control Valves
Water control valves consist mainly of a valve body, one or two valve disks or plugs, one or two
valve seats, a valve stem, and a seal packing. Based on their structure, water control valves can be
classified into the following types:
1. Single-seated. A single-seated valve has only a single valve disk and seat, as shown in Fig. 5.15b
and c. It is usually used for water systems that need a tight shutoff.
2. Double-seated. A double-seated valve has two valve disks connected to the same valve
stem and is designed so that the fluid pressure exerted on the valve disks is always balanced.
Consequently, less force is required for the operation of a double-seated valve, as shown in
Fig. 5.15a.
3. Butterfly. A butterfly valve consists of a cylindrical body, a shaft, and a disk that rotates on an
axis, as shown in Fig. 5.15d. When the valve closes, it seats against a ring inside the body. A
butterfly valve exhibits low flow resistance when it is fully opened. It is compact and is usually
used in large water pipes.
According to the pattern of the water flow, water control valves can again be classified as twoway
valves or three-way valves. A two-way valve has one inlet port and one outlet port. Water flows
straight through the two-way valve along a single passage, as shown in Fig. 5.15a.
In a three-way valve, there are three ports: two inlet ports and one common outlet port for a
three-way mixing valve, as shown in Fig. 5.15b, and one common inlet port and two outlet ports for
a three-way diverting valve, as shown in Fig. 5.15c. In a three-way mixing valve, the main water
stream flows through the coil or boiler, and the bypass stream mixes with the main stream in the
common mixing outlet port. In a three-way diverting valve, the supply water stream divides into
two streams in the common inlet port. The main water stream flows through the coil, and the bypass
stream mixes with the main water stream after the coil. A three-way mixing valve is always located
downstream of the coil. But a diverting valve is always located upstream of the coil.
A diverting valve should never be used as a mixing valve. The unbalance pressure difference between
the two inlet ports and the outlet port at a closed position may cause disk bouncing and valve
wear when the valve disk travels between the two extremes.
Valve Characteristics and Ratings
The different types of control valves and the characteristics that are important during the selection
of these valves are as follows:
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.27
1. Equal-percentage valve. This control valve changes the water flow rate by a certain percentage
for that same percentage of lift in the valve stem when the upstream versus downstream water
pressure difference across the valve (its pressure drop) is constant.
2. Linear valve. This control valve shows a directly proportional relationship between the flow rate
and the lifting of the valve stem for a constant pressure drop.
5.28 CHAPTER FIVE
FIGURE 5.15 Various types of control valves: (a) Double-seated two-way valve; (b) single-seated three-way mixing valve; (c) singleseated
three-way diverting valve; (d) butterfly valve.
3. Quick-opening valve. This control valve gives the maximum possible flow rate when the valve
disk or plug is just lifted from its seat.
Rangeability is defined as the ratio of the maximum flow rate to the minimum flow rate under
control. An equal-percentage valve may have a very good rangeability of 50 : 1. A linear valve may
have a rangeability of 30 :1.
The flow characteristics of equal-percentage, linear, and quick-open valves are shown in
Fig. 5.16.
The following control valve ratings should be considered during the selection and sizing of a
valve:
1. Body rating. The nominal body rating of the valve is the theoretical rating of the valve body
only, in psig. The actual body rating is the permissible safe water pressure for the valve body, in
psig (kPa g), at a specific water temperature.
2. Close-off rating. That is the maximum pressure difference between the inlet and outlet ports that
a valve can withstand without leakage when the valve is fully closed, in psi (kPa).
3. Maximum pressure and temperature. These are the maximum pressure and temperature of water
that the whole valve, including body, disk, seat, packing, etc., can withstand.
Valve Selection
Proper selection of water control valves depends on water system performance, load variations, pipe
size, control modes, etc. Today, the use of scaling factors on analog outputs in a DDC system permits
a nonlinear device to provide an output of linear response. However, select a control valve having a
linear relationship between a change in the controlled variable and the amount of travel of the valve
stem, or a linear system control characteristic over the operating range is still desirable when it is costeffective.
Hence, a linear valve is often used for the water system for which the controlled variable has
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.29
FIGURE 5.16 Typical flow characteristics of various types of control
valve.
a linear relationship with the water flow or valve opening; or for applications that do not have wide
load variations.
When a control valve is used to modulate the water flow rate of a hot or chilled water coil, a
large reduction in the flow rate causes only a small reduction in the heating or cooling output of the
coil. Given such circumstances, the nonlinear behavior of an equal-percentage valve combined with
the nonlinear output performance of a hot or chilled water coil will provide more linear system
behavior.
When three-way valves are used, the water flow rate before or after the common port is approximately
constant, no matter how wide the openings of the various ports in the three-way valves. As
such, three-way valves are used in constant or approximately constant water flow rate systems, even
the coil load changes.
As a two-way valve closes, the flow rate of the water system decreases and its pressure drop
across the two-way valve increases. A two-way valve is thus used for water systems that have variable
volume flow during a variation in system load, as shown in Fig. 5.17.
Valve Sizing
The size of a control valve affects the controllability of a water system. If a control valve is oversized,
then the smallest increment possible may overshoot the controlled variable. An undersized
control valve needs great pumping power. The size of a control valve is also closely related to its
5.30 CHAPTER FIVE
FIGURE 5.17 A typical chilled water system.
design water flow rate , gpm (L/ s), and the pressure drop across the valve pvv, psi (kPa), when
the control valve is fully opened. Their relationship can be expressed as
(5.8)
where Cv is the flow coefficient for a flow rate of 1 gpm at a pressure drop of 1 psi. The flow coeffi-
cient of control valves can be found in manufacturers’ catalogs.
For a water system that has several cooling coils between supply and return mains, as shown in
Fig. 5.17, modulating the water flow is effective only when the opening and closing of the control
valve V1 in piping section EH has a significant effect on the change in the pressure drop pEH
across piping section EH between the supply and return mains. As control valve V1 opens wider
and pEH decreases, if there is no concurrent change in flow resistance of piping sections JM and
NQ, then a greater water flow will pass through piping section EH. If pEH increases, less water
will flow through EH. If pEH, pJM, and pNQ all increase, based on the characteristic curve of the
centrifugal pump, the water flow rate through pumps P2 and P3 will decrease for an increase in the
system head. If pEH, pJM, and pNQ all drop, the water flow rate through pump P2 or P3 will
increase.
Because of the use of DDC systems with PI or PID control modes and the variable-speed pumping
systems, as well as the results of the previous analysis the following hold:
1. The size of a control valve should be determined according to the flow coefficient calculated
from Eq. (5.8) and listed in the manufacturer’s catalog.
2. The assumed pressure drop pvv across the control valve should be appropriate. It is affected by
the type of control valve used. It is also a compromise between a higher pvv value to provide
desirable controllability and a lower pvv value to save energy.
3. For a water system using variable-speed pumping and DDC systems with PI or PID control
modes, a pressure drop across an equal-percentage control valve pvv  5 to 10 ft WC (15 to
30 kPa) and a rangeability of 30 : 1 or greater is recommended for energy saving.
Example 5.1. A water system using variable-speed pumping with DDC supplies chilled water to
the cooling coils of three air-handling units (AHUs), as shown in Fig. 5.17. If the chilled water
flowing through two-way valve V1 is at 100 gpm (6.31 L/s), select and size control valve V1.
Solution.
1. To achieve a nearly linear relationship between the coil load and the travel of the valve stem, an
equal-percentage valve is selected for valve V1.
2. For a water system using variable-speed pumping with DDC, it is assumed that the pressure
drop across the control valve pvv  5 ft WC, or 5 0.433  2.165 psi (15 kPa). From
Eq. (5.8), the flow coefficient is
From one of the manufacturers’ catalog, for Cv  68, the size of equal-percentage control valve V1
is 3 in. (76 mm).
3. From the friction chart of water in steel pipes, for a chilled water flow rate of 100 gpm (6.31
L/ s) and a piping head loss of 2.5 ft /100 ft (2.5 m/100 m) of length, the diameter of the branch
piping section EH is also 3 in. (76 mm). The size of the control valve and the diameter of the
branch pipe are the same.
Cv 
V
?pvv

100
?2.165
 68
Cv 
V?
?pvv
V?
 Cv?pvv
V?
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.31
5.7 DAMPERS AND DAMPER ACTUATORS
A damper is a device that controls the airflow in an air system or ventilating system by changing the
angle of the blades and therefore the area of its flow passage. In HVAC&R systems, dampers can be
divided into volume control dampers and fire dampers. Fire dampers are covered in a later section.
In this section, only volume control dampers are discussed.
Types of Volume Control Dampers
Volume control dampers can be classified as single-blade dampers or multiblade dampers according
to their construction. Various types of volume control dampers are shown in Fig. 5.18.
Butterfly Dampers. A butterfly damper is a single-blade damper. A butterfly damper is made from
either a rectangular sheet mounted inside a rectangular duct or a round disk placed in a round duct,
as shown in Fig. 5.18a. It rotates about an axle and is able to modulate the air volume flow rate of
the duct system by varying the size of the opening of the passage for air flow.
Gate Dampers. A gate damper is a single-blade damper. It also may be rectangular or round. It
slides in and out of a slot in order to shut off or open up a flow passage, as shown in Fig. 5.18b.
Gate dampers are mainly used in industrial exhaust systems with high static pressure.
Split Dampers. A split damper is also a single-blade damper. It is a piece of movable sheet metal
that is usually installed at the Y connection of a rectangular duct system, as shown in Fig. 5.18c.
5.32 CHAPTER FIVE
FIGURE 5.18 Various types of volume control dampers: (a) Butterfly damper; (b) gate damper; (c) split
damper; (d) opposed-blade damper; (e) parallel-blade damper.
The movement of the split damper from one end to the other modulates the volume of air flowing
into the two legs or branches. A split damper is usually modulated only during air balancing after
installation or during periodic air balancing.
Opposed-Blade Dampers. An opposed-blade damper is a type of multiblade damper that is often
rectangular, as shown in Fig. 5.18d. It is usually used for a flow passage of large cross-sectional
area. The damper blades may be made of galvanized steel, aluminum alloy, or stainless-steel sheets,
usually not exceeding 10 in. (25.4 cm) in width. Rubber or spring seals can be provided at the fully
closed position to control the air leakage rating, which often does not exceed 6 cfm/ ft2 (30 L/ s m2)
at a pressure drop across the damper of 4 in. WC (1000 Pa). The bearing used for supporting the
blade axle should be made of a corrosion-resistant material such as copper alloy or nylon. Tefloncoated
bearings may also be used to ensure smooth operation of the damper. Lever linkages are
used to open and close the damper blades. The characteristics of the opposed-blade dampers are
covered later in this section.
The maximum static pressure drop across closed opposed-blade dampers is 6 in. WC (1500 Pa)
for a 36-in.- (914-mm-) long damper (the length of the damper blade) and 4 in. WC for a 48-in.-
(1219-mm-) long damper.
Parallel-Blade Dampers. A parallel-blade damper is also a type of multiblade damper used
mainly for large cross-sectional areas, as shown in Fig. 5.18e. The blade material and the requirement
for the seals and bearings are the same as those for opposed-blade dampers.
Damper Actuators (Motors)
Damper actuators, also called damper motors, are used to position dampers according to a signal
from the controller. As with valve actuators, damper motors can be classified as either electric or
pneumatic.
Electric Damper Motors. These either are driven by electric motors in reversible directions or are
unidirectional and spring-returned. A reversible electric motor is used often for more precise control.
It has two sets of motor windings. When one set is energized, the motor’s shaft turns in a
clockwise direction; and when the other set is energized, the motor’s shaft turns in a counterclockwise
direction. If neither motor winding is energized, the shaft remains in its current position. Such
an electric motor can provide the simplest floating control mode, as well as other modes if required.
Pneumatic Damper Motors. Their construction is similar to that of pneumatic valve actuators,
but the stroke of a pneumatic damper motor is longer. They also have lever linkages and crank arms
to open and close the dampers.
Volume Flow Control between Various Airflow Paths
For air conditioning control systems, most of the dampers are often installed in parallel connected
airflow paths to control their flow volume, as shown in Fig. 5.19. The types of airflow volume
control are as follows:
Mixed-Air Control. In Fig. 5.19a, there are two parallel airflow paths: the recirculating path um
in which a recirculating air damper is installed and the exhaust and intake path uom, in which exhaust
and outdoor air dampers are installed. The outdoor air and the recirculating air are mixed together
before entering the coil. Both the outdoor damper and the recirculating damper located just
before the mixing box (mixed plenum) are often called mixing dampers. The openings of the outdoor
and recirculating dampers can be arranged in a certain relationship to each other. When the
outdoor damper is at minimum opening for minimum outdoor air ventilation, the recirculating
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.33
5.34 CHAPTER FIVE
FIGURE 5.19 Airflow paths: (a) mixed- air control, (b) bypass control, and (c) branch flow control.
damper is then fully opened. If the outdoor damper is fully opened for free cooling, the recirculating
damper is closed.
Bypass Control. In the flow circuit for bypass control, as shown in Fig. 5.19b, the entering air at
the common junction m1 is divided into two parallel airflow paths: the bypass path, in which a
bypass damper is installed, and the conditioned path, in which the coil face damper is installed in
series with the coil, or the washer damper with the air washer. The bypass and the conditioned
airstreams are then mixed together at the common junction m2. The face and bypass dampers can
also be arranged in a certain relationship to each other.
Branch Flow Control. In a supply main duct that has many branch take offs, as shown in Fig.
5.19c, there are many parallel airflow path combinations: paths b1s1 and b1b2s2, b2s2 and b2b3s3,
etc. In each branch flow path, there is a damper in the VAV box, and points s1, s2, s3, etc., are the
status points of the supply air.
Parallel airflow paths such as those shown in Fig. 5.19 have the following characteristics:
1. The total pressure losses of the two airflow paths that connect the same endpoints are always
equal; for example, , etc.
2. The relationship between total pressure loss p, in. WC (Pa); flow resistance R, in. WC/ (cfm)2
(Pa s2/m6); and volume flow rate , cfm (m3 / s), can be expressed as
(5.9)
Flow resistance is covered in greater detail in Chap. 10.
3. If the total pressure loss p remains constant and the flow resistance Rn of one parallel path
increases, from Eq. (5.9), the airflow through this path V must be reduced. The airflow in other
parallel paths remains the same.
4. The total pressure loss of an airflow path between two common junctions p determines the volume
flow rate of air passing through that path and can be calculated from Eq. (5.9) as
5. When the flow resistances in most of the branches increase because of the closing of the
dampers to a small opening in their VAV boxes, the flow resistance of the supply duct system
Rsys and the system total pressure loss psys both tend to increase, and thus the total air volume
flow of the supply duct system sys will reduce accordingly.
Flow Characteristics of Opposed- and Parallel-Blade Dampers
A parallel-blade or an opposed-blade damper that is installed in a single airflow path to modulate
airflow is often called a volume control damper (or throttling damper). For volume control dampers,
a linear relationship between the percentage of the damper opening and the percentage of full flow
is desirable for better controllability and cost effectiveness. (Full flow is the air volume flow rate
when the damper is fully opened at design conditions.) The actual relationship is given by the installed
characteristic curves of parallel-blade and opposed-blade dampers shown in Fig. 5.20a and
b. For the sake of energy savings, it is also preferable to have a lower pressure drop when air flows
through the damper at the fully open condition.
In Fig. 5.20,  is called the damper characteristic ratio and may be calculated as
(5.10)


ppath pod
pod

ppath
pod
 1 
pp-od
 pod
V?
?
p
R
V?

p  RV?2
V?
pum  puom, pm1 bym2  pm1 conm2
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.35
5.36 CHAPTER FIVE
FIGURE 5.20 Flow characteristic curves for dampers: (a) parallel-blade
and (b) opposed-blade.
where ppath  total pressure loss of airflow path, in. WC (Pa)
pod  total pressure loss of the damper when it is fully opened, in. WC (Pa)
pp-od  total pressure loss of air flow path excluding damper, in. WC (Pa)
Damper Selection
Butterfly dampers are usually used in ducts of small cross-sectional area or in places like VAV boxes.
For volume control dampers in a single airflow path, in order to have better controllability, an
opposed-blade damper is recommended if many dynamic losses other than the damper itself (such
as coil or air washer, heat exchanger, and louvers) exist in the airflow path. If the damper is the primary
source of pressure drop in the airflow path, a parallel-blade damper is often used.
For mixing dampers, a parallel-blade damper is recommended for the recirculating damper as the
pressure drop across the damper is often the primary source in its airflow path. An opposed-blade
damper is recommended for the outdoor damper and exhaust (relief) damper for better controllability.
The parallel blades of the recirculating damper should be arranged so that the recirculating airstream
will blow toward the outdoor airstream, resulting in a more thorough mixing. Many packaged units
also use parallel-blade outdoor dampers for smaller pressure drop and less energy consumption.
For face and bypass dampers, an opposed-blade coil face damper in an airflow path of greater pressure
drop and a parallel-blade bypass damper will give better linear system control characteristics.
For two-position control dampers, a parallel-blade damper is always used because of its lower
price.
Damper Sizing
Damper sizing should be chosen to provide better controllability (such as a linear relationship between
damper opening and airflow), to avoid airflow noise if the damper is located in the ceiling
plenum, and to achieve an optimum pressure drop at design flow to save energy.
The face area of the damper Adam, ft2 (m2), in most cases is smaller than the duct area Ad, in
ft2 (m2). Based on Alley (1988) paper, the local loss coefficient Cdam of the damper for different setups
can be determined from Fig. 5.21. Then the pressure drop across the damper when the damper
is fully opened pod, in. WC (Pa), can be calculated as
(5.11)
(5.12)
where vdam  face velocity of the damper, fpm.
1. The damper is generally sized when the air flowing through the damper is at a maximum. For
an outdoor damper, the maximum airflow usually exists when the free cooling air economizer cycle
is used. For a recirculating damper, its maximum airflow occurs when the outdoor air damper is at
minimum opening position, to provide outdoor air ventilation.
2. The face velocity of dampers vdam is usually 1000 to 3000 fpm (5 to 15 m/ s), except that the
face velocity of a butterfly damper in a VAV box may drop to only 500 fpm (2.5 m/ s) for energy
savings and to avoid airflow noise. The ratio Adam /Ad is often between 0.5 and 0.9.
3. The outdoor damper may be either made in a one-piece damper or split into two dampers, a
larger and a smaller, to match the needs at free cooling and minimum outdoor ventilation.
4. For a bypass damper, its face area should be far smaller than that of an air washer or than a
water heating or cooling coil’s face damper. When the air washer or coil’s face damper is closed, the
area of the bypass damper should provide an airflow that does not exceed the system design airflow.
vdam 
V?
dam
Adam
pod  Cdamvdam
40052
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.37
5.8 SYSTEM ARCHITECTURE
Architecture of a Typical EMCS with DDC
Figure 5.22 shows the system architecture of a typical energy management and control system with
direct digital control (EMCS with DDC) for a medium or large building.
Operating Levels. Such an EMCS has mainly two operating levels:
1. Unit level. This level is controlled by unit controllers. A unit controller is a small and specialized
direct digital controller which is used to control a specific piece of HVAC&R equipment or device
such as a VAV box, a fan-coil unit, a water-source heat pump, an air-handling unit, a packaged
unit, a chiller, or a boiler. For HVAC&R, most of the control operations are performed at the unit
level. Since the software is often factory-loaded, only the time schedules, set points, and tuning
constants can be changed by the user. Some of the most recently developed unit controllers are also
5.38 CHAPTER FIVE
FIGURE 5.21 Local loss coefficient Cdam of air damper. (Source: ASHRAE Transactions 1988, Part I.
Reprinted by permission.)
programmable to various degrees. Sometimes the manufacturer provides a variety of preprogrammed
control sequences, such as monitoring, and diagnostics, and designers can specify the
required control sequence that best fits their designs.
2. System/building level. This level is controlled by system controllers. Since a system controller
has an onboard capacity, programmed by an operator or factory-preprogrammed software, to
execute complicated HVAC&R and other programs, they are the brain of an EMCS. Generic control
software such as scheduling, trending, alarming, diagnostics, and security is also provided in system
controllers. Generic control is covered in detail in a later section.
A system controller is used to coordinate the control operations of an HVAC&R system, such as
the coordination between the duct static pressure and the total air volume of VAV boxes in an air
system, or the sequencing of three centrifugal chillers and cooling towers in a refrigeration plant. A
system controller may interface with sensors/transmitters by means of input/ output (I /O) connections
directly. Unit controllers are also configured on a separate subnetwork and connected to a
system controller.
Operator Personal Computer (PC) Workstation. The operator may interface with the EMCS
primary through an operator’s PC workstation or purpose-built device, either handheld or fixed to
the system controller. Each workstation shall consist of a PC, autodial telephone modems, and
printers. The central processing unit (CPU) shall be a minimum of an Intel 80486 and operated at a
minimum of 33 MHz. Its memory includes a minimum of 8 Mbytes RAM and 212 Mbytes hard
disk. The communication ports connected to system controllers and other control systems should be
provided. A 14-in. (356-mm) color monitor also shall be provided.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.39
Operator
interface
System
Unit
Gateway Unit controller System controller
Proprietary Network
PC
GW
BAC net
BAC net
BAC net
BAC net
GW
SC
AHU
Chiller
UC UC UC
BAC net
SC
UC UC UC UC
UC SC
Printer Modem
FIGURE 5.22 System architecture of a typical large EMCS.
The software in the workstation shall do the following:
 Accommodate processes as well as prioritize applications based on their input/output priority level.
 Provide system graphics including the HVAC&R equipment (such as a display of up to 10
graphic screens at once for comparison) to monitor the operating status of the system. The operator
with the proper password is able to add, delete, or change the set points, time scheduling, etc.
 Support the editing of all system applications including the generic control software provided in
the system controllers. The edited or custom programmed software shall be downloaded and executed
at one or more of the system controllers.
 Automatically save the database and restore the database that has been lost in one of the system
controllers.
 Provide scheduling, trends, totalization, alarm processing, security to view and edit data, and
system diagnostics.
Communication Network. A peer-to-peer data communication of a local-area network (LAN)
adopting either Ethernet or ARCNET will be used between system controllers and between system
controllers and PC workstations (or other system). A peer-to-peer communication means that all
system controllers or work stations have equitable access to communication resources. For the communication
subnetwork between system controllers and unit controllers, and between unit
controllers themselves, a master-slave token-passing (MS/TP) technology is often used. A system
controller also acts as a medium to provide data communication between the work station and the
unit controllers. The network technology is covered in detail in a later section.
Power Source. The temperature sensors and humidity sensors may need up to 12 V dc and
24 V ac as a power source. Many DDC units have a power source of 24 V ac or 120 V ac line voltage.
Most of the valve actuators and damper motors need a power source of 24 V ac.
Size of EMCS. The size of an EMCS depends on the number of points (or objects) that belong to
its DDC units. An EMCS of 100 points or less can be considered a small project. An EMCS that
has 1000 points or more can be considered a large project.
System Characteristics
An architecture of EMCS incorporating DDC such as that in Fig. 5.22 has the following characteristics:
1. All DDC units are independent and stand-alone controllers. If any of the controller fails, there
is only a limited effect.
2. The system architecture shows a distributed processing model. Since most of the control operations
are performed at the unit controller level, and partly at the system controller level, such an
architecture has the advantage that it tremendously reduces the data communication between the
unit controllers and the system controller, as well as between the DDC system controller and the PC
workstation.
3. Zimmerman (1996) noted that since the introduction of DDC in the 1980s, “. . . microprocessor
and memory have declined rapidly in cost while wiring and installation costs have not
declined at the same rate.” To provide more powerful unit controllers, moving the controllers nearer
to the sensors and control devices will reduce a lot of wiring and installation costs as well as the
overall system cost.
4. If each HVAC&R piece of equipment has its own unit controller, it surely will be beneficial to
the HVAC&R equipment manufacturers to fabricate controllers and other control system components
themselves.
5.40 CHAPTER FIVE
Future Development
The development of more powerful, programmable unit controllers using a modular configuration,
will reduce the difference between a system controller and a unit controller. As predicted by
Hartman (1993), the future architecture of an EMCS may have only a single tier of various kinds of
unit controllers. The operator’s PC workstation, modems, and unit controllers will all be connected
to a peer-to-peer data communication trunk. Such an architecture will simplify the DDC units and
the communication network, move the intelligence nearer to the control devices, and finally create
more effective control at lower system cost.
5.9 INTEROPERABILITY AND OPEN PROTOCOL BACnet
Interoperability
Turpin (1999) defined interoperability as “the ability of systems including equipment and components
from different manufacturers to share data and information for the purpose of operation with
plug-and-play connectivity.” Interoperability is one of the necessary conditions for system integration.
System integration is a strategy to integrate various HVAC&R systems of various manufacturers
together, and to integrate HVAC&R systems with lighting, fire protection, security, elevator, and
electrical systems in a building together.
Robertson and Moult (1999) note that the advantages of system integration include the following:
 It reduces the installation cost of the building automation system.
 It enhances energy management.
 It can apply building automation system features.
 It provides building operator training.
 There is a single user interface.
The greater the interoperability of a single system and the more systems you try to integrate into a
single system the more complicated and costly the process will be. Interoperability and system integration
is one of the goals of the HVAC&R industry. More and more engineers, facility owners, and
manufacturers recognize this need. It takes time to accomplish such a complicated process.
BACnet—Open Data Communication Protocol
BACnet means building automation and control network. It is an open data communication protocol.
Open means that an independent institution governs its development. All contents are known,
fixed, and accessible. A protocol refers to the rules by which two or more devices communicate
data to each other that must be obeyed. BACnet enables that building automation devices from various
manufacturers can talk to one another, share data, and work together following a standard way.
BACnet defines all the elements of data communication between devices in a building automation
control system. It is specifically tailored for HVAC&R control equipment, but it also provides a basis
for integrating lighting, security, and fire-detecting systems. BACnet was developed from 1987
to 1995 by ASHRAE and was adopted as a national standard in 1995 by ANSI, as ANSI/ASHRAE
Standard 135-1995. For details, refer to ASHRAE’s BACnet.
BACnet will assist building owners, designers, contractors, and operators in three areas:
1. It gives more freedom to select the best equipment and components from different manufacturers
in order to have a more efficient system at lower cost.
2. It operates the control system from a single workstation; i.e., it is more effective to operate and
easier to maintain, and there is only one system to learn.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.41
3. It collects data from different systems and offers greater flexibility for extended systems in
retrofit projects.
Most of the EMCS manufacturers agreed to fabricate BACnet-compliant products from the late 1990s.
There are many other data communication protocols. One was developed by Echelon Corporation,
called LonTalk Protocol, and it became a working system in the mid-1990s and was favored
by members of the LonMark Interoperability Association. LonTalk protocol is neither an opento-
public protocol nor a standard.
An independent consortium called the OPC foundation, formed as a nonprofit organization in
1996 in Boca Raton, Florida, has dedicated itself to provide interoperability with Microsoft technologies
to develop a global specification and multivendor interoperability in industries. In 1990,
OPC is leading 140 member companies including Honeywell , Johnson Controls, and Siemens.
Application Layer
Layered Structure. A data communication system often adopts a hierarchical layered structure so
that a complex problem is broken into smaller and more easily solved problems. BACnet is based
on a four-layer collapsed architecture that corresponds to application, network, data link, and physical
layers in an International Organization for Standardization (ISO) model. This is the result of
careful consideration of the characteristics and requirements of the building automation control
(including HVAC&R) together with a constraint that protocol overhead be as low as possible.
An application layer is the highest layer in BACnet. It serves to define the objects and services
(including control operations, information exchange, and control devices) in a building automation
control (BAC) system. It also provides communication services and data encoding schemes
required by applications to perform monitoring and control functions.
Object Types, Properties, and Devices. The BACnet defines a set of standard object types instead
of conventional points. Analog input, analog output, binary value, command, file, program, schedule,
etc., grouped in 18 types are standard object types, for every device in an EMCS must have a
device object. There are 123 properties that have been identified by BACnet which fully describe
the device, or object type, in the network. Certain properties are required to be specified whereas
others are optional. An object identifier specifies its object name, object type, etc., and optional
properties such as description and device type. In BACnet, a device is defined as any device, real or
virtual, that supports digital communication using the BACnet protocol.
Services. In BACnet, services are the operations by which one device acquires information from
another device, commands another device to do something, or announces that some event
happened. BACnet defines 32 services that can be grouped into five categories:
 Alarm and event services refer to changes in conditions detected by a device, such as acknowledgment
of an alarm and confirmed change of value notification.
 File access services are used to read and manipulate files that are kept in devices, such as only
one read or write operation at a time.
 Object access service provides the means to read, to write, and to modify properties, such as to
add one or more items to a property.
 Remote device management offers disparate operations such as to tell a device to stop accepting
messages.
 Virtual terminal services are used by an operator to establish a bidirectional connection with an
application program implemented in a remote device.
Services are classified as confirmed when a reply is usually expected with data and unconfirmed
when no reply is expected. In BACnet, a given device is not required to implement every service.
However, “read property” is required to be executed by all the devices.
5.42 CHAPTER FIVE
Conformance Class, Functional Groups, and Protocol Implementation Conformance Statement
(PICS). BACnet defines six levels of conformance for all devices, classes 1 to 6, that are hierarchical
to indicate the difference in requirements that must be met to conform to BACnet. The
requirement of a class includes all the requirements of all the other classes having a lower number.
At the lowest level, conformance class 1 requires only that a device be able to execute (respond to)
a “read property service” request, such as a sensor. For conformance class 6, a device is required to
implement 21 types of service requests, such as a PC workstation.
A functional group defines a combination of services and object types that are required to perform
certain BAC functions. In BACnet, there are altogether 13 functional groups, such as clock,
workstation, and event initiation.
The protocol implementation conformance statement (PICS) is a document provided by the
manufacturer of a device to identify those options implemented by a particular device. The PICS
covers the conformance class of the device, supported functional groups, standard and proprietary
services executed and initiated, etc.
Data Encoding. In BACnet, application-layer protocol data units (APDUs) are used to convey the
information contained in the service primitives and associated parameters. ISO Standard 8824,
Specification of Abstract Syntax Notation One (ASN.1), has been chosen as the method to represent
the data content of BACnet services. Each data element consists of three components: (1) identifier
octets, (2) length octets, and (3) content octets. The fixed portion of each APDU providing protocol
control information is encoded implicitly, and the variable portion of each APDU providing servicespecific
information is encoded explicitly.
Network Layer
In BACnet, the purpose of the network layer is to provide the means from which messages can be
relayed from one BACnet network to another in the internetwork. Two or more BACnet networks
are connected by routers to form a BACnet internetwork. A router, a BACnet device, is used to interconnect
two disparate BACnet local-area networks (LANs). A network layer directs the messages
to a single remote device or broadcasts the messages on a remote network, or to all devices
on all networks. A device is located by a network number and a medium access control (MAC) address.
Another network layer function is message segmentation and reassembly.
In an EMCS, there are often two networks: one uses Ethernet or ARCNET for high-speed
message transmission between system controllers and the PC workstation, and another adopts lowspeed
message transmission between unit controllers and system controllers. Consequently, a network
layer is required for the BACnet protocol.
Data Link/Physical Layer—Network Technology
In BACnet, a data link layer has the capability to address messages to a single device or to all
devices. At the data link layer, only incoming BACnet messages received from the physical layer
are passed on to the network layer, and a code is added to the outgoing messages to identify them as
BACnet messages before they are passed to the physical layer. A physical layer provides the physical
medium for message transmission.
For local-area networks, BACnet supports Ethernet, ARCNET, MS/TP, PTP, and LonTalk as
alternatives. Various media are used as physical transmission entities. Typical media are twistedpair
wire, fiber-optic cable, and coaxial cable.
Ethernet. It is one of the commonly used LAN technologies. At any time, multiple devices may
access the network simultaneously, i.e., multiple access with collision detection. Ethernet is using
a peer-to-peer communication with a bus network configuration. Ethernet often runs at a speed of
10 Mbits/ s. Two types of coaxial cable are often used in Ethernet: thick-wire and thin-wire. A
thick-wire Ethernet segment has a maximum length of 1600 ft (488 m), and up to 100 nodes can be
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.43
attached. Thick-wire Ethernet is more expensive. A thin-wire Ethernet segment has a maximum
length of 600 ft (183 m) and 50 attached nodes.
ARCNET (Attached Resources Computer Network). This is also a commonly used LAN technology
and is lower in cost than Ethernet. ARCNET adopts a token-passing network access method.
A token which indicates the permission to use the physical medium is passed from one network
node to the next node in a circular manner. As in Ethernet, ARCNET also uses a peer-to-peer communication
network, and its network nodes reside on a bus.
Both coaxial cable and twisted-pair wire are used in ARCNET and run at a speed of
2.5 Mbits/s. For each ARCNET segment, up to 8 nodes can be connected. All together, up to
255 nodes can be communicated over an ARCNET network.
Master-Slave/Token-Passing (MS/TP). In BACnet, MS/TP divides all the nodes on the network
into two categories: masters and slaves. Only masters can initiate data communication, whereas
slaves cannot initiate. Slaves can only respond to requests from masters. The MS/ TP also adopts a
token-passing network access method. A master node may access the network only when the token
(permission to use the medium) is passed to it from the previous master node. The token never
passes to the slave nodes. Master nodes in an MS/ TP network are at a peer-to-peer communication.
The MS/TP network often uses twisted-pair wires. It can operate at a speed of 9600 bits / s, as well
as 19.2, 38.4, and 76.8 kbits/ s.
Point-to-Point (PTP). This is a data link layer protocol which provides serial communication
between two devices. Such a point-to-point communication typically involves a dial-up phone modem
or hardwired connection between two nodes. PTP has a simpler medium access mechanism
and is often temporary in nature. PTP is much slower than a LAN. Both devices can receive and
transmit simultaneously.
LonTalk LAN. This is an option of the physical medium in BACnet and is at the base of the Lon
Mark protocol. LonTalk supports a number of choices of physical media.
Connection between BACnet and Proprietary Network
For an extension project, it is possible that the original building is still intended to keep the proprietary
network and the extended part is to construct a BACnet network. The proprietary network
needs a gateway to connect to the BACnet network, as shown in Fig. 5.22. A proprietary network
does not open to others. Typically, information can only be exchanged between EMCS components
and the proprietary network of the same manufacturer. For two networks and their computers using
different protocols to communicate, some translation must take place. The device that provides this
translation is called a gateway.
LonTalk Protocol
According to Glinke (1997), the Local Operators Network (LonTalk) protocol is a seven-layer proprietary
protocol developed by Echelon Corp. Data communication is implemented on a neuron
chip. There are actually three microprocessors within the chip; two are used for the network and
one is for specific functions. These devices, which allow different types of existing networks to
communicate, use multiple, low-cost media (Ethernet, ARCNET, fiber optics) and provide flexibility.
Since this technology is essentially peer-to-peer communication without the need for a central
supervisory control node, Lon Mark-certified unit controllers can exist on the same network as
equals.
The two promising and widely used protocols, BACnet and LonTalk, were able to communicate
with each other in the late 1990s. They will improve themselves through actual operation in the
future and produce better choices for engineers and users.
5.44 CHAPTER FIVE
5.10 CONTROL LOGIC AND ARTIFICIAL INTELLIGENCE
In an EMCS, the software in the DDC units determines the control functionality. Since the logic is
separate from the hardware, the control functions and sequences are now limited only by the knowledge
and innovation of the designer and operators. The development of artificial intelligence in the
DDC software—fuzzy logic, expert systems, artificial neural network, etc.—tremendously expands
the control flexibility and improves the operating quality to meet the more complex requirements
for an HVAC&R energy management and control system.
Fuzzy Logic
Basics. Since many systems have become more and more complex over the past decades, an accurate
analytical model based on rigorous and nonlinear mathematics is difficult to develop and to be
accepted in the daily management by operators. In 1965 Professor Lotfi Zadeh of the University of
California, Berkeley, developed fuzzy set theory which provides a new control logic for a system.
Fuzzy logic has the ability to deal with the imprecision that happens in everyday life. Conventional
digital technology is based on bivalence: yes or no, on or off, 1 or 0, and black or white.
Fuzzy logic is multivalent. Things can be partly yes to some degree, partly no to some degree, and
fuzzy logic deals with shades of gray. Conventional set theory observes that a fact must be either
true or false, whereas in fuzzy logic set theory, a fact can be partly true and partly false, can belong
to a set and also belong to another set.
According to Lehr (1996) and Scholten (1995), fuzzy logic offers simplicity in the midst of
complexity and is a real alternative for system operation and control. The obvious benefits of
fuzzy logic are that it makes things more human and more friendly to a person who is less trained,
and it is more easily maintainable. Fuzzy logic really is not a vague theory. It is comprised of a set
of precise rules based on rigorous mathematics which governs the behavior of a system by means
of words and phrases instead of nonlinear models. Fuzzy logic controllers (FLCs) have been
widely used in air conditioners, humidifiers, refrigerators, and many other devices such as elevators.
There are also many applications for which conventional control logic is better than fuzzy
logic controller.
Fuzzy Sets and Membership Function. Fuzzy sets, membership function, and production rules
are three primary elements of fuzzy logic. Figure 5.23 shows the fuzzy sets, membership functions,
and a diagram of fuzzy logic control. In conventional bivalent crisp set theory, “sets” of thermal
comfort of an occupant in an air conditioned space can be categorized according to the indoor temperature
Tr exactly as cool, a range of 62 to 72°F (16.7 to 22.2°C); and just right, a range of 72 to
76°F (22.2 to 24.6°C), etc., as shown in Fig. 5.23. At Tr  74°F (23.3°C) and Tr  75°F (23.9°C),
both are “just right”—have a membership value of 1 with regard to the just-right set and a membership
value of 0 with regard to the cool set.
On the other hand, the fuzzy set “just right” ranges between a membership value of 0 at 70°F
(21.1°C), a membership value of 1 at 74°F (23.3°C), and a membership value of 0 again at 78°F
(25.6°C). For Tr  75°F (23.9°C), the membership value of fuzzy sets would be described as
75 percent just right and 25 percent warm. In another way, the assertion of thermal comfort of just
right is 75 percent true and 25 percent false.
If Tr  76°F (24.4°C), then the conventional crisp set will be difficult to determine. In addition,
for Tr  74°F (23.3°C) and Tr  75°F (23.9°C), the assertion of the conventional crisp set for both
is “just right,” so there is no difference between them. And the assertion of fuzzy sets for Tr  74°F
(23.3°C) is 100 percent just right, which is different from 75 percent just right and 25 percent warm
when Tr  75°F (23.9°C).
Production Rules. When the fuzzy set of an occupant’s thermal comfort is integrated with another
fuzzy set of percentage of fan speed variation (positive large, positive small, zero, negative small,
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.45
and negative large), then the production rules, or fuzzy logic rules, can be described as follows:
 If the indoor temperature is hot, then the fan speed rises a lot.
 If the indoor temperature is warm, then the fan speed rises a little.
 If the indoor temperature is just right, then the fan speed stays unchanged.
 If the indoor temperature is cool, then the fan speed reduces a little.
 If the indoor temperature is cold , then fan speed reduces a lot.
5.46 CHAPTER FIVE
PL
NL
100
80
60
40
32
20
0
0 0.5
Fan speed variation range, %
If hot,
then
decrease
a lot If warm,
then
decrease
a little
If just right,
then do
nothing
If cool,
then increase
a little If cold,
then
increase
a lot
0
62 66 7071
Fuzzy sets
Positive large PL
Positive small PS
Zero ZE
Negative small NS
Negative large NL
Indoor temperature, F
Indoor temperature, F Conventional crisp sets
74 78 82 86 90
62 66 70 74 78 82 86 90
Cold
Cold
Cool
Cool
Warm
Warm
Hot
Hot Just
right
0.5
0.2 0.45
1
0
0.5
1
Centroid
20
40
60
80
100
1
PS
ZE
NS
Just
right
FIGURE 5.23 Fuzzy sets, membership function, and fuzzy logic control.
Fuzzy Logic Controller. An FLC consists of three parts: a fuzzifier converts ordinary inputs to
fuzzy variables, a fuzzy reasoning unit produces fuzzy control signals based on input fuzzy variables,
and a defuzzifier converts fuzzy control signals to conventional control outputs.
If the indoor temperature Tr is 71°F (21.7°C), then the membership function of the fuzzy sets of
thermal comfort in Fig. 5.23 would be described as 0.2 (20 percent) just right and 0.45 (45 percent)
cool. From the production rules, then, “reduce the fan speed a little” has a membership value of
0.75 and “fan speed remains unchanged” has a membership value of 0.2. One way to interpret these
two fuzzy outputs in a conventional crisp output is to determine the centroid based on the area of
the two truncated triangles. From Fig. 5.23, the output value is 32 (reduce the fan speed by 32 percent
of its range).
Huang and Nelson (1994) noted that an FLC is characterized by a set of linguistic fuzzy logic
rules. The initial set of these rules is often based on past experience or analysis of the control
process. The initial set of rules can be modified by analyzing the performance trajectory on the linguistic
plane to obtain an optimal rule set. This is called rule refinement. The second significant in-
fluence on the behavior of an FLC is the choice of membership functions. The overlap of the fuzzy
sets should be moderate to allow for reasoning with uncertainty and the need for completeness of
the control rules. Refer to Huang and Nelson’s paper for details.
Knowledge-Based Systems and Expert Systems
Basics. A knowledge-based system (KBS) is a knowledge-rich, logic-oriented computer program
that mimics human knowledge and reasoning in a specific domain to assist in solving more complex
problems. Human knowledge includes expert knowledge, common facts, and knowledge in
any form, whereas an expert system, strictly speaking, mimics only an expert’s expertise in a given
domain to solve specific problems. An expert system can be considered as the core part of a knowledge-
based system.
In comparing a knowledge-based system with conventional programmed software, Hall and
Deringer (1989) noted that the benefits of a knowledge-based system are due to the following superior
abilities:
 Logical reasoning. Logical problems are solved by means of sorting, comparing, searching, and
reasoning as well as to evaluate alternative choices to a specific problem. More often, several
causes (or solutions) of a problem are identified.
 Resolution of uncertainty. When precise knowledge is not available and for problems that involve
qualitative facts and incomplete data, the certainty of the conclusions of the KBS using heuristic
rules depends on the reliability of the heuristic rules themselves.
 Multiple approaches. Multiple experts are often used to develop a knowledge base. Experts will
rarely agree exactly on the proper approach to solve a problem. Solution of the problem with multiple
approaches is one aspect of KBS.
 Justification of results. A well-developed KBS has the capability to recall the basis of each decision
during the problem-solving process. It can provide a decision trail that will explain the
sequence of decisions used to produce the conclusion.
The limitations of a KBS are as follows. First, a KBS can make no decision that is not explicitly
contained in the knowledge base. Second, two different KBSs developed using the expertise
from different experts may give different advice for the same problem. Third, the KBS is timeconsuming
and expensive.
Knowledge-based systems will not replace experts and engineering professionals. Instead, the
systems will play the role of specialized knowledge-based advisers or consultants.
KBS Structure. An expert system comprises the following four modules, as shown in Fig. 5.24.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.47
1. Knowledge base. This is the most important part of an expert system. It contains the common
facts and inference rules, often in the form of if-then, and is domain-specific. Every rule in the
knowledge base is extracted from the expert’s knowledge and experience by the developer of the
KBS, the knowledge engineer, or is based on data and information from published handbooks,
manufacturers’ engineer manuals, and field survey results. Frames are another common knowledge
representation method. Hall and Deringer (1989) defined frames that “. . . facilitate the description
of knowledge about complex objects, which have many subelements, . . .” The content of the
knowledge base should be tailored to the user’s task and requirements. The user’s task can be determined
through direct interview with potential end users.
2. Inference engine. The inference engine asks for inputs from the user interface, executes
reasoning algorithms by applying the knowledge from the knowledge base, and arrives at the conclusion
based on the rules. There are two kinds of reasoning: (1) data-driven forward chaining or
forward reasoning and (2) goal-driven backward chaining or backward reasoning. They are categorized
according to how new information is inferred. Forward reasoning proceeds forward from the
user inputs about probable outcomes. In most engineering design problems, forward reasoning is
adopted. For instance, knowledge of the building envelope leads to load calculations, which forwards
next to determining the equipment capacity, and so on.
Backward reasoning proceeds backward from the user inputs about probable outcomes. In troubleshooting
or diagnostics, backward reasoning is used. For an example, collected facts lead to
determining the cause of the problem. Forward and backward reasoning were often used in conjunction
to control the flow of questioning as well as provide a dialog between the system and the
end user.
3. User interface. This includes the facilities to support smooth and convenient interaction with
the user. Only simple instruction is needed to start the system, end a session, save the contents, and
5.48 CHAPTER FIVE
FIGURE 5.24 Structure of a knowledge-based system for an EMCS and HVAC&R design.
give a printout. A friendly dialog between the KBS and the user and a simple question-and-answer
format with a provided menu of possible answers are expected. An intelligent operating system
which accesses various application tools is recommended. Modular new application tools can be
plugged into the operating system as required. A user interface with built-in intelligence, graphically
based high-resolution display with standardized menus and format may be the preferable answer.
4. Knowledge acquisition. A knowledge acquisition module provides strategies to capture the
experts’ knowledge to develop a KBS. Sometimes, it also checks for consistency and completeness.
Knowledge acquisitions are usually accomplished through personal interviews with experts and review
of the application literature. Most HVAC&R-related knowledge systems have long-term plans
to continue the process of knowledge acquisition. A new trend is machine learning; i.e., the computer
learns from the experience how to capture and manipulate new knowledge.
Development of KBSs. Many KBSs are developed by using commercially available development
tools called shells. A shell consists of mainly a rule editor, a knowledge base, an inference engine,
and user interfaces. Usually, the inference engine and user interface are fully developed within a
shell. Thus, the user can focus on the collection and the input of the knowledge. He or she can
change the existing knowledge base and does not need to change the entire system. KBS is expected
to operate on a PC.
The performance of a KBS and the resources needed during development are highly affected by
its knowledge engineer. The knowledge base is built through the cooperation of the knowledge engineer
and experts in a specific domain. The knowledge engineer has the responsibility to choose an
appropriate inference strategy, and a suitable shell and to ensure compliance of the system with the
task.
The development cycle of a KBS is an iterative and incremental process. It begins with the initial
prototype. The next is an improved, and expanded, one. The process is repeated and may take years.
How the KBS Works. For example, if the space air is too humid and the space cooling load is
only one-half of the design load, it is required to find the general cause of these symptoms. The
problem is a diagnosis problem. Brothers and Cooney (1989) stated that if-then rules are composed
of parameters (symptom, set point) and values (too hot, too humid, correct). In the inference
engine, the computer program of the KBS will begin to search for an if-then rule through the
knowledge base that will give a value (too humid) for a general cause. If the values do not match
those required by the rule, the computer program will search for the next rule, until the following
if-then rule has been found:
IF symptom is too humid, and
cold supply air temperature set point is correct, and
cold supply air relative humidity is O.K., and
sensible cooling load is 50 percent of design load
THEN general cause is size of supply airflow rate CF 90
The cause of a too humid space air is that the volume flow of supply air is too great. The term CF is
the abbreviation of certainty factor. That means the confidence in the answer to the problem is 90
percent.
Testing, Verification, and Validation. In Jafar et al. (1991), testing often detects logical errors related
to the knowledge base, syntactic errors, and missing knowledge. Testing only shows errors
and does not explain their cause. Verification and validation should be performed at each stage of
development, to check the knowledge base for internal inconsistencies, mismatches, etc.
Applications. There are three primary areas of HVAC&R-related applications of KBS, as reported
in the paper by Hall and Deringer:
 Monitoring—interpretation of measured data in comparison to expected behavior
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.49
5.50
Outputs
Output
Neuron
1n
Neuron
W12,1
W12,n
Wi1,n
WLo,n
WLo,1
Wi1,1
Wi1,1
Wi1,2
Wi1,n
(a)
0
	1
i
(b)
Neuron
Hidden layer
Input
Input layer Hidden layer
1 i 2...L
Neuron
f(x)
Output layer
o
FIGURE 5.25 An artificial neural network (ANN): (a) Structure of an ANN; (b) sigmoid function.
 Diagnostics—assistance in identifying solutions to complex technical problems
 Design—assistance in the selection of the HVAC&R systems and subsystems
Artificial Neural Networks
Basics. An artificial neural network (ANN) is massive interconnected, parallel processing,
dynamic system of interacting processing elements that are in some aspect similar to the human
brain. The fundamental processing element is called the neuron, which is analogous to the neural
cell in human brain. The neurons are set in layers, and thus a network is formed as shown in
Fig. 5.25. Inputs representing the variables that affect the output of the network are feeding forward
to each of the neurons in the following layers with an activation depending on their weighted sum.
Finally, an output can be calculated as a function of the weighted sum of the inputs and an additional
factor, the biases.
The ability to learn is one of the outstanding characteristics of an ANN. The weights of the
inputs are adjusted to produce a predicted output within specified errors. ANNs have been increasingly
used in recent years to predict or to improve nonlinear system performance in HVAC&R. An
ANN system is characterized by its net topology, neuron activations transfer, and learning method.
Net Topology. The structure of the network of an ANN, or net topology, depends on the data flow
mode, the number of layers, and the number of hidden neurons.
 In Miller and Seem (1991), there are two types of data flow modes: state and feed-forward models.
In the state models, all the neurons are connected to all the other neurons. In feed-forward
models, the neurons are connected between layers, as shown in Fig. 5.25a, and the information
flows from one layer to the next. Feed-forward models are the most popular and most often analyzed
models.
 In an ANN, there is always an input layer with the number of inputs equal to the number of parameters
(variables) that affect the output.
 There may be one or more hidden layers of neurons next to the input layer. The selection of number
of hidden layers and the number of neurons in each hidden layer remains an art. Curtiss et al.
(1996) noted that too many hidden layers and hidden neurons tend to memorize data rather than
learning. The hidden layers and hidden neurons must be sufficient to meet the requirement during
the learning process for more complex nonlinear systems. More hidden layers and hidden units
need more calculations and become a burden.
Among the 10 papers published from 1993 to 1996 in ASHRAE Transactions regarding developed
ANNs, most have only one hidden layer, some have two layers, and only one has three hidden
layers. None exceeds three hidden layers. Kawashima (1994) recommended that in an ANN only
one hidden layer is sufficient for load prediction.
 If the relationship between the inputs and output is more complex, i.e., nonlinear, and more inputs
are involved, then more neurons are needed in each hidden layer. Kawashima (1994) also suggested
that the number of neurons in each hidden layer exceed 2m 	 1. Here m indicates the
number of inputs.
 There is always an output layer next to the hidden layer(s). It is preferable to have one neuron
(single output) in the output layer for simplicity. There may be two or more neurons for multiple
outputs.
Neuron Activation Transfer. In Miller and Seem (1991) and Curtiss et al. (1996), for each neuron
in the hidden and output layers:
1. The input activations to a neuron in the first hidden layer h, denoted by i1n, can be
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.51
calculated as
(5.13)
where in  normalized input
Wi1,n  weights of connection between inputs and neurons in first hidden layer
B  biases
(5.14)
where io  original input data
imax, imin  maximum and minimum of original input data
2. The output from the neurons in the first hidden layer o1n, is expressed as a selected sigmoid
activation function:
(5.15)
There is only one output o for a neuron. This output is transmitted through output connections in
which it usually splits into multiple connections with identical activations.
3. The input activation to a neuron in the hidden layer L, denoted iLn, or a neuron in the output
layer o, denoted by ion, is equal to the output of the neuron in the previous layer which split identically
to all the neurons in hidden layer L. The output of a neuron in hidden layer L, denoted by oLn,
or output layer oon can be calculated as
(5.16)
and
(5.17)
Learning Method. The learning method, also called the training process, determines the connection
weights Wxyn through known sets of input/output pairs. The initially assigned connection
weights are adjusted repeatedly during the learning process until the error is within the specified
values. After training, the ANN can predict the outputs from given inputs. Backpropagation is the
most often used systematic method to train multilayer ANNs. Curtiss et al. (1996) and Miller and
Seem (1991) recommended the following training procedure:
1. Assign initial random values for connection weights, often between 0.5 and 	0.5. Select a
training input output pair; calculate the normalized input activations in and the input activation to
the neurons in the first hidden layer according to Eqs. (5.13) and (5.14).
2. Calculate the outputs of the 1, 2, . . ., L hidden layers o1n, o2n, . . ., oLn and then of the output
layer oon from Eqs. (5.15) and (5.16).
V  	iLnWLo,n
T  	i1nW1L,n
oon 
1
1 	 e(VB)
oLn 
1
1 	 e(TB)
o1n 
1
1 	 e(SB)
in 
io imin
imax imin
i1n  	inWi1, n 	 B  S 	 B
5.52 CHAPTER FIVE
3. Evaluate the error  between the calculated oon and the selected training output (target output)
ot by:
(5.18)
4. Adjust the connection weights from Wj to Wj	1 to minimize the error by using the following
rule:
(5.19)
where  is the learning rate whose value lies between 0 and 1.
5. Repeat the previous steps for all the input / output pairs in the training set until the error for the
entire training set is lower than the preset training tolerance. The training set should cover the
operating range of the inputs and outputs.
Applications. An ANN can model multiple parameters simultaneously for nonlinear systems. It
can also be periodically trained to update the weights. ANNs are now widely used for predictive
control, such as energy use prediction, energy optimization, adaptive control, data trending, and
optimum start and stop.
5.11 PROGRAMMING OF DDC SYSTEMS
Evolution of DDC Programming
The EMCSs having more complex functional controls are heavily software-driven. From the 1960s
to the 1970s, the software for HVAC&R control was mostly performed in a workstation computer
and programmed in the control manufacturer’s factory. Because of the trend to use distributed
microprocessor-based DDC units in the late 1970s, many EMCS manufacturers provide some type
of operator control language (OCL) for field-programmable line programming in the controllers using
BASIC-type language. This meets the requirements of computer program development, periodic
updating, and necessary modifications and improvements.
Since the late 1980s, there is a trend toward the function and object-orientated graphical programming
to provide software for DDC units and the PC in the operator’s workstation. Today, both
traditional line programming using BASIC-like language (sometimes Pascal and C++ languages are
also used) and graphical programming are used in DDC systems.
Graphical Programming
Davison (1992) described that graphical programming is a schematic drawing of a desirable control
scheme (functional control), such as mechanical cooling or economizer control, using symbols
called templates. The templates are displayed on the computer screen interconnected by lines that
direct the flow of data. The control scheme shows the inputs, through control operations, to outputs.
After the diagram of the graphical programming is completed, it is converted to a program usable
by the DDC units and PC through a computer program.
Compared with traditional line programming using a BASIC-like language, graphical programming
has the following advantages: First, it is intuitive and easier to use and understand by field
HVAC&R engineers and operators. Second, the user is more familiar with it because of meaningful
visual symbols represent functions and devices. It is simpler in documentation and easier in troubleshooting.
Third, it provides a time-saving tool for the specification and documentation of the
HVAC&R energy management and control systems. Finally, graphical programming enables one to
improve programming quality at a reduced cost. The limitations of graphical programming are that
Wj	1  Wj 	
oon

 oon(1 oon)(ot oon)
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.53
it is bulky to display on screen and that it is new to us and needs to develop required supports such
as a portable interface that acts directly to the DDC units.
Templates
In Davison’s (1992) paper, a template is a graphical symbol (icon) in graphical programming that
describes a single or combined control scheme (specific functional control), as shown in Fig. 5.26.
A template consists of input and output connections, a small section of computer code, and private
data storage registers. A template is function- and object-oriented. Each template performs a small
portion of a specific functional control. If several templates are connected by lines, they form a
complete computer program for a specific functional control. The template symbol provides a
visual reminder or memory of the function that the template accomplished.
A template maintains its own set of private data and variables that cannot be altered by the action
of any other template. The state of an instance of a template is contained in its private variables. They
can be only modified by the program code contained in that instance. The private variables of an instance
of a template are hidden from the action from other templates.
In a template, inputs are used to receive data from other templates or constants. Outputs are used
to send data to other templates. Inputs and outputs between two or more templates are connected by
lines that direct the data flow between the templates. An input may receive a value from only one
output in order to prevent overwriting of data at the input. On the other hand, an output data may be
sent and used by any number of inputs.
For a more complex control function that needs several templates, a combination of templates
called a macrotemplate is often used to simplify a graphical programming diagram.
5.54 CHAPTER FIVE
10K
Templates Dead band DB Set point SP
Cooling control
Outdoor temperature
lockout
Average
temperature
2
I
SP
Cool
Cool SP
DB 	
Zone
temperature
west
Zone schedule
10K
Zone temperature
east
10K
Outdoor
temperature

85
74
180
T
T
T
on off Min Min
Economizer
Compressor
0
1
0.5
Time
clock
on off Min Min
I
SP
Cool
DB
3
SO
FIGURE 5.26 Graphical programming for mechanical cooling control in a small packaged unit.
Graphical Programming for Mechanical Cooling Control
Figure 5.26 shows the graphical programming for mechanical cooling control for a small packaged
unit. This packaged unit has a supply fan, a DX coil, a single-scroll compressor, and an economizer.
The graphical programming for the control of this packaged unit includes mechanical cooling,
economizer, heating, and fan control.
During the scheduled occupied time, the zone schedule produces a 1.0 signal, and the cooling
control template SP selects the input labeled 1.0 and passes the value 74°F (23.3°C) as the cooling
set point. During unoccupied periods, SP passes the value 85°F (29.4°C). The average temperature
of east and west zones is compared with the current set point of the cooling dead-band control template.
If the average temperature exceeds the set point plus the dead-band value, 0.5°F (0.28°C), or
74.5°F (23.6°C), then the output of the cooling control template goes to 1.0, and the compressor is
turned on. If the zone average temperature drops below the set point minus the dead-band value,
73.5°F (23.1°C), then the output of the cooling control template goes to 0.0 and the compressor is
turned off.
When the outdoor air temperature drops below 50°F (10.0°C) minus the dead-band value
3°F (1.7°C), or 47°F (8.3°C), then the outdoor temperature lockout template produces 1.0, and the
compressor is locked out (turned off). Minimum on and minimum off timer templates prevent the
compressor from operating at short cycles.
Graphical programming of DDC systems may become supplemental to the sequence of operation
and control diagrams, and may become a part of the DDC design documentation.
5.12 TUNING DDC UNITS
Most of the DDC units use PI or PID control modes. Tuning determines the gains and control
parameters which have a direct impact on the steady-state error and transient characteristics of the
DDC unit and the control system. A well-tuned DDC unit minimizes the steady-state error, or offset
from the set point; shows a quick response to disturbance; and provides operating stability at all
operating conditions. HVAC&R processes are nonlinear, and system characteristics change when
seasons are varied. The DDC unit tuned at one condition may not be appropriate at other operating
conditions.
Tuning PI Controllers
For DDC units using PI control modes, proportional gain and integral gain should be properly
selected. A high gain decreases the control stability and the offset or error, whereas a low gain will
produce a slow response, which increases the control stability and offset. In tuning PI controllers,
trial and error is often used by the EMCS subcontractor. The trial-and-error method adjusts the gain
until the desirable response to a set-point change is shown. This response should start with a small
overshoot and rapidly damp to steady-state conditions. The trial-and-error method is time-consuming.
Other tuning methods for PI controllers, such as closed- and open-loop process identification,
are also used. Refer to ASHRAE Handbook 1995, HVAC Applications, for details. Bekker et al.
(1991) recommended a root locus tuning method for first-order processes like most temperature and
humidity control for PI controllers to achieve a critically damped response.
Self-Tuning PI and PID Controllers
During the 1980s, different schemes were developed including Astrom and Hagglund’s (1984)
automatic tuning of PI and PID controllers. The tuning procedure suggested by Astrom and
Hagglund is based on the identification of one point on the Nyquist curve of an open-loop system
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.55
with relay feedback. In the earlier 1990s, commercial products of automatic tuning DDC units
started to appear in the EMCS in buildings.
Based on Astrom and Hagglund’s principle, Wallenborg (1991) accomplished automatic tuning
of PID controllers in supply air temperature and duct static pressure control experiments. After automatic
tuning, its system performance was improved. The autotuner used in these experiments was
different from a “true adaptive controller.” An autotuner is operated with fixed parameters during
normal operation. The tuning experiment is initiated between the normal operation periods by the
operator. The required parameters are calculated from the results of the tuning experiments.
Adaptive Control
The controller has the ability to adapt to the control system by determining the optimum PID parameters
and adjusting itself accordingly. An adaptive controller, or self-tuning controller, continuously
updates its parameter during operation based on some on-line process identifier and computer
programs. Self-tuning adaptive controllers are also available as commercial products now.
5.13 FACTORS AFFECTING CONTROL PROCESSES
As defined in Sec. 5.1, the function of an air conditioning (HVAC&R) control system is to modulate
the air conditioning system capacity to match the off-design condition, load variation, and
climate change, to maintain the indoor environment within desirable limits at optimum energy use.
Load
The term load refers to the magnitude of the space load, coil load, refrigeration load, or boiler load
that determines the amount of the supply air, chilled water, or hot water needed to control and
maintain the controlled variable at the desirable value(s). Load variation and disturbances affect the
controlled variable in three circumstances: part-load operation, intermittent operation, and disturbances.
Part-Load Operation. The sizes of an air conditioning system and its components are always selected
according to the magnitude of the load at the design condition, which is often called the design
load or full load. In actual operation, air conditioning or HVAC&R systems operate at part load
most of the time. Many air systems even spend 85 to 90 percent of their annual operating hours at
part-load operation. The load drops below the design load because of
1. Changes in the outdoor climate
2. Changes in the internal loads at the time of operation
During part-load operation, the capacity of an air conditioning system must be appropriately reduced
so that a desirable indoor environment can be maintained; at the same time, the energy use of
the HVAC&R equipment can be saved.
Climate Change
Outdoor climate change affects not only the space load, but also the performance of HVAC&R
systems. Outdoor climate parameters for HVAC&R include the dry-bulb temperature, wet-bulb
temperature, solar radiation, wind speed, and wind direction. They are usually specified at design
conditions for an HVAC&R system or at a rated condition for certain equipment. Outdoor climate
5.56 CHAPTER FIVE
changes during off-design conditions. The general trend in summer when the outdoor dry- and wetbulb
temperatures fall is that the space cooling load drops accordingly. Because each system has its
own characteristics, sometimes it is simpler and more convenient to consider the load ratio and climate
parameter as two different factors during the calculation and analysis of system performance.
For instance, the analysis of the energy use of a water-cooled chiller requires both the load ratio and
the entering condenser water temperature (ECWT) to determine the power input to the compressor.
Intermittent Operation. Many air systems do not operate continuously. Some operate only few
hours within a diurnal cycle. Others operate only in the daytime and shut down at night when the
building is unoccupied. During the warm-up and cool-down periods, space loads vary a great deal
from the design load as well as from that of a continuous part-load operation. The space loads represent
transient loads of a dynamic model. Optimum starting and stopping of intermittently operated
air systems is important for better indoor environmental control and energy savings.
Disturbance. These can be sudden load changes or set-point changes within a short time, say, a
fraction of an hour, that affect the controlled variable. The offset of the controlled variable resulting
from a disturbance can be eliminated by a DDC system with a proportional-integral control mode.
For disturbances resulting from a sudden climate change that affects a 100 percent outdoor airhandling
unit, or from a sudden switch-on of the spotlights in a conditioned space, a DDC system
that incorporates proportional-integral-derivative control mode may be more suitable.
System Capacity
First, the capacity of an air conditioning (HVAC&R) system and its components should be adequate.
An oversized electric heater produces a greater overshoot of space temperature and wastes
more energy than an appropriately sized heater. An undersized chiller always causes a higher space
temperature during the cooling season and the EMCS can become out of control.
Second, from the point of view of capacity control and energy conservation, a heating and
cooling plant installed with multiple units is always better than a single-unit plant. A piece of equipment
will operate more efficiently at a higher turndown ratio than at a lower turndown ratio. The
turndown ratio RTD, often called part-load capability, is usually expressed as the percentage of design
capacity:
(5.20)
Third, modulation of capacity continuously is always better for matching the load to outdoor climate
change than two-position or step controls. Finally, capacity modulation of large fans, pumps,
and compressors using adjustable-frequency variable-speed drives is now often energy-efficient and
cost-effective in many circumstances.
Performance of Control Processes
The performance of an EMCS is often represented by the performance of its control processes, or
more directly its control loops that perform the control functions. HVAC&R control processes are
mostly nonlinear processes whose characteristics cannot be expressed by first-order equations.
Nordeen (1995) listed five criteria to use to evaluate the performance of an HVAC&R control
process:
 Stability. A control system is stable if the controlled variable does not show a continuing trend
away from the set point following a disturbance, as shown in Fig. 5.9a. As described previously, a
wider throttling range and a slower response both are beneficial to system stability. The decay of
RTR 
minimum capacity
design capacity
 100
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.57
the oscillation of the controlled variable is called damping. The more damped a control system is,
the less it oscillates.
 Response time. The length of time needed for the controlled variable to reach the set point after a
step change in set point or load is the response time.
 Overshoot. The difference between the maximum controlled variable and the set point following a
change in set point or load is the overshoot.
 Settling time. The length of time needed for the controlled variable to reach steady-state following
a disturbance.
 Offset. As defined previously, this is the constant error that exists between the controlled variable
and the set point when a steady state is reached.
The problems associated with poor performance include the time lag or dead time, which is large
compared to the reaction rate or time constant of a control system; poor measurements of the controlled
variables; inappropriate capacity control; and hysteresis within control and control components
which create time lag. Time lag or dead time is the delay in time between the change in the
controlled variable and when that change is sensed by the sensor, or when the controlled device is
modulated, or when the capacity of the process plant varies.
During the sizing of control valves and air dampers, a linear relationship between the coil load
and the valve stem travel was the primary design consideration during the past decades. However,
many microprocessor-based DDC systems now permit scaling factors to be applied to the analog
outputs of the DDC units, and thus an inherently nonlinear system will respond in a linear relationship.
Thermal Capacitance
Thermal capacitance, which is sometimes called thermal inertia, is related to the mass of the building
envelope, equipment, and system components. Capacitance is usually calculated by the product
of specific heat of the material and the mass of the material. The thermal capacitance of the
HVAC&R process and the building envelope affects the performance of a control system as follows.
First, the high thermal capacitance of the building envelope, the equipment, chilled and hot water,
etc., reduce the effect of a disturbance to vary the controlled variable. For example, consider a
sudden increase in the lighting load in a conditioned space. As the space air temperature increases
because of heat released from the electric lights, a large portion of heat at the same time will be
transferred to the building envelope because of the additional temperature difference between the
space air and the building structure. The increase in the space temperature resulting from the disturbance
is thus significantly reduced.
Second, during the warm-up and cool-down periods, because of the high thermal capacitance of
the building envelope, heat exchanger, ducts, and pipes, a greater system capacity and longer period
of time are needed to raise or cool down the space air temperature.
Third, the thermal influence of the weather and of external load changes on the conditions of the
indoor space air through the building envelope is rather slow, often needing several hours to reach
full effect.
5.14 FUNCTIONAL CONTROLS
The control functions of a microprocessor-based EMCS in a large high-rise building become more
and more complex and demanding. To provide the required functional controls with satisfactory
system performance is always the primary target of an EMCS. In an EMCS, there are four kinds of
functional controls: generic, specific, safety, and diagnostics.
5.58 CHAPTER FIVE
Generic Controls
Generic controls are usually needed by most of the systems, units (equipment), and components.
The software of generic controls is usually provided in the PC workstation and in the system controllers
except system graphical displays. The editing of the generic software usually takes place in
the PC workstation and can be downloaded to any one of the system controllers. In many EMCSs,
data and information can be monitored and collected from any point in the system and analyzed in
the working station PC and in the system controllers. Graphics displays, trending, totalization,
scheduling, alarming, etc., are examples of the generic controls. The following are features provided
by various manufacturers (or vendors).
Graphical Displays. The graphical displays provided in the PC workstation become the
showpiece of an EMCS. Some EMCS manufacturers demand that the software in the operator PC
workstation be graphically oriented. The graphical displays include the following:
 Building floor plan graphics show the selected floor plan and the space temperatures.
 Equipment graphics are provided for each major piece of equipment such as packaged unit, airhandling
unit, and chiller with status of all points.
 Schematic graphics show the detailed system drawings which can be created, modified, and saved.
 High-resolution digitized photo-quality displays are also accommodated in some EMCSs.
The graphical display system can allow a display of up to 10 graphical screens for comparison and
monitoring of system status. Graphics combined with color coding, such as temperature that is assigned
with different colors at various values, will increase viewer’s effectiveness. A graphical display
system should provide navigation from a facility map down to a specific floor plan. Associated
temperature and related parameters are provided on the floor plan. An additional “click” of the
mouse on any questionable area will cause the serving HVAC&R system with its operating information
to appear on the screen.
Speed and convenience of editing are two key issues in graphical displays. If a 486 processor is
used, graphics may take 5 to 7 s to paint. Including the navigating through 5 to 10 screens in search
of a problem, too much time will be needed. Because of the use of the Pentium chip, it is possible
to reduce the graphical response to 2 to 3 s per screen paint. The operator must create, modify, or
edit the graphics easily.
Graphical displays play a small role in increasing the effectiveness of any functional control in
an EMCS. However, they enable the user to interact more efficiently with the data and information
from the EMCS and to make proper decisions to operate the EMCS and the associated HVAC&R
system effectively and efficiently.
Trending. Trending is the ability to provide continuous track of certain parameter(s), or operating
status of a piece of equipment. The commonly used time interval for the trend log is once for every
30 min. Trending is mostly used for troubleshooting. For an example, if the output capacity of a heat
exchanger is gradually reduced, scale may form on the heat-exchange surface. Totalization is another
trend log which records the total accumulated operated time for each of the units in a plant employing
multiple units. Totalization is usually helpful for maintenance and troubleshooting purposes.
The trend graph displays the trend data in graphical form. When the user asks for a trend graph,
select the point(s). The Y axis of the graph is often automatically scaled; the X axis to indicate time
is appropriately labeled. The user can also choose the capacity of trend points, time interval between
two trend data other than the standard 30 min, and selection of both changes of value (COV)
that exceed the defined value and the timed trend. The trend graph is a powerful tool. It is very
helpful for tuning of PID control loops and analyzing operating problems.
Scheduling. Prior to the DDC systems, the use of a time clock to schedule the start and stop of
equipment automatically was most widely adopted. A DDC system usually starts and stops the
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.59
equipment according to the predetermined schedule in the system controllers. They can be in the
form of day, week, month, and season schedules, and the users can define the executing periods.
The graphical schedules are often used for clarification. Modifications of schedules are easily made
at the PC workstation and downloaded to the system controllers. Optimum start and stop are actually
run optimizing of scheduling.
Some manufacturers provide temporary schedules that will be used only once and supplement
the existing schedule. Sometimes an override schedule which supersedes the existing schedule is
also used only once.
Alarming. This is the software for tracking and reporting the alarm conditions, i.e., abnormal
conditions. The operator needs to determine the limits or the difference from the set points that
causes the abnormal condition. The operator also should determine the reactions, logging, printing,
displaying messages and graphics, and producing audible announcements that are taken during an
alarm. An alarm that is not acknowledged by the operator within a specific time will move to a
higher level of priority. The system must automatically lock out alarms when an alarmed system is
turned off and appropriate reactions have been taken.
Discriminator Control. Discriminator control is a kind of optimum control which searches for the
required lowest cooling supply air temperature and the highest heating supply temperature. This
control also minimizes the amount of mixing of cold and warm supply air, terminal reheat, and resets
the cold deck supply temperature to reduce the mechanical cooling and the amount of reheat.
Fault Detection and Diagnostics. Fault detection and diagnostics should monitor the operation of
HVAC&R equipment, components, and control devices; analyze the data; and identify performance
problems to be corrected. Fault detection and diagnostics are discussed in detail in the next section.
Specific Controls
Specific controls including capacity controls are controls for a specific function in an HVAC&R
system, a unit (or equipment), or a zone. The following are specific controls that are discussed in
the corresponding sections later:
 Zone temperature control
Fan-coil unit
VAV reheat
Fan-powered VAV box
 Air system control
Economizer control
Discharge air temperature control
Minimum ventilation control
Duct static pressure control
Space humidity control
Warm-up or cool-down control
 Water system control
Differential pressure control
Chilled water temperature reset
Condenser water temperature control
Low-temperature heating system control
5.60 CHAPTER FIVE
 Central plant control
Multiple-chiller optimizing control
Multiple-boiler optimizing control
Condenser fan cycling
Demand limit control
 Safety controls
Commissioning and Maintenance
A properly designed and installed EMCS needs commissioning to test and tune its controllers and
system components according to the design specification. A sufficient, clear, and well-followed
operations manual and a well-implemented maintenance schedule are key factors for an effective
and efficient EMCS. Commissioning and maintenance are covered in detail in Chap. 32.
5.15 FAULT DETECTION AND DIAGNOSTICS
Basics
Modern air-handling units (AHUs), packaged units (PUs), and chillers become more and more
complicated because of the IAQ, thermal comfort and energy efficiency requirements, and the use
of DDC. An HVAC&R operator is hardly able to monitor and detect the fault operations of the
AHUs, PUs, and chillers; find their causes; and correct them. An automatic fault detection and diagnostic
system is important for the effective operation and control, for the monitoring and maintenance,
and for the optimizing of utilization and continuous improving of the HVAC&R systems.
As of the late 1990s, fault detection and diagnostics are already a standard component in many
large PUs.
In HVAC&R operations, faults occur when the actual measured operating parameters deviate
from the normal operating values. There are two types of detected faults: complete failures and
performance degradations. Complete failures are abrupt faults that often cause discontinuation of
the operation of a system or component. Symptoms of abrupt faults can be easily observed. Peformance
degradation is the result of an evolving fault accumulated during a certain time.
The differences between the actually measured values of an operating parameter, such as temperature
T, °F (°C); pressure p, in. WG (Pa); volume flow rate , cfm (m3 / s); or mass flow rate ,
lb / s (kg/s) and the expected values (estimated, simulated, or set points) of temperature Texp, pressure
pexp, volume flow rate , or mass flow rate under normal operating conditions are
called residual. A fault can be detected by investigating and analyzing residuals.
A temperature residual Tres, °F (°C), a pressure residual pres, in. WC (Pa), or a volume flow rate
residual , cfm (m3 / s), can be calculated as
(5.21)
where the subscript exp indicates expected (predicted) values and the units of Texp, pexp, and are
the same as those of T, p, and . Most of the measured operating parameters in a fault detection and
diagnostic system are the same monitored parameters (sensed or measured) in an EMCS.
Residuals are often normalized so that the dominant symptom may have approximately the same
V?
V?
exp
V?
res  V? V?exp
pres  p pexp
Tres  T Texp
V?
res
m?exp V?exp
m? V?
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.61
magnitude for different types of faults. The residual R can be normalized as
(5.22)
where Rnor  normalized residual
Rmax, Rmin  maximum and minimum residuals
In the late 1990s, a large rooftop packaged unit made by one U. S. manufacturer offered a standard
fault detection and diagnostic device to review active and historical lists of diagnostic conditions.
A total of 49 different diagnostics can be read at the human interface panel, and the last 20
diagnostics can be held in an active history buffer log at the panel. A human interface panel provides
2-line 40-character liquid-crystal display and a 16-button keypad for monitoring, setting,
controlling, and diagnostics.
In the late 1980s, the earlier development of diagnostics of HVAC&R system operation was
mainly rule-based expert systems. During the late 1990s, the development of automating fault detection
and diagnosis was emphasized. Inputs and outputs of an HVAC&R operating process can
be mathematically related by using autoregressive models with exogenous inputs (ARX), artificial
neural network (ANN) models, and many other developing models. Both ARX and ANN are
called black-box because they require less physical knowledge of the operating process. These
technologies are expected to be commercially available after laboratory and field tests in the
early 2000s.
Expert System Rule-Based Diagnostics
Expert systems are discussed in Sec. 5.10, and diagnostics is one of the applications used in
HVAC&R system operations. Brambley et al. (1998) reported a tool—the outdoor air /economizer
(OAE) diagnostician—which monitors the performance of an air-handling unit to provide outdoor
air as a constant fraction of its supply flow rate and detects problems of outdoor ventilation air control
and economizer operation. An OAE diagnostician uses sensors that are commonly installed for
control purposes and diagnoses the operating problems based on rules derived from engineering
models of proper and improper performance of the AHUs. These rules are implemented in a decision
tree structure in computer software. The diagnostician collects data periodically from the
Building Automation System (BAS) to navigate the decision tree and produce conclusions. At each
point of the tree, a rule is evaluted according to the collected data, and the results determine in
which branch the diagnosis should be. At the end of the branch, a conclusion is reached corresponding
to the current operating condition of the AHU. The operator or installer of the diagnostician enters
data only once during setup.
The OAE diagnostician was installed and operated on three AHUs in a newly constructed and
occupied 200,000-ft2 (18,600-m2) DOE William R. Wiley Environmental Molecular Sciences Laboratory
and on four AHUs in a 72,700-ft2 (6760-m2) Technical Management Center, both in Rishland,
Washington.
For each AHU, data were recorded hourly from sensors in the BAS for outside temperature, return
air temperature, mixed air temperature, discharge air temperature, on/off status of the supply
fan, and open/closed status of the chilled water and hot water valves. No sensors were installed specially
for the diagnostician; all are used by the BAS for the control purpose. The data were automatically
transferred hourly from the BAS to the diagnostician’s database. The diagnostician then
processed these data and produced the diagnostic results that can be viewed on the display.
Shortly after initial processing of data, four of the seven AHUs monitored were found to have
problems. The problems included a sensor malfunction, a return damper that was not closed fully, a
mixed-air sensor problem, and a chilled water control problem. All problems were confirmed after
inspection.
Rnor 
R Rmin
Rmax Rmin
5.62 CHAPTER FIVE
The diagnostician used color coding on a display to alert the HVAC&R system operator when
the problem happened and provided assistance in modifying the causes as well as making the
correction. The OAE diagnostician has proved itself effective in identifying installation and operating
problems during the field testing of the outdoor ventilation control and economizer cycle.
Dexter and Benouarets (1996) introduced a fault diagnosis of HVAC&R plant items based on
semiqualitative generic reference models. The scheme uses reference models describing fault-free
and fault operations that are collected from data producd by simulating a number of HVAC&R systems
(or subsystems) of the same type. Results have showed that a fuzzy-logic-based fault diagnosis
requiring no training on the actual plant can successfully identify faults in a simulated HVAC&R
subsystem if the parameter denoting faults is sufficiently large. The proposed fuzzy method of fault
diagnosis is computationally simple enough to be used for more complex subsystems, such as
AHUs.
Stylianou and Nikanpour (1996) present a methodology that uses thermodynamic modeling,
pattern recognition, and expert knowledge to determine whether a reciprocating chiller is fault-free
and to diagnose selected faults. The status of this chiller consists of the off cycle, start-up, and
steady-state operation. The off cycle deals with those faults that are more easily detected when the
chiller is off. Start-up deals with faults which are related to refrigerant flow characteristics and are
generally more apparent during the transient period. During steady-state operation, faults of performance
deterioration are detected.
This methodology requires training data. Data can be collected during commissioning and online
measurement of variables to build up normal steady-state linear regression models and data
collected from manufacturers. This methodology also needs additional treatment to improve the de-
finition of the thresholds to classify emerging patterns and to establish the range of applicability of
these patterns. Further study is required to produce a list of faults and their associated transient
characteristics and steady-state patterns.
ARX and ANN Model-Based Diagnosis
Peitsman and Bakker (1996), Peitsman and Soethout (1997), Lee et al. (1996), and Yoshida et al.
(1996) discussed the application of ARX and ANN model-based fault detection and diagnosis and
the related laboratory testing results. Model-based diagnosis is a technique capable of finding diagnoses
based on the behavior of the system and components. The behavior is best understood as
the interaction of observation and prediction expressed in the difference between the measured
(observed) and predicted parameter values as previously covered, called residual. The behavior of the
system and components is defined using ARX or ANN models. One advantage of a model-based diagnosis
over rule-based diagnosis is that the model-based diagnosis does not rely on symptom-fault
patterns. Such patterns are incomplete, and it is difficult for an expert to anticipate all possible faults
and to predict their symptoms. Another advantage is that a model-based diagnosis can be used for a
new system even when there is no repair experience or only when a system model is available.
ARX Model. Consider a process in an autoregressive model with exogenous inputs (ARX) having
an input signal u(t) and an output signal y(t) and described by a simple input/output relationship as
(5.23)
where t  time
e(t)  enter as a direct error
a  autoregressive parameter (order n)
b  exogenous parameter (order m)
The model relating output and input and extra (exogenous) variables is linear. The actual performance
of a system is nonlinear. However, such an approximation is allowable in fault detection and
y(t) 	 a1y(t 1) 	    	 any(t n)  b1u(t 1) 	    	 bmu(t m) 	 e(t)
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.63
diagnosis. Extra inputs are added, and the use of historical measured inputs (t 1, t 2, . . . )
will make it more nearly a nonlinear relationship.
Procedure to Identify Parameters. After the ARX model has been selected, according to Peitsman
and Bakker (1996), the procedure to identify parameters a and b in Eq. (5.23) is as follows:
 Measure input/output data. Input and output values of the process are collected in a data set that
consists of measured and design (predicted) data under “healthy” (fault-free) conditions. The
measured data sets should contain sufficient dynamic conditions of the process.
 Model order (m and n) should be selected as the optimum values by trial and error.
 From the collected input/ output data, estimate parameters a and b.
 Validate the selected model. The accuracy of the model is tested by comparing the predicted value
with the measured values.
ANN Models. For an ANN model of three layers with an input layer, a hidden layer, and an output
layer, as discussed in Sec. 5.10, the outputs can be calculated as a function of the weighted sum
of the inputs, and an additional factor, the biases B as expressed in Eqs. (5.16) and (5.17):
where iLn  input activation to a neuron in hidden layer
WLon  weights of connection between hidden layer and outputs
ANN models are nonlinear. A learning or training process in an ANN model is necessary. The initially
assigned connection weights are adjusted repeatedly during the learning and training process
until the error is within specified values, as discussed in Sec. 5.10. After training, an ANN model
can predict the outputs from given inputs.
Reliability of ARX and ANN Models. Peitsman and Bakker (1996) recognized that the reliability
of the ARX and ANN models (the use of mathematic equations to forecast the outputs) is good.
One of the conditions needed to produce a reliable model is a healthy and dynamic data set that is
measured under optimal and good operating conditions over a wide working range. Another necessary
condition is the availability of a fixed time interval (time step) because the regression models
use values of the previous time interval of the inputs and outputs to predict the output.
System and Component Models. In Peitsman and Bakker (1996), two types of models are produced
with the measured and predicted learned data:
 System models. A chiller or an AHU is considered a black box in which multiple independent inputs
estimate outputs. Independent variables are necessary to form a system model. This model
checks the whole system performance. In a system model if the possibility of a malfunction is detected,
the next step is to use the component model to localize this malfunction.
 Component models. Major components are considered as black boxes. The configuration of component
models is accomplished after studying the structure of the modeled HVAC&R systems.
The model becomes a little bit gray (with some physical knowledge) instead of black (without
physical knowledge). After the system model has detected a malfunction, the component model
can pinpoint the cause of the malfunction with a greater accuracy than the system model can.
Fault Detection and Diagnosis. After the parameters (variables) of a system model or component
model are simulated using an ARX or ANN model, the predicted outputs of the model must be
V  	iLnWLon
oon 
1
1 	 e(VB)
5.64 CHAPTER FIVE
compared with the measured outputs of the system or the components (temperature, pressure, volume
flow) as shown in Fig. 5.27. Only when the measured output is not within the threshold of the
predicted output may there be a detected fault or an incorrect model used.
The threshold is usually determined from the statistical properties of the process. A residual
(margin) of is used. Here x indicates the measured value, xexp is the assumed predicted
or expected mean value, and  is the standard deviation of the predicted values. The probability
that x falls within is 0.9973. If , a detected fault may
occur, as shown in Fig. 5.27, in which the difference between the measured and predicted discharge
air temperatures is greater than 3 . The possible causes of the fault include block of chilled water
flow, high chilled water temperature, and a fouled cooling coil.
Comparison of ARX and ANN Models. According to Peitsman and Bakker (1996), the comparison
of ARX and ANN models for fault detection and diagnosis is as follows:
ARX ANN
Prediction of process variables Linear Nonlinear
Results ANN gives slightly
better results than
ARX models
Training period Shorter Longer
 x xexp   3 
[xexp 3 , xexp 	 3 ]
 x xexp   3 
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.65
Measured
Predicted
Detected
fault
0
5
100 200 300 400
Time, s
TresTTexp
500 600 700 800
Discharge air temperature Tdis,F
Discharge air temperature Tdis,C
40
50
60
70
10
15
20
FIGURE 5.27 Comparison of the measured and predicted discharge air temperatures for a VAV system in an ARX
system model. (Source: Peitsman and Bakker (1996) ASHRAE Transactions Part I 6637. Reprinted with permission.)
5.16 CONTROLS IN ASHRAE/IESNA STANDARD 90.1-1999
ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, specifies the following mandatory provisions for HVAC&R controls. For exceptions and
details refer to Standard 90.1-1999.
General
 Thermostatic controls. The supply of heating and cooling energy to each zone shall be controlled
individually by thermostatic controls using the zone temperature as the response. A dwelling unit
is permitted to be considered as a single zone.
 Dead band. When both heating and cooling are controlled, zone thermostatic controls shall be capable
of providing a temperature range or dead band of at least 5°F (2.8°C). Within that range the
heating and cooling energy supplied to the zone is shut off.
 Set-point overlap. When heating and cooling supply to a zone are controlled by separate zone
thermostatic controls, means such as limit switches, mechanical stops, and software for DDC systems,
shall be provided to prevent the heating set point exceeding the cooling set point minus any
proportional band.
Off-Hour Controls
HVAC systems with a design heating or cooling capacity greater than 65,000 Btu/h (19,050 W) and
fan system power greater than hp (0.56 kW) shall have all the following off-hour controls: automatic
shutdown, setback controls, optimum start controls, shutoff damper controls, and zone isolation.
 Automatic shutdown. HVAC systems shall be equipped with at least one of the following:
1. Controls that can start and stop the system under different time schedules, are capable of retaining
programming and time setting during loss of power for a period at least 10 hours, and
include an accessible manual override or equivalent that allows temporary operation of the
system for up to 2 hours.
2. An occupancy sensor that can shut the system off when no occupant is sensed for a period of
up to 30 minutes.
3. A manual operated timer that can adjust the system for two hours.
4. An interlock to a security system which shuts the system off when the security system is
activated.
 Setback controls. A heating system located where the heating outdoor design temperature is 40°F
(4.4°C) or less shall be equipped with controls that can automatically restart and temporarily operate
the system as required to maintain zone temperature above a heating set point adjustable
down to 55°F (12.8°C) or lower. A cooling system located where the cooling outdoor design temperature
is greater than 100°F (37.8°C) shall be equipped with controls that can automatically
restart and temporarily operate the system as required to maintain zone temperature below a cooling
set point adjustable up to 90°F (32.2°C) or higher or to prevent high space humidity.
 Optimum start controls. An air system with a supply volume flow rate exceeding 10,000 cfm
(4720 L/ s), served by one or more supply fans, shall have optimum controls. The optimum start
shall be at least a function of the difference between space temperature and occupied setpoint, and
the amount of time prior to scheduled occupancy.
 Shutoff damper controls. Both outdoor air supply and exhaust system shall be installed with motorized
dampers that will automatically shut off when the systems or the space are not in use. Out-
3	4
5.66 CHAPTER FIVE
door dampers for ventilation can be automatically shut off during morning warm-up and cooldown,
or nighttime setback except when ventilation reduces energy costs, or ventilation must be
supplied. Both outdoor air or exhaust dampers shall have a maximum air leakage of 3 cfm/ ft2
(15 L/ s m2) at a pressure difference of 1.0 in WC (250 Pa). As an exception, gravity dampers
(nonmotorized) are acceptable in buildings less than three stories in height or for buildings of any
height in climates with HDD65 less than 2700. Also excepted are systems with a design outside
air intake of 300 cfm (142 L/s) or less that are equipped with motor operated dampers.
 Zone isolation. HVAC systems serving zones that are intended to operate or be occupied nonsimultaneously
shall be divided into isolation areas. Each isolation area shall be no larger than
25,000 ft2 (2320 m2) of floor area nor include more than one floor. Each isolation area shall be
equipped wih isolation devices that can automatically shut off the supply of conditoned air, outside
air, and exhaust air from the area. Each isolation area can be controlled independently.
REFERENCES
Alexander, J., Aldridge, R., and O’Sullivan, D.,Wireless Zone Sensors, Heating/Piping/Air Conditioning,
no. 5, 1993, pp. 37–39.
Alley, R. L., Selecting and Sizing outside and Return Air Dampers for VAV Economizer Systems, ASHRAE
Transactions, 1988, Part I, pp. 1457–1466.
Alpers, R., and Zaragoza, J., Air Quality Sensors for Demand Controlled Ventilation, Heating/Piping/Air
Conditioning, no. 7, 1944, pp. 89–91.
Anderson, R., Gems to Look for in EMCS, Heating/Piping/Air Conditioning, no. 11, 1991, pp. 47–52.
ANSI/ASHRAE, Standard 135–1995, BACnet: A Data Communication Protocol for Building Automation and
Control Networks, ASHRAE Inc., Atlanta, GA, 1995.
Asbill, C. M., Direct Digital vs. Pneumatic Controls, Heating/Piping/Air Conditioning, November 1984,
pp. 111–116.
ASHRAE, ASHRAE Handbook 1995, HVAC Applications, Atlanta, GA, 1995.
ASHRAE/IESNA, Standard 90.1–1999, Energy Standard for Buildings Except Low-Rise Residential Buildings,
ASHRAE Inc., Atlanta, GA, 1999.
Astrom, K. J., and Hagglund, T., A New Auto-Tuning Design, IFAC International Symposium on Adaptive
Control of Chemical Processes, Copenhagen, 1984.
Avery, G., Selecting and Controlling Economizer Dampers, Heating/Piping/Air Conditioning, no. 8, 1996,
pp. 73–78.
Becker, H. P., How Much Sense Do Room Occupancy Sensor Controls Make? ASHRAE Transactions, 1986,
Part I, pp. 333–342.
Bekker, J. E., Meckl, P. H., and Hittle, D. C., A Tuning Method for First Order Processes with PI Controllers,
ASHRAE Transactions, 1991, Part II, pp. 19–23.
Brambley, M., Pratt, R., Chassin, D., Katipamula, S., and Hatley, D., Diagnostics for Outdoor Air Ventilation
and Economizers, ASHRAE Journal no. 10, 1998, pp. 49–55.
Brothers, P., and Cooney, K., A Knowledge-Based System for Comfort Diagnostics, ASHRAE Journal, no. 9,
1989, pp. 60–67.
Bushby, S. T., and Newman, M., BACnet: A Technical Update, ASHRAE Journal, no. 1, 1994, pp. S72–S84.
Curtiss, P. S., Shavit, G., and Kreider, J. F., Neural Networks Applied to Buildings—A Tutorial and Case Studies
in Prediction and Adaptive Control, ASHRAE Transactions, 1996, Part I, pp. 1141–1146.
Davison, F. G., Direct Digital Control Documentation Employing Graphical Programming, ASHRAE Journal,
no. 9, 1992, pp. 46–52.
Dexter, A. L., and Benouarets, M., A Generic Approach to Identifying Faults in HVAC Plants, ASHRAE Transactions,
1996, Part I, pp. 550–556.
Elyashiv, T., Beneath the Surface: BACnet Data Link and Physical Layer Options, ASHRAE Journal, no. 11,
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.67
1994, pp. 32–36.
French, J. C., Object-Oriented Programming of HVAC Control Devices, ASHRAE Journal, no. 12, 1991, pp.
33–41.
Gibson, G. L., and Kraft, T. T., Electric Demand Prediction Using Artificial Neural Network Technology,
ASHRAE Journal, no. 3, 1993, pp. 60–68.
Glinke, T. J., The Open Protocol Choice: A Market’s Decision in Process, ACH&R News, Oct. 6, 1997, pp. 6–8.
Grimm, N. R., and Rosaler, R. C., Handbook of HVAC Design, McGraw-Hill, New York, 1990.
Haines, R. W., Proportional plus Integral Control, Heating/Piping/ Air Conditioning, January 1984,
pp. 131–132.
Haines, R. W., Reset Schedules, Heating/Piping/Air Conditioning, September 1985, pp. 142–146.
Hall, J. D., and Deringer, J. J., Computer Software Invades the HVAC Market, ASHRAE Journal, no. 7, 1989,
pp. 32–44.
Hartman, T. B., Direct Digital Controls for HVAC Systems, McGraw-Hill, New York, 1993.
Hayner, A. N., Engineering for Energy Management, Engineered Systems, no. 10, 1993, pp. 23–26.
Honeywell, Engineering Manual of Automatic Control for Commercial Buildings, Honeywell Inc., Minneapolis,
MN, 1988.
Huang, S., and Nelson, R. M., Rule Development and Adjustment Strategies of a Fuzzy Logic Controller for an
HVAC system, Part One—Analysis and Part Two—Experiment, ASHRAE Transactions, 1994, Part I, pp.
841–856.
Jafar, M., Bahill, A. T., and Osborn, D. E., A Knowledge-Based System for Residential HVAC Applications,
ASHRAE Journal, no. 1, 1991, pp. 20–26.
Judson, K. W., An Argument for VOC Sensors, Engineered Systems, no. 4, 1995, pp. 34–36.
Kammers, B. K., Have You Seen BACnet Yet? Engineered Systems, no. 6, 1996, pp. 24–44.
Kawashima, M., Artificial Neural Network Backpropagation Model with Three-Phase Annealing Developed for
the Building Energy Predictor Shootout, ASHRAE Transactions, 1994, Part II, pp. 1096–1103.
Kovacs, M., Intelligent Network Sensors, Engineered Systems, no. 9, 1996, pp. 28–34.
Lee, J., and Russell, D., BACnet: Agent for Change, Engineered Systems, no. 6, 1996, pp. 24–40.
Lee,Won-Yong, Park, C., and Kelly, G. E., Fault Detection in an Air-Handling Unit Using Residual and Recursive
Parameter Identification Methods, ASHRAE Transactions, 1996, Part I, pp. 528–539.
Lehr, V. A., Fuzzy Logic: A Technology and Design Philosophy, Heating/Piping/Air Conditioning, no. 5, 1996,
pp. 41–46.
Miller, R. C., and Seem, J. E., Comparison of Artificial Neural Networks with Traditional Methods of Predicting
Return Time from Night or Weekend Setback, ASHRAE Transactions, 1991, Part II, pp. 500–508.
Nordeen, H., Fundamentals of Control From a System Perspective, Heating/Piping/Air Conditioning, no. 8,
1995, pp. 33–38.
Peitsman, H. C., and Bakker, V. C., Application of Black-Box Models to HVAC Systems for Fault Detection,
ASHRAE Transactions, 1996, Part I, pp. 628–640.
Peitsman, H. C., and Soethout, L. L., ARX Models and Real-Time Model-Based Diagnosis, ASHRAE Transactions,
1997, Part I, pp. 657–671.
Petze, J., Understanding Temperature Sensing Methods and Myths, Heating/Piping/Air Conditioning, November
1986, pp. 193–208.
Petze, J., Modularity and the Design of Building Automation Systems, Heating/Piping/Air Conditioning, no. 8,
1995, pp. 43–46.
Robertson, R., and Moult, R.,Working Together, Engineered Systems, no.7, 1999, pp. 74–79.
Schell, M., Making Sense Out of Sensors, Engineered Systems, no. 2, 1996, pp. 108–116.
Scholten, A., Fuzzy Logic Control, Engineered Systems, no. 6, 1995, pp. 72–78.
Shams, H., Nelson, R. M., Maxwell, G. M., and Leonard, C., Development of a Knowledge-Based System for
HVAC Type Selection, ASHRAE Journal, no. 8, 1995, pp. 165–171.
Shavit, G., The Evolution of Control during the Past 100 Years, ASHRAE Transactions, 1995, Part I, pp.
538–544.
Shinn, K. E., A Specifier Guide to BACnet , ASHRAE Journal, no. 4, 1994, pp. 54–58.
5.68 CHAPTER FIVE
Skaer, M., Sensor Savvy, Engineered Systems, no. 7, 1996, pp. 24–32.
Stylianou, M., and Nikanpour, D., Performance Monitoring, Fault Detection, and Diagnosis of Reciprocating
Chillers, ASHRAE Transactions, 1996, Part I, pp. 615–627.
Swan, B., The Language of BACnet, Engineered Systems, no. 7, 1996, pp. 24–32.
The Trane Company, Building Control Unit Sizing for Tracer Summit Systems, Engineering Bulletin, American
Standard Inc., 1993.
The Trane Company, Guide Specifications ICS-GS-3, American Standard Inc., 1993.
Turpin, J. R., The Great Divide, The Experts Offer Their Opinions on Open Systems Standards, Engineered Systems,
no. 5, 1996, pp. 35–36.
Turpin, J. R., Interoperability,Where Art Thou? Engineered Systems, July 1999, pp. 56–70.
Wallenborg, A. O., A New Self–Tuning Controller for HVAC Systems, ASHRAE Transactions, 1991, Part I,
pp. 19–25.
Yoshida, H., Iwami, T.,Yuzawa, H., and Suzuki, M., Typical Faults of Air–Conditioning Systems and Fault
Detection by ARX Model and Extended Kalman Filter, ASHRAE Transactions, 1996, Part I, pp. 557–564.
Zhang, Z. J., Another Look at Traditional Control Valve Sizing Practice, ASHRAE Journal, no. 2, 1993,
pp. 38–41.
Zimmerman, A. J., Fundamentals of Direct Digital Control, Heating/Piping/Air Conditioning, no. 5, 1996,
pp. 49–59.
ENERGY MANAGEMENT AND CONTROL SYSTEMS 5.69
6.1
CHAPTER 6
LOAD CALCULATIONS
6.1 SPACE LOAD CHARACTERISTICS 6.2
Space, Room, and Zone 6.2
Convective and Radiative Heat 6.2
Space and Equipment Loads 6.3
Night Shutdown Operating Mode 6.3
Influence of Stored Heat 6.6
6.2 COOLING LOAD AND COIL LOAD
CALCULATIONS 6.6
Components of Cooling Load 6.6
Components of Cooling Coil Load 6.7
Difference between Cooling Load and
Cooling Coil Load 6.8
Load Profile 6.9
Peak Load and Block Load 6.9
Characteristics of Night Shutdown
Operating Mode 6.10
Moisture Transfer from the Building
Structures 6.10
6.3 HISTORICAL DEVELOPMENT OF
COOLING LOAD CALCULATIONS 6.11
6.4 METHODOLOGY—HEAT
BALANCE 6.12
The Physical Model 6.12
Heat Balance Equations 6.12
Heat Balance of Space Air 6.13
Characteristics of Heat
Balance Method 6.14
6.5 METHODOLOGY—TRANSFER
FUNCTION 6.14
Basics 6.14
Transfer Function and Time Function 6.14
Calculation Procedure 6.15
CLTD/SCL/CLF Method 6.15
TETD/TA Method 6.15
6.6 DETAILED CALCULATION
PROCEDURES FOR TFM 6.16
Conduction Heat Gain through Exterior
Walls and Roofs 6.16
Heat Gain through Ceilings, Floors, and
Interior Partition Walls 6.16
Solar Heat Gain and Conductive Heat
Gain through Window Glass 6.17
Internal Heat Gain 6.17
Infiltration 6.24
Cooling Load Conversion Using Room
Transfer Function 6.24
Space Cooling Load Calculation 6.25
Heat Extraction Rate and Space Air
Transfer Function 6.25
Heat Loss to Surroundings 6.25
6.7 DETAILED CALCULATION PROCEDURE
USING CLTD/SCL/CLF METHOD 6.26
Space Cooling Load due to Heat Gain
through Exterior Walls and Roofs and
Conductive Gain through Glass 6.26
Space Cooling Load due to Solar Heat
Gain through Fenestration 6.28
Space Cooling Load due to Heat Gain
through Wall Exposed to
Unconditioned Space 6.28
Calculation of Internal Cooling Loads and
Infiltration 6.29
Space Cooling Load of Night Shutdown
Operating Mode 6.32
6.8 COOLING COIL LOAD 6.32
Basics 6.32
Fan Power 6.33
Duct Heat Gain 6.33
Temperature of Plenum Air and
Ventilation Load 6.34
6.9 COOLING LOAD CALCULATION BY
FINITE DIFFERENCE METHOD 6.34
Finite Difference Method 6.34
Simplifying Assumptions 6.36
Heat and Moisture Transfer at
Interior Nodes 6.36
Heat and Moisture Transfer at
Surface Nodes 6.37
Space Air Temperature and
Cooling Loads 6.38
6.10 HEATING LOAD 6.39
Basic Principles 6.39
Transmission Loss 6.39
Adjacent Unheated Spaces 6.40
Latent Heat Loss and Heat Loss from
Products 6.41
Infiltration 6.41
Setback of Night Shutdown
Operation 6.41
6.11 LOAD CALCULATION
SOFTWARE 6.42
Introduction 6.42
TRACE 600—Structure and Basics 6.42
TRACE 600 Input—Load
Methodology 6.43
TRACE 600 Input—Job 6.44
TRACE 600 Input—External Loads 6.45
TRACE 600 Input—Schedules 6.45
TRACE 600 Input—Internal Loads 6.46
TRACE 600 Input—Partition and Shading
Devices 6.47
TRACE 600—Minimum Input
Requirements, Run, and Outputs 6.47
REFERENCES 6.49
6.1 SPACE LOAD CHARACTERISTICS
Space, Room, and Zone
Space indicates either a volume or a site without a partition or a partitioned room or group of
rooms. A room is an enclosed or partitioned space that is usually treated as a single load. A conditioned
room often has an individual control system. A zone is a space, or several rooms, or units of
space having some sort of coincident loads or similar operating characteristics. A zone may or may
not be an enclosed space, or it may consist of many partitioned rooms. It could be a conditioned
space or a space that is not air conditioned. A conditioned zone is always equipped with an individual
control system. A control zone is the basic unit of control.
Convective and Radiative Heat
Whether heat enters the conditioned space from an external source or is released to the space from
an internal source, the instantaneous heat gains of the conditioned space can be classified into two
categories: convective heat and radiative heat, as shown in Fig. 6.1. When solar radiation strikes the
outer surface of a concrete slab, most of its radiative heat is absorbed by the slab; only a small portion
is reflected. After absorption, the outer surface temperature of the slab increases. If the slab and
the conditioned space are in thermal equilibrium originally, there is then a convective heat and
radiative heat transfer from the surface of the slab to the space air and other surfaces. Meanwhile,
heat transfer due to conduction takes place from the surface to the inner part of the slab. Heat is
then stored inside the slab. The stored heat is released to the space air when the surface temperature
falls below the temperature of the inner part of the slab.
6.2 CHAPTER SIX
FIGURE 6.1 Convective and radiant heat in a conditioned space and the temperatures of the interior surfaces.
To maintain the preset space air temperature, the heat that has been convected or released to the
conditioned space should be removed from the space instantaneously.
Space and Equipment Loads
The sensible and latent heat transfer between the space air and the surroundings can be classified as
follows:
1. Space heat gain qe, in Btu/h (W), represents the rate at which heat enters a conditioned space
from an external source or is released to the space from an internal source during a given time interval.
2. Space cooling load, often simply called the cooling load Qrc, Btu /h (W), is the rate at which
heat must be removed from a conditioned space so as to maintain a constant temperature and acceptable
relative humidity. The sensible cooling load is equal to the sum of the convective heat transfer
from the surfaces of the building envelope, furnishings, occupants, appliances, and equipment.
3. Space heating load Qrh, Btu /h (W), is the rate at which heat must be added to the conditioned
space to maintain a constant temperature and sometimes a specified relative humidity.
4. Space heat extraction rate Qex, Btu /h (W), is the rate at which heat is actually removed from
the conditioned space by the air system. The sensible heat extraction rate is equal to the sensible
cooling load only when the space air temperature remains constant.
5. Coil load Qc, Btu /h (W), is the rate of heat transfer at the coil. The cooling coil load Qcc,
Btu/h (W), is the rate at which heat is removed by the chilled water flowing through the coil or is
absorbed by the refrigerant inside the coil.
6. The heating coil load Qch, Btu /h (W), is the rate at which heat is added to the conditioned air
from the hot water, steam, or electric heating elements inside the coil.
7. Refrigerating load Qrl, Btu /h (W), is the rate at which heat is absorbed by the refrigerant at
the evaporator. For central hydronic systems, the refrigerating load is the sum of the coil load plus
the chilled water piping heat gain, pump power heat gain, and storage tank heat gain. For most
water systems in commercial buildings, the water piping and pump power heat gain is only about
5 to 10 percent of the coil load. In an air conditioning system using DX coil(s), the refrigerating
load is equal to the DX coil load.
The instantaneous sensible heat gain of a conditioned space is not equal to the instantaneous
sensible cooling load because of storage of part of the radiative heat inside the building structures.
Such phenomenon results in a smaller instantaneous cooling load than that of the heat gain when it
is at its maximum value during a diurnal cycle, as shown in Fig. 6.2.
If the space relative humidity is maintained at an approximately constant value, the storage
effect of the moisture in the building envelope and furnishings can be ignored. Then the instantaneous
space latent heat gain will be the same as the instantaneous space latent cooling load.
Night Shutdown Operating Mode
In many commercial buildings, air systems are often shut down during the nighttime or during unoccupied
periods. The operating characteristics of the conditioned space in a 24-h diel cycle can
then be divided into three periods, as shown in Fig. 6.3.
Night Shutdown Period. This period commences when the air system is switched off and ends
when the air system is switched on again. When the air-handling unit, packaged unit, or terminal
unit is turned off during summer in a hot and humid area, the infiltrated air (through the lift, pipe
shafts, and window cracks) and any heat transfer from the window glass, external wall, or roofs will
cause a sudden increase in indoor temperature Tr of a few degrees Fahrenheit (Celsius). After that,
LOAD CALCULATIONS 6.3
Tr will rise farther or drop slightly, depending on the difference between outdoor and indoor temperatures.
At the same time, the indoor space relative humidity increases gradually because of the infiltration
of hot and humid outdoor air and the moisture transfer from the wetted surfaces in the air
system. Higher space relative humidity causes a moisture transfer from the space air to the building
envelopes and furnishings.
During winter, after the air system is turned off, the indoor temperature drops because of the
heat loss through external windows, walls, and roofs, as well as the infiltration of outdoor cold air.
Meanwhile, the space relative humidity increases mainly owing to the fall of the indoor space temperature.
Cool-down or Warm-up Period. This period commences when the air system begins to operate
and ends when the space temperature or other controlled variables have attained predetermined
limits. During summer, the supply of cold and dehumidified air after the air system is turned on
causes a sudden drop in space air temperature and relative humidity. Both heat and moisture are
transferred from the building envelope to the space air because of the comparatively higher temperature
and moisture content of the building envelope. These heat and moisture transfers form
the cool-down cooling load. Sometimes, the cool-down load can be the maximum summer design
cooling load.
6.4 CHAPTER SIX
FIGURE 6.2 Solar heat gain and heat gain from electric lights and the corresponding space cooling loads
for a night shutdown air system.
6.5
FIGURE 6.3 Operating characteristics of an air conditioned space operated at night shutdown mode: (a) summer
cooling mode (hot and humid area) and (b) winter mode.
If refrigeration is used for cooling during the cool-down period, then it usually lasts less than
1 h, depending mainly on the tightness of the building and the differences of the space temperatures
and humidity ratios between the shutdown and cool-down periods. If the cool-down of the space air
and the building envelope is by means of the free cooling of outdoor air, the cool-down period may
last for several hours.
During winter, the supply of warm air during the warm-up period raises the space temperature
and lowers the space relative humidity. Because of the higher space temperature, heat transfer from
the space air to the colder building envelope and furnishings, and the heat energy required to raise
the temperature of the building envelope, forms the warm-up heating load.
Conditioning Period. This period commences when the space air temperature has fallen or risen
to a value within the predetermined limits. It ends when the air system is shut down. In summer,
cold and dehumidified air is usually supplied to the space to offset the space cooling load and to
maintain a required temperature and relative humidity. During winter, warm air is supplied to
compensate for heat losses from the conditioned space. Space temperature is often controlled and
maintained within predetermined limits by control systems.
In commercial buildings located in areas with a cold winter, the air system may operate in dutycycling
mode during nighttime, unoccupied periods in winter, or intermediate seasons to maintain a
night setback temperature. In duty-cycling mode, the fan and heater turn on and off to maintain a
desired temperature. Space temperature is set back at night, e.g., to 55 or 60°F (12.8 or 15.6°C), to
prevent freezing of water pipes and to produce a comparatively smaller temperature lift during a
warm-up period. The operation of the air system in a 24-h diel cycle in winter is then divided into
night setback period, warm-up period, and conditioning period. The required load on the heating
coil during a morning warm-up period at winter design conditions is usually the winter design heating
load.
Influence of Stored Heat
The curve of solar heat gain from a west window is shown in Fig. 6.2, as well as the cooling load
curve due to this solar heat gain for a conditioned space operated at night shutdown mode in summer.
The difference between the maximum solar heat gain qhgm and the maximum cooling load Qclm
during the conditioning period indicates the amount of heat stored inside the building structures.
This difference significantly affects the size of air conditioning equipment required. The amount
of heat stored depends mainly on the mass of the building envelope (whether it is heavy, medium,
or light), the duration of the operating period of the air system within a 24-h cycle, and the characteristics
of heat gain, whether radiant heat or convective heat predominates. For solar radiation
transmitted through a west window, Qclm may have a magnitude of only 40 to 60 percent of qhgm.
ASHRAE Handbook divides the mass of the building construction into the following three
groups:
Heavy construction: approximately 130 lb/ ft2 (634 kg/m2) floor area
Medium construction: approximately 70 lb/ ft2 (342 kg/m2) floor area
Light construction: approximately 30 lb/ ft2 (146 kg/m2) floor area
6.2 COOLING LOAD AND COIL LOAD CALCULATIONS
Components of Cooling Load
Cooling load calculations for air conditioning system design are mainly used to determine the volume
flow rate of the air system as well as the coil and refrigeration load of the equipment—to size
6.6 CHAPTER SIX
the HVAC&R equipment and to provide the inputs to the system for energy use calculations in
order to select optimal design alternatives. Cooling load usually can be classified into two categories:
external and internal.
External Cooling Loads. These loads are formed because of heat gains in the conditioned space
from external sources through the building envelope or building shell and the partition walls.
Sources of external loads include the following cooling loads:
1. Heat gain entering from the exterior walls and roofs
2. Solar heat gain transmitted through the fenestrations
3. Conductive heat gain coming through the fenestrations
4. Heat gain entering from the partition walls and interior doors
5. Infiltration of outdoor air into the conditioned space
Internal Cooling Loads. These loads are formed by the release of sensible and latent heat from
the heat sources inside the conditioned space. These sources contribute internal cooling loads:
1. People
2. Electric lights
3. Equipment and appliances
If moisture transfers from the building structures and the furnishings are excluded, only infiltrated
air, occupants, equipment, and appliances have both sensible and latent cooling loads. The remaining
components have only sensible cooling loads.
All sensible heat gains entering the conditioned space represent radiative heat and convective
heat except the infiltrated air. As in Sec. 6.1, radiative heat causes heat storage in the building
structures, converts part of the heat gain into cooling load, and makes the cooling load calculations
more complicated. Latent heat gains are heat gains from moisture transfer from the occupants,
equipment, appliances, or infiltrated air. If the storage effect of the moisture is ignored, all release
heat to the space air instantaneously and, therefore, they are instantaneous cooling loads.
Components of Cooling Coil Load
If the conductive heat gain from the coil’s framework and the support is ignored, the cooling coil
load consists of the following components, as shown in Fig. 6.4 by the summer air conditioning
cycle O-m-cc-s-r-rf-m of a constant volume of supply air, single supply duct, and serving a
single zone.
1. Space cooling load Qrc, including sensible and latent load
2. Supply system heat gain qss, because of the supply fan heat gain qsf , and supply duct heat
gain qsd
3. Return system heat gain qrs because of heat gains of recessed electric lights and ceiling plenum
qrp, of return duct qrd , and return fan qrf , if any
4. Sensible and latent load because of the outdoor ventilation rates Qo to meet the requirements of
the occupants and others
In Fig. 6.4, the summer air conditioning cycle O-m-cc-s-r-rf-m consists of an adiabatic mixing
process O-m-rf, a cooling and dehumidifying process m-cc, a supply system heat gain process cc-s,
a space conditioning process r-s, and a return system heat gain process r-rf. Here, O indicates
the status of outdoor air, m the mixture of outdoor air and recirculating air, cc the conditioned air
LOAD CALCULATIONS 6.7
leaving the coil, s the supply air, r the conditioned space, and rf the recirculating air. All the cooling
loads and heat gains are in Btu/h (W). Usually, both the supply and return system heat gains are
sensible loads. These components are absorbed by the supply air and return air and appear as cooling
and dehumidifying loads at the cooling coil during summer.
Difference between Cooling Load and Cooling Coil Load
For the same air conditioning cycle shown in Fig. 6.4, note the following:
1. The space cooling load is represented by Qrc, and the cooling coil load is represented by Qc.
Since supply system heat gain qss and return system heat gain qrs are both instantaneous cooling
loads, then
Qc  Qrc  qss  qrs  Qo (6.1)
where Qo  load due to the outdoor ventilation air intake, Btu/h (W).
2. The space cooling load is used to determine the supply volume flow rate , whereas the coil
load is used to determine the size of the cooling coil in an air-handling unit or DX coil in a
packaged unit.
3. A cooling load component influences both and the size of the cooling coil, whereas a cooling
coil load component may not affect .
4. Infiltration heat gain is an instantaneous cooling load. From Fig. 6.4, it is apparent that the load
due to the outdoor ventilaion air Qo, sometimes called the ventilation load, is a coil load. If Qo is
considered a cooling load, the volume flow rate of the air system will be oversized.
V?
s
V?
s
V?
s
6.8 CHAPTER SIX
FIGURE 6.4 Difference between cooling load and cooling coil load on
psychrometric chart.
Load Profile
A load profile shows the variation of space load within a certain time period, such as a 24-h operating
cycle or an annual operating cycle, as shown in Fig. 6.5. In a load profile, the space load is
always plotted against time. For a space cooling load, the magnitude of the curve is positive; for a
space heating load, the magnitude is negative. A load profile may be used to illustrate the load variation
of an air conditioned space—a room, a zone, a floor, a building, or a project. The shape of a
load profile depends on the outdoor climate and, therefore, the latitude, orientation, and structure of
the building. It is also affected by the operating characteristics and the variation of the internal
loads.
The load duration curve is the plot of number of hours versus the load ratio. The load ratio is
defined as the ratio of cooling or heating load to the design full load, both in Btu/ h, over a certain
period. The period may be a day, a week, a month, or a year. The load duration curve is helpful in
many operating and energy consumption analyses.
Peak Load and Block Load
The zone peak load is the maximum space cooling load in a load profile of a control zone of the
same orientation and similar internal load characteristics calculated according to the daily outdoor
dry-bulb and coincident wet-bulb temperature curves containing summer or winter outdoor design
conditions. For a zone cooling load with several components, such as solar load through window
glass, heat transfer through roofs, or internal load from electric lights, the zone peak load is always
LOAD CALCULATIONS 6.9
FIGURE 6.5 Load profile, peak load, and block load.
the maximum sum of these zone cooling load components at a given time. The block load is the
maximum sum of the zone cooling loads of various load profiles of the control zones of a building
floor or a building at the same time. The block load of a space, room, floor, or building is the maximum
sum of cooling load components in that space, room, floor, or building at a given time.
For air systems, the supply volume flow rate required for a control zone is calculated based
on the zone peak load; and the supply volume flow rate for a specific area, room, floor, or building
should be calculated based on the block load (cooling) of this specific area, room, floor, or
building. For conditioned space using variable-air-volume systems and space air conditioning
systems, the required cooling coil load or refrigeration load can be calculated based on block
load of the corresponding specific area that air system serves.
The heating load is usually the peak heating load of a space, room, zone, floor, or building in a
load profile. Block load is not needed in heating load calculations because the solar radiation and
internal loads are not taken into account. Design heating load depends entirely on indoor-outdoor
temperature difference.
Suppose a typical floor of a multistory building has five zones: a north, a south, an east, and a
west perimeter zone and an interior zone. For each zone, there is a corresponding daily cooling load
profile curve and a zone peak load QN,13, Qs,12, QE,8, QW,16, and QI,14. Here, subscripts N, S, E, W,
and I indicate north, south, east, west, and interior, respectively; 8, 12, 13, 14, and 16 represent time
at 8, 12, 13, 14, and 16 h, respectively.
The block load of this typical floor Qmax,t is not calculated as the sum of the zone peak loads
QN,13, Qs,12, QE,8, QW,16, and QI,14. Rather the block load is calculated for a specific time. For example,
block load Qmax.14, Btu /h (W), is the block load at 14 h, i.e.,
Qmax.14  QN,14  QS,14  QE,14  QW,14  QI,14 (6.2)
Characteristics of Night Shutdown Operating Mode
Compared with the continuous operating mode, night shutdown operating mode has the following
characteristics:
1. A greater peak load, therefore a higher air system and initial cost
2. A higher maximum heat extraction rate
3. A smaller accumulated heat extraction rate over the 24-h operating cycle
4. A lower power consumption for the fans, compressors, and pumps
Moisture Transfer from the Building Structures
In hot and humid climates, the frequency of outdoor wet-bulb temperatures higher than 73°F
(22.8°C) exceeds 1750 h during the six warmest consecutive months annually. At the same time, the
air infiltrated through the elevator and pipe shafts during shutdown periods often causes a temperature
increase of more than 7°F (3.9°C) and a relative humidity increase exceeding 10 percent. These
result in a significant increase of the latent load because of the moisture transfer from the building
envelope and furnishings to the space air during the cool-down period in summer. Such an increase
in the coil’s latent load not only necessitates a greater refrigerating capacity, but also lowers
the sensible heat ratio of the space conditioning process. A space conditioning process removes
heat and moisture from the space by the supply air, or sometimes adds moisture and heat to
compensate for space heat losses. In industrial applications where a very low relative humidity is
6.10 CHAPTER SIX
maintained in the conditioned space, the latent load due to the moisture transfer from the building
envelope should be added to the space cooling load calculations during night shutdown operation
mode.
6.3 HISTORICAL DEVELOPMENT OF COOLING LOAD
CALCULATIONS
In the 1930s, Houghton et al. introduced the analysis of heat transmission through the building
envelope and discussed the periodic heat flow characteristics of the building envelope. In 1937,
ASHVE Guide introduced a systematic method of cooling load calculation involving the division of
various load components. In the ASHVE Guide, solar radiation factors were introduced and their
influence on external walls and roofs was taken into consideration. Both the window crack and
number-of-air-changes methods were used to calculate infiltration.
Mackey and Wright first introduced the concept of sol-air temperature in 1944. In the same
paper, they recommended a method of approximating the changes in inside surface temperature
of walls and roofs due to periodic heat flow caused by solar radiation and outside temperature
with a new decrement factor. In 1952, Mackey and Gay analyzed the difference between the
instantaneous cooling load and the heat gain owing to radiant heat incident on the surface of the
building envelope.
In 1964, Palmatier introduced the term thermal storage factor to indicate the ratio between the
rate of instantaneous cooling load in the space and rate of heat gain. One year after, Carrier Corporation
published a design handbook in which the heat storage factor and equivalent temperature
difference (ETD) were used to indicate the ratio of instantaneous cooling load and heat gain because
of the heat storage effect of the building structure. This cooling load calculation method was
widely used by many designers until the current ASHRAE methods were adopted.
In 1967, ASHRAE suggested a time-averaging (TA) method to allocate the radiant heat over
successive periods of 1 to 3 h or 6 to 8 h, depending on the construction of the building structure.
Heat gains through walls and roofs are tabulated in total equivalent temperature differentials
(TETDs). In the same year, Stephenson and Mitalas recommended the thermal response factor,
which includes the heat storage effect for the calculation of cooling load. The thermal response
factor evaluates the system response on one side of the structure according to random temperature
excitations on the other side of the structure. This concept had been developed and forms the basis
of the weighting factor method (WFM) or transfer function method (TFM) in the 1970s. In 1977,
ASHRAE introduced a single-step cooling load calculation procedure that uses the cooling load
factor (CLF) and cooling load temperature difference (CLTD); these are produced from the simpli-
fied TFM.
In 1963, Kusuda and Achenbach used a digital computer to analyze the thermal environment of
occupied underground space. The use of computers to design building mechanical systems was first
accepted in 1965 when a group of mechanical engineering consultants organized Automated Procedures
for Engineers Consultants Inc. (APEC) because of sharing of software and development
costs.
Because of the need for computerized load and energy calculations, ASHRAE established a
Committee on Energy Consumption in 1965 and renamed the Task Group on Energy Requirements
(TGER) for Heating and Cooling in 1967. In the mid-1970s, ASHRAE and the National Bureau of
Standards (NBS) published the computerized calculation of heating and cooling loads in energyestimating
programs. In 1980, the U.S. Department of Energy sponsored a computer program for
energy estimation and load calculation through hour-by-hour detailed system simulation, called
DOE-2, which was published through Los Alamos National Laboratory and Lawrence Berkeley
Laboratory. In this program, a custom weighting factor method for various room configurations is
used for heating and cooling load calculations. Many computerized thermal load and energy calculating
software programs had been developed in the 1980s. Since the 1980s, because of the wide
LOAD CALCULATIONS 6.11
adoption of personal computers, the use of computer-aided HVAC&R design was rapidly increased
and many thermal load and energy analysis programs were developed in this period.
6.4 METHODOLOGY—HEAT BALANCE
The Physical Model
The exact method to calculate the space cooling load is to use heat balance equations to determine
the temperature of the interior surfaces of the building structure at time t simultaneously and then to
calculate the space sensible cooling load, which is equal to the sum of the convective heat transfer
from these surfaces, latent cooling loads, and the cooling load due to infiltrated air at time t.
Consider a typical air conditioned room, as shown in Fig. 6.1. In this room, the building envelope
consists of mainly walls, window, ceiling, and floor. There is also heat transfer from occupants,
electric lights, and equipment. The heat transfer between various surfaces takes place under the
following assumptions:
 Only one-dimensional transient heat flow through the building envelope is considered.
 The room air is perfectly mixed with the supply air so that the resulting room air temperature is
uniform.
 The materials of the building envelope are homogeneous. The surface temperature, the surface
heat-transfer coefficient, and the absorptivity for each surface are uniform values. Reflectivity is
very small and can be ignored.
 The cooling load is calculated based on the mean value of a fixed time interval, such as 1 h.
Heat Balance Equations
Any interior surface of this air conditioned room may receive conduction heat transfer at time t,
denoted by qi,t, Btu /h (W), from the adjacent building material. Each interior surface receives shortwave
solar radiation from the window glass and longwave radiative heat transfer from other interior
surfaces and from the surfaces of the lighting fixtures, appliances, equipment, and occupants. Convective
heat transfer is also present between these interior surfaces and room air.
The sensible heat balance at the ith surface (for i  1, 2, . . ., m) at time t can be expressed as
follows:
(6.3)
where hci  convective heat-transfer coefficient of ith surface, Btu /h ft2 °F (W/m2 °C)
hij  radiative heat-transfer coefficient between interior surface i and j, i  j, Btu/h ft2 °F
(W/m2 °C)
Tr,t  room air temperature at time t, °F (°C)
Ti,t  temperature of ith surface at time t, °F (°C)
Tj,t  temperature of jth surface at time t, °F (°C)
Ai  area of ith surface, ft2 (m2)
Sir, t  solar radiation transmitted through window glass and absorbed by ith surface at time t,
Btu/h (W)
Lir, t  radiative energy from electric lights and absorbed by ith surface at time t, Btu /h (W)
Eir, t  radiative energy from equipment and absorbed by ith surface at time t, Btu /h (W)
Oir, t  radiative energy from occupants and absorbed by ith surface at time t, Btu /h (W)
qi, t  [hci (Tr, t  Ti, t)  k
j1
hij (Tj, t  Ti, t)] Ai  Sir, t  Lir, t  Eir, t  Oir, t
6.12 CHAPTER SIX
Because the conductive heat and radiative heat in Eq. (6.3) are all continuous functions of time f (t),
the transient conductive heat transfer qi, t either can be solved by a partial differential equation using
numerical solutions (finite difference method) or can be solved by means of the conduction transfer
function [Eq. (6.10)].
Usually, the room air temperature Tr,t is considered as constant, and Eq. (6.3) and the conduction
transfer function equation can be solved simultaneously to determine the interior surface
temperatures Ti, t. Then the sensible cooling load at time t, denoted by Qrs,t, Btu /h (W), can be
calculated as
(6.4)
where  volume flow rate of infiltrated air at time t, cfm [m3/(60 s)]
 air density, lb/ ft3 (kg/m3)
cpa  specific heat of moist air, Btu / lb°F (J /kg°C)
To,t  temperature of outdoor air at time t, °F(°C)
Sc,t  solar heat coming through windows and convected into room air at time t, Btu /h (W)
Lc,t  sensible heat from electric lights and convected into room air at time t, Btu /h (W)
Ec,t  sensible heat from equipment and convected into room air at time t, Btu /h (W)
Oc,t  sensible heat from occupants and convected into room air at time t, Btu /h (W)
Since the latent heat gains convert to latent cooling loads instantaneously, therefore the space latent
cooling load at time t, or Qrl,t, Btu /h (W), can be calculated as
(6.5)
where qil, t  latent heat from ith interior surface and convected into room air at time t, Btu /h (W)
wo, t  humidity ratio of infiltrated air at time t, lb/ lb (kg /kg)
wr, t  humidity ratio of room air at time t, lb/ lb (kg /kg)
hfg  latent heat of condensation, Btu/ lb (J /kg)
El, t  latent heat from equipment at time t, Btu /h (W)
Ol, t  latent heat from occupants at time t, Btu /h (W)
Then the space cooling load at time t, denoted by Qrc, t, Btu /h (W), is
Qrc,t  qrs,t  qrl,t (6.6)
Heat Balance of Space Air
The conditioned air for this room or space is supplied from the ceiling diffuser at a mass flow rate
at time t of . Here, indicates the volume flow rate of supply air at time t, cfm
(m3/min), and s is the density of supply air, lb/ ft3 (kg/m3). The supply air is then mixed with the
space air. The resulting mixture absorbs convective sensible heat and latent heat from various surfaces.
At time t, the outdoor air infiltrates into the space at a volume flow rate , cfm [m3 / (60 s)].
The sensible heat balance of the space air can be expressed as
(6.7)
where Ts,t  supply air temperature at time t, °F (°C). If the space air temperature is allowed to
float, Eq. (6.3), the transient conductive equation, and Eq. (6.7) must be solved simultaneously.
 Sc, t  Lc, t  Ec, t  Oc, t  m
i1
hci Ai(Ti, t  Tr, t)  60V?if, t cpa (To, t  Tr, t)
60 V?s, t scpa (Tr, t  Ts,t)
V?
if, t
V?
s,t m?s,t  V?s, t s
Qrl, t  qil, t  60V?if  (wo, t  wr, t)hfg  El, t  Ol, t
V?
if, t
Qrs, t  m
i1
hci (Ti, t  Tr, t) Ai  60V?if,t cpa(To, t  Tr, t)  Sc, t  Lc, t  Ec, i  Oc, t
LOAD CALCULATIONS 6.13
Characteristics of Heat Balance Method
The heat balance method is more direct and clear in load calculation methodology. Using the heat
balance method, the assumption of linear supposition is not required, and the changing of certain
parameters, such as the surface convective heat-transfer coefficient, can be modeled as required. If
moisture transfer should be included in the cool-down period in nighttime shutdown mode in a
location where the outdoor climate is hot and humid in summer, then the heat balance method will
give comparatively more accurate results.
However, the heat balance method demands laborious work, more computing time, complicated
computer programs, and experienced users. Only expensive mainframe computers could run
computer programs adopting the heat balance methodology in the 1970s and early 1980s. The heat
balance method is generally used for research and analytical purposes.
6.5 METHODOLOGY—TRANSFER FUNCTION
Basics
The transfer function method or weighting factor method is a simplification of the laborious heat
balance method. The wide application of the TFM is due to the user-friendliness of the inputs and
outputs of the TFM software and the saving of computing time. In the transfer function method,
interior surface temperatures and the space cooling load were first calculated by the exact heat
balance method for many representative constructions. The transfer function coefficients (weighting
factors) were then calculated which convert the heat gains to cooling loads. Sometimes, transfer
function coefficients were also developed through test and experiments.
The room transfer function coefficients (weighting factors ) were originally generated based on
a typical room configuration of 15 ft 	 15 ft with 10-ft (4.5 m 	 4.5 m with 3-m) ceiling and one
exposure of 30 percent glass in the early 1970s . In the late 1980s, the introduction of 14 influential
parameters of zone characteristics by Sowell (1988) enabled the adopted room transfer function
coefficients to more closely match the room type to be used. Today, TFM is the most widely
adopted computer-aided load calculation method in HVAC&R consulting firms.
Transfer Function and Time Function
The transfer function K of an element or a system is the ratio of the Laplace transform of the output
Y to the Laplace transform of the input or driving force G, or
Y  KG (6.8)
When a continuous function of time f (t) is represented at regular intervals t and its magnitudes are
f (0), f (), f (2), . . ., f (n), the Laplace transform is given as
(6.9a)
where   time interval, h
z  es
The preceding polynomial in z1 is called the z transform of the continuous function f (t).
In Eq. (6.8), Y, K, and G can all be represented in the form of a z transform. Because of the
radiative component and the associated heat storage effect, the space sensible cooling load at time t
can be related to the sensible heat gains and previous cooling loads in the form of a continuous
function of time f (t), which can be expressed as a z transform.
	(z)  f (0)  f ()z1  f (2)z2      f (n)zn
6.14 CHAPTER SIX
Weighting factors are transfer function coefficients presented in the form of z transform functions.
Weighting factors are so called because they are used to weight the importance of current and
historical values of heat gain and cooling load on currently calculated cooling loads.
Calculation Procedure
The calculation of space cooling load using the transfer function method consists of two steps.
First, heat gains or heat loss from exterior walls, roofs, and floors is calculated using response
factors or conduction transfer function coefficients; and the solar and internal heat gains are calculated
directly for the scheduled hour. Second, room transfer function coefficients or room weighting
factors are used to convert the heat gains to cooling loads, or the heat losses to heating loads.
As described in Sec. 6.2, the sensible infiltration heat gain is the instantaneous sensible cooling
load. All latent heat gains are instantaneous latent cooling loads.
The TFM is limited because the cooling loads thus calculated depend on the value of transfer
function coefficients as well as the characteristics of the space and how they are varied from those
used to generate the transfer function coefficients. In addition, TFM assumed that the total cooling
load can be calculated by simply adding the individual components—the superposition principle.
However, this assumption can cause some errors.
CLTD/SCL/CLF Method
The cooling load temperature difference (CLTD)/solar cooling load (SCL)/cooling load factor
(CLF) method first calculates the sensible cooling load based on the TFM. The result is divided by
the U value, shading coefficient, or sensible heat gain to generate the CLTD, SCL, or CLF. Thus, it
provides a direct, one-step space cooling load calculation instead of a heat gain–cooling load conversion,
a two-step calculation in TFM. Cooling load calculation using the CLTD/SCL/CLF
method can be either computer-aided or performed manually for a check or rough estimate. The
CLTD/SCL/CLF method is one of the members of the TFM family.
In the CLTD/SCL/CLF method, the CLTD is used to calculate the sensible cooling load for the
exterior wall and roofs. Recently, an SCL factor has been added which represents the product of the
solar heat gain at that hour and the fraction of heat storage effect due to various types of room
construction and floor coverings. CLF is used to calculate internal sensible cooling loads.
The limitations of the TFM are also carried through to the CLTD/SCL/CLF results. Furthermore,
the grouping of CLTD/SCL/CLF may cause additional errors.
TETD/TA Method
In the total eqivalent temperature difference (TETD)/time-averaging (TA) method, heat gains of a
number of representative exterior wall and roof assemblies qw, Btu /h (W), are calculated as
qw  AU(TETD) (6.9b)
where A  area of wall or roof, ft2 (m2)
U  overall heat-transfer coefficient of wall or roof, Btu /h ft2 °F (W/m2 °C)
In Eq. (6.9b), TETD, in °F (°C), can be evaluated by:
 Using the conduction transfer function as in TFM to determine qw. Then it is divided by the U
value to generate TETD values.
 Using the following relationship:
TETD  Tsol, a  Tr  (Tsol, 
 Tsol, a) (6.9c)
LOAD CALCULATIONS 6.15
where Tsol, a  daily average sol-air temperature, °F (°C)
Tsol, 
 sol-air temperature at time lag 
 h, °F (°C)
 effective decrement factor 
The internal heat gains and conductive heat gain are calculated in the same manner as in the TFM.
The radiant fraction of each of the sensible heat gains is then allocated to a period including the current
and successive hours, a total of 1 to 3 h for light construction and 6 to 8 h for heavy construction.
The TETD/ TA method is also a member of the TFM family and is developed primarily for
manual calculation. TETD/TA is simpler in the conversion of heat gains to cooling loads. However,
the time-averaging calculation procedure is subjective—it is more an art than a rigorous scientific
method. Also the TETD/TA method inherits the limitations that a TFM possesses if the TFM is
used to calculate the TETD.
6.6 DETAILED CALCULATION PROCEDURES FOR TFM
Most of the widely adopted computer software programs for space load calculations are based on
the transfer function method, i.e., the weighting factors method. The following are the detailed
calculation procedures for TFM.
Conduction Heat Gain through Exterior Walls and Roofs
The sensible heat gain through an exterior wall or a roof using TFM can be calculated by the conduction
transfer function. It uses a sol-air temperature at time t, denoted by Tsol, t, to represent the combined
temperature excitation of the outdoor temperature and solar heat on the exterior surface of an exterior
wall or roof, and a constant indoor temperature Tr. Conduction-transfer function coefficients are calculated
based on an outdoor surface heat-transfer coefficient ho  3.0 Btu/h ft2 °F (17 W/m2 °C), and
an indoor surface heat-transfer coefficient hi  1.46 Btu/h ft2 °F (8.3 W/m2 °C).
The external heat gain through an exterior wall or roof at time t, denoted by qe,t, Btu /h (W), can
be calculated as
(6.10)
where t  time, h
  time interval, h
n  summation index of number of terms
Tsol,t – n  sol-air temperature at time tn, °F (°C)
qe,t – n  conduction heat gain at time tn, Btu /h (W)
bn,cn,dn  conduction-transfer function coefficients; refer to ASHRAE Handbook for details
A  interior surface area of wall or roof, ft2 (m2)
Harris and McQuiston (1988) provide the conduction-transfer function coefficients for 41 representive
wall assemblies and 42 roof assemblies with variations in components, insulation, and mass.
Heat Gain through Ceilings, Floors, and Interior Partition Walls
If the temperature of the adjacent space Tad, °F (°C), is constant, or the variation of Tad is small, the
sensible heat gain due to the heat transfer from the adjacent space through the ceiling, floor, or interior
partition wall at time t, denoted by qp,t, Btu /h (W), can be calculated as
qp,t  UA(Tad  Tr) (6.11)
qe,t   
n0
bnTsol,t – n  
n1
dn
qe,t – n
A
 Tr
n0
cn A
6.16 CHAPTER SIX
where U  overall heat-transfer coefficient of ceiling, floor, or partition wall, Btu/h ft2 °F
(W/m2 °C)
A  surface area of ceiling, floor, or partition wall, ft2 (m2)
If the temperature of the adjacent space Tad,t varies with time, then the sensible heat gain transferring
through the ceiling, floor, or partition wall at time qp, t, Btu /h (W), can be calculated from
Eq. (6.10), except qe,t should be replaced by qp,t, qe,t – n replaced by qp,t – n, and Tsol,t replaced by
Tad,t. Here qp,t – n indicates the sensible heat gain of the ceiling, floor, or partition wall at time
t  n, Btu /h (W); and Tad,t indicates the adjacent space temperature, °F (°C).
Solar Heat Gain and Conductive Heat Gain through Window Glass
Although the inward heat flow from the solar radiation absorbed by the window glass and the heat
flow due to the outdoor and indoor temperature difference are actually combined, it is simpler and
acceptably accurate to separate this composite heat gain into solar heat gain and conductive heat
gain; both are sensible heat gains.
Solar heat gain qso,t, Btu /h (W), is given as
qso,t  As, t (SC)(SHGFt)  Ash, t (SC)(SHGFsh, t) (6.12)
where As,t  sunlit area of window glass at time t, ft2 (m2)
Ash,t  shaded area of window glass at time t, ft2 (m2)
SC  shading coefficient
SHGFt  solar heat gain factor at time t considering orientation, latitude, month, and hour,
Btu/h ft2 (W/m2)
SHGFsh,t  solar heat gain factor for shaded area at time t, considering latitude, month, and hour,
Btu/h ft2 (W/m2)
Usually, the SHGFt of the vertical surface facing north orientation is taken as SHGFsh,t. Conductive
heat gain due to the outdoor-indoor temperature difference qwin,t, Btu /h (W), is given as
qwin,t  UwinAwin (To,t  Tr) (6.13)
where Uwin  overall heat-transfer coefficient of window including glass and frame (Table 3.7),
Btu/h ft2 °F (W/m2 °C)
Awin  gross area of window including glass and frame, ft2 (m2)
To,t  outdoor temperature at time t considering month, hour, and location, °F (°C)
Internal Heat Gain
The internal sensible heat gain consists of the sensible heat gain from people, from electric lights,
and from equipment and appliances.
People. Human beings release both sensible heat and latent heat to the conditioned space. The
radiative portion of the sensible heat gain is about 70 percent when the indoor environment of
the conditioned space is maintained within the comfort zone. The space sensible heat gain for
occupants staying in a conditioned space at time t, denoted by qsp,t, Btu /h (W), can be calculated as
qsp,t  Np,t (SHGp) (6.14)
where Np,t  number of occupants in conditioned space at time t
SHGp  sensible heat gain of each person, Btu/h (W)
LOAD CALCULATIONS 6.17
Space latent heat gain for occupants staying in a conditioned space at time t, denoted by qlp,t, Btu/h
(W), is given as
qlp,t  Np,t(LHGp) (6.15)
where LHGp  latent heat gain of each person, Btu/h (W). Table 6.1 lists the heat gain from occupants
in conditioned space, as abridged from ASHRAE Handbook 1989, Fundamentals. In Table
6.1, total heat is the sum of sensible and latent heat. The adjusted heat is based on a normally
distributed percentage of men, women, and children among the occupants.
Lighting. The sensible heat gains from the electric lights depend on the types of installation, as
follows:
Inside Conditioned Space. For electric lights installed inside the conditioned space, such as
light fixtures hung below the ceiling, the sensible heat gain released from the electric lights, the
emitting element, and light fixtures qs, l is equal to the sensible heat released to the conditioned
space qes, l, Btu /h (W); both depend mainly on the criteria of illumination and the type and effi-
ciency of electric lights and can be calculated as
qs.l  3.413 Wlamp FuslFal
 3.413 WAAfl (6.16)
where Wlamp  rated input of electric lights,W
WA  wattage per ft2 of floor area,W/ ft2 (W/m2)
Fusl  ratio of wattage in use to installation wattage
In Eq. (6.16), Fal indicates an allowance factor for light fixtures, such as Ballast losses. For rapidstart
40-W fluorescent fixtures, Fal varies from 1.18 to 1.3 with a recommended value of 1.2
(ASHRAE Handbook 1993, Fundamentals).
Recess-Mounted Fixtures Using Return Plenum. For situations in which electric lights are
recess-mounted on the ceiling and the ceiling plenum is used as a return plenum, the fraction of the
6.18 CHAPTER SIX
TABLE 6.1 Rates of Heat Gain from Occupants of Conditioned Spaces*
Total heat Total heat Sensible
of adults, adjusted,† heat, Latent heat,
Degree of activity Typical application male, Btu /h Btu/h Btu /h Btu/h
Seated at theater Theater—matinee 390 330 225 105
Seated at theater Theater—evening 390 350 245 105
Seated, very light work Offices, hotels, apartments 450 400 245 155
Moderately active office work Offices, hotels, apartments 475 450 250 200
Standing, light work; walking Department store, retail store 550 450 250 200
Walking; standing Drugstore, bank 550 500 250 250
Light bench work Factory 800 750 275 475
Moderate dancing Dance hall 900 850 305 545
Walking 3 m/h; light machine work Factory 1000 1000 375 625
Heavy work Factory 1500 1450 580 870
Heavy machine work; lifting Factory 1600 1600 635 965
Athletics Gymnasium 2000 1800 710 1090
*Tabulated values are based on 75°F room dry-bulb temperature. For 80°F room dry-bulb temperature, the total heat remains the same, but the
sensible heat values should be decreased by approximately 20 percent and the latent heat values increased accordingly. All values are rounded to
nearest 5 Btu/ h.
†Adjusted heat gain is based on normal percentage of men, women, and children for the application listed, with the postulate that the gain from
an adult female is 85 percent of that for an adult male, and that the gain from a child is 75 percent of that for an adult male.
Sources: Adapted with permission from ASHRAE Handbook 1989, Fundamentals.
sensible heat gain of electric lights that enters the conditioned space qes,l, Btu /h (W), is closely
related to the type of lighting fixture, the ceiling plenum, and the return system.
If the ceiling plenum is used as a return plenum in a multistory building, as shown in Fig. 6.6,
the heat released from recessed fluorescent light fixtures and the heat transfer between the outdoor
air, plenum air, return air, and space air are as follows:
 Radiative and convective heat transfer from the lighting fixture downward directly into the conditioned
space qld, Btu /h (W), which is calculated as
qld  qs,l  qlp  (1  Flp)qs,l (6.17)
where qlp  heat released by electric lights to return air, Btu /h (W)
Flp  fraction of heat released from light fixture to plenum air
The fraction that enters conditioned space qld depends on the volume flow rate of the return air
flowing through the lighting fixture and the type of fixture. Its magnitude should be obtained from
the lighting fixture manufacturer. In Fig. 6.7 is shown the relationship between Flp and the intensity
of volume flow rate of return air if the lighting fixture is ventilated. Usually, Flp varies
between 0.4 and 0.6 for a ventilated fixture in a return air plenum. For unventilated lighting
fixtures recess-mounted in a return air plenum, Flp varies between 0.15 and 0.5.
V?
r / Afl
LOAD CALCULATIONS 6.19
FIGURE 6.6 Heat released from a recess-mounted ventilated lighting fixture.
 Heat carried away by return air from the ceiling plenum qret, Btu /h (W), which is calculated as
(6.18)
where volume flow rate of return air, cfm [m3 / (60 s)]
r  density of return air, lb/ ft3 (kg/m3)
cpa  specific heat of return air, Btu / lb°F (J /kg°C)
Tp, Tr  temperatures of plenum air and space air, °F (°C)
 Heat transfer from the plenum air to the conditioned space through the suspended ceiling qcl and
heat transfer from the plenum air to the conditioned space through the composite floor qfl, both in
Btu/h (W), can be calculated as
qcl  UclAcl (Tp  Tr) (6.19a)
qfl  UflAfl (Tp  Tr) (6.19b)
where Ucl, Ufl  overall heat-transfer coefficient of ceiling and composite floor, Btu /h ft2 °F
(W/m2 °C)
Acl, Afl  area of ceiling and composite floor, ft2 (m2)
 Heat transfer between the outdoor air and the plenum air through the exterior wall of the ceiling
plenum qwp,t, Btu /h (W), can be calculated by conductive transfer function as shown in Eq. (6.10).
Because qwp,t is comparatively small, it is often ignored.
 For a multistory building where all the floors are air conditioned, heat gain from the electric lights
that enters the conditioned space, heat to space, qes,l is given as
qes,l  qld  qcl  qfl (6.20)
Of this, about 50 percent is radiative, and the rest is convective. Heat released from the electric
lights to the return air including radiative transfer upward, heat to plenum, qlp, Btu /h (W), is
V?
r 
qret  60 V?r r cpa (Tp  Tr)
6.20 CHAPTER SIX
FIGURE 6.7 Relationship between Flp and the intensity of return air
/ Afl. V?r
given as
qlp  qret  qcl  qfl (6.21)
In a return air plenum, precise calculation of the plenum air temperature Tp is rather complicated.
A simplified method is to assume a steady-state heat transfer between the plenum air and the conditioned
space air and to use a cooling load temperature difference CLTDwp (CLTDwp can be
found from Table 6.2) to calculate the heat transfer through the exterior wall of the plenum. Then,
based on the heat balance at the plenum air,
(6.22)
where Uwp  overall heat-transfer coefficient of exterior wall, Btu/h ft2 °F (W/m2 °C)
Awp  area of exterior wall, ft2 (m2)
CLTDwp  cooling load temperature difference of exterior wall of ceiling plenum, °F (°C)
Plenum air temperature Tp can thus be determined.
Surface-Mounted Fixtures under Ceiling. If the lighting fixtures are surface-mounted under the
suspended ceiling, then the fraction of heat gain downward that enters the conditioned space qes,l,
Btu/h (W), can be calculated as
qes,l  Fes,l qs,l (6.23 )
where Fes,l  fraction of sensible heat gain entering the conditioned space for surface-mounted
lighting fixtures, which varies from approximately 0.8 to 0.95.
Equipment and Appliances. With equipment and appliances, all energy inputs are converted to
heat energy, e.g., in the motor windings, combustion chamber, rubbing surfaces, even at components
where mechanical work is performed. A portion of heat released may be exhausted locally by
a mechanical ventilation system.
In many industrial applications, the space sensible heat gain due to the machine load when a
motor is located inside the conditioned space qs,e, Btu /h (W), can be calculated as
(6.24)
where Php  rated horsepower of machine, hp
Fload  load factor indicating ratio of actual power required to rated power
Fuse  use factor indicating ratio of actually used equipment and appliance to total installed
Fexh  heat removal factor due to mechanical exhaust system
mo  motor efficiency
In Table 6.3 are listed the efficiencies of motors from a fraction of a horsepower to 250 hp
(187 kW). The bigger the motor, the higher the motor efficiency. A high-efficiency motor is often
cost-effective. The Energy Policy Act of 1992 (EPACT), after a phase-in period of 5 years which
ended on October 24 1997, requires that all covered motors meet increased minimum efficiency
levels. If the motor is located outside the conditioned space, then in Eq. (6.24), mo  1.
In many types of equipment and appliances installed with exhaust hoods in the conditioned
space, energy input and Fexh are better determined according to the actual performance of similar
projects. For example, for food preparation appliances equipped with exhaust hoods, only radiation
from the surface of the appliances, which as a fraction of the total heat input is between 0.1
and 0.5, should be counted; or Fexh is between 0.9 and 0.5. If there is no mechanical exhaust
system, Fexh  0.
qs,e  2546Php Fload Fuse
1  Fexh
mo
qlp  Uwp Awp (CLTDwp)  (60 V?r rcpa  Ucl Acl  Ufl Afl) (Tp  Tr)
LOAD CALCULATIONS 6.21
6.22
TABLE 6.2 CLTD for Calculating Sensible Cooling Loads from Sunlit Walls of North Latitude, °F
Solar time, h
Facing 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Group C walls: typical, outside 1-in. stucco, 2-in. insulation (5.7lb/ft3), 4-in. concrete, 0.75-in. plaster or gypsum, inside U  0.119 Btu/hft2°F; mass, 63 lb/ft2
N 15 14 13 12 11 10 9 8 8 7 7 8 8 9 10 12 13 14 15 16
NE 19 17 16 14 13 11 10 10 11 13 15 17 19 20 21 22 22 23 23 23
E 22 21 19 17 15 14 12 12 14 16 19 22 25 27 29 29 30 30 30 29
SE 22 21 19 17 15 14 12 12 12 13 16 19 22 24 26 28 29 29 29 29
S 21 19 18 16 15 13 12 10 9 9 9 10 11 14 17 20 22 24 25 26
SW 29 17 25 22 20 18 16 15 13 12 11 11 11 13 15 18 22 26 29 32
W 31 29 27 25 22 20 18 16 14 13 12 12 12 13 14 16 20 24 29 32
NW 25 23 21 20 18 16 14 13 11 10 10 10 10 11 12 13 15 18 22 25
Group D walls: typical, outside 1-in. stucco, 4-in. concrete, 1- or 2-in. insulation (2 lb/ft3), 0.75-in. plaster or gypsum, inside U  0.119–0.20 Btu/hft2°F; mass, 63 lb/ft2
N 15 13 12 10 9 7 6 6 6 6 6 7 8 10 12 13 15 17 18 19
NE 17 15 13 11 10 8 7 8 10 14 17 20 22 23 23 24 24 25 25 24
E 19 17 15 13 11 9 8 9 12 17 22 27 30 32 33 33 32 32 31 30
SE 20 17 15 13 11 10 8 8 10 13 17 22 26 29 31 32 32 32 31 30
S 19 17 15 13 11 9 8 7 6 6 7 9 12 16 20 24 27 29 29 29
SW 28 25 22 19 16 14 12 10 9 8 8 8 10 12 16 21 27 32 36 38
W 31 27 24 21 18 15 13 11 10 9 9 9 10 11 14 18 24 30 36 40
NW 25 22 19 17 14 12 10 9 8 7 7 8 9 10 12 14 18 22 27 31
Group G walls: typical, outside 1-in. stucco, airspace; 1-, 2-, or 3-in. insulation (2 lb/ft3); 0.75-in. plaster or gypsum, inside U  0.081–0.78 Btu/hft2°F; mass, 16 lb/ft2
N 3 2 1 0 1 2 7 8 9 12 15 18 21 23 24 24 25 26 22 15
NE 3 2 1 0 1 9 27 36 39 35 30 26 26 27 27 26 25 22 18 14
E 4 2 1 0 1 11 31 47 54 55 50 40 33 31 30 29 27 24 19 15
SE 4 2 1 0 1 5 18 32 42 49 51 48 42 36 32 30 27 24 19 15
S 4 2 1 0 1 0 1 5 12 22 31 39 45 46 43 37 31 25 20 15
SW 5 4 3 1 0 0 2 5 8 12 16 26 38 50 59 63 61 52 37 24
W 6 5 3 2 1 1 2 5 8 11 15 19 27 41 56 67 72 67 48 29
NW 5 3 2 1 0 0 2 5 8 11 15 18 21 27 37 47 55 55 41 25
Direct applications and adjustments are stated in the text.
Source: Abridged with permission from ASHRAE Handbook 1989, Fundamentals.
Hours of
maxi- Mini- Maxi- Difference
mum mum mum in
21 22 23 24 CLTD CLTD CLTD CLTD
17 17 17 16 22 7 17 10
23 22 21 20 20 10 23 13
28 27 26 24 18 12 30 18
28 27 26 24 19 12 29 17
25 25 24 22 20 9 26 17
33 33 32 31 22 11 33 22
35 35 35 33 22 12 35 23
27 27 27 26 22 10 27 17
19 19 18 16 21 6 19 13
23 22 20 18 19 7 25 18
28 26 24 22 16 8 33 25
28 26 24 22 17 8 32 24
27 26 24 22 19 6 29 23
38 37 34 31 21 8 38 30
41 40 38 34 21 9 41 32
32 32 30 27 22 7 32 25
11 9 7 5 18 1 26 27
11 9 7 5 9 1 39 40
12 10 8 6 10 1 55 56
12 10 8 6 11 1 51 52
12 10 8 5 14 1 46 47
17 13 10 8 16 0 63 63
20 15 11 8 17 1 72 71
17 13 10 7 18 0 55 55
Because of the widespread installation of microcomputers, display terminals, printers,
copiers, calculators, and facsimile machines, the heat released from machines in office buildings
has increased considerably in recent years. Komor (1997) listed the measured office plug loads
from 17 U.S. office buildings in the 1990s which varied from 0.44 to 1.11 W/ ft2. Precise office
equipment heat gain can be calculated from manufacturers’ data using a load factor Fload between
0.3 and 0.5.
The latent heat gain from the equipment and appliances ql,e, Btu /h (W), can be calculated from
the mass flow rate of water vapor evaporated , as
(6.25) ql,e  1075 m? w
m? w, lb / h (kg / h)
LOAD CALCULATIONS 6.23
TABLE 6.3 Heat Gain from Typical Electric Motors
Location of motor and driven equipment with
respect to conditioned space or airstream
Motor Full-load Motor in, Motor out, Motor in,
nameplate motor driven driven driven
or rated Nominal effciency, equipment in, equipment in, equipment out,
horsepower rpm percent Btu/h Btu /h Btu/h
Motor type: shaded pole
0.05 1,500 35 360 130 240
0.125 1,500 35 900 320 590
Motor type: split-phase
0.25 1,750 54 1,180 640 540
0.33 1,750 56 1,500 840 660
0.50 1,750 60 2,120 1,270 850
Motor type: three-phase
0.75 1,750 72 2,650 1,900 740
1 1,750 75 3,390 2,550 850
1 1,750 77 4,960 3,820 1,140
2 1,750 79 6,440 5,090 1,350
3 1,750 81 9,430 7,640 1,790
5 1,750 82 15,500 12,700 2,790
7.5 1,750 84 22,700 19,100 3,640
10 1,750 85 29,900 24,500 4,490
15 1,750 86 44,400 38,200 6,210
20 1,750 87 58,500 50,900 7,610
25 1,750 88 72,300 63,600 8,680
30 1,750 89 85,700 76,300 9,440
40 1,750 89 114,000 102,000 12,600
50 1,750 89 143,000 127,000 15,700
60 1,750 89 172,000 153,000 18,900
75 1,750 90 212,000 191,000 21,200
100 1,750 90 283,000 255,000 28,300
150 1,750 91 420,000 382,000 37,800
200 1,750 91 569,000 509,000 50,300
250 1,750 91 699,000 636,000 62,900
For motors operating more than 750 h/year, it is usually cost-effective to use a high-efficiency motor. Typical efficiency ratings
are as follows: 5 hp, 89.5%; 10 hp, 91.0%; 50 hp, 94.1%; 100 hp, 95.1%; 200 hp, 96.2%.
Source: Abridged with permission from ASHRAE Handbook 1989, Fundamentals.
Infiltration
Infiltration is the uncontrolled inward flow of outdoor air through cracks and openings in the building
envelope due to the pressure difference across the envelope. The pressure difference may be
caused by any of the following:
1. Wind pressure
2. Stack effect due to the outdoor and indoor temperature difference
3. Mechanical ventilation
In summer, for low-rise commercial buildings that have their exterior windows well sealed, and if a
positive pressure is maintained in the conditioned space when the air system is operating, normally
the infiltration can be considered zero.
For high-rise buildings, infiltration should be considered and calculated in both summer and
winter. Infiltration is discussed again in Sec. 20.4.
As soon as the volume flow rate of infiltrated air , cfm (m3 /min), is determined, the space
sensible heat gain from infiltration qinf, Btu /h (W), can be calculated as
(6.26)
where o  density of outdoor air, lb/ ft3 (kg/m3). The space latent heat gain from infiltration ql, inf,
Btu/h (W), can be calculated as
(6.27)
where wo,wr  humidity ratio of outdoor and space air, respectively, lb/ lb (kg /kg)
hfg, 32  latent heat of vaporization at 32°F, Btu/ lb (J /kg)
Cooling Load Conversion Using Room Transfer Function
The conversion of space sensible heat gains qrs,t, Btu /h (W), having radiative only or radiative and
convective components to space sensible cooling loads Qrs,t, Btu /h (W), using the transfer function
method and room transfer function coefficients can be expressed as follows:
(6.28)
where i  number of heat gain components in same group
  time interval
tn  time at tn
Here v0, v1, v2, . . ., w1, w2, . . . are the coefficients of the room transfer function; refer to the
ASHRAE Handbook for details. Their relationship can be expressed from Eqs. (6.8) and (6.9) as
(6.29)
The magnitude of the coefficients depends on the duration of the time interval, fraction of
the radiative component, and heat storage capacity because of the 14 influential parameters
of zone characteristics, such as zone geometry, height, exterior wall construction, interior shade,
furniture, zone location, glass percentage, and type of partition, midfloor, slab, ceiling, roof,
and floor.
K(z) 
v0  v1z1  v2z2    
1  w1z1  w2z2    
Qrs,t  
i1
(v0 qs,t  v1qs,t –   v2qs,t – 2    )  (w1 Qrs,t –   w2 Qrs,t – 2    )
ql,inf  60V?inf o (wo  wr)hfg, 32
qs, inf  60V?inf ocpa (To  Tr)
V?
inf
6.24 CHAPTER SIX
Space Cooling Load Calculation
The types of conversion of space heat gains to space cooling loads can be grouped into the following
two categories:
1. Space sensible cooling loads Qrs,t that are only a fraction of the space sensible heat gain qrs,t.
These kinds of sensible heat gains have both radiative and convective components, and it is diffi-
cult to separate the convective component from the radiative component, such as sensible heat
gains from exterior walls and roofs, and solar heat gains through windows. Equation (6.28) will
be used to convert these types of sensible heat gains qrs,t to cooling loads Qrs,t.
2. Space heat gains qin,t are instantaneous space cooling loads Qin,t, both in Btu/h (W), or
Qin, t  qin, t (6.30)
This category includes all latent heat gains qrl, convective sensible heat gains, infiltration sensible
heat gain qinf, s, and sensible heat gains whose convective component can be separated from
the radiative component, such as air-to-air heat gain from windows, lights, and people.
The space cooling load Qrc, t, Btu /h (W), is their sum, or
Qrc,t  Qrs,t  Qin, t  Ql,t  Qs,t  Ql,t (6.31)
where Qs,t, Ql,t  space sensible cooling load and latent load, respectively, Btu /h (W).
As mandated in ASHRAE/IESNA Standard 90.1-1999, Optimum Start Controls, pickup loads
either for cooling or heating depend on the difference between space temperature and occupied set
point and the amount of time prior to scheduled occupancy, and is often determined by computer
software.
Heat Extraction Rate and Space Air Transfer Function
In the calculation of the conduction heat gain through the exterior wall and roofs by using the
conduction transfer function [Eq. (6.10)] and then converting to sensible cooling load from the
room transfer function, the space temperature Tr is found to be a constant. Most direct-digital control
(DDC) zone control systems are now adopting proportional-integral control mode. When the
air system and the DDC system are in operation, Tr is a constant once it has achieved a stable
condition.
However, for many air systems operated at nighttime shutdown mode, Tr will drift away from the
set point during the shutdown period. The space air transfer function relates the heat extraction rate at
time t, denoted by Qex, t, Btu/h (W), to the space air temperature at time t, denoted by Tr, t, °F (°C ), as
(6.32)
where pi, gi  space air transfer function coefficients, refer to ASHRAE Handbook for details
Qex,t  heat extraction rate at time t, Btu /h (W)
Qrc,ti  calculated space cooling load at time ti, Btu /h (W)
Tr, con  assumed constant space air temperature, °F (°C)
Heat Loss to Surroundings
There is always a radiant heat loss from the outer surface of the building to the sky vault without
clouds because atmospheric temperature is lower at high altitudes, as described in Sec. 3.12. In
many locations in the United States, there are also radiant heat losses to the surroundings at
1
i0
pi (Qex, t –   Qrc,t – i)  2
i0
gi (Tr,con  Tr, t – i)
LOAD CALCULATIONS 6.25
nighttime in summer due to the lower ground temperature. ASHRAE Handbook 1993, Fundamentals,
recommends a simplified calculation procedure so that the sum of sensible cooling loads from
heat gain by conduction through exterior walls and roofs, from conduction and solar heat gain
through windows, and from heat gain through interior partitions, ceilings, and floors at time t, denoted
by Qex, t, Btu /h (W), plus the sensible cooling loads from the radiant component of internal
heat gain at time t, denoted by Qin, t, Btu /h (W), will be multiplied by a factor Fsur to take into
account these heat losses. The corrected sensible cooling load at time t, denoted by Qrs, cor, t,
Btu/h (W), is
Qrs, cor, t  Fsur (Qex, t  Qin, t) (6.33)
The factor of heat loss to surroundings Fsur can be calculated as
(6.34)
where Lex  length of space exterior wall, ft
U  overall heat-transfer coefficient (subscript win for window and part for interior
partitions), Btu/h ft2 °F (W/m2 °C)
A  area of component of building envelope, ft2 (m2)
6.7 DETAILED CALCULATION PROCEDURE USING
CLTD/SCL/CLF METHOD
The following sections describe in greater detail the principles of the CLTD/SCL/CLF method.
They also provide a simple, manual cooling load calculation procedure in case an estimate or a
rough check of computer-aided cooling load calculation is required.
Space Cooling Load due to Heat Gain through Exterior Walls and Roofs
and Conductive Gain through Glass
If the ratio of sensible cooling load to sensible heat gain through the exterior wall or roof
Qrs,w /qrs,w  CLTD/T, for an exterior sunlit wall or roof under the combined effect of solar radiation
and the outdoor-indoor temperature difference, the one-step calculation of space sensible cooling
load Qrs,w, Btu /h (W), can be performed as
Qrs,w  UA (CLTD) (6.35)
where U  overall heat-transfer coefficient of exterior wall or roof, Btu /h ft2 °F (W/m2 °C)
A  area of exterior wall, roof, or window including frame or sash, ft2 (m2)
CLTD  cooling load temperature difference, °F (°C)
The CLTD values recommended by ASHRAE for calculating the space sensible cooling load through
flat roofs and sunlit walls of various constructions are listed in Tables 6.2 and 6.4, respectively. The
values in both Table 6.2 and Table 6.4 were calculated under the following conditions. In other words,
these are the conditions under which the listed data can be applied directly without adjustments:
 Indoor temperature of 78°F (25.6°C)
 Outdoor maximum temperature of 95°F (35°C) with an outdoor daily mean of 85°F (29.4°C) and
an outdoor daily range of 21°F (11.7°C)
 Solar radiation of 40° north latitude on July 21
 Roof with dark, flat surface
Fsur  1  0.02  1
Lex
 (Uroof Aroof  Uwall Awall  Uwin Awin  Upart Apart)
6.26 CHAPTER SIX
6.27
TABLE 6.4 CLTD for Calculating Sensible Cooling Loads from Flat Roofs, °F
Solar time, h
Description of Weight, U value,
construction lb/ft2 Btu/hft2°F 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
Without suspended ceiling
2.5-in. wood with 13 0.093 30 26 23 19 16 13 10 9 8 9 13 17 23 29 36 41 46
2-in. insulation
4-in. wood with 18 0.078 38 36 33 30 28 25 22 20 18 17 16 17 18 21 24 28 32
2-in. insulation
With suspended ceiling
1-in. wood with 10 0.083 25 20 16 13 10 7 5 5 7 12 18 25 33 41 48 53 57
2-in. insulation
2.5-in. wood with 15 0.096 34 31 29 26 23 21 18 16 15 15 16 18 21 25 30 34 38
1-in. insulation
8-in. lightweight 33 0.093 39 36 33 29 26 23 20 18 15 14 14 15 17 20 25 29 34
concrete
4-in. heavyweight 54 0.090 30 29 27 26 24 22 21 20 20 21 22 24 27 29 32 34 36
concrete with
2-in. insulation
2.5-in. wood with 15 0.072 35 33 30 28 26 24 22 20 18 18 18 20 22 25 28 32 35
2-in. insulation
Roof terrace 77 0.082 30 29 28 27 26 25 24 23 22 22 22 23 23 25 26 28 29
system
6-in. heavyweight 77 0.088 29 28 27 26 25 24 23 22 21 21 22 23 25 26 28 30 32
concrete with
2-in. insulation
4-in. wood with 19 0.082 35 34 33 32 31 29 27 26 24 23 22 21 22 22 24 25 27
2-in. insulation 20 0.064
Conditions of direct application and adjustments are stated in the text.
Source: Abridged with permission from ASHRAE Handbook 1989, Fundamentals.
Hours of
maximum Minimum Maximum Difference
18 19 20 21 22 23 24 CLTD CLTD CLTD in CLTD
49 51 50 47 43 39 35 19 8 51 43
36 39 41 43 43 42 40 22 16 43 27
57 56 52 46 40 34 29 18 5 57 52
41 43 44 44 42 40 37 21 15 44 29
38 42 45 46 45 44 42 21 14 46 32
38 38 38 37 36 34 33 19 20 38 18
38 40 41 41 40 39 37 21 18 41 23
31 32 33 33 33 33 32 22 33 22 11
33 34 34 34 33 32 31 20 21 34 13
30 32 34 35 36 37 36 23 21 37 16
 Outer surface R value Ro  0.333 h ft2 °F/Btu (0.06 m2 °C/W) and inner surface Ri  0.685
h ft2 °F/Btu (0.123 m2 °C/W)
 No attic fans or return air ducts in suspended ceiling space
The following formula can be used for adjustments when the conditions are different from those
mentioned:
CLTDcorr  CLTD  78  Tr  Tom  85 (6.36)
where 78  Tr  indoor temperature correction; Tr is indoor temperature, °F (°C)
Tom  85  outdoor temperature correction; Tom is outdoor mean daily temperature, °F (°C)
In Table 6.4, the roof terrace system includes the following:
 2-in. (50-mm) lightweight concrete
 Airspace
 2-in. (50-mm) insulation [5.7 lb / ft3 (91.2 kg/m3)]
 0.5-in. (13-mm) slag or stone
 0.375-in. (9.5-mm) felt and membrane
For a pitched roof with a suspended ceiling, the area A in Eq. (6.35) should be the area of the suspended
ceiling. If a pitched roof has no suspended ceiling under it, then the actual CLTD is slightly
higher than the value listed in Table 6.4 because a greater area is exposed to the outdoor air.
Space Cooling Load due to Solar Heat Gain through Fenestration
The space sensible cooling load from solar heat gain transmitted through the window facing a specific
direction Qrs, s, Btu /h (W), can be calculated as follows:
Qrs, s  Qsun  Qsh
 As 	 SCLs 	 SC  Ash 	 SCLsh 	 SC (6.37)
where Qsun  space cooling load from solar heat gain through sunlit area of window glass, Btu/h (W)
Qsh  space cooling load from solar heat gain through shaded area of window glass, Btu/h (W)
As, Ash  sunlit and shaded area, ft2 (m2)
SC  shading coefficient
SCLs  solar cooling load for sunlit glass facing specific direction, Btu/h ft2 (W/m2)
SCLsh  solar cooling load for shaded area as if glass is facing north, Btu/h ft2 (W/m2)
Zone types for use with SCLs and SCLsh tables, single-story building, are listed in Table 6.5.
July solar cooling loads for sunlit glass 40° north latitude are listed in Table 6.6. Refer to ASHRAE
Handbook 1993, Fundamentals, for zone types for multistory buildings and other details.
In Eq. (6.37), at a given time, As  Ash  Aglass. Here Aglass indicates the glass area of the window,
in ft2 (m2). In the northern hemisphere for a conditioned space with southern orientation, the
maximum SCLs may occur in December instead of June.
Space Cooling Load due to Heat Gain through Wall Exposed to
Unconditioned Space
When a conditioned space is adjacent to an area that is unconditioned, and if the temperature fluctuation
in this area is ignored, then the sensible heat gain qrs transferred through the partitioned walls
and interior windows and doors, in Btu/h (W), can be calculated as
6.28 CHAPTER SIX
qrs  AU(Tun  Tr) (6.38)
where Tun,Tr  daily mean air temperature of adjacent area that is unconditioned and space temperature,
respectively, °F (°C). For an adjacent area that is not air conditioned and has heat sources
inside, such as a kitchen or boiler room, Tun may be 15°F (8.3°C) higher than the outdoor temperature.
For an adjacent area without any heat source other than electric lights, Tun  Tr may be
between 3 and 8°F (1.7 and 4.4°C). For floors built directly on the ground or located above a basement
that is neither ventilated nor conditioned, the space sensible cooling load from the heat gain
through the floor can often be ignored.
Calculation of Internal Cooling Loads and Infiltration
The calculation of the internal heat gains of people, lights, equipment, and appliances qint, s and heat
gains of infiltration qinf, all in Btu /h (W), using the CLTD/SCL/CLF method is the same as that
which uses TFM. The sensible internal heat gain that contains the radiative component is then multiplied
by a cooling load factor CLFint to convert to space sensible cooling load Qint, s, Btu /h (W),
and can be calculated as
Qint, s  qint, s (CLFint) (6.39)
In conditioned spaces in which an air system is operated at nighttime shutdown mode, CLFint is
equal to 1 during the occupied period when the air system is operating. Refer to ASHRAE Handbook
1993, Fundamentals, for CLFint when the air system is operated 24 h continuously.
Since internal latent heat gains qint, l are instantaneous internal latent cooling loads Qint, l both in
Btu/h (W),
Qint,l  qint,l (6.40)
LOAD CALCULATIONS 6.29
TABLE 6.5 Zone Types for Use with SCL and CLF Tables, Single-Story Building
Zone parameters Zone type Error band
No. Floor Partition Inside Glass People and
walls covering type shade solar equipment Lights Plus Minus
1 or 2 Carpet Gypsum * A B B 9 2
1 or 2 Carpet Concrete block * B C C 9 0
1 or 2 Vinyl Gypsum Full B C C 9 0
1 or 2 Vinyl Gypsum Half to none C C C 16 0
1 or 2 Vinyl Concrete block Full C D D 8 0
1 or 2 Vinyl Concrete block Half to none D D D 10 6
3 Carpet Gypsum * A B B 9 2
3 Carpet Concrete block Full A B B 9 2
3 Carpet Concrete block Half to none B B B 9 0
3 Vinyl Gypsum Full B C C 9 0
3 Vinyl Gypsum Half to none C C C 16 0
3 Vinyl Concrete block Full B C C 9 0
3 Vinyl Concrete block Half to none C C C 16 0
4 Carpet Gypsum * A B B 6 3
4 Vinyl Gypsum Full B C C 11 6
4 Vinyl Gypsum Half to none C C C 19 1
A total of 14 zone parameters are defined. Those not shown in this were selected to achieve the minimum error band shown in
the right-hand column for solar cooling load.
*The effect of inside shade is negligible in this case.
Source: Adapted from ASHRAE Handbook 1997, Fundamentals. Reprinted with permission.
6.30
TABLE 6.6 July Solar Cooling Load for Sunlit Glass, 40° North Latitude
Glass Solar time, h
face 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
Zone type A
N 0 0 0 0 1 25 27 28 32 35 38 40 40 39 36
NE 0 0 0 0 2 85 129 134 112 75 55 48 44 40 37
E 0 0 0 0 2 93 157 185 183 154 106 67 53 45 39
SE 0 0 0 0 1 47 95 131 150 150 131 97 63 49 41
S 0 0 0 0 0 9 17 25 41 64 85 97 96 84 63
SW 0 0 0 0 0 9 17 24 30 35 39 64 101 133 151
W 1 0 0 0 0 9 17 24 30 35 38 40 65 114 158
NW 1 0 0 0 0 9 17 24 30 35 38 40 40 50 84
Horiz. 0 0 0 0 0 24 69 120 169 211 241 257 259 245 217
Zone type B
N 2 2 1 1 1 22 23 24 28 32 35 37 38 37 35
NE 2 1 1 1 2 73 109 116 101 73 58 52 48 45 41
E 2 2 1 1 2 80 133 159 162 143 105 74 63 55 48
SE 2 2 1 1 1 40 81 112 131 134 122 96 69 58 49
S 2 2 1 1 1 8 15 21 36 56 74 86 87 79 63
SW 6 5 4 3 2 9 16 22 27 31 36 58 89 117 135
W 8 6 5 4 3 9 16 22 27 31 35 37 59 101 139
NW 6 5 4 3 2 9 16 22 27 31 34 37 37 46 76
Horiz. 8 6 5 4 3 22 60 104 147 185 214 233 239 232 212
Zone type C
N 5 5 4 4 4 24 23 24 27 30 33 34 35 34 32
NE 7 6 6 5 6 75 106 107 88 61 49 47 45 43 40
E 9 8 8 7 8 83 130 148 145 124 89 62 56 52 47
SE 9 8 7 6 6 45 82 107 121 121 107 82 59 51 47
S 7 7 6 5 5 12 18 23 36 54 70 79 79 70 54
SW 14 12 11 10 9 15 21 26 29 33 36 57 86 110 124
W 17 15 13 12 11 17 22 27 31 34 36 37 59 98 132
NW 12 11 10 9 8 14 20 25 29 32 34 36 36 44 73
Horiz. 24 21 19 17 16 34 68 107 144 175 199 212 215 207 189
Zone type D
N 8 7 6 6 6 21 21 21 24 27 29 31 32 31 30
NE 11 10 9 8 9 63 87 90 77 58 49 48 46 44 42
E 15 13 12 11 11 70 107 123 124 110 85 65 60 57 53
SE 14 13 11 10 10 39 68 90 102 104 95 78 60 55 51
S 11 10 9 8 7 12 17 21 32 46 59 67 69 63 52
SW 21 19 17 15 14 18 22 25 28 31 34 51 74 94 106
W 25 23 20 18 17 21 24 28 30 33 34 35 53 84 112
NW 18 16 15 13 12 17 21 24 27 30 32 33 34 41 64
Horiz. 37 33 30 27 24 38 64 95 124 150 171 185 191 188 176
Notes:
1. Values are in Btu /hft2.
2. Apply data directly to standard double-strength glass with no inside shade.
3. Data apply to 21st day of July.
4. For other types of glass and internal shade, use shading coefficients as multiplier. For externally shaded glass, use north orientation.
Source: ASHRAE Handbook 1997, Fundamentals. Reprinted with permission.
16 17 18 19 20 21 22 23 24
31 31 36 12 6 3 1 1 0
32 26 18 7 3 2 1 0 0
33 26 18 7 3 2 1 0 0
34 27 18 7 3 2 1 0 0
42 31 20 8 4 2 1 0 0
152 133 93 35 17 8 4 2 1
187 192 156 57 27 13 6 3 2
121 143 130 46 22 11 5 3 1
176 125 70 29 14 7 3 2 1
32 31 35 16 10 7 5 4 3
36 30 23 13 9 6 5 3 3
41 34 25 15 10 7 5 4 3
42 35 26 15 10 8 6 4 3
46 37 27 16 11 8 6 4 3
138 126 94 46 31 21 15 11 8
166 173 147 66 43 30 21 15 11
108 128 119 51 33 22 16 11 8
180 137 90 53 37 27 19 14 11
29 29 34 14 10 8 7 6 6
36 31 25 16 13 11 10 9 8
43 37 30 20 17 15 13 12 11
42 36 29 19 16 14 13 11 10
40 33 26 16 13 12 10 9 8
125 111 80 37 28 23 20 17 15
153 156 128 50 35 28 24 21 19
102 118 107 39 26 21 17 15 13
160 123 83 53 44 38 34 30 27
28 29 32 17 14 12 11 10 9
39 35 29 22 19 17 15 14 12
48 43 37 29 25 22 20 18 16
47 42 35 27 24 21 19 17 16
41 36 30 22 19 17 15 14 12
109 100 78 45 37 33 29 26 23
130 135 116 57 46 39 35 31 28
87 101 94 42 33 29 25 22 20
156 128 96 72 63 56 50 45 41
Internal load density (ILD), W/ ft2 (W/m2), indicates the total internal heat gains of people, lights,
and equipment, and it can be calculated as
(6.41)
where SHGp, LHGp  sensible and latent heat gains for occupants, respectively, Btu /h (W)
Afl  floor area, ft2 (m2)
WA,l, WA,e  lighting and equipment power density, respectively,W/ ft2 (W/m2)
Both infiltration sensible heat gain qinf, s and infiltration latent heat gain qinf, l, in Btu/h (W), are instantaneous
space cooling loads; they also can be expressed as
Qinf, s  qinf, s Qinf, l  qinf, l (6.42)
where Qinf, s, Qinf, l  infiltration sensible and latent cooling loads, respectively, Btu /h (W).
Example 6.1. A return air plenum in a typical floor of a multistory building has the following
construction and operating characteristics:
ILD 
SHGp  LHGp
3.413 Afl
 WA,l  WA,e
LOAD CALCULATIONS 6.31
1.5 W/ ft2 (16.1 W/m2)
11,800 cfm (334 m3 /min)
0.073 lb/ ft3 (1.168 kg/m3)
0.5
0.2 Btu/h ft2 °F (1.136 W/m2 °C)
0.32 Btu/h ft2 °F (1.817 W/m2 °C)
0.21 Btu/h ft2 °F (1.192 W/m2 °C)
11,900 ft2 (1106 m2)
1920 ft2 (178 m2)
1.0
24°F (13.3°C)
75°F (23.9°C)
Wattage of electric lights
Return air volume flow rate
Density of return air r
Fraction of heat to plenum Flp
U values: exterior wall of plenum
suspended ceiling
floor
Area of ceiling and floor
Area of exterior wall of plenum
CLF of electric lights
CLTD of exterior wall of plenum
Space temperature
Determine the return air plenum temperature and the space cooling load from the electric lights
when they are recess-mounted on the ceiling and the ceiling plenum is used as a return plenum.
Solution. From Eq. (6.17),
qlp  Flpqs, l  0.5 	 1.5 	 11,900 	 3.413  30,461 Btu/h (W)
Since , and given Eq. (6.22), the
temperature of return air in the return plenum is
 75 
39,933
18,866
 77.12F (25.07C)
 75 
30,717  0.2 	 1920 	 24
12,559  0.32 	 11,900  0.21 	 11,900
Tp  Tr 
qlp  Uwp AwpCLTDwp
60V?r r cpa  Ucl Acl  UflAfl
60V?r r cpa  60 	 11,800 	 0.073 	 0.243  12,559 Btu / hF
From Eqs. (6.19a) and (6.19b), heat transfer from the plenum air to conditioned space through the
ceiling and floor is calculated as
Then, from Eq. (6.20), heat to space is calculated as
qes, l  qld  qcl  qfl
 30,461  7997  5248  43,706 Btu/h (W)
From Eq. (6.39), the space sensible cooling load from electric lights is
Qrs, l  CLFint qes, l  1.0 	 43,692  43,692 Btu/h (12,802 W)
Space Cooling Load of Night Shutdown Operating Mode
In commercial buildings, air systems are often operated in night shutdown mode during unoccupied
hours in summer. The accumulated stored heat because of the external heat gains increases the
space cooling load during cool-down and conditioned periods. On the other hand, heat losses to the
surroundings due to the radiant heat exchange between the outer surface of the building and the sky
vault and surroundings decrease the accumulated stored heat as well as the space cooling load, although
the radiative heat losses to the sky vault and surroundings partly compensate the stored heat
released to the space. However, overlooking the remaining stored heat released to the space during
cool-down and conditioned periods in summer is the limitation of the CLTD/SCL/CLF method
compared to TFM, especially when peak load occurs during the cool-down period. As in TFM, an
increase of up to 10 percent is recommended by ASHRAE/IES Standard 90.1-1989 for pickup load
during the cool-down period for air systems operated at nighttime shutdown mode.
6.8 COOLING COIL LOAD
Basics
Based on the principle of heat balance, the cooling coil load is given as
Total enthalpy of entering air  total enthalpy of leaving air
 cooling coil load (or cooling capacity)  heat energy of condensate
Since the heat energy of the condensate is small and can be ignored, the cooling coil load Qcc,
Btu/h (W), can be calculated as
(6.43)
where volume flow rate of supply air, cfm [m3 / (60 s)]
s  density of supply air, lb/ ft3 (kg/m3)
hae, hcc  enthalpy of entering air and conditioned air leaving coil, respectively, Btu / lb (J /kg)
Of this, the sensible cooling coil load Qcs, Btu /h (W), is
(6.44) Qcs  60V?s s cpa(Tae  Tcc)
V?
s 
Qcc  60V?s s (haehcc)
 0.21 	 11,900(77.1  75)  5248 Btu / h (W)
qfl  Ufl Afl (Tp  Tr)
 0.32 	 11,900(77.1  75)  7997 Btu / h (W)
qcl  Ucl Acl (Tp  Tr)
6.32 CHAPTER SIX
where Tae, Tcc  temperature of entering air and conditioned air leaving coil, respectively, °F (°C).
And the latent coil load Qcl, Btu /h (W), is
(6.45)
where wae, wcc  humidity ratio of entering air and conditioned air leaving coil, respectively, lb/ lb
(kg/ kg). Also,
Qcc  Qcs  Qcl (6.46)
From Fig. 6.4, alternatively, the sensible cooling coil load can be calculated as
Qcs  Qrs  qs, s  qr, s  Qo, s (6.47)
where qs, s, qr, s  supply and return system heat gain (as mentioned in preceding section, both are
instantaneous cooling loads), Btu/h (W)
Qo, s  sensible load from outdoor air intake, Btu/h (W)
And the latent coil load can be calculated as
Qcl  Qrl  Qo, l (6.48)
where Qo, l  latent load from outdoor air intake, Btu/h (W). The supply system heat gain consists
of mainly the supply fan power heat gain qsf and supply duct heat gain qsd; and the return system
heat gain comprises the return fan power heat gain qrf, return duct heat gain qrd, and ceiling plenum
heat gain qrp, all in Btu /h (W).
Fan Power
In the air duct system, the temperature increase from the heat released to the airstream because of
frictional and dynamic losses is nearly compensated by the expansion of air from the pressure drop
of the airstream. Therefore, it is usually assumed that there is no significant temperature increase
from frictional and dynamic losses when air flows through an air duct system.
Fan power input is almost entirely converted to heat energy within the fan. If the fan motor is
located in the supply or return airstream, the temperature increase across the supply (or return fan)
Tf, °F (°C), can be calculated as
(6.49)
where pt  fan total pressure, in. WC
f, m  total efficiency of fan and efficiency of motor
If the motor is located outside the airstream, then, in Eq. (6.49), m  1.
The pt of the return fan for a central hydronic air conditioning system in commercial buildings
is usually 0.25 to 0.5 of the pt of the supply fan. Therefore, the temperature increase of the return
fan is far smaller than that of the supply fan. The temperature increase of the relief fan or exhaust
fan affects only the relief or exhaust airstream. It is not a part of the supply and return system heat
gain. A relief fan is used to relieve excess space pressure when 100 percent outdoor air is flowing
through the supply fan for free cooling.
Duct Heat Gain
Duct heat gain is the heat transfer caused by the temperature difference between the ambient air and
the air flowing inside the air duct. Duct heat gain is affected by this temperature difference, the
Tf 
0.37pt
f m
Qcl  60V?s s (wae  wcc)hfg, 32
LOAD CALCULATIONS 6.33
thickness of the duct insulation layer, air volume flow rate, and the size and shape of the duct.
Detailed calculations are presented in Secs. 17.4 and 20.16. A rough estimate of the temperature increase
of the supply air for an insulated duct is as follows:
Supply air velocity Air temperature rise
 2000 fpm (10 m/ s) 1°F/ 100 ft (0.6°C/30 m) duct length
 2000 fpm (10 m/ s) 0.75°F/ 100 ft (0.45°C/30 m) duct length
Temperature of Plenum Air and Ventilation Load
For a ceiling plenum using a return plenum, the plenum air temperature can be calculated from
Eq. (6.22). The temperature increase of the plenum air, caused by the heat released from the electric
lights (Tp  Tr), is affected by their power input, type of lighting fixture, return air volume flow
rate, and construction of the ceiling plenum. The temperature increase of plenum air Tp  Tr is usually
between 1 and 3°F (0.6 and 1.7°C).
From Eqs. (6.26) and (6.27), the sensible and latent loads Qo, s and Qo, l, Btu /h (W), which are
attributable to the outdoor air intake, can be similarly calculated, except in Eqs. (6.26) and
(6.27) should be replaced by the volume flow rate of outdoor air , cfm (m3 /min).
System heat gains are mainly due to convective heat transfer from the surfaces. For simplification,
they are considered instantaneous cooling coil loads.
6.9 COOLING LOAD CALCULATION BY FINITE
DIFFERENCE METHOD
When both heat and moisture transfer from the surface of the walls, ceiling and carpet or floors
should be considered in the space cooling load calculation during the cool-down period in summer
in a location where the outdoor climate is hot and humid, the finite difference method might be the
best choice.
Finite Difference Method
Because of the rapid increase in the use of microcomputers in the HVAC&R calculations, it is now
possible to use a finite difference method, a numerical approach, to solve transient simultaneous
heat- and moisture-transfer problems in heating and cooling load calculations and energy
estimations. The finite difference method divides the building structures into a number of sections.
A fictitious node i is located at the center of each section or on the surface, as shown in Fig. 6.8. An
energy balance or a mass balance at each node at selected time intervals results in a set of algebraic
equations that can be employed to determine the temperature and moisture for each node in terms
of neighboring nodal temperatures or moisture contents, nodal geometry, and the thermal and moisture
properties of the building structure. The stored heat energy and moisture are expressed as an
increase of internal energy and moisture content at the nodes.
Heat conduction can be approximated by using the finite difference form of the Fourier law, as
(6.50)
where k  thermal conductivity, Btu / h ft °F (W/m°C)
Ai  area of building structure perpendicular to direction of heat flow, ft2 (m2)
qi1:i 
kAi(T ti
1  T t
1)
x
V?
o
V?
inf
6.34 CHAPTER SIX
LOAD CALCULATIONS 6.35
FIGURE 6.8 Building structures and nodes for a typical room.
T  temperature, °F (°C)
x  spacing between the nodal points, ft (m)
In Eq. (6.50), superscript t denotes at time t. Each nodal equation is solved explicitly in terms of the
future temperature of that node. The explicit method is simpler and clearer than the more complex
implicit method. The time derivative is then approximated by a forward finite difference in time, or
(6.51)
Compared with the transfer function method to calculate the cooling load, the finite difference
method has the following benefits:
1. It solves heat- and moisture-transfer load calculations simultaneously.
2. The concept and approach are easily understood.
3. It permits custom-made solutions for special problems.
4. It allows direct calculation of cooling loads and energy estimates.
Its drawbacks are mainly due to the great number of computerized calculations and comparatively
fewer computer programs and less information and experience are available.
Simplifying Assumptions
When the finite difference method is used to calculate space cooling loads, simplifications are often
required to reduce the number of computer calculations and to solve the problem more easily. The
errors due to simplification should be within acceptable limits.
For a typical room in the building, as shown in Fig. 6.8, the following are the simplifying
assumptions:
 Heat and moisture flow are one-dimensional.
 Thermal properties of the building materials are homogeneous.
 The properties of the airstream flowing over the surface of the building structures are homogeneous.
 The surface temperature differences between the partition walls, ceiling, and floors are small;
therefore, the radiative exchange between these surfaces can be ignored.
 The radiative energy received on the inner surface of the building structures can be estimated as
the product of the shape factor and the radiative portion of the heat gains, and the shape factor is
approximated by the ratio of the receiving area to the total zone area.
 During the operating period, the heat capacity of the space air is small compared with other heat
gains; therefore, it can be ignored.
 When the air system is not operating during the night shutdown period, the heat capacity of the
space air has a significant influence on the space air temperature; therefore, it should be taken into
account.
 Different heat- and mass-transfer coefficients and analyses are used for the operating period and
shutdown periods.
Heat and Moisture Transfer at Interior Nodes
Consider an interior node i as shown in the upper part of Fig. 6.8. For a one-dimensional heat flow,
if there is no internal energy generation, then according to the principle of heat balance,
Ti
t

T i
t t  T i
t
t
6.36 CHAPTER SIX
Conduction heat from node i  1  conduction heat from node i  1
 rate of change of internal energy of node i
(6.52)
where Ui  internal energy of node i, Btu/lb (J/kg)
b , cb  density and specific heat of building material, lb / f t 3 (kg/m3) and Btu/ lb°F
(J/kg°C)
t  selected time interval, s or min
Substituting Eq. (6.50) into (6.52) and solving for Ti
tt, we have
(6.53)
In Eq. (6.53), Fo is the Fourier number and is defined as
(6.54)
Subscript b indicates the building material.
The choice of spacing x and the time interval t must meet some criterion to ensure convergence
in the calculations. The criterion is the stability limit, or
(6.55)
Similarly, for moisture transfer at the interior nodes,
(6.56)
and
(6.57)
where X  moisture content, dimensionless
Dlv  mass diffusivity of liquid and vapor, ft2/s (m2/s)
Heat and Moisture Transfer at Surface Nodes
For a one-dimensional heat flow, the energy balance at surface node i of the partition wall as shown
in Fig. 6.8 is
Conductive heat from node i  1  convective heat transfer from space air
 latent heat of moisture transfer from space air  radiative heat from internal loads
 rate of change of internal energy of node i
 F1:iLr  Fp:iOr  Fm:i Mr 
bcb Ai
x
2
T i
t t  T i
t
t
kAi (T i1
t  T ti
)
x
 hci Ai(T t
r  T ti
) 
ahmi Ai Xti (wt
r  wt
is) hfg
Fomass 
Dlv t
(x)2  1
2
Xi tt  Fomass (Xti
1  Xti
1)  (1  2 Fomass) Xti
Fo  1
2
Fo 
kb
bcb
t
(x)2
T i
tt  Fo(T ti
1  T ti
1)  (1  2Fo)T ti

bcb Aix (T i
tt  T i
t)
t
qi1:i  qi1:i 
Ui
t
LOAD CALCULATIONS 6.37
where hci  convective heat-transfer coefficient of surface i, Btu/h ft2 °F (W/m2 °C)
a  density of space air, lb/ f t 3 (kg/m3)
hmi  convective mass-transfer coefficient of surface i, ft/s (m/s)
wt
is  humidity ratio corresponding to surface i at time t, lb/ lb (kg /kg)
hf g  latent heat of vaporization at surface temperature, Btu/lb (J /kg)
 shape factor between surface of lights, occupants, appliances and surface i
Lr,Or, Mr  radiative portion of heat energy from lights, occupants, and appliances and machines,
Btu/ h (W)
Solving for gives
(6.58)
In Eq. (6.58) Bi is the Biot number and can be expressed as
(6.59)
The stability limit of the surface nodes requires that
(6.60)
Similarly, according to the principle of conservation of mass, the moisture content at surface node i
is given as
(6.61)
The temperature and moisture content at other surface nodes such as the floor, ceiling, exterior
walls, glass, and Plexiglas of the lighting fixture can be found in the same manner.
According to ASHRAE Handbook 1993, Fundamentals, the radiative and convective portions of
the heat gains are as follows:
Radiative, percent Convective, percent
Fluorescent lights 50 50
People 33 67
External walls and roofs 60 40
Appliance and machines 20–80 80–20
Space Air Temperature and Cooling Loads
If the infiltrated air is ignored, the heat balance on the space air or the plenum air can be described
by the following relationship:
Internal energy of supply air  convective heat transfer from building structures
 convective heat transfer from internal loads  internal energy of space air
60V?s scpaT s
t t  n
k1
hck Ak(T i
t t  T r
t t)  Lc  Oc  Mc  60V?s scpaT r
t t
Xi tt  (1  2Fomass)Xi t  2Fomass Xti
1 
2 ahmi t Xi t (wt
r  wt
is)
b x
Fo(1  Bi)  1/2
Bi 
hc x
k

F1:i Lr
hci Ai

Fp:i Or
hci Ai

Fm:i Mr
hci Ai 	  [1  2Fo(1  Bi)]T i
t
Ti
tt  2FoT ti
1  Bi	T r
t 
a hmi hfgXi t (wr t  wis
t )
hci
T i
tt
F1:i, Fp:i, Fm:i
6.38 CHAPTER SIX
Solving for Tr
tt, we have
(6.62)
where ,Ts  volume flow and temperature of supply air, cfm (m3 / (60 s)) and °F (°C)
Lc, Oc, Mc  convective heat from lights, occupants, and appliances, Btu/h (W)
The temperature of the return plenum air can be similarly calculated. The space sensible cooling
loads, therefore, can be calculated as
(6.63)
The latent heat gains are instantaneous latent cooling loads.
6.10 HEATING LOAD
Basic Principles
The design heating load, or simply the heating load, is always the maximum heat energy that might
possibly be required to supply to the conditioned space at winter design conditions to maintain the
winter indoor design temperature. The maximum heating load usually occurs before sunrise on the
coldest days. The following are the basic principles of heating load calculation that are different
from those for the cooling load calculation:
 All heating losses are instantaneous heating loads. The heat storage effect of the building structure
is ignored.
 Solar heat gains and the internal loads are usually not taken into account except for those internal
loads Qin, Btu /h (W), that continuously release heat to the conditioned space during the operating
period of the whole heating season.
 Only that latent heat Ql , Btu /h (W), required to evaporate liquid water for maintaining necessary
space humidity is considered as heating load.
For a continuously operated heating system, the heating load Qrh, Btu /h (W), can be calculated as
(6.64)
where Qtran  transmission loss, Btu/h (W)
Qinf, s  sensible heat loss from infiltrated air, Btu /h (W)
Qmat  heat added to entering colder product or material, Btu/ h (W)
Transmission Loss
Transmission loss Qtran, Btu /h (W), is the sum of heat losses from the conditioned space through
the external walls, roof, ceiling, floor, and glass. If the calculation is simplified to a steady-state
heat flow, then
(6.65) Qtran   AU(Tr  To)
Qrh  Qtran  Qif,.s  Ql  Qmat  Qin
Qrs
tt  n
k1
hck Ak(T i
t t  T r
tt )  Lc  Oc  Mc
V?
s
T r
tt 
60V?s scpaT s
t t  n
k1
hck AkT i
t t  Lc  Oc  Mc
60V?s scpa  n
k1
hck Ak
LOAD CALCULATIONS 6.39
where A  area of walls, roof, ceiling, floor, or glass, ft2 (m2)
U  overall heat-transfer coefficient of walls, roof, ceiling, floor, or glass, Btu/h ft2 °F
(W/m2 °C)
When the winter outdoor design temperature is used for To, the heat loss calculated by transient
heat transfer will be less than that from Eq. (6.65) because of the cyclic fluctuations of the outdoor
temperature and the heat storage in the building structures.
For concrete slab floors on a grade, heat loss Qfl, Btu /h (W), is mostly through the perimeter instead
of through the floor and the ground. It can be estimated as
(6.66)
where P  length of perimeter, ft (m)
Cfl  heat loss coefficient per foot (meter) of perimeter length, Btu/h ft °F (W/m°C)
For areas having an annual total of heating degree-days HDD65  5350 and for a concrete wall
with interior insulation in the perimeter having an R value of 5.4 h ft2 °F/Btu (0.97 m2 °C/W),
Cfl  0.72 Btu/h ft °F (1.25 W/m°C). Refer to ASHRAE Handbook 1989, Fundamentals, chapter
25, for more details.
For basement walls, the paths of the heat flow below the grade line are approximately concentric
circular patterns centered at the intersection of the grade line and the basement wall. The thermal
resistance of the soil and the wall depends on the path length through the soil and the construction
of the basement wall. A simplified calculation of the heat loss through the basement walls and floor
Qb,g, Btu /h (W), is as follows:
(6.67)
where Ab,g  area of basement wall or floor below grade, ft2 (m2)
Ub,g  overall heat-transfer coefficient of wall or floor and soil path, Btu/h ft2 °F
(W/m2 °C)
The values of Ub,g are roughly given as follows:
0 to 2 ft below grade Lower than 2 ft
Uninsulated wall 0.35 0.15
Insulated wall 0.14 0.09
Basement floor 0.03 0.03
Refer to ASHRAE Handbook 1989, Fundamentals, for details.
The space heating load in the perimeter zone, including mainly transmission and infiltration
losses, is sometimes expressed in a linear density qh, ft, in Btu/ h per linear foot of external wall, or
Btu/hft (W/m).
Adjacent Unheated Spaces
Heat loss from the heated space to the adjacent unheated space Qun, Btu /h (W), is usually assumed
to be balanced by the heat transfer from the unheated space to the outdoor air, and this can be calculated
approximately by the following formula:
(6.68) Qun  n
i1
AiUi(Tr  Tun)  (m
j1
AjUj  V?inf o cpa)(Tun  To)
Qb, g  Ab, gUb, g(Tbase  To)
Qfl  PCfl(Tr  To)
6.40 CHAPTER SIX
where Ai,Ui  area and overall heat-transfer coefficient of partitions between heated space and
unheated space, ft2 (m2) and Btu/h ft2 °F (W/m2 °C)
Aj,Uj  area and overall heat-transfer coefficient of building structures exposed to outdoor
air in unheated space, ft2 (m2) and Btu/h ft2 °F (W/m2 °C)
The temperature of the unheated space Tun, °F (°C), can be calculated as
(6.69)
Latent Heat Loss and Heat Loss from Products
In Eq. (6.64), Ql, Btu /h (W), represents the heat required to evaporate the liquid water to raise the
relative humidity of the space air or to maintain a specific space relative humidity, i.e.,
(6.70)
where  mass flow of water evaporated, lb/h (kg/ s)
 volume flow rate of outdoor ventilation air and infiltrated air, cfm [m3 / (60 s)]
 mass flow of water evaporated from minimum number of occupants that always stay
in conditioned space when heating system is operated, lb /h (kg / s)
In Eq. (6.70), hfg,57 indicates the latent heat of vaporization at a wet-bulb temperature of 57°F
(13.9°C), that is, 72°F (22.2°C) dry-bulb temperature and a relative humidity of 40 percent. Its
value can be taken as 1061 Btu/ lb (2.47 	 106 J / kg).
For factories, heat added to the products or materials that enter the heated space within the occupied
period Qmat, Btu /h (W), should be considered part of the heating load and can be calculated as
(6.71)
where  mass flow rate of cold products and cold material entering heated space, lb/h (kg/ s)
cpm  specific heat of product or material, Btu/ lb °F (J /kg°C)
Infiltration
Infiltration can be considered to be 0.15 to 0.4 air changes per hour (ach) at winter design conditions
only when (1) the exterior window is not well sealed and (2) there is a high wind velocity. The
more sides that have windows in a room, the greater will be the infiltration. For hotels, motels, and
high-rise domicile buildings, an infiltration rate of 0.038 cfm / ft2 (0.193 L/ s m2) of gross area of
exterior windows is often used for computations for the perimeter zone.
As soon as the volume flow rate of infiltrated air , cfm (m3 /min), is determined, the sensible
heat loss from infiltration Qinf, s, Btu /h (W), can be calculated as
(6.72)
where o  density of outdoor air, lb/ ft3 (kg/m3).
Setback of Night Shutdown Operation
During a nighttime or unoccupied period, when the space temperature is set back lower than the
indoor temperature during the operating period, it is necessary to warm up the conditioned space
Qinf, s  V?inf
ocpa(Tr  To)
V?
inf
m? mat
Qmat  m? matcpm(Tr  To)
m? p
V?
oinf
m? w
Ql  m? w hfg,57  [60V?oinf
o(wr  wo)  m? p]hfg,57
Tun  n
i1
AiUiTr 
(60V?inf ocpa  m
j1
AjUj)To
n
i1
AiUi  60V?inf ocpa  m
j1
AjUj
LOAD CALCULATIONS 6.41
the next morning before the arrival of the occupants in offices or other buildings. The warm-up or
pickup load depends on the pickup temperature difference, the outdoor-indoor temperature difference,
the construction of the building envelope or building shell, and the time required for warm-up.
There are insufficient data to determine the oversizing factor of the capacity of the heating plant for
morning warm-up. According to tests by Trehan et al. (1989), one can make a rough estimate of the
energy required to warm up or raise a space temperature by 12°F (6.7°C) for a single- or two-story
building with a roof-ceiling U value of 0.03 Btu/h ft2 °F (0.17 W/m2 °C) and a U value for external
walls of 0.08 Btu /h ft2 °F (0.45 W/m2 °C) as follows:
Outdoor-indoor temperature difference, °F Warm-up period, h Oversizing factor, percent
35 1 40
55 2 40
55 1 100
As discussed in Sec. 5.16, heating pickup load during warm-up period depends on the difference
between space temperature and occupied set point, and the amount of time prior to scheduled occupancy,
and is often determined by computer software.
6.11 LOAD CALCULATION SOFTWARE
Introduction
Today, most of the load calculations are performed by personal computers. Among the widely
adopted load calculation and energy analysis software, only Building Load Analysis and System
Thermodynamics (BLAST) developed by the University of Illinois adopts the heat balance method.
All the others are based on the transfer function or weighting factors method.
Load calculation software can also be divided into two categories. The first includes those
developed by government or public institutions, such as the Department of Energy (DOE-2.0),
National Bureau of Standards Load Program (NBSLP), and BLAST, which are “white box,” or
transparent to the user, and called public domain software. The second category consists of
those programs developed by the private sector, such as TRACE 600, developed by The Trane
Company; HAP E20-II, developed by Carrier Corporation; and HCC (loads) and ESP (energy),
developed by Automatic Procedures for Engineering Consultants Inc. (APEC). The privatesector-
developed software programs were based on the published literature of government,
research, and public institutions such as ASHRAE. The most widely used, reliable, user-friendly,
and continuously supported load and energy calculation programs in design are TRACE 600,
HAP E20-II, and DOE-2.1E. BLAST is the most elaborate load calculation program developed in
the United States and is usually considered a research tool. Most HVAC&R designers do care
about the accuracy of the computational results of the software; however, the priority is the userfriendly
inputs and outputs.
Trace 600—Structure and Basics
In this section we include the basics and inputs of the most widely used software program in load
calculations, TRACE 600 written for DOS, in the United States. Recently, Windows 95-based
TRACE 700 Load Design has become available. TRACE 600 Load and TRACE 700 Load Design
are similar in structure, basics, and engineering capabilities. Most of the load and energy calculation
software consists of four principal programs: loads, systems, plant, and economics (LSPE).
TRACE 600 divides into five phases:
6.42 CHAPTER SIX
 Load. In the load phase, all external and internal heat gains are calculated. The heat gain profile is
then converted to a cooling load profile.
 Design. In the design phase, based on the maximum block space sensible load, the volume flow
and the size of the air system are determined from the psychrometric analysis. In design phase it
also calculates the coil’s load and selects the size of the coils.
 System. In the system simulation, the program predicts the load that is imposed on the equipment
according to the space load profile and the type of air system selected.
 Equipment. In the equipment (plant) phase, the energy consumption of the fans, furnaces or boilers,
and refrigeration systems is determined based on the hourly coil loads.
 Economic. During the economic analysis, utility cost is calculated from the energy use. When it
is combined with installation cost and maintenance cost, a life cycle cost is available for
comparison.
Only the load calculation program is covered in this section. The available documentation for
TRACE 600 includes User’s Manual, Engineering Manual, Quick Reference, Cook Book, Software
Bulletin—Getting Started, etc.
The hardware requirements for TRACE 600 are as follows:
LOAD CALCULATIONS 6.43
IBM AT 286 processor or higher
Program: 7.5 Mbytes hard disk space (3 M
bytes for load calculation)
Job: 3 to 10 Mbytes depending on the size of
job
640 kbytes
Personal computer
Hard disk
Ram requirement
DOS Version 3.1 or higher
The TRACE 600 data sheet can be created by using the input editor. The input file is organized
in card format. In each input card, every input is shown by a given field. Sometimes, the value and
unit of input are separated into two fields. Entering the inputs can be done either by using field
mode or by using full-screen mode. In field mode, the user can access help and selection to determine
and select a specified input. Field mode input is preferable. In full-screen mode, the user can
view a full screen of raw input. Each line shows the inputs of a card and is prefixed by a twocharacter
code. Each field in a card is separated by a forward slash.
Input data are subdivided into five groups: job, load, system, equipment, and economic. Many
input cards and fields inside the input cards are optional, i.e., may be left blank.
Trace 600 Input—Load Methodology
In TRACE 600, card 10 lists various methods of cooling and heating load calculations for the user’s
selection. There are five cooling load calculating methods:
 CEC-DOE2. This method adopts the transfer function method for both heat gain and space cooling
load calculations. The space load calculations adopt the precalculated weighting factors listed
in DOE 2.1 Engineering Manual. This is a comparatively exact cooling load and energy calculation
method, especially for air systems operated at nighttime shutdown mode. CEC-DOE2 needs
more computational calculations. However, it is often not a primary problem when the computations
are performed by a powerful PC.
 CLTD-CLF. This method also adopts the transfer function method for both heat gain and space
cooling load calculations. Cooling load temperature difference (CLTD) and cooling load factor
(CLF) tables are prepared based on the exact transfer function coefficients or weighting factors
from the TFM method. For many commercial buildings, when the air system is shut down during
an unoccupied period, the external heat gains entering the space cannot be reasonably allocated
over the cool-down period and the successive operating hours.
 TETD-TA1 and TETD-TA2. The TETD-TA1 method adopts the transfer function method to calculate
heat gain, and TETD-TA2 uses an approximate TETD to calculate heat gain which adopts a
decrement factor and a time lag 
to describe the amplitude and time-delay characteristics of the
heat wave inside an exterior wall or a concrete roof slab. Both TETD-TA1 and TETD-TA2 use the
time-averaging (TA) method to convert the heat gain to space cooling load. The TA technique
lacks scientific support.
 TETD-PO. This method also adopts an approximate TETD to calculate the heat gain. It uses Post
Office RMRG weighting factors to convert a heat gain of 100 percent radiative to space cooling
load, and Post Office RMRX weighting factors to convert a heat gain which is not 100 percent
radiative, such as heat gains from people, lights, and equipment.
There are six heating load calculation methods. Five of them—CEC-DOE2, CLTD-CLF,
TETD-TA1, TETD-TA2, and TETD-PO—have already been described in the cooling load calculation
methodology. The sixth heating load calculation method is called the UATD method. In the
UATD method, heat losses are calculated based on the U value 	 area 	 temperature difference,
which is the temperature difference of the outdoor and indoor design temperatures. Heat losses are
also considered as instantaneous heating load.
For peak load calculation at winter design conditions, internal heat gains are not taken into
account in the UATD method. However, for energy use calculation, internal heat gains need to be
taken into account according to the load schedules. Otherwise the heating energy use becomes too
conservative.
Trace 600 Input—Job
There are 11 job cards: Cards 01 to 05 are used to describe the name of the project, its location, the
client of the project, the program user, and comments. Card 08, climatic information, lists the name
of weather file, summer and winter clearness numbers, outdoor dry- and wet-bulb temperatures,
ground reflectance, as well as the building orientation. Card 09, load simulation periods, covers the
first and last month of cooling design, summer period, and daylight savings time. It also covers the
peak cooling load hour.
In card 10, load simulation parameters (optional), the selection of cooling and heating load
methodology has been described in preceding paragraphs. In addition, the input data cover airflow
input and output units, percentage of wall load included in the return air, and room circulation rate
when the transfer function method is used in the conversion of heat gain to space cooling load. Outdoor
air dry-bulb temperature and humidity ratio are also required to determine the state of the supply
air during psychrometric analysis.
The sensible and latent loads due to the ventilation outdoor air intake are a component of cooling
coil load. They do not affect the supply airflow and the size of the supply fan if the mixture of
outdoor and recirculating air is extracted by the supply fan.
Card 11, energy simulation parameters (optional), describes the first and last months of energy
simulation, holiday and calendar type, and the conditioned floor area. It also determines the input
data of calculation level whether it is at room, zone, or air system level.
Card 12, resource utilization factors (optional), covers the input data of energy utilization
factors which indicate the inefficiency of producing and transfering energy, of electricity, gas,
oil, steam, hot water, chilled water, and coal. Card 13, daylighting parameters (optional),
describes the atmospheric moisture, atmospheric turbidity, or a measure of aerosols that
affects daylighting, inside visible reflectivity, and geometry method, such as glass percentage
(GLAS-PCT).
6.44 CHAPTER SIX
Trace 600 Input—External Loads
There are seven cards for external load input.
Card 19, load alternative description, will save separate files for each of the Load, Design, System,
and Equipment alternatives.
Card 20, general room parameters, determines the room number, zone reference number, room
description, floor length and width, construction type (a 2- to 12-in. (50- to 300-mm), light- or
heavyweight concrete construction), plenum height, R value of the acoustic ceiling, floor-to-
floor height, perimeter length, and duplicate floor and duplicate room/zone multipliers to save
input data for a similar floor or room/zone.
Card 21, thermostat parameters, lists input data of room design cooling and heating dry-bulb
temperatures and relative humidities; the highest and lowest room temperatures allowable to
drift up during low occupancy or nonoccupancy; cooling and heating thermostat schedules;
location of the thermostat; light, medium, or heavy room construction and the corresponding
2 to 8 h of time averaging; and carpet covering.
Card 22, roof parameters, describes the room number, alternate roof numbers, whether roof area
is equal to floor area, roof length and width, roof U value, roof construction type, roof direction,
roof tilting angle, and roof absorptivity .
Card 23, skylight parameters (optional), covers the room number, roof number, length and width of
skylight, number of skylights or percentage of glass of roof area, skylight U value, skylight shading
coefficient, external and internal shading types, percentage of solar load to be picked up by return
plenum air because a portion of skylight is exposed to the plenum air, visible light transmissivity,
and the fraction of inside visible light being reflected on the inside surface of the glazing.
Card 24, wall parameters, lists the input data of room number; wall number; the length, height,
U value, and construction type of the wall; wall direction; wall tilting angle; wall absorptivity;
and ground reflectance.
Card 25, wall glass parameters, describes the room number, wall number, glass length and height,
number of windows or percentage of glass (of gross wall area), glass U value, shading coefficient,
external and internal shading type, percentage of solar load to return plenum air, visible light
transmissivity, and the inside visible light being reflected on the inside surface of the glazing.
Trace 600 Input—Schedules
In TRACE 600, card 26 specifies the load and operating schedules of the internal loads, fans, reheating,
and daylighting (optional). The operating schedule of internal loads and equipment affects
the design load (maximum block load), especially the annual energy use.
1. For the design load calculation of internal loads include people, lighting, and miscellaneous
equipment:
 TRACE 600 assumes that the internal loads are scheduled at 100 percent, 24 h/day during
cooling design months operation and at 0 percent, 24 h/day during heating design months, a
schedule of cooling only (CLGONLY).
 If the pickup, cool-down, or warm-up load is greater than the design load, the user should
consider oversizing the cooling capacity and heating capacity to handle the pickup loads.
 For the energy use calulation of internal loads, choose or create a schedule at which the internal
load varies 24 h /day.
2. For the outdoor ventilation air and infiltration loads:
 TRACE 600 assumes that the design outdoor ventilation air and infiltration are scheduled at
100 percent, 24 h/day for both cooling and heating design months.
LOAD CALCULATIONS 6.45
 For the calculation of the energy use, only when the outdoor ventilation air system is separated
from the main supply system is it then possible to choose and to create a schedule in
which the airflow of outdoor ventilation air varies according to people’s occupancy.
 During the calculation of energy use for a variable-air-volume (VAV) system, the airflow of
the outdoor ventilation air may be varied based on its control systems into the following
schedule: 100 percent, 24 h/day; varying according to supply volume flow of the air system;
varying according to people’s occupancy.
 For the calculation of energy use due to infiltration, choose and create a schedule in which
the infiltration varies inversely with supply fan schedule, such as for hours fan schedule
is 100 percent, infiltration is zero; then when fan schedule is 0 percent, the infiltration is
100 percent.
3. Operating schedules of fans that affect the design load and energy use calculations are as
follows:
 At design load, TRACE 600 assumes that the main supply fan and auxiliary fan, if any, are
scheduled at 100 percent, 24 h/day as the default value during cooling mode operation, and
fan heat will be included in the cooling coil load. Similarly, at design load, main supply and
auxiliary fans are scheduled at 0 percent, 24 h/day during heating mode operation, and no
fan heat will be taken into account.
 In the energy use calculation, main supply and auxiliary fans in a constant-volume air system
are scheduled at 100 percent, 24 h/day. Or choose or create a schedule at which fans will be
cycling (on and off) depending on the space cooling and heating load.
 In the energy use calculation, main supply and auxiliary fans in a VAV system are scheduled
depending on the space cooling and heating load.
 For a room exhaust fan operated at cooling design load, it is scheduled at 100 percent,
24 h/day as the default value. For the energy use calculation, choose or create an appropriate
operating schedule.
4. The reheat minimum percentage determines the ratio of reheat minimum airflow for that room
to the design supply airflow.
5. Daylighting schedule describes the availability of daylighting each hour. Choose an available
schedule or create an appropriate daylighting schedule.
Trace 600 Input—Internal Loads
Four cards cover the input data of internal loads and airflows.
Card 27, people and lights, describes the room number, people’s density and corresponding unit,
people’s sensible and latent loads, lighting heat gain and corresponding unit, type of lighting
fixture, ballast factor, and daylighting reference points 1 and 2, to show whether there is
daylighting control or one or two control areas. According to the types of lighting fixtures and
the return airflow per unit floor area, the percentage of lighting load to return air should be
determined more precisely. Refer to Sec. 6.6 for details.
Card 28, miscellaneous equipment (optional), covers the room number, miscellaneous equipment
reference number, equipment description, energy consumption value and corresponding
units, schedule and energy metering code, percentage of sensible load, percentage of miscellaneous
equipment load to space and to air path, radiant fraction, and whether the plenum air,
exhaust air, or return air path picks up the miscellaneous load.
Card 29, room airflows, covers the flow rates of outdoor ventilation air during cooling and heating
operation and their corresponding units, the flow rates of infiltration air for cooling and
heating and their corresponding units, and the amount of minimum reheat and its units.
Card 30, fan flow rates, describes the room number, airflow of the main and auxiliary cooling and
heating supply fans and their corresponding units, and airflow of room exhaust fan and its unit.
6.46 CHAPTER SIX
Trace 600 Input—Partition and Shading Devices
Five cards cover the input data of partition and shading devices.
Card 31, partition parameters (optional), describes the room number, adjacent room number,
partition number, and the length, height, U value, and construction type of the partition wall.
The adjacent space temperature during cooling and heating periods can follow these profiles:
constant, sine fit, prorate the adjacent space temperature against the outdoor air temperature,
vary as the outdoor air temperature, or there is no heat transfer from the adjacent space through
the partition.
Card 32, exposed floor parameters (optional), lists the room number, adjacent room number, exposed
floor number, the perimeter length and loss coefficient of the slab on grade; the area, U
value, and construction type of the exposed floor; and adjacent space temperature during cooling
and heating periods whether it is constant, sine curve fit, or prorated curve fit.
Card 33, external shading devices (optional), describes its constructional type, glass height and
glass width, the height above glass and the horizontal projection (projection out) of the overhang,
the projection left and right, the left and right projection out, and the shading due to adjacent
buildings. TRACE 600 does not allow for shading because of both adjacent building and
overhang.
Card 34, internal shading devices, covers the types, overall U value, overall shading coefficient,
location, operating schedule, and overall visible light transmittance of internal shading
device; requires that below the minimum outdoor air dry-bulb temperature and above the
maximum solar heat gain, the internal shading device is deployed; and sun control and glare
control probability.
Card 35, daylight sensor (optional), describes the daylight sensor reference number, percentage
of space affected, lighting set point in footcandles (luxes), type of control (continuous or
stepped), minimum power or minimum light percentage, light control steps, manually operated
control probability, height of the reference point and its distance from the glass, window sill
height, ratio of the glass length seen by a sensor to the distance between the reference point and
the glass, and skylight length and distance from the reference point.
Trace 600—Minimum Input Requirements, Run, and Outputs
For load calculation and energy analysis, the following are the minimum input requirements:
Card 08: climatic information, field 2: weather filename.
Card 20: general room parameters, field 1: room number.
Card 40: system type, field 1: system number, and field 2: system type.
Card 41: zone assignment, field 1: system number, field 2: system serving these zones begins at,
and field 3 : end at.
After the user has completed all the input required for the job, he or she must edit for errors,
check for possible data loss, and run and save. Choose output from the TRACE 600 menu. Select
the desired section of the output. Print the compressed report in a desired order.
Example 6.2. The plan of a typical floor in a multistory office building in New York City is
shown in Fig. 6.9. The tenants and the partition walls of this floor are unknown during the design
stage. The core part contains the restrooms, stairwells, and mechanical and electrical service
rooms, and it has mechanical exhaust systems only. This building has the following construction
characteristics:
LOAD CALCULATIONS 6.47
This typical floor has the following internal loads: The occupant density is 7 persons per 1000
ft2 (93 m2). For each person, there is 250 Btu/h (73 W) sensible heat gain, and 250 Btu/h (73 W)
latent heat gain. The lighting load, including ballast allowance, is 1 W/ ft2 (10.7 W/m2). The miscellaneous
equipment load due to personal computers and appliances is 1 W/ ft2 (10.7 W/m2).
6.48 CHAPTER SIX
FIGURE 6.9 Typical plan of a high-rise office building.
6300 ft2 (586 m2)
5600 ft2 (520 m2)
13 ft (4 m)
9 ft (2.74 m)
0.345
0.67 Btu/h ft2 °F (3.8 W/m2 °C)
0.36
120 ft (36.6 m)
13 ft (2.74 m)
0.12 Btu/ hft2 °F (0.68 W/ (m2 °C)
200 ft (61 m)
9 ft (2.74 m)
0.12 Btu/h ft2 °F (0.68 W/m2 °C)
Area of perimeter zone
Area of interior zone
Floor-to-floor height
Floor-to-ceiling height
Window:
Percentage of glass/exterior wall
U value
Interior shading, SC
Exterior wall:
Length
Height
U value
Partition wall:
Length
Height
U value
The summer indoor design temperature is 75°F (23.9°C). If the infiltration load and the shading
due to the adjacent building are ignored, calculate the maximum space cooling load during summer
outdoor design conditions of this typical floor by using TRACE 600 Load program with the CLTDCLD
method.
Solution. The maximum space cooling loads for this typical floor, calculated by using TRACE
600 Load, with CLTD-CLF method on July 7, 5 P.M., are as follows:
Manually calculated in
TRACE 600, Btu/h Example 7.2 of first edition, Btu/h
Envelope loads
Glass solar 36,846
Glass conduction 11,095
Wall conduction 9,927
Partition wall 1,296
Ceiling load 4,226
63,390
Internal loads
Lights 36,147
People 31,446
Miscellaneous 37,994
105,587
Grand total 168,977 198,770
Compare the TRACE 600 calculated space cooling load of 168,977 Btu/h (49,520 W) with the
manually calculated space cooling load of the same typical floor with the same building and operating
characteristics (in Example 7.2 of Handbook of Air Conditioning and Refrigeration, 1st ed.) of
198,770 Btu/h (58,240 W). The TRACE 600 calculated space cooling load is only 85 percent of the
manually calculated value. This is due to a lower solar CLF and therefore lower solar cooling loads
in the TRACE 600 calculation.
REFERENCES
Amistadi, H., Energy Analysis Software Review, Engineered Systems, no. 10, 1993, pp. 34–45.
ASHRAE, ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, ASHRAE Inc., Atlanta, GA, 1989.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
Ayres, J. M., and Stamper, E., Historical Development of Building Energy Calculations, ASHRAE Journal,
no. 2, 1995, pp. 47–53.
Carrier Air Conditioning Co., Handbook of Air Conditioning System Design, 1st ed., McGraw-Hill, New York, 1965.
Deringer, J. J., An Overview of Standard 90.1: Building Envelope, ASHRAE Journal, February 1990, pp. 30–34.
Harris, S. M., and McQuiston, F. C., A Study to Categorize Walls and Roofs on the Basis of Thermal Response,
ASHRAE Transactions, 1988, Part II, pp. 688–715.
Johnson, C. A., Besent, R.W., and Schoenau, G. J., An Economic Parametric Analysis of the Thermal Design of
a Large Office Building under Different Climatic Zones and Different Billing Schedules, ASHRAE Transactions,
1989, Part I, pp. 355–369.
Kerrisk, J. F., Schnurr, N. M., Moore, J. E., and Hunn, B. D., The Custom Weighting-Factor Method for Thermal
Load Calculations in the DOE-2 Computer Program, ASHRAE Transactions, 1981, Part II, pp. 569–584.
Kimura, K. I., and Stephenson, D. G., Theoretical Study of Cooling Loads Caused by Lights, ASHRAE Transactions,
1968, Part II, pp. 189–197.
LOAD CALCULATIONS 6.49
Komor, P., Space Cooling Demands from Office Plug Loads, ASHRAE Journal, no. 12, 1997, pp. 41–44.
Kreith, F., and Black,W. Z., Basic Heat Transfer, Harper & Row, New York, 1980.
Mackey, C. O., and Gay, N. R., Cooling Load from Sunlit Glass, ASHVE Transactions, 1952, pp. 321–330.
Mackey, C. O., and Wright, L. T., Periodic Heat Flow—Homogeneous Walls or Roofs, ASHVE Transactions,
1944, pp. 293–312.
McQuiston, F. C., and Spitler, J. D., Cooling and Heating Load Calculation Manual, 2d ed., ASHRAE Inc.,
Atlanta, GA, 1992.
Mitalas, G. P., Transfer Function Method of Calculating Cooling Loads, Heat Extraction Rate and Space Temperature,
ASHRAE Journal, no. 12, 1972, p. 52.
Palmatier, E. P., Thermal Characteristics of Structures, ASHRAE Transactions, 1964, pp. 44–53.
Persily, A. K., and Norford, L. K., Simultaneous Measurements of Infiltration and Intake in an Office Building,
ASHRAE Transactions, 1987, Part II, pp. 42–56.
Romine, T. B., Cooling Load Calculation: Art or Science? ASHRAE Journal, no. 1, 1992, pp. 14–24.
Rudoy,W., and Duran, F., Development of an Improved Cooling Load Calculation Method, ASHRAE Transactions,
1975, Part II, pp. 19–69.
Rudoy,W., and Robins, L. M., Pulldown Load Calculations and Thermal Storage during Temperature Drift,
ASHRAE Transactions, 1977, Part I, pp. 51–63.
Snelling, H. J., Duration Study for Heating and Air Conditioning Design Temperature, ASHRAE Transactions,
1985, Part II B, p. 242.
Sowell, E. F., Classification of 200,640 Parametric Zones for Cooling Load Calculations, ASHRAE
Transactions, 1988, Part II, pp. 754–777.
Sowell, E. F., and Chiles, D. C., Zone Descriptions and Response Characterization for CLF/CLTD Calculations,
ASHRAE Transactions, 1985, Part II A, pp. 179–200.
Sowell, E. F., and Hittle, D. C., Evolution of Building Energy Simulation Methodology, ASHRAE Transactions,
1995, Part I, pp. 851–855.
Stephenson, D. G., and Mitalas, G. P., Cooling Load Calculations by Thermal Response Factor Method,
ASHRAE Transactions, 1967, Part III, pp. 1.1–1.7.
Sun, T. Y., Air Conditioning Load Calculation, Heating/Piping/Air Conditioning, January 1986, pp. 103–113.
The Trane Company, TRACE 600: Engineering Manual, The Trane Company, LaCrosse, WI, 1992.
The Trane Company, TRACE 600: User’s Manual, LaCrosse, WI, 1992.
Trehan, A. K., Fortmann, R. C., Koontz, M. D., and Nagda, N. L., Effect of Furnace Size on Morning Pickup
Time, ASHRAE Transactions, 1989, Part I, pp. 1125–1129.
Wang, S. K., Air Conditioning, vol. 1, Hong Kong Polytechnic, Hong Kong, 1987.
Williams, G. J., Fan Heat: Its Source and Significance, Heating/Piping/Air Conditioning, January 1989,
pp. 101–112.
6.50 CHAPTER SIX
CHAPTER 7
WATER SYSTEMS
7.1
7.1 FUNDAMENTALS 7.2
Types of Water System 7.2
Volume Flow and Temperature
Difference 7.4
Water Velocity and Pressure Drop 7.5
7.2 WATER PIPING 7.7
Piping Material 7.7
Piping Dimensions 7.7
Pipe Joints 7.13
Working Pressure and Temperature 7.13
Expansion and Contraction 7.14
Piping Supports 7.15
Piping Insulation 7.15
7.3 VALVES, PIPE FITTINGS, AND
ACCESSORIES 7.16
Types of Valves 7.16
Valve Connections and Ratings 7.17
Valve Materials 7.18
Piping Fittings and Water System
Accessories 7.18
7.4 WATER SYSTEM PRESSURIZATION
AND THE PRESENCE OF AIR 7.19
Water System Pressurization
Control 7.19
Open Expansion Tank 7.20
Closed Expansion Tank 7.21
Size of Diaphragm Expansion Tank 7.21
Pump Location 7.23
Air in Water Systems 7.23
Penalties due to Presence
of Air and Gas 7.24
Oxidation and Waterlogging 7.24
7.5 CORROSION AND DEPOSITS IN
WATER SYSTEM 7.25
Corrosion 7.25
Water Impurities 7.25
Water Treatments 7.26
7.6 CLOSED WATER SYSTEM
CHARACTERISTICS 7.27
System Characteristics 7.27
Changeover 7.28
7.7 CENTRIFUGAL PUMPS 7.30
Basic Terminology 7.30
Performance Curves 7.32
Net Positive Suction Head 7.33
Pump Selection 7.33
7.8 PUMP-PIPING SYSTEMS 7.34
System Curve 7.34
System Operating Point 7.34
Combination of Pump-Piping
Systems 7.35
Modulation of Pump-Piping
Systems 7.36
Pump Laws 7.37
Wire-to-Water Efficiency 7.37
7.9 OPERATING CHARACTERISTICS OF
CHILLED WATER SYSTEM 7.38
Coil Load and Chilled Water
Volume Flow 7.38
Chiller Plant 7.39
Variable Flow for Saving Energy 7.40
Water Systems in Commercial
Buildings 7.40
7.10 PLANT-THROUGH-BUILDING
LOOP 7.40
Bypass Throttling Flow 7.40
Distributed Pumping 7.41
Variable Flow 7.41
7.11 PLANT-BUILDING LOOP 7.43
System Description 7.43
Control Systems 7.43
System Characteristics 7.45
Sequence of Operations 7.46
Low T between
Chilled Water Supply and Return Temperatures
7.49
Variable-Speed Pumps Connected in
Parallel 7.49
Use of Balancing Valves 7.49
Common Pipe and Thermal
Contamination 7.51
7.12 PLANT-DISTRIBUTED PUMPING 7.52
7.13 CAMPUS-TYPE WATER
SYSTEMS 7.53
Plant-Distribution-Building Loop 7.54
Plant-Distributed Building Loop 7.56
Multiple Sources-Distributed
Building Loop 7.57
Chilled and Hot Water Distribution
Pipes 7.58
7.14 COMPUTER-AIDED PIPING DESIGN
AND DRAFTING 7.58
General Information 7.58
Computer-Aided Drafting
Capabilities 7.58
Computer-Aided Design
Capabilities 7.59
REFERENCES 7.60
7.1 FUNDAMENTALS
Types of Water System
Water systems that are part of an air conditioning system and that link the central plant,
chiller / boiler, air-handling units (AHUs), and terminals may be classified into the following categories
according to their use:
Chilled Water System. In a chilled water system, water is first cooled in the water chiller—the evaporator
of a reciprocating, screw, or centrifugal refrigeration system located in a centralized plant—to
a temperature of 40 to 50°F (4.4 to 10.0°C). It is then pumped to the water cooling coils in AHUs and
terminals in which air is cooled and dehumidified. After flowing through the coils, the chilled water
increases in temperature up to 60 to 65°F (15.6 to 18.3°C) and then returns to the chiller.
Chilled water is widely used as a cooling medium in central hydronic air conditioning systems.
When the operating temperature is below 38°F (3.3°C), inhibited glycols, such as ethylene glycol or
propylene glycol, may be added to water to create an aqueous solution with a lower freezing point.
Evaporative-Cooled Water System. In arid southwestern parts of the United States, evaporativecooled
water is often produced by an evaporative cooler to cool the air.
Hot Water Systems. These systems use hot water at temperatures between 450 and 150°F (232
and 66°C) for space and process heating purposes. Hot water systems are covered in greater detail
in Chap. 8.
Dual-Temperature Water System. In a dual-temperature water system, chilled water or hot water
is supplied to the coils in AHUs and terminals and is returned to the water chiller or boiler mainly
through the following two distribution systems:
 Use supply and return main and branch pipes separately.
 Use the common supply and return mains, branch pipe, and coil for hot and chilled water supply
and return.
The changeover from chilled water to hot water and vice versa in a building or a system depends
mainly on the space requirements and the temperature of outdoor air. Hot water is often produced
by a boiler; sometimes it comes from a heat recovery system, which is discussed in later chapters.
Condenser Water System. In a condenser water or cooling water system, the latent heat of condensation
is removed from the refrigerant in the condenser by the condenser water. This condenser
water either is from the cooling tower or is surface water taken from a lake, river, sea, or well. For
an absorption refrigeration system, heat is also removed from the solution by cooling water in the
absorber. The temperature of the condenser water depends mainly on the local climate.
Water systems also can be classified according to their operating characteristics into the following
categories:
Closed System. In a closed system, chilled or hot water flowing through the coils, heaters,
chillers, boilers, or other heat exchangers forms a closed recirculating loop, as shown in Fig. 7.1a.
In a closed system, water is not exposed to the atmosphere during its flowing process. The purpose
of recirculation is to save water and energy.
Open System. In an open system, the water is exposed to the atmosphere, as shown in Fig. 7.1b.
For example, chilled water comes directly into contact with the cooled and dehumidified air in the
air washer, and condenser water is exposed to atmosphere air in the cooling tower. Recirculation of
water is used to save water and energy.
7.2 CHAPTER SEVEN
7.3
FIGURE 7.1 Types of water systems. (a) Closed system; (b) open system; (c) once-through system.
Open systems need more water treatments than closed systems because dust and impurities in
the air may be transmitted to the water in open systems. A greater quantity of makeup water is required
in open systems to compensate for evaporation, drift carryover, or blow-down operation.
Once-Through System. In a once-through system, water flows through the heat exchanger only
once and does not recirculate, as shown in Fig. 7.1c. Lake, river, well, or seawater used as condenser
cooling water represents a once-through system. Although the water cannot recirculate to the
condenser because of its rise in temperature after absorbing the heat of condensation, it can still be
used for other purposes, such as flushing water in a plumbing system after the necessary water
treatments, to conserve water. In many locations, the law requires that well water be pumped back
into the ground.
Volume Flow and Temperature Difference
The heating and cooling capacity of water when it flows through a heat exchanger Qw, Btu /h (W),
can be calculated as
(7.1)
where volume flow of water, ft3 /h (m3/s)
volume flow rate of water, gpm (L/ s)
density of water, lb/ ft3 (kg/m3)
cpw  specific heat of water, Btu / lb °F (J /kg°C)
Twe ,Twl  temperature of water entering and leaving heat exchanger,°F (°C)
temperature drop or rise of water after flowing through heat exchanger,°F (°C)
Here, the equivalent of pounds per hour is gpm  60 min/h  0.1337 ft3/gal  62.32 lb/ ft3  500.
Equation 7.1 also gives the relationship between Tw and Vgal during the heat-transfer process.
The temperature of water leaving the water chiller should be no lower than 37°F (2.8°C) to
prevent freezing. If the chilled water temperature is lower than 37°F (2.8°C), brine, ethylene glycol,
or propylene glycol should be used. Brine is discussed in a later chapter. For a dual-temperature
water system, the hot water temperature leaving the boiler often ranges from 100 to 150°F (37.8 to
65.6°C), and returns at a between 20 and 40°F (11.1 and 22.2°C).
For most dual-temperature water systems, the value of and the pipe size are determined
based on the cooling capacity requirement for the coils and water coolers. This is because chilled
water has a smaller than hot water does. Furthermore, the system cooling load is often higher
than the system heating load. For a chilled water system to transport each refrigeration ton of cooling
capacity, a of 8°F (4.4°C) requires a of 3 gpm (0.19 L/ s), whereas for a of 24°F
(13.3°C), is only 1.0 gpm (0.063 L/ s).
The temperature of water entering the coil Twe, the temperature of water leaving the coil Twl, and
the difference between them are closely related to the performance of a chilled
water system, air system, and refrigeration system:
 Temperature Twe directly affects the power consumption in the compressor.
 The temperature differential is closely related to the volume flow of chilled water and
thus the size of the water pipes and pumping power.
 Both Twe and influence the temperature and humidity ratio of air leaving the coil.
If the chilled water temperature leaving the water chiller and entering the coil is between 44 and
45°F (6.7 and 7.2°C), the off-coil temperature in the air system is usually around 55°F (12.8°C) for
conventional comfort air conditioning systems. In low-temperature cold air distribution systems,
Tw
V?
gal Tw
Tw  Twl  Twe
V?
gal
Tw V?gal Tw
Tw
V?
gal
T

Tw 
	w 
V?
gal 
V?
w 
 500V?gal T
 500V?gal (Twe  Twl)
Qw  V?w 	wcpw (Twe  Twl)
7.4 CHAPTER SEVEN
WATER SYSTEMS 7.5
chilled water leaving the chiller may be as low as 34°F (1.1°C), and the off-coil temperature is
often between 42 and 47°F, typically 44°F (5.6 and 8.3°C, typically 6.7°C).
The greater the value of for chilled water, the lower the amount of water flowing through
the coil. Current practice is usually to use a value of between 10 and 18°F (5.6 and 10.0°C) for
chilled water systems in buildings. Kelly and Chan (1999) noted a greater results at lower total
power consumption in water pumps, cooling tower fans, and chillers. However, a greater
means a larger coil and air-side pressure drop. For chilled water systems in a campus-type central
plant, a value of between 16 and 24°F (8.9 and 13.3°C) is often used.
Water Velocity and Pressure Drop
The maximum water velocity in pipes is governed mainly by pipe erosion, noise, and water hammer.
Erosion of water pipes is the result of the impingement of rapidly moving water containing air
bubbles and impurities on the inner surface of the pipes and fittings. Solden and Siegel (1964)
increased their feedwater velocity gradually from 8 ft/ s (2.4 m/ s) to an average of 35.6 ft / s (10.8
m/ s). After 3 years, they found no evidence of erosion in the pipe or a connected check valve. Erosion
occurs only if solid matter is contained in water flowing at high velocity. Velocity-dependent
noise in pipes results from flow turbulence, cavitation, release of entrained air, and water hammer
that results from the transient pressure impact on a sudden closed valve. Ball and Webster (1976)
performed a series of tests on -in. copper tubes with elbows. At a water velocity of 16.4 ft / s (5.0
m/ s), the noise level was less than 53 dBA. Tests also showed that cold water at a speed up to 21
ft / s (6.4 m/ s) did not cause cavitation. In copper and steel pipes, water hammer at a water velocity
of 15 ft / s (4.6 m/ s) exerted a pressure on 2-in.- (50-mm-) diameter pipes that was less than 50 percent
of their design pressure.
Given the above results, excluding the energy cost for the pump power, the maximum water velocity
in certain short sections of a water system may be raised to an upper limit of 11 ft/s (3.35 m/s)
for a special purpose, such as enhancing the heat-transfer coefficients.
Normally, water flow in coils and heat exchangers becomes laminar and seriously impairs the
heat-transfer characteristics only when its velocity drops to a value less than 2 ft / s (0.61 m/s) and
its corresponding Reynolds number is reduced to about 10,000 (within the transition region). In
evaporators, Redden (1996) found that at low tube water velocity at 1.15 ft / s (0.35 m/s) and low
heat flux, instability of heat transfer occurred and caused the chilled water leaving temperature to
fluctuate by 4°F (2.2°C). In condensers, condensing operation is not affected even if the condenser
water velocity in the tube was about 1 ft / s (0.31 m/ s). Water velocity should also be maintained at
not less than 2 ft / s (0.61 m/ s) in order to transport the entrained air to air vents.
When pipes are being sized, the optimum pressure drop , commonly expressed in feet
(meters) of head loss of water per 100 ft of pipe length (p in pascals of pressure drop per meter
length), is a compromise between energy costs and investments. At the same time, the agecorrosion
of pipes should be considered.
Generally, the pressure drop for water pipes inside buildings is in a range of 1 ft / 100 ft to
4 ft /100 ft (100 to 400 Pa/m), with a mean of 2.5 ft /100 ft (250 Pa/m) used most often. Because of
a lower increase in installation cost for smaller-diameter pipes, it may be best to use a pressure
drop lower than 2.5 ft /100 ft (250 Pa/m) when the pipe diameter is 2 in. or less.
Age corrosion results in an increase in the friction factor and a decrease in the effective diameter.
The factors that contribute to age corrosion are sliming, caking of calcareous salts, and corrosion.
Many scientists recommended an increase in friction loss of 15 to 20 percent, resulting in a design
pressure drop of 2 ft/100 ft (200 Pa/m), for closed water systems; and a 75 to 90 percent increase in
friction loss, or a design pressure drop of 1.35 ft/100 ft (135 Pa/m), for open water systems.
Figures 7.2, 7.3, and 7.4 show the pressure drop charts for steel, copper, and plastic pipes, respectively,
for closed water systems. Each chart shows the volume flow (gpm), pressure drop
(ft / 100 ft), water velocity vw (ft /s), and water pipe diameter D (in.). Given any two of these
parameters, the other two can be determined. For instance, for a steel water pipe that has a water
volume flow of 1000 gpm, if the pressure drop is 2 ft / 100 ft, the diameter is 8 in. and the corresponding
velocity is about 8 ft / s.
Hf
V?
gal
Hf
Hf
3
8
Tw
Tw
Tw
Tw
Tw
It is a common practice to limit the water velocity to no more than 4 ft/ s (1.2 m/s) for water
pipes 2 in. (50 mm) or less in diameter in order to prevent an excessive Hf. The pressure drop
should not exceed 4 ft /100 ft (400 Pa/m) for water pipes of greater than 2-in. (50-mm) diameter.
An open water system or a closed water system that is connected with an open expansion tank,
all the pressure differences between two points or levels, and pressure drops across a piece of
equipment or a device are expressed in feet of water column or psi (head in meters of water column
or pressure loss in kPa). The total or static pressure of water at a certain point in a water system is
actually measured and expressed by that part of pressure which is greater or smaller than the atmospheric
pressure, often called gauge pressure, in feet of water column gauge or psig (meters gauge or
kPa  g). The relationships between the steady flow energy equation and the fluid pressure, and between
the pressure loss and fluid head, are discussed in Sec.17.1.
7.6 CHAPTER SEVEN
FIGURE 7.2 Friction chart for water in steel pipes (Schedule 40). (Source: ASHRAE Handbook 1989 Fundamentals. Reprinted with
permission.)
FIGURE 7.3 Friction chart for water in copper tubing (types K, L, and M). (Source: ASHRAE Handbook 1989 Fundamentals.
Reprinted with permission.)
7.2 WATER PIPING
Piping Material
For water systems, the piping materials most widely used are steel, both black (plain) and galvanized
(zinc-coated), in the form of either welded-seam steel pipe or seamless steel pipe; ductile iron
and cast iron; hard copper; and polyvinyl chloride (PVC). The piping materials for various services
are shown below:
WATER SYSTEMS 7.7
FIGURE 7.4 Friction chart for water in plastic pipes (Schedule 80). (Source: ASHRAE Handbook 1989 Fundamentals. Reprinted with
permission.)
Black and galvanized steel
Black steel, hard copper
Black steel, galvanized ductile iron, PVC
Chilled water
Hot water
Cooling water and drains
Copper, galvanized steel, galvanized ductile iron, and PVC pipes have better corrosion resistance
than black steel pipes. Technical requirements, as well as local customs, determine the selection of
piping materials.
Piping Dimensions
The steel pipe wall thicknesses currently used were standardized in 1930. The thickness ranges
from Schedule 10, light wall, to Schedule 160, very heavy wall. Schedule 40 is the standard for a
pipe with a diameter up to 10 in. (250 mm). For instance, a 2-in. (50-mm) standard pipe has an outside
diameter of 2.375 in. (60.33 mm) and an inside diameter of 2.067 in. (52.50 mm). The nominal
pipe size is only an approximate indication of pipe size, especially for pipes of small diameter.
Table 7.1 lists the dimensions of commonly used steel pipes.
The outside diameter of extruded copper is standardized so that the outside diameter of the copper
tubing is 1/8 in. (3.2 mm) larger than the nominal size used for soldered or brazed socket joints.
As in the case with steel pipes, the result is that the inside diameters of copper tubes seldom equal
the nominal sizes. Types K, L, M, and DWV designate the wall thickness of copper tubes: type K is
the heaviest, and DWV is the lightest. Type L is generally used as the standard for pressure copper
tubing. Type DWV is used for drainage at atmospheric pressure.
7.8
TABLE 7.1 Dimensions of Commonly Used Steel Pipes
Working pressure†
Nominal Surface area Cross-sectional Weight of ASTM A538B to 400°F
size and Schedule Wall Inside Metal Flow
pipe OD number thickness diameter Outside, Inside, area, area, Pipe, Water, Mfr. Joint
D, in. or weight* t, in. d, in. ft2/ft ft2/ft in.2 in.2 lb/ft lb/ft process type psig
40 ST 0.088 0.364 0.141 0.095 0.125 0.104 0.424 0.045 CW Thrd 188
D  0.540 80 XS 0.119 0.302 0.141 0.079 0.157 0.072 0.535 0.031 CW Thrd 871
40 ST 0.091 0.493 0.177 0.129 0.167 0.191 0.567 0.083 CW Thrd 203
D  0.675 80 XS 0.126 0.423 0.177 0.111 0.217 0.141 0.738 0.061 CW Thrd 820
40 ST 0.109 0.622 0.220 0.163 0.250 0.304 0.850 0.131 CW Thrd 214
D  0.840 80 XS 0.147 0.546 0.220 0.143 0.320 0.234 1.087 0.101 CW Thrd 753
40 ST 0.113 0.824 0.275 0.216 0.333 0.533 1.13 0.231 CW Thrd 217
D  1.050 80 XS 0.154 0.742 0.275 0.194 0.433 0.432 1.47 0.187 CW Thrd 681
1 40 ST 0.133 1.049 0.344 0.275 0.494 0.864 1.68 0.374 CW Thrd 226
D  1.315 80 XS 0.179 0.957 0.344 0.251 0.639 0.719 2.17 0.311 CW Thrd 642
1 40 ST 0.140 1.380 0.435 0.361 0.669 1.50 2.27 0.647 CW Thrd 229
D  1.660 80 XS 0.191 1.278 0.435 0.335 0.881 1.28 2.99 0.555 CW Thrd 594
1 40 ST 0.145 1.610 0.497 0.421 0.799 2.04 2.72 0.881 CW Thrd 231
D  1.900 80 XS 0.200 1.500 0.497 0.393 1.068 1.77 3.63 0.765 CW Thrd 576
2 40 ST 0.154 2.067 0.622 0.541 1.07 3.36 3.65 1.45 CW Thrd 230
D  2.375 80 XS 0.218 1.939 0.622 0.508 1.48 2.95 5.02 1.28 CW Thrd 551
2 40 ST 0.203 2.469 0.753 0.646 1.70 4.79 5.79 2.07 CW Weld 533
D  2.875 80 XS 0.276 2.323 0.753 0.608 2.25 4.24 7.66 1.83 CW Weld 835
3 40 ST 0.216 3.068 0.916 0.803 2.23 7.39 7.57 3.20 CW Weld 482
D  3.500 80 XS 0.300 2.900 0.916 0.759 3.02 6.60 10.25 2.86 CW Weld 767
4 40 ST 0.237 4.026 1.178 1.054 3.17 12.73 10.78 5.51 CW Weld 430
D  4.500 80 XS 0.337 3.826 1.178 1.002 4.41 11.50 14.97 4.98 CW Weld 695
6 40 ST 0.280 6.065 1.734 1.588 5.58 28.89 18.96 12.50 ERW Weld 696
D  6.625 80 XS 0.432 5.761 1.734 1.508 8.40 26.07 28.55 11.28 ERW Weld 1209
1
2
1
2
1
4
3
4
1
2
3
8
1
4
7.9
8 30 0.277 8.071 2.258 2.113 7.26 51.16 24.68 22.14 ERW Weld 526
D  8.625 40 ST 0.322 7.981 2.258 2.089 8.40 50.03 28.53 21.65 ERW Weld 643
80 XS 0.500 7.625 2.258 1.996 12.76 45.66 43.35 19.76 ERW Weld 1106
30 0.307 10.136 2.814 2.654 10.07 80.69 34.21 34.92 ERW Weld 485
10 40 ST 0.365 10.020 2.814 2.623 11.91 78.85 40.45 34.12 ERW Weld 606
D  10.75 XS 0.500 9.750 2.814 2.552 16.10 74.66 54.69 32.31 ERW Weld 887
80 0.593 9.564 2.814 2.504 18.92 71.84 64.28 31.09 ERW Weld 1081
30 0.330 12.090 3.338 3.165 12.88 114.8 43.74 49.68 ERW Weld 449
12 ST 0.375 12.000 3.338 3.141 14.58 113.1 49.52 48.94 ERW Weld 528
D  12.75 40 0.406 11.938 3.338 3.125 15.74 111.9 53.48 48.44 ERW Weld 583
XS 0.500 11.750 3.338 3.076 19.24 108.4 65.37 46.92 ERW Weld 748
80 0.687 11.376 3.338 2.978 26.03 101.6 88.44 43.98 ERW Weld 1076
30 ST 0.375 13.250 3.665 3.469 16.05 137.9 54.53 59.67 ERW Weld 481
14 40 0.437 13.126 3.665 3.436 18.62 135.3 63.25 58.56 ERW Weld 580
D  14.00 XS 0.500 13.000 3.665 3.403 21.21 132.7 72.04 57.44 ERW Weld 681
80 0.750 12.500 3.665 3.272 31.22 122.7 106.05 53.11 ERW Weld 1081
16 30 ST 0.375 15.250 4.189 3.992 18.41 182.6 62.53 79.04 ERW Weld 421
D  16.00 40 XS 0.500 15.000 4.189 3.927 24.35 176.7 82.71 76.47 ERW Weld 596
ST 0.375 17.250 4.712 4.516 20.76 233.7 70.54 101.13 ERW Weld 374
18 30 0.437 17.126 4.712 4.483 24.11 230.3 81.91 99.68 ERW Weld 451
D  18.00 XS 0.500 17.000 4.712 4.450 27.49 227.0 93.38 98.22 ERW Weld 530
40 0.562 16.876 4.712 4.418 30.79 223.7 104.59 96.80 ERW Weld 607
20 ST 0.375 19.250 5.236 5.039 23.12 291.0 78.54 125.94 ERW Weld 337
D  20.00 30 XS 0.500 19.000 5.236 4.974 30.63 283.5 104.05 122.69 ERW Weld 477
40 0.593 18.814 5.236 4.925 36.15 278.0 122.82 120.30 ERW Weld 581
*Numbers are schedule number per ASTM B36.10; ST  standard weight; XS  extra strong.
†Working pressures have been calculated per ASME/ANSI B31.9 using furnace butt weld (continuous weld, CW) pipe through 4 in. and electric resistance weld (ERW)
thereafter. The allowance A has been taken as (a) 12.5 percent of t for mill tolerance on pipe wall thickness, plus (b) an arbitrary corrosion allowance of 0.025 in. for pipe sizes
through NPS 2 and 0.065 in. from NPS 2 through 20 plus (c) a thread cutting allowance for sizes through NPS 2.
Because the pipe wall thickness of threaded standard weight pipe is so small after deducting the allowance A, the mechanical strength of the pipe is impaired. It is good practice
to limit standard-weight threaded pipe pressures to 90 psig for steam and 125 psig for water.
Source: ASHRAE Handbook 1988, Equipment. Reprinted with permission.
1
2
TABLE 7.1 Dimensions of Commonly Used Steel Pipes (Continued)
Working pressure†
Nominal Surface area Cross-sectional Weight of ASTM A538B to 400°F
size and Schedule Wall Inside Metal Flow
pipe OD number thickness diameter Outside, Inside, area, area, Pipe, Water, Mfr. Joint
D, in. or weight* t, in. d, in. ft2/ft ft2/ft in.2 in.2 lb/ft lb/ft process type psig
7.10
TABLE 7.2 Dimensions of Copper Tubes
Working pressure*
Surface area Cross-sectional Weight of ASTMB 888 to 250°F
Nominal Wall Outside Inside
diameter, thickness diameter diameter Outside, Inside, Metal Flow Tube, Water, Annealed, Drawn,
in. Type t, in. D, in. d, in. ft2/ft ft2/ft area, in.2 area, in.2 lb/ft lb/ft psig psig
K 0.035 0.375 0.305 0.098 0.080 0.037 0.073 0.145 0.032 851 1596
L 0.030 0.375 0.315 0.098 0.082 0.033 0.078 0.126 0.034 730 1368
K 0.049 0.500 0.402 0.131 0.105 0.069 0.127 0.269 0.055 894 1676
L 0.035 0.500 0.430 0.131 0.113 0.051 0.145 0.198 0.063 638 1197
M 0.025 0.500 0.450 0.131 0.008 0.037 0.159 0.145 0.069 456 855
K 0.049 0.625 0.527 0.164 0.138 0.089 0.218 0.344 0.094 715 1341
L 0.040 0.625 0.545 0.164 0.143 0.074 0.233 0.285 0.101 584 1094
M 0.028 0.625 0.569 0.164 0.149 0.053 0.254 0.203 0.110 409 766
K 0.049 0.750 0.652 0.196 0.171 0.108 0.334 0.418 0.144 596 1117
L 0.042 0.750 0.666 0.196 0.174 0.093 0.348 0.362 0.151 511 958
K 0.065 0.875 0.745 0.229 0.195 0.165 0.436 0.641 0.189 677 1270
L 0.045 0.875 0.785 0.229 0.206 0.117 0.484 0.455 0.209 469 879
M 0.032 0.875 0.811 0.229 0.212 0.085 0.517 0.328 0.224 334 625
1 K 0.065 1.125 0.995 0.295 0.260 0.216 0.778 0.839 0.336 527 988
L 0.050 1.125 1.025 0.295 0.268 0.169 0.825 0.654 0.357 405 760
M 0.035 1.125 1.055 0.295 0.276 0.120 0.874 0.464 0.378 284 532
1 K 0.065 1.375 1.245 0.360 0.326 0.268 1.217 1.037 0.527 431 808
L 0.055 1.375 1.265 0.360 0.331 0.228 1.257 0.884 0.544 365 684
M 0.042 1.375 1.291 0.360 0.338 0.176 1.309 0.682 0.566 279 522
DWV 0.040 1.375 1.295 0.360 0.339 0.168 1.317 0.650 0.570 265 497
1 K 0.072 1.625 1.481 0.425 0.388 0.351 1.723 1.361 0.745 404 758
L 0.060 1.625 1.505 0.425 0.394 0.295 1.779 1.143 0.770 337 631
M 0.049 1.625 1.527 0.425 0.400 0.243 1.831 0.940 0.792 275 516
DWV 0.042 1.625 1.541 0.425 0.403 0.209 1.865 0.809 0.807 236 442
2 K 0.083 2.125 1.959 0.556 0.513 0.532 3.014 2.063 1.304 356 668
L 0.070 2.125 1.985 0.556 0.520 0.452 3.095 1.751 1.339 300 573
M 0.058 2.125 2.009 0.556 0.526 0.377 3.170 1.459 1.372 249 467
DWV 0.042 2.125 2.041 0.556 0.534 0.275 3.272 1.065 1.416 180 338
1
2
1
4
3
4
5
8
1
2
3
8
1
4
7.11
2 K 0.095 2.625 2.435 0.687 0.637 0.755 4.657 2.926 2.015 330 619
L 0.080 2.625 2.465 0.687 0.645 0.640 4.772 2.479 2.065 278 521
M 0.065 2.625 2.495 0.687 0.653 0.523 4.889 2.026 2.116 226 423
3 K 0.109 3.125 2.907 0.818 0.761 1.033 6.637 4.002 2.872 318 596
L 0.090 3.125 2.945 0.818 0.771 0.858 6.812 3.325 2.947 263 492
M 0.072 3.125 2.981 0.818 0.780 0.691 6.979 2.676 3.020 210 394
DWV 0.045 3.125 3.035 0.818 0.795 0.435 7.234 1.687 3.130 131 246
3 K 0.120 3.625 3.385 0.949 0.886 1.321 8.999 5.120 3.894 302 566
L 0.100 3.625 3.425 0.949 0.897 1.107 9.213 4.291 3.987 252 472
M 0.083 3.625 3.459 0.949 0.906 0.924 9.397 3.579 4.066 209 392
4 K 0.134 4.125 3.857 1.080 1.010 1.680 11.684 6.510 5.056 296 555
L 0.110 4.125 3.905 1.080 1.022 1.387 11.977 5.377 5.182 243 456
M 0.095 4.125 3.935 1.080 1.030 1.203 12.161 4.661 5.262 210 394
DWV 0.058 4.125 4.009 1.080 1.050 0.741 12.623 2.872 5.462 128 240
5 K 0.160 5.125 4.805 1.342 1.258 2.496 18.133 9.671 7.846 285 534
L 0.125 5.125 4.875 1.342 1.276 1.963 18.665 7.609 8.077 222 417
M 0.109 5.125 4.907 1.342 1.285 1.718 18.911 6.656 8.183 194 364
DWV 0.072 5.125 4.981 1.342 1.304 1.143 19.486 4.429 8.432 128 240
6 K 0.192 6.125 5.741 1.603 1.503 3.579 25.886 13.867 11.201 286 536
L 0.140 6.125 5.845 1.603 1.530 2.632 26.832 10.200 11.610 208 391
M 0.122 6.125 5.881 1.603 1.540 2.301 27.164 8.916 11.754 182 341
DWV 0.083 6.125 5.959 1.603 1.560 1.575 27.889 6.105 12.068 124 232
8 K 0.271 8.125 7.583 2.127 1.985 6.687 45.162 25.911 19.542 304 570
L 0.200 8.125 7.725 2.127 2.022 4.979 46.869 19.295 20.280 224 421
M 0.170 8.125 7.785 2.127 2.038 4.249 47.600 16.463 20.597 191 358
DWV 0.109 8.125 7.907 2.127 2.070 2.745 49.104 10.637 21.247 122 229
10 K 0.338 10.125 9.449 2.651 2.474 10.392 70.123 40.271 30.342 304 571
L 0.250 10.125 9.625 2.651 2.520 7.756 72.760 30.054 31.483 225 422
M 0.212 10.125 9.701 2.651 2.540 6.602 73.913 25.584 31.982 191 358
12 K 0.405 12.125 11.315 3.174 2.962 14.912 100.554 57.784 43.510 305 571
L 0.280 12.125 11.565 3.174 3.028 10.419 105.046 40.375 45.454 211 395
M 0.254 12.125 11.617 3.174 3.041 9.473 105.993 36.706 45.863 191 358
*When using soldered or brazed fittings, the joint determines the limiting pressure.
Working pressures calculated using ASME B31.9 allowable stresses. A 5 percent mill tolerance has been used on the wall thickness. Higher tube ratings can be calculated using allowable
stress for lower temperatures.
If soldered or brazed fittings are used on hard-drawn tubing, use the annealed ratings. Full-tube allowable pressures can be used with suitably rated flare or compression-type fittings.
Source: ASHRAE Handbook 1988, Equipment. Reprinted with permission.
1
2
1
2
TABLE 7.2 Dimensions of Copper Tubes (Continued)
Working pressure*
Surface area Cross-sectional Weight of ASTMB 888 to 250°F
Nominal Wall Outside Inside
diameter, thickness diameter diameter Outside, Inside, Metal Flow Tube, Water, Annealed, Drawn,
in. Type t, in. D, in. d, in. ft2/ft ft2/ft area, in.2 area, in.2 lb/ft lb/ft psig psig
Copper tubes are also categorized as hard and soft copper. Soft pipes should be used in applications
for which the pipe will be bent in the field. Table 7.2 lists the dimensions of copper tubes.
Thermoplastic plastic pipes are the most widely used plastic pipes in air conditioning. They are
manufactured with dimensions that match steel pipe dimensions. The advantages of plastic pipes include
resistance to corrosion, scaling, and the growth of algae and fungi. Plastic pipes have smooth
surfaces and negligible age allowance. Age allowance is the allowance for corrosion and scaling for
plastic pipes during their service life. Most plastic pipes are low in cost, especially compared with
corrosion-resistant metal tubes.
The disadvantages of plastic pipes include the fact that their pressure ratings decrease rapidly
when the water temperature rises above 100°F (37.8°C). PVC pipes are weaker than metal pipes
and must usually be thicker than steel pipes if the same working pressure is to be maintained. Plastic
pipes may experience expansion and contraction during temperature changes that is 4 times
greater than that of steel. Precautions must be taken to protect plastic pipes from external damage
and to account for its behavior during fire. Some local codes do not permit the use of some or allplastic
pipes. It is necessary to check with local authorities.
Pipe Joints
Steel pipes of small diameter (2 in. or 50 mm less) threaded through cast-iron fittings are the most
widely used type of pipe joint. For steel pipes of diameter 2 in. (50 mm) and more, welded joints,
7.12 CHAPTER SEVEN
TABLE 7.3 Maximum Allowable Pressures at Corresponding Temperatures
System
Maximum
allowable
Fitting
pressure at
Temperature, temperature,
Application Pipe material Weight Joint type Class Material °F psig
Recirculating water
2 in. and smaller Steel (CW) Standard Thread 125 Cast iron 250 125
Copper, hard Type L 95-5 solder — Wrought copper 250 150
PVC Sch. 80 Solvent Sch. 80 PVC 75 350
CPVC Sch. 80 Solvent Sch. 80 CPVC 150 150
PB SDR-11 Heat fusion — PB 160 115
Insert crimp — Metal 160 115
2.5–12 in. A53 B ERW Standard Weld Standard Wrought steel 250 400
steel Flange 150 Wrought steel 250 250
Flange 125 Cast iron 250 175
Flange 250 Cast iron 250 400
Groove — MI or ductile iron 230 300
PB SDR-11 Heat fusion PB 160 115
Refrigerant
Copper, hard Type L or K Braze — Wrought copper — —
A53 B SML
steel Standard Weld Wrought steel — —
Note: Maximum allowable working pressures have been derated in this table. Higher system pressures can be used for lower temperatures and
smaller pipe sizes. Pipe, fittings, joints, and valves must all be considered.
Note: A53 ASTM Standard A53
PVC Polyvinyl chloride
CPVC Chlorinated polyvinyl chloride
PB Polybutylene
Source: Abridged with permission from ASHRAE Handbook 1988, Equipment.
bolted flanges, and grooved ductile iron joined fittings are often used. Galvanized steel pipes are
threaded together by galvanized cast iron or ductile iron fittings.
Copper pipes are usually joined by soldering and brazing socket end fittings. Plastic pipes are
often joined by solvent welding, fusion welding, screw joints, or bolted flanges.
Vibrations from pumps, chillers, or cooling towers can be isolated or dampened by means of
flexible pipe couplings. Arch connectors are usually constructed of nylon, dacron, or polyester and
neoprene. They can accommodate deflections or dampen vibrations in all directions. Restraining
rods and plates are required to prevent excessive stretching. A flexible metal hose connector
includes a corrugated inner core with a braided cover. It is available with flanged or grooved
end joints.
Working Pressure and Temperature
In a water system, the maximum allowable working pressure and temperature are not limited to the
pipes only; joints or the pipe fittings, especially valves, may often be the weak links. Table 7.3 lists
types of pipes, joint, and fittings and their maximum allowable working pressures for specified
temperatures.
Expansion and Contraction
During temperature changes, all pipes expand and contract. The design of water pipes must take
into consideration this expansion and contraction. Both the temperature change during the operating
period and the possible temperature change between the operating and shutdown periods should
WATER SYSTEMS 7.13
FIGURE 7.5 Expansion loops. (a) U bends; (b) L bends; (c) Z bends.
also be considered. For chilled and condenser water, which has a possible temperature change of 40
to 100°F (4.4 to 37.8°C), expansion and contraction cause considerable movement in a long run of
piping. Unexpected expansion and contraction cause excess stress and possible failure of the pipe,
pipe support, pipe joints, and fittings.
Expansion and contraction of hot and chilled water pipes can be better accommodated by using
loops and bends. The commonly used bends are U bends, Z bends, and L bends, as shown in
Fig. 7.5. Anchors are the points where the pipe is fixed so that it will expand or contract between
them. Reaction forces at these anchors should be considered when the support is being designed.
ASHRAE Handbook 1992, HVAC Systems and Equipment, gives the required calculations and data
for determining these stresses. Guides are used so that the pipes expand laterally.
Empirical formulas are often used instead of detailed stress analyses to determine the dimension
of the offset leg Lo [ft (m)]. Waller (1990) recommended the following formulas:
U bends: Lo  0.041D0.48Lac
0.46 T
Z bends: Lo  (0.13DLacT)0.5 (7.2)
L bends: Lo  (0.314DLacT)0.5
where D  diameter of pipe, in. (mm)
Lac  distance between anchors, hundreds of ft (m)
T  temperature difference, °F (°C)
If there is no room to accommodate U, Z, or L bends (such as in high-rise buildings or tunnels),
mechanical expansion joints are used to compensate for movement during expansion. Packed
expansion joints allow the pipe to slide to accommodate movement during expansion. Various types
of packing are used to seal the sliding surfaces in order to prevent leakage. Another type of mechanical
joint uses bellows or flexible metal hose to accommodate movement. These types of joints
should be carefully installed to avoid distortion.
7.14 CHAPTER SEVEN
TABLE 7.4 Recommended Pipe Hanger Spacing, ft
Standard-weight
Nominal pipe steel pipe Copper Rod size,
diameter, in. (water) tube (water) in.
7 5
7 5
1 7 6
1 9 8
2 10 8
2 11 9
3 12 10
4 14 12
6 17 14
8 19 16
10 20 18
12 23 19
14 25 1
16 27 1
18 28 1
20 30 1
Note: Spacing does not apply where concentrated loads are placed between
supports such as flanges, valves, and specialties.
Source: ASHRAE Handbook 1988, Equipment. Reprinted with permission.
1
4
1
4
7
8
3
4
5
8
1
2
1
2
3
8
3
8
1
2
3
8
3
8
1
2
1
4
1
4
3
4
1
4
1
2
Piping Supports
Types of piping support include hangers, which hang the pipe from above; supports, which usually
use brackets to support the pipe from below; anchors to control the movement of the piping; and
guides to guide the axial movement of the piping. Table 7.4 lists the recommended spacing of pipe
hangers. Piping support members should be constructed based on the stress at their point of connection
to the pipe as well as on the characteristics of the structural system. Pipe supports must have
sufficient strength to support the pipe, including the water inside. Except for the anchors, they
should also allow for expansion movement.
Pipes should be supported around the connections to the equipment so that the pipe’s weight and
expansion or contraction do not affect the equipment. For insulated pipes, heavy-gauge sheet-metal
half-sleeves are used between the hangers and the insulation. Corrosion protection should also be
carefully considered.
Piping Insulation
External pipe insulation should be provided when the inside water temperature 105°F  Tw  60°F
(41°C  Tw  15.6°C) for the sake of energy saving, surface condensation, and high-temperature
safety protection. The optimum thickness of the insulation of pipes depends mainly on the operating
temperature of the inside water, the pipe diameter, and the types of service. There is a compromise between
initial cost and energy cost. ASHRAE/IESNA Standard 90.1-1999 specifies the minimum pipe
insulation thickness for water systems, as listed in Table 7.5. Insulation shall be protected from damage
including that because of sunlight, moisture, equipment maintenance, and wind.
WATER SYSTEMS 7.15
TABLE 7.5 Minimum Pipe Insulation Thickness*, in.
Fluid design
Insulation conductivity
Nominal pipe or tube size, in.
operating temperature Conductivity, Mean rating
range, °F Btuin./h  ft2°F temp. °F <1 1 to <1 1 to <4 4 to <8 ?8
Heating systems (steam, steam condensate, and hot water)†‡
>350 0.32–0.34 250 2.5 3.0 3.0 4.0 4.0
251–350 0.29–0.32 200 1.5 2.5 3.0 3.0 3.0
201–250 0.27–0.30 150 1.5 1.5 2.0 2.0 2.0
141–200 0.25–0.29 125 1.0 1.0 1.0 1.5 1.5
105–140 0.22–0.28 100 0.5 0.5 1.0 1.0 1.0
Domestic and service hot water systems
105+ 0.22–0.28 100 0.5 0.5 1.0 1.0 1.0
Cooling systems (chilled water, brine, and refrigerant)§
40–60 0.22–0.28 100 0.5 0.5 1.0 1.0 1.0
<40 0.22–0.28 100 0.5 1.0 1.0 1.0 1.5
*For insulation outside the stated conductivity range, the minimum thickness T shall be determined as follows:
where T  minimum insulation thickness (in.), r  actual outside radius of pipe (in.), t  insulation thickness listed in this table
for applicable fluid temperature and pipe size, K  conductivity of alternate material at mean rating temperature indicated for
the applicable fluid temperature (Btu in.[h ft2°F]); and k  the upper value of the conductivity range listed in this table for the
applicable fluid temperature.
†These thicknesses are based on energy efficiency considerations only. Additional insulation is sometimes required relative
to safety issues/surface temperature.
‡Piping insulation is not required between the control valve and coil on run-outs when the control valve is located within 4
ft of the coil and the pipe size is 1 in. or less.
§These thicknesses are based on energy efficiency considerations only. Issues such as water vapor permeability or surface
condensation sometimes require vapor retarders or additional insulation.
Source: ASHRAE/IESNA Standard 90.1-1999. Reprinted with permission.
T  r[(1 	 t/r)K/k  1]
1
2
1
2
Insulation exposed to weather shall be suitable for outdoor service, such as, protected by aluminum,
sheet metal, painted canvas, or plastic cover. Cellular foam insulation shall be protected as
above or with a painted coating which itself is a water retardant and also provides shielding from
solar radiation.
Insulation of chilled water piping or refrigerant suction piping shall include an exterior vapor retardant
covering the insulation (unless the insulation is inherently vapor retardant). All penetrations
and joints of the vapor retardant shall be sealed.
7.3 VALVES, PIPE FITTINGS, AND ACCESSORIES
Types of Valve
Valves are used to regulate or stop the water flow in pipes either manually or by means of automatic
control systems. Valves used in automatic control systems are called control valves, discussed in
Chap. 5. In this section, only manually operated valves, or simply valves, are discussed.
Hand-operated valves are used to stop or isolate flow, to regulate flow, to prevent reverse flow,
and to regulate water pressure. The basic construction of a valve consists of the following (see
Fig. 7.6): a disk to open or close the water flow; a valve body to seat the disk and provide the
flow passage; a stem to lift or rotate the disk, with a handwheel or a handle and corresponding
mechanism to make the task easier; and a bonnet to enclose the valve from the top.
Based on the shape of the valve disk, the valve body, or its function, commonly used valves can
be classified into the following types:
Gate Valves. The disk of a gate valve is in the shape of a “gate” or wedge, as shown in Fig. 7.6a.
When the wedge is raised at the open position, a gate valve does not add much flow resistance. The
wedge can be either a solid wedge, which is most commonly used, or a split wedge, in which two
disk halves being forced outward fit tightly against the body seat. Gate valves are used either fully
opened or closed, an on/off arrangement. They are often used as isolating valves for pieces of
equipment or key components, such as control valves, for service during maintenance and repair.
Globe Valves. They are so named because of the globular shape of the valve body, as shown in
Fig. 7.6b. Globe valves have a round disk or plug-type disk seated against a round port. Water flow
enters under the disk. Globe valves have high flow resistances. They can be opened or closed
7.16 CHAPTER SEVEN
FIGURE 7.6 Types of valves. (a) Gate valve; (b) Globe valve; (c) Check valve, swing check.
substantially faster than gate valves. Angle valves are similar to globe valves in their seats and
operation. The basic difference is that the valve body of an angle valve can also be used as a 90°
elbow at that location.
Globe valves are used to throttle and to regulate the flow. They are sometimes called balancing
valves. They are deliberately designed to restrict fluid flow, so they should not be used in applications
for which full and unobstructed flow is often required.
Check Valves. Check valves, as their name suggests, are valves used to prevent, or check, reverse
flow. There are basically two types of check valves: swing check and lift check. A swing check
valve has a hinged disk, as shown in Fig. 7.6c. When the water flow reverses, water pressure pushes
the disk and closes the valve. In a lift check valve, upward regular flow raises the disk and opens the
valve, and reverse flow pushes the disk down to its seat and stops the backflow. A swing check
valve has a lower flow resistance than a lift check valve.
Plug Valves. These valves use a tapered, cylindrical plug disk to fit the seat. They vary from fully
open to fully closed positions within a quarter-turn. Plug valves may be used for throttling control
during the balancing of a water system.
Ball Valves. These valves use a ball as the valve disk to open or close the valve. As with plug
valves, they vary from fully open to fully closed positions within a quarter-turn. As with gate valves,
ball valves are usually used for open/ close service. They are less expensive than gate valves.
Butterfly Valves. A butterfly valve has a thin rotating disk. Like a ball or plug valve, it varies
within a quarter-turn from fully open to fully closed. As described in Sec. 5.6, a butterfly valve
exhibits low flow resistance when it is fully opened. The difference between a butterfly valve
used for control purposes and a hand-operated butterfly valve is that the former has an actuator
and can be operated automatically. Butterfly valves are lightweight, easy to operate and install,
and lower in cost than gate valves. They are primarily used as fully open or fully closed, but they
may be used for throttling purposes. Butterfly valves are gaining in popularity, especially in large
pipes.
Balance Valves. These valves are used to balance the water flow in a water system. There are two
kinds of balancing valves: manual balance valves and automatic balance valves. A globe valve can
be used as a manual balance valve. A manual balance valve can also be a valve with integral pressure
taps for flow measurement and a calibrated port to adjust the flow. An automatic balancing
valve is also called an automatic flow-limiting valve. There is a moving element that adjusts the
flow passage area according to the water pressure differential across the valve.
Pressure Relief Valves. These valves are safety valves used to prevent a system that is overpressurized
from exceeding a predetermined limit. A pressure relief valve is held closed by a spring or
rupture member and is automatically opened to relieve the water pressure when it rises above the
system design working pressure.
Valve Connections and Ratings
The type of connection used between a valve and the pipes is usually consistent with the type of
joint used in the pipe system. A water piping system with flanged joint requires a valve with flanged
ends. The commonly used types of valve connection are as follows:
 Threaded ends. These connections are mainly used for small pipes with diameters from to 2 in.
(6 to 50 mm). Threaded-end valves are usually inexpensive and simpler to install.
1
4
WATER SYSTEMS 7.17
 Flanged ends. These connections are commonly used for larger pipes ( in. or 63 mm and
above). Flanged ends are more easily separated when necessary.
 Welded ends. Steel valves, when used at higher pressure and temperature, are often connected
with welded ends. Welded ends exhibit the fewest instances of leakage.
 Grooved ends. These connections use circumferential grooves in which a rubber gasket fits and
are enclosed by iron couplings. Butterfly valves are often connected with grooved ends.
 Soldered ends. Bronze valves in copper piping systems use soldered ends. Tin-alloy soldering is
the type of soldering commonly used. Lead soldering cannot be used in a potable water system
because it will contaminate the water.
Valves are usually rated according to their ability to withstand pressure at a specified temperature.
Metal valves have two different ratings, one for steam [working steam pressure (WSP)], which
should correspond to its operating temperature, and the other for cold water, oil, or gas (WOG). The
following are the commonly used ratings:
 125 psig (862 kPa g) WSP, 250 psig (1724 kPag) WOG
 150 psig (1034 kPa g) WSP, 300 psig (2068 kPag) WOG
 300 psig (2069 kPa g) WSP, 600 psig (4138 kPag) WOG
Here psig represents gauge pressure in pounds per square inch (kPag gauge pressure in kPa). As
listed in Table 7.3, a wrought steel valve flange joint with a 150 psig (1034 kPag) rating can be used
for a hot or chilled water system with a maximum allowable pressure of 250 psig (1724 kPa g) at a
temperature below 250°F (121°C), for pipes of diameter between 2.5 and 12 in. (63 and 300 mm).
Valve Materials
Valve materials are selected according to their ability to withstand working pressure and temperature,
their resistance to corrosion, and their relative cost. The most commonly used materials for
valves are as follows:
 Bronze. It has a good corrosive resistance and is easily machined, cast, or forged. Bronze is
widely used for water valves up to a size of 3 in. (75 mm) because of its high cost. For valves
above 3 in. (75 mm), bronze is still often used for sealing elements and stems because it is
machinable and corrosion-resistant.
 Cast iron and ductile iron. These materials are used for pressure-containing parts, flanges, and
glands in valves 2 in. (50 mm) and larger. Ductile iron has a higher tensile strength than cast iron.
 Steel. Forged or cast steel provides a higher tensile strength as well as toughness in the form of resistance
to shock and vibration than do bronze, cast iron, and ductile iron. Steel is used in applications
that require higher strength and toughness than bronze and ductile iron can provide.
 Trim materials. These include the elements and components that are easily worn as well as those
parts that need to be resistant to corrosion, such as the disk, seating elements, and stem. Stainless
steel, stellite (a kind of cobalt-chromium-tungsten alloy), and chromium-molybdenum steel are
often used for trim material in valves.
Pipe Fittings and Water System Accessories
Water pipe fittings include elbows, tees, and valves. Water pipe elbows and tees are often made of
cast iron, ductile iron, or steel. Pressure losses due to the water pipe fittings are usually expressed in
terms of an equivalent length of straight pipe, for the sake of convenience. Table 7.6 allows calculation
of the pressure losses for various types of piping fittings, in terms of an equivalent length of
21
2
7.18 CHAPTER SEVEN
straight pipe. The equivalent length for a fitting can be estimated by multiplying the elbow equivalent
to that fitting by the equivalent length for a 90° elbow.
Water system accessories include drains, strainers, and air vents. Drains should be equipped at
all low points of the system. Arrangements should be made so that a part of the system or individual
components can be drained rather than draining the entire system. A condensate drain pipe is
always required for cooling and dehumidifying coils. Galvanized steel is often used for this
purpose. It is usually piped to a plumbing drain or other suitable location. A condensate drain pipe
should be insulated so as to avoid surface condensation.
Water strainers are often installed before the pumps, control valves, or other components to protect
them from dirt and impurities. Air vents are discussed in the next section.
7.4 WATER SYSTEM PRESSURIZATION AND
THE PRESENCE OF AIR
Water System Pressurization Control
For an open water system, the maximum operating gauge pressure is the pressure at a specific point
in the system where the positive pressure exerted by the water pumps, to overcome the pressure
WATER SYSTEMS 7.19
TABLE 7.6 Pressure Losses for Pipe Fittings and Valves, Expressed in Terms of an Equivalent Length (in ft) of Straight Pipe
Equivalent length, ft of pipe, for 90° elbows
Pipe size, in.
Velocity, ft/s 1 1 1 2 2 3 3 4 5 6 8 10 12
1 1.2 1.7 2.2 3.0 3.5 4.5 5.4 6.7 7.7 8.6 10.5 12.2 15.4 18.7 22.2
2 1.4 1.9 2.5 3.3 3.9 5.1 6.0 7.5 8.6 9.5 11.7 13.7 17.3 20.8 24.8
3 1.5 2.0 2.7 3.6 4.2 5.4 6.4 8.0 9.2 10.2 12.5 14.6 18.4 22.3 26.5
4 1.5 2.1 2.8 3.7 4.4 5.6 6.7 8.3 9.6 10.6 13.1 15.2 19.2 23.2 27.6
5 1.6 2.2 2.9 3.9 4.5 5.9 7.0 8.7 10.0 11.1 13.6 15.8 19.8 24.2 28.8
6 1.7 2.3 3.0 4.0 4.7 6.0 7.2 8.9 10.3 11.4 14.0 16.3 20.5 24.9 29.6
7 1.7 2.3 3.0 4.1 4.8 6.2 7.4 9.1 10.5 11.7 14.3 16.7 21.0 25.5 30.3
8 1.7 2.4 3.1 4.2 4.9 6.3 7.5 9.3 10.8 11.9 14.6 17.1 21.5 26.1 31.0
9 1.8 2.4 3.2 4.3 5.0 6.4 7.7 9.5 11.0 12.2 14.9 17.4 21.9 26.6 31.6
10 1.8 2.5 3.2 4.3 5.1 6.5 7.8 9.7 11.2 12.4 15.2 17.7 22.2 27.0 32.0
Iron and copper elbow equivalents
Source: ASHRAE Handbook 1997, Fundamentals. Reprinted with permission.
Fitting Iron pipe Copper tubing
Elbow, 90° 1.0 1.0
Elbow, 45° 0.7 0.7
Elbow, 90° long turn 0.5 0.5
Elbow, 90° welded 0.5 0.5
Reduced coupling 0.4 0.4
Open return bend 1.0 1.0
Angle radiator valve 2.0 3.0
Radiator or convector 3.0 4.0
Boiler or heater 3.0 4.0
Open gate valve 0.5 0.7
Open globe valve 12.0 17.0
1
2
1
2
1
2
1
4
3
4
1
2
drops across the equipment, components, fittings, and pipes plus the static head due to the vertical
distance between the highest water level and that point, is at a maximum.
In a closed chilled or hot water system, a variation in the water temperature will cause an expansion
of water that may raise the water pressure above the maximum allowable pressure. The purposes
of system pressurization control for a closed water system are as follows:
 To limit the pressure of the water system to below its allowable working pressure
 To maintain a pressure higher than the minimum water pressure required to vent air
 To assist in providing a pressure higher than the net positive suction head (NPSH) at the pump
suction to prevent cavitation
 To provide a point of known pressure in the system
Expansion tanks, pressure relief valves, pressure-reducing valves for makeup water, and corresponding
controls are used to achieve water system pressurization control. There are two types of
expansion tanks for closed water systems: open and closed.
Open Expansion Tank
An expansion tank is a device that allows for the expansion and contraction of water contained in a
closed water system when the water temperature changes between two predetermined limits. Another
function of an expansion tank is to provide a point of known pressure in a water system.
An open expansion tank is vented to the atmosphere and is located at least 3 ft (0.91 m) above
the highest point of the water system, as shown in Fig. 7.7. Makeup water is supplied through a
7.20 CHAPTER SEVEN
FIGURE 7.7 Open expansion tank.
float valve, and an internal overflow drain is always installed. A float valve is a globe or ball valve
connected with a float ball to regulate the makeup water flow according to the liquid level in the
tank. An open expansion tank is often connected to the suction side of the water pump to prevent
the water pressure in the system from dropping below the atmospheric pressure. The pressure of the
liquid level in the open tank is equal to the atmospheric pressure, which thus provides a reference
point of known pressure to determine the water pressure at any point in the water system. The minimum
tank volume should be at least 6 percent of the volume of water in the system Vs, ft3 (m3). An
open expansion tank is simple, more stable in terms of system pressure characteristics, and low in
cost. If it is installed indoors, it often needs a high ceiling. If it is installed outdoors, water must be
prevented from freezing in the tank, air vent, or pipes connected to the tank when the outdoor temperature
is below 32°F (0°C). Because the water surface in the tank is exposed to the atmosphere,
oxygen is more easily absorbed into the water, which makes the tank less resistant to corrosion than
a diaphragm tank (to be described later). Because of these disadvantages, an open expansion tank
has only limited applications.
Closed Expansion Tank
A closed expansion tank is an airtight tank filled with air or other gases, as shown in Fig. 7.8. When
the temperature of the water increases, the water volume expands. Excess water then enters the
tank. The air in the tank is compressed, which raises the system pressure. When the water temperature
drops, the water volume contracts, resulting in a reduction of the system pressure.
To reduce the amount of air dissolved in the water so as to prevent corrosion and prevent air
noise, a diaphragm, or a bladder, is often installed in the closed expansion tank to separate the filled
air and the water permanently. Such an expansion tank is called a diaphragm, or bladder, expansion
tank. Thus, a closed expansion tank is either a plain closed expansion tank, which does not have a
diaphragm to separate air and water, or a diaphragm tank.
For a water system with only one air-filled space, the junction between the closed expansion
tank and the water system is a point of fixed pressure. At this point, water pressure remains
constant whether or not the pump is operating because the filled air pressure depends on only the
volume of water in the system. The pressure at this point can be determined according to the ideal
gas law, as given by Eq. (2.1): pv  RTR. The pressure in a closed expansion tank during the initial
filling process or at the minimum operating pressure is called the fill pressure pfil, psia. The fill
pressure is often used as the reference pressure to determine the pressure characteristics of a water
system.
Size of Diaphragm Expansion Tank
If a closed expansion tank with its filled volume of air is too small, the system pressure will easily
exceed the maximum allowable pressure and cause water to discharge from the pressure relief
valve, thus wasting water. If the closed tank is too large, when the water temperature drops, the system
pressure may decrease to a level below the minimum allowable value and cause trouble in the
air vent. Therefore, accurate sizing of a closed expansion tank is essential.
For diaphragm expansion tanks, the minimum volume of the water tank, Vt, gal (m3), can be calculated
by the following formula, recommended by ASHRAE Handbook 1996, HVAC Systems and
Equipment:
(7.3)
where T1  lower temperature, °F (°C)
T2  higher temperature, °F (°C)
Vs  volume of water in system, gal (m3 )
p1  absolute pressure at lower temperature, psia (kPa abs.)
Vt  Vs  v2 / v1  1  3(T2  T1)
1  p1 / p2 
WATER SYSTEMS 7.21
7.22 CHAPTER SEVEN
FIGURE 7.8 Closed expansion tank for a water system. (a) Diaphragm expansion tank in a chilled water system.
(b) Diaphragm expansion tank in a hot water system. (c) Plain closed expansion tank.
p2  absolute pressure at higher temperature, psia (kPa abs.)
v1, v2  specific volume of water at lower and higher temperature, respectively, ft3 /lb (m3 /kg)
 linear coefficient of thermal expansion; for steel,  6.5  106 in. / in °F (1.2 
105 per °C); for copper,  9.5  106 in. / in. °F (1.7  105 per °C)
In a chilled water system, the higher temperature T2 is the highest anticipated ambient temperature
when the chilled water system shuts down during summer. The lower temperature in a heating
system is often the ambient temperature at fill conditions (for example, 50°F or 10°C).
Pump Location
The location of the pump in a water system that uses a diaphragm expansion tank should be
arranged so that the pressure at any point in the water system is greater than the atmospheric pressure.
In such an arrangement, air does not leak into the system, and the required net positive suction
head (NPSH) can be maintained at the suction inlet of the water pump. NPSH is discussed in
detail in Sec. 7.7.
A water pump location commonly used for hot water systems with diaphragm expansion tanks
is just after the expansion tank and the boiler, as shown in Fig. 7.8b. In this arrangement, the pressure
at the pump suction is the sum of the water pressure and the fill pressure. In another often-used
arrangement, the diaphragm expansion tank is moved to the highest point of the water system, and
the pump is still located after the boiler. In a chilled water system, the location of the chilled water
pump is usually before the water chiller, and the diaphragm expansion tank is usually connected to
the suction side of the water pump.
Air in Water Systems
In a closed recirculated water system, air and nitrogen are present in the following forms: dissolved
in water, free air or gas bubbles, or pockets of air or gas. The behavior of air or gas dissolved in liquids
is governed and described by Henry’s equation. Henry’s equation states that the amount of gas
dissolved in a liquid at constant temperature is directly proportional to the absolute pressure of that
gas acting on the liquid, or
(7.4)
where x  amount of dissolved gas in solution, percent by volume
p  partial pressure of that gas, psia
H  Henry’s constant; changes with temperature
The lower the water temperature and the higher the total pressure of the water and dissolved gas,
the greater the maximum amount of dissolved gas at that pressure and temperature.
When the dissolved air or gas in water reaches its maximum amount at that pressure and temperature,
the water becomes saturated. Any excess air or gas, as well as the coexisting water vapor,
can exist only in the form of free bubbles or air pockets. A water velocity greater than 1.5 ft / s
(0.45 m/ s) can carry air bubbles along with water. When water is in contact with air at an air-water
interface, such as the filled airspace in a plain closed expansion tank, the concentration gradient
causes air to diffuse into the water until the water is saturated at that pressure and temperature. An
equilibrium forms between air and water within a certain time. At specific conditions, 24 h may be
required to reach equilibrium.
The oxygen in air that is dissolved in water is unstable. It reacts with steel pipes to form oxides
and corrosion. Therefore, after air has been dissolved in water for a long enough time, only nitrogen
remains as a dissolved gas circulating with the water.
x 
p
H
WATER SYSTEMS 7.23
Penalties due to Presence of Air and Gas
The presence of air and gas in a water system causes the following penalties for a closed water system
with a plain closed expansion tank:
 Presence of air in the terminal and heat exchanger, which reduces the heat-transfer surface
 Corrosion due to the oxygen reacting with the pipes
 Waterlogging in plain closed expansion tanks
 Unstable system pressure
 Poor pump performance due to gas bubbles
 Noise problems
There are two sources of air and gas in a water system. One is the air-water interface in a plain
closed expansion tank or in an open expansion tank, and the other is the dissolved air in a city water
supply.
Oxidation and Waterlogging
Consider a chilled water system that uses a plain closed expansion tank without a diaphragm, as
shown in Fig. 7.8. This expansion tank is located in a basement, with a water pressure of 90 psig
(620 kPa  g) and a temperature of 60°F (15.6°C) at point A. At such a temperature and pressure, the
solubility of air in water is about 14.2 percent. The chilled water flows through the water pump, the
chiller, and the riser and is supplied to the upper-level terminals. During this transport process, part
of the oxygen dissolved in the water reacts with the steel pipes to form oxides and corrosion. At
upper-level point B, the water pressure is only 10 psig (69 kPa g) at a chilled water temperature of
about 60°F (15.6°C). At this point, the solubility of air in water is only about 3.3 percent. The
difference in solubility between point A and B is 14.23.310.9 percent. This portion of air,
containing a higher percentage of nitrogen because of the formation of oxides, is no longer
dissolved in the chilled water, but is released from the water and forms free air, gas bubbles, or
pockets. Some of the air pockets are vented through air vents at the terminals, or high points of the
water system. The chilled water returns to point A again and absorbs air from the air-water interface
in the plain closed expansion tank, creating an air solubility in water of about 14.2 percent. Of
course, the actual process is more complicated because of the formation of oxides and the presence
of water vapor.
Such a chilled water recirculating process causes the following problems:
 Oxidation occurs because of the reaction between dissolved oxygen and steel pipes, causing corrosion
during the chilled water transport and recirculating process.
 The air pockets vented at high levels originally come from the filled air in the plain closed expansion
tank; after a period of recirculation of the chilled water, part of the air charge is removed to
the upper levels and vented. The tank finally waterlogs and must be charged with compressed air
again. Waterlogging also results in an unstable system pressure because the amount of filled air in
the plain closed expansion tank does not remain constant. Oxidation and water logging also exist
in hot water systems, but the problems are not as pronounced as in a chilled water system.
Oxidation and waterlogging can be prevented or reduced by installing a diaphragm expansion
tank instead of a plain closed expansion tank. Air vents, either manual or automatic, should be
installed at the highest point of the water system and on coils and terminals at higher levels if a
water velocity of not less than 2 ft / s (0.61 m/ s) is maintained in the pipes, in order to transport the
entrained air bubbles to these air vents.
In a closed chilled water system using a diaphragm expansion tank, there is no air-water interface
in the tank. The 3.3 percent of dissolved air, or about 2.6 percent of dissolved nitrogen, in
7.24 CHAPTER SEVEN
water returning from point B to A cannot absorb more air again from the diaphragm tank. If there is
no fresh city water supply to the water system, then after a period of water recirculation the only
dissolved air in water will be the 2.6 percent nitrogen. No further oxidation occurs after the initial
dissolved oxygen has reacted with the steel pipe. Waterlogging does not occur either.
Because of the above concerns, a closed water system should have a diaphragm or bladder
expansion tank. An open expansion tank at high levels causes fewer problems than a plain closed
expansion tank. A diaphragm tank may be smaller than an equivalent plain tank. An air eliminator
or air separator is usually required for large water systems using a diaphragm tank to separate
dissolved air from water when the water system is charged with a considerable amount of city
water.
7.5 CORROSION AND DEPOSITS IN WATER SYSTEM
Corrosion
Corrosion is a destructive process that acts on a metal or alloy. It is caused by a chemical or electrochemical
reaction of a metal. Galvanic corrosion is the result of contact between two dissimilar
metals in an electrolyte. The corrosion process involves a flow of electricity between two areas of a
metal surface in a solution that conducts the electric current. One area acts as the anode and
releases electrons, whereas the other area acts as the cathode, which accepts electrons and forms
negative ions. Corrosion, or the formation of metal ions by means of oxidation and disintegration
of metal, occurs only at the anodes. In iron and steel, ferrous ions react with oxygen to form ferric
hydroxide (the rust).
Moisture encourages the formation of an electrolyte, which is one of the basic elements that give
rise to corrosion. Oxygen accelerates the corrosion of ferrous metals by means of a reaction with
hydrogen produced at the cathode. This creates the reaction at the anode. Some alloys, such as
those of stainless steel and aluminum, develop protective oxide films to prevent further corrosion.
For iron and steel, solutions such as those containing mineral acids accelerate the corrosion, and
solutions such as those containing alkalies retard it. Because the corrosion reaction at the cathode
depends on the concentration of hydrogen ions, the more acidic the solution, the higher the concentration
of hydrogen ions and the greater the corrosion reaction. Alkaline solutions have a much
higher concentration of hydroxyl ions than hydrogen, and as such the ions decrease the corrosion
rate.
Water Impurities
In hot and chilled water systems, the problems associated with water mainly concern water’s dissolved
impurities, which cause corrosion and scale, and the control of algae, bacteria, and fungi.
Typical samples of dissolved impurities in public water supplies are listed in Table 7.7.
Calcium hardness, sulfates, and silica all contribute to the formation of scale. Scale is the
deposit formed by the precipitation of water-insoluble constituents on a metal surface. Chlorides
cause corrosion. Iron may form deposits on a surface through precipitation. All these increase the
fouling factor of water.
In addition to dissolved solids, unpurified water may contain suspended solids, which may be either
organic or inorganic. Organic constituents may be in the form of colloidal solutions. At high
water velocities, hard suspended solids may abrade pipes and equipment. Particles that settle at the
bottom may accelerate corrosion.
In open water systems, bacteria, algae, and fungi cause many operating problems. The possibility
of bacteria existing in the cooling tower and causing Legionnaires’ disease necessitates microbiological
control.
WATER SYSTEMS 7.25
Water Treatments
Scale and Corrosion Control. One effective corrosion control method is to reduce the oxygen
composition in water systems. In past years, acids and chromates were the chemical compounds
commonly used to eliminate or to reduce scale and corrosion. On January 3, 1990, however, “Proposed
Prohibition of Hexavalent Chromium Chemicals in Comfort Cooling Towers” was posted by
the Environmental Protection Agency because chromates are suspected carcinogens and disposal
problems are associated with these chemicals.
There has been a significant improvement in water treatment chemistry in recent years. Currently
used chemical compounds include crystal modifiers and sequestering chemicals. Crystal modifiers
cause a change in the crystal formation of scale. As a result, scale ions cannot interlace with ions of
other scale-forming elements. Another important characteristic of these crystal modifiers and sequestering
chemicals is that they can be applied to water systems that have a wide range of pH values.
(The pH value indicates the acidity or alkalinity of a solution. It is the negative logarithm of the hydrogen
ion concentration of a solution.) Even if the chemical is over- or underfed, it will not cause operating
problems. Crystal modifiers and sequestering chemicals create fewer environmental problems.
Microbiological Control. The growth of bacteria, algae, and fungi is usually treated by biocides
to prevent the formation of an insulating layer on the heat-transfer surface, which would promote
corrosion and restrict water flow. Chlorine and chlorine compounds have been effectively and
widely used. Bromine has the same oxidizing power as chlorine and is effective over a wide pH
range. Biocide chemicals are detrimental to the environment if they are used in excess, however.
Biostat is a new type of chemical used in algae growth control. It prevents the algae spores from
maturing, which is an approach different from that of a biocide.
Blow-down or bleed-off operation is an effective water treatment. It should be considered as
important as treatments that use chemicals.
7.26 CHAPTER SEVEN
TABLE 7.7 Analyses of Typical Public Water Supplies
Location or area*,†
Substance Unit (1) (2) (3) (4) (5) (6) (7) (8) (9)
Silica SiO2 2 6 12 37 10 9 22 14 —
Iron Fe2 0 0 0 1 0 0 0 2 —
Calcium Ca 6 5 36 62 92 96 3 155 400
Magnesium Mg 1 2 8 18 34 27 2 46 1,300
Sodium Na 2 6 7 44 8 183 215 78 11,000
Potassium K 1 1 1 — 1 18 10 3 400
Bicarbonate HCO3 14 13 119 202 339 334 549 210 150
Sulfate SO4 10 2 22 135 84 121 11 389 2,700
Chloride Cl 2 10 13 13 10 280 22 117 19,000
Nitrate NO3 1 — 0 2 13 0 1 3 —
Dissolved solids 31 66 165 426 434 983 564 948 35,000
Carbonate hardness CaCO3 12 11 98 165 287 274 8 172 125
Noncarbonate hardness CaSO4 5 7 18 40 58 54 0 295 5,900
*All values are ppm of the unit cited to nearest whole number.
†Numbers indicate location or area as follows:
(1) Catskill supply, New York City
(2) Swamp water (colored), Black Creek, Middleburg, FL
(3) Niagara River (filtered), Niagara Falls, NY
(4) Missouri River (untreated), average
(5) Well waters, public supply, Dayton, OH, 30 to 60 ft (9 to 18 m)
(6) Well water, Maywood, IL, 2090 ft
(7) Well water, Smithfield, VA, 330 ft
(8) Well water, Roswell, NM
(9) Ocean water, average
Source: ASHRAE Handbook 1987, HVAC Systems and Applications. Reprinted with permission.
7.27 WATER SYSTEMS
Chemical Feeding. Improper chemical feeding causes operating problems. A water treatment program
with underfed chemicals results in an ineffective treatment, whereas an overfed program not
only increases the operating cost but also may cause environmental problems. Generally, a continuous
feeding of very small amounts of chemicals often provides effective and economical water
treatment.
System Characteristics
Closed water systems, including hot, chilled, and dual-temperature systems, can be characterized as
follows:
Constant or Variable Flow. A constant-flow water system is a system for which the volume flow
at any cross-sectional plane in the supply or return mains remains constant during the operating
period. Three-way mixing valves are used to modulate the water flow rates to the coils. In a variable-
flow system, all or part of the volume flow varies when the system load changes during the operating
period. Two-way valves are used to modulate the water flow rates to the coils or terminals.
Direct Return or Reverse Return. In a direct-return water system, the various branch piping circuits,
such as ABGBA and ABCFGBA, are not equal in length (see Fig. 7.9a). Careful balance is often
required to establish the design flow rates for a building loop when a direct-return distribution
loop is used as described in later sections. In a reverse-return system, the piping lengths for each
branch circuit, including the main and branch pipes, are almost equal (see Fig. 7.9b).
FIGURE 7.9 Direct-return and reverse-return water systems. (a) Direct-return; (b) reverse-return
Two Pipe or Four Pipe. In a dual-temperature water system, the water piping from the boiler or
chiller to the coils and the terminals, or to various zones in a building, can be either a two-pipe system,
with a supply main and a return main, as shown in Fig. 7.10a; or a four-pipe system, with a hot
water supply main, a hot water return main, a chilled water supply main, and a chilled water return
main, as shown in Fig. 7.10b. For a two-pipe system, it is impossible to heat and cool two different
coils or terminals in the same zone simultaneously. Changeover from summer cooling mode operation
to winter heating mode operation is required. A four-pipe system does not need changeover operation.
Chilled and hot water can be supplied to the coils or terminals simultaneously. However, a
four-pipe system requires a greater installation cost.
Several decades earlier, there was also a three-pipe system with a hot water supply main, a chilled
water supply main, and a common return main. ASHRAE/IESNA Standard 90.1-1999 clearly speci-
fies that hydronic systems that use a common return system for both hot water and chilled water shall
not be used. This is because of the energy loss during the mixing of the hot and chilled water.
Changeover
In a dual-temperature two-pipe system, changeover refers to when the operation of one zone or the
entire water system in a building changes from heating mode to cooling mode, or vice versa. During
changeover, the water supplied to the terminals is changed from hot water to chilled water, or
vice versa. The changeover temperature Tco, °F (°C), is the outdoor temperature at which the space
sensible cooling load can be absorbed and removed by the combined effect of the conditioned
outdoor air, the primary air, and the space transmission and infiltration loss. Such a relationship can
be expressed as:
(7.5)
where Tr  space temperature, °F (°C)
sum of internal sensible loads from electric lights, occupants, and appliances,
Btu/h (W)
sum of external sensible loads through building shell, Btu/h (W)
volume flow rate and density of conditioned outdoor air, cfm (m3/min) and lb/ ft3
(kg/m3)
cpa  specific heat of air, Btu / lb °F (J/kg °C)
Tso  supply temperature of outdoor air or primary air, °F (°C)
qtl  transmission and infiltration losses per 1°F of outdoor-indoor temperature difference,
Btu/h °F (W/°C)
Changeover usually takes from 3 to 8 h to complete. The greater the size of the water system, the
longer the changeover period. To prevent more than one changeover per day, the changeover temperature
Tco may have a tolerance of 
2°F (1.1°C).
Changeover may cause a sudden flow of a large amount of hot water into the chiller or of chilled
water into the boiler. Such a rapid change in temperature imposes a thermal shock on the chiller or
boiler and may damage the equipment. For chillers, the temperature of water entering the chiller
should be no higher than 80°F (26.7°C) to prevent excessive refrigerant pressure in the evaporator.
For boilers, a temperature control system bypasses most of the low-temperature water until the water
temperature can be gradually increased.
Changeover may be performed either manually or automatically. Manual changeover is simple
but may be inconvenient during periods when daily changeover is required. With sufficient
safety controls, automatic changeover reduces the operating duties significantly. A compromise is
V?
so, 	so 
Qres 
Qris 
K  60 	socpa
Tco  Tr 
Qris 	 Qres  KV?so(Tr  Tso)
qtl
7.28 CHAPTER SEVEN
7.29
FIGURE 7.10 Multiple-zone, dual-temperature water systems. (a) Two-pipe system; (b) four-pipe system.
a semiautomatic changeover system in which the changeover temperature is set by a manual
switch.
Outdoor reset control is often used to vary the supply water temperature Tws [°F(°C)], in response
to the outdoor temperature To for a hot water system. Typically, Tws is 130°F (54.4°C) at the
winter design temperature and drops linearly to 80°F (26.7°C) at the changeover temperature.
ASHRAE/IESNA Standard 90.1-1999 specifies that two-pipe changeover systems are acceptable
when the following requirements are met:
 The designed deadband width between changeover from one mode to the other is of at least 15°F
(8.3°C) outdoor temperature.
 System controls will allow operation in one mode for at least 4 h before changing over to
another mode.
 At the changeover point, reset controls allow heating and cooling supply temperatures to be no
more than 30°F (16.7°C) apart.
7.7 CENTRIFUGAL PUMPS
Centrifugal pumps are the most widely used pumps for transporting chilled water, hot water, and
condenser water in HVAC&R systems because of their high efficiency and reliable operation.
Centrifugal pumps accelerate liquid and convert the velocity of the liquid to static head. A typical
centrifugal pump consists of an impeller rotating inside a spiral casing, a shaft, mechanical seals
and bearings on both ends of the shaft, suction inlets, and a discharge outlet, as shown in Fig. 7.11.
The impeller can be single-stage or multistage. The vanes of the impeller are usually backwardcurved.
The pump is usually described as standard-fitted or bronze-fitted. In a standard-fitted construction,
the impeller is made of gray iron, and in a bronze-fitted construction, the impeller is made of
bronze. For both constructions, the shaft is made of stainless steel or alloy steel, and the casing is
made of cast iron.
Three types of centrifugal pumps are often used in water systems in HVAC&R systems: doublesuction
horizontal split-case, frame-mounted end suction, and vertical in-line pumps, as shown in
the upper part of Fig. 7.12. Double-suction horizontal split-case centrifugal pumps are the most
widely used pumps in large central hydronic air conditioning systems.
Basic Terminology
Volume flow rate [gpm (m3/s)] is the capacity handled by a centrifugal pump. On a cross-sectional
plane perpendicular to fluid flow in a water system, the static head Hs [ft (m)] is the pressure
expressed in feet (meters) of water column that is exerted on the surrounding fluid and enclosure.
On a cross-sectional plane, velocity head Hv [ft (m)] can be calculated as
(7.6)
where vo  velocity of water flow at pump outlet, ft / s (m/ s)
g  gravitational acceleration, 32.2 ft / s2 (9.81 m/ s2 )
Total head Ht [ft (m)] is the sum of static head and velocity head, i.e.,
Ht  Hs 	 Hv (7.7)
Hv 
Vo 2
2g
V?
p
7.30 CHAPTER SEVEN
WATER SYSTEMS 7.31
(a)
(b)
Suction
inlet
Enclosed double-suction
impeller
Ball bearings
Shaft
Case wear rings
Bronze shaft sleeve
FIGURE 7.11 A double-suction, horizontal split-case, single-stage centrifugal pump. (a) Sectional view;
(b) centrifugal pump and motor assembly.
Net static head [ft (m)] is the head difference between the discharge static head Hdis and suction
static head Hsuc, both in feet (meter), as shown in Fig. 7.13.
Pump power Pp (hp) is the power input on the pump shaft; and pump efficiency p is the ratio of
the energy output from water to the power input on the pump shaft, and both can be calculated as
(7.8)
where gs  specific gravity, i.e., the ratio of the mass of liquid handled by the pump to the mass of
water at 39°F (4°C).
Performance Curves
Pump performance is often illustrated by a head-capacity curve and a power-capacity
curve, as shown in Fig. 7.12. The head-capacity curve illustrates the performance of a centrifugal
pump from maximum volume flow to the shutoff point. If the total head at shutoff point Hso is 1.1
to 1.2 times the total head at the point of maximum efficiency Hef , the pump is said to have a flat
head-capacity curve. If Hso  1.1Hef to 1.2Hef, it is a steep-curve pump.
Pp-V?p Ht-V?p
p 
V?
p t gs
3960Pp
Pp 
V?
p Ht gs
3960 p
Hs
7.32 CHAPTER SEVEN
FIGURE 7.12 Performance curves for centrifugal pumps.
Net Positive Suction Head
The lowest absolute water pressure at the suction inlet of the centrifugal pump (shown in Fig. 7.11)
must exceed the saturated vapor pressure at the corresponding water temperature. If the absolute
pressure is lower than the saturated vapor pressure, the water evaporates and a vapor pocket forms
between the vanes in the impeller. As the pressure increases along the water flow, the vapor pocket
collapses and may damage the pump. This phenomenon is called cavitation.
The sum of the velocity head at the suction inlet and the head loss (due to friction and
turbulence) between the suction inlet and the point of lowest pressure inside the impeller is
called the net positive suction head required (NPSHR), in feet (meters). This factor is determined by
the pump manufacturer for a given centrifugal pump. For a specific water system using a centrifugal
pump, the net positive suction head available (NPSHA), in feet (meters), can be calculated as
NPSHA  Hat 	 Hsuc  Hf  2.31pvap (7.9)
where Hat  atmopheric pressure, usually expressed as 34 ft (10.3 m or 101 kPa) of water
column
Hsuc  static suction head, ft (m)
Hf  head loss due to friction and dynamic losses of suction pipework and fittings, ft (m)
pvap  saturated water vapor pressure corresponding to water temperature at suction inlet,
psia (kPa abs.)
NPSHA must be greater than NPSHR to prevent cavitation.
Pump Selection
First, the selected pump must satisfy the volume flow and total head requirements and should operate
near maximum efficiency most of the time. Second, for comfort air conditioning systems, quiet operation
is an important consideration. A noise generated in a water system is very difficult to isolate and
WATER SYSTEMS 7.33
FIGURE 7.13 Net static head.
remove. In most cases, the lowest-speed pump that can provide the required and pump head Ht is
the quietest and often the most economical choice. Third, today a variable-speed pump is often costeffective
for a variable-flow system. However, when a constant-speed pump is used to serve a variable-
flow system that operates with minor changes of head, a flat-curve pump should be selected.
7.8 PUMP-PIPING SYSTEMS
System Curve
When a pump is connected with a pipe system, it forms a pump-piping system. A water system may
consist of one pump-piping system or a combination of several pump-piping systems.
The speed of a variable-speed pump in a variable-flow water system is often controlled by a
pressure-differential transmitter installed at the end of the supply main, with a set point normally
between 15 and 20 ft WC (4.5 and 6 m WC). This represents the head loss resulting from the control
valve, pipe fittings, and pipe friction between the supply and return mains at the farthest branch
circuit from the variable-speed pump. Therefore, the head losses of a pump-piping system can be
divided into two parts:
 Constant, or fixed, head loss Hfix, which remains constant as the water flow varies. Its magnitude
is equal to the set point of the pressure-differential transmitter Hset, or the head difference between
the suction and the discharge levels of the pump in open systems Hsd [ft WC (m WC)].
 Variable head loss Hvar, which varies as the water flow changes. Its magnitude is the sum of the head
losses caused by pipe friction Hpipe, pipe fittings Hfit, equipment Heq (such as the pressure drop
through the evaporator, condenser, and coils), and components Hcp, all in ft WC (m WC), that is,
Hvar  Hpipe 	 Hfit 	 Heq 	 Hcp (7.10)
Head losses Hfix and Hvar are shown in Fig. 7.14. The relationship between the pressure loss
p [ft WC (kPa)]; flow head Hvar [ft WC (m WC)]; flow resistance of the water system Rvar [ft
WC/(gpm)2 (m WC s 2/m6)]; and water volume flow rate [gpm (m3/s)], can be expressed as
(7.11)
where 	w  density of water, lb/ft3 (kg/m3)
g  gravitational acceleration, ft / s2 (m/s2)
gc  dimensional constant, 32.2 lbmft / lbf  s2
The curve that indicates the relationship between the flow head, flow resistance, and water volume
flow rate is called the system curve of a pump-piping system, or a water system.
System Operating Point
The intersection of the pump performance curve and the water system curve is the system operating
point of this variable-flow water system, as shown by point P in Fig. 7.14. Its volume flow rate is
represented by [gpm (m3/s)], and its total head is HP  Hfix 	 Hvar [ft WC (m WC)].
Usually, the calculated system head loss is overestimated, and the selected pump is oversized
with a higher pump head, so that the actual system operation point is at point P. Therefore, for a
V?
P
Hvar  R
var V?
w 2
p  Rvar V?w
2
Hvar  p
gc
	wg
V?
w
V?
p
7.34 CHAPTER SEVEN
variable-flow water system installed with a constant-speed pump, the design system operating
point is preferably located to the left of the region of pump maximum efficiency, because the system
operating point of an oversized pump moves into or nearer to the region of pump maximum
efficiency.
Combination of Pump-Piping Systems
When two pump-piping systems 1 and 2 are connected in series as shown in Fig. 7.15a, the volume
flow rate of the combined pump-piping system, [gpm (m3/s)] is
(7.12)
where and are the volume flow rate of pump-piping systems 1 and 2, gpm (m3/s). The total
head lift of the combined system Hcom [ft WC (m WC)] is
Hcom  H1 	 H2 (7.13)
where H1 and H2 are the head of pump-piping systems 1 and 2, ft WC (m WC).
It is simpler to use one system curve to represent the whole system, i.e., to use a combined system
curve. The system operating point of the combined pump-piping system is illustrated by point
P with a volume flow of and head of HP. The purpose of connecting pump-piping systems in
series is to increase the system head.
When a pump-piping system has parallel-connected water pumps, its volume flow rate [gpm
(m3/s)] is the sum of the volume flow rates of the constituent pumps , , etc. The head of each
constituent pump and the head of the combined pump-piping system are equal. It is more convenient
to draw a combined pump curve and one system curve to determine their intersection, the
system operating point P, as shown in Fig. 7.15b. The purpose of equipping a water system with
parallel-connected water pumps is to increase its volume flow rate.
V?
2 V?1
V?
V?
P
V?
2 V?1
V?
com  V?1  V?2
V?
com
WATER SYSTEMS 7.35
FIGURE 7.14 Water system curve and system operating point.
Modulation of Pump-Piping Systems
Modulation of the volume flow rate of a pump-piping system can be done by means of the following:
 Throttle the volume flow by using a valve. As the valve closes its opening, the flow resistance of
the pump-piping system increases. A new system curve is formed, which results in having a new
system operating point that moves along the pump curve to the left-hand side of the original
curve, with a lower volume flow rate and higher total head, as shown in Fig. 7.16a. Such behavior
is known as riding on the curve. Using the valve to modulate the volume flow rate of a
pump-piping system always wastes energy because of the head loss across the valves Hval in
Fig. 7.16a.
 Turn water pumps on or off in sequence for pump-piping systems that have multiple pumps in a
parallel connection. Modulation of the volume flow rate by means of turning water pumps on and
off often results in a sudden drop or increase in volume flow rate and head, as shown by system
operating points P, Q, and T in Fig. 7.16b.
 Vary the pump speed to modulate the volume flow and the head of a pump-piping system. When
the speed of the pump is varied from n1 to n2 and then to n3, new pump curves P2 and P3 are
formed, as shown in Fig. 7.16b. The system operating point will move from point P to Q and then
to T along the system curve, with a lower volume flow rate, head, and input pump power. The system
curve becomes the modulating curve and approaches Hfix  Hset when the volume flow
rate is zero. Here Hset is the set point of the pressure-differential transmitter, ft WC (m WC).
Varying the pump speed requires the lowest pump power input in comparison with other modulation
methods.
7.36 CHAPTER SEVEN
FIGURE 7.15 Combination of pump-piping systems. (a) Two pump-piping systems connected in series; (b) three parallel-connected
pumps.
Pump Laws
The performance of geometrically and dynamically similar pump-piping systems 1 and 2 can be expressed
as follows:
(7.14a)
(7.14b)
(7.14c)
where  volume flow rate of pump-piping system, gpm (m3/s)
Ht  total head lift, ft WC (m WC)
P  pump power input at shaft, hp (kW)
D  outside diameter of pump impeller, ft (m)
n  speed of pump impeller, rpm
Equations (7.14a) through (7.14c) are known as the pump laws. They are similar to the fan laws and
are discussed in detail in Chap. 17.
Wire-to-Water Efficiency
A pump may be directly driven by a motor, or it may be driven by a motor and belts. When the energy
cost of a water system is evaluated, the pump total efficiency p, the motor efficiency mot, and
the efficiency of the variable-speed drives dr should all be considered.
V?
P2
P1

n2 3
n1 3
Ht2
Ht1

n2 2
n1 2
V?2
V?
1

D2 3n2
D1 3n1
WATER SYSTEMS 7.37
FIGURE 7.16 Modulation of pump-piping systems. (a) Using a valve; (b) varying the pump speed.
The wire-to-water efficiency of a water system ww, expressed either in dimensionless form or as
a percentage, is defined as the ratio of energy output from water to the energy input to the electric
wire connected to the motor. It can be calculated as
ww  p dr mot (7.15)
The total efficiency of the centrifugal pump p can be obtained from the pump manufacturer or calculated
from Eq. (7.8). The pump efficiency p depends on the type and size of pump as well as the
percentage of design volume flow rate during operation. Pump efficiency usually varies from 0.7 to
0.85 at the design volume flow rate. Drive efficiency dr indicates the efficiency of a direct drive,
belt drive, and various types of variable-speed drives. For direct drive, dr  1. Among variablespeed
drives, an adjustable-frequency alternating-current (ac) drive has the highest drive efficiency.
For a 25-hp (18.7-kW) motor, dr often varies from 0.96 at design flow to 0.94 at 30 percent design
flow to 0.80 at 20 percent design flow. Motor efficiency mot depends on the type and size of motor.
It normally varies from 0.91 for a 10-hp (7.5-kW) high-efficiency motor to 0.96 for a 250-hp (187-
kW) motor as listed in Table 6.2.
7.9 OPERATING CHARACTERISTICS OF CHILLED
WATER SYSTEM
Many chilled and hot water systems used in commercial central hydronic air conditioning systems
often have their central plant located in the basement, rooftop, or equipment floors of the building.
The hot/chilled water from the boiler/chiller in the central plant is then supplied to the coils and terminals
of various zones in one building or in adjacent buildings by means of supply main pipes.
Water returns from the coils and terminals to the central plant via the return mains.
Coil Load and Chilled Water Volume Flow
In AHUs or fan coils, two-way control valves are currently widely used to modulate the water
volume flow rate so as to maintain a predetermined air discharge temperature or space temperature
at reduced system loads. Coils, especially oversized coils, operate at design load usually less than 5
percent of their total operating time. For a typical coil, nearly 60 percent of the operating time may
correspond to a coil load of 35 to 65 percent of the design value.
During part-load operation, the required fraction of design volume flow rate of chilled water
flowing through a coil is not equal to the fraction of design sensible coil load Qcs Btu/h (W)
which is the sensible heat transfer from the coil to the conditioned air, as shown in Fig. 7.17a. In
Fig. 7.17a, indicates the design chilled water volume flow rate, gpm (m3/min), and Qcs,d the
design sensible coil load, Btu/h (W). This is because of the characteristics of sensible heat transfer
described by
Qcs  AoUoTm (7.16)
where Ao  outer surface area of coil, ft2 (m2)
Uo  overall heat-transfer coefficient based on outer area, Btu/h  f t 2 °F (W/m2°C)
Tm  logarithmic temperature between conditioned air and chilled water,°F (°C)
When the volume flow rate of chilled water is reduced, the decrease in the product of AoUoTm
is not the same as the reduction in the chilled water volume flow rate . When drops, the outer
surface area Ao remains the same and Uo is slightly reduced. Only a considerable rise in chilled
water temperature across the coil Tw,c  Twl  Twe, as shown in Fig. 7.17b, can reduce Tm suffi-
V?
w V?w
V?
w
V?
w,d
V?
w
7.38 CHAPTER SEVEN
ciently to match the reduction of Qcs. Figure 7.17 is obtained for entering water and entering air
temperatures that remain constant at various fractions of the design flow.
Theoretically, when the sensible coil load Qcs is reduced to 0.6 of the design value, the chilled
water volume flow rate should be decreased to about 0.25 of the design volume flow rate to match
the reduction of Qcs. Meanwhile, the power input at the shaft of the variable-speed pump is only
about 8 percent of its design brake horsepower. There is a tremendous savings in pump power for a
variable-flow system compared to a constant-flow system.
A two-way control valve for the coil must be carefully selected. First, an equal-percentage contour
valve should be used. As described in Sec. 5.6 and shown in Figs. 5.16 and 7.17a, when a coil
is equipped with an equal-percentage valve, the sensible coil load is directly related to the valve
stem travel or control output signal through the percentage water flow rate and thus provides better
control quality. Second, the control valve closes its opening to provide both the pressure drop for
the modulated required flow and an additional pressure drop for the coils nearer to the pump for
water flow balance if it is a direct-return piping system.
Chiller Plant
For chilled water systems, a central plant is often installed with multiple chillers, typically two to
four chillers. Multiple chillers are usually connected in parallel. Each chiller is often installed with
a chilled water pump that has the same volume flow rate as the water chiller. In such an arrangement,
it is more convenient to turn the chillers on or off in sequence. ASHRAE/IESNA Standard
90.1-1999 specifies that when a chilled water plant is equipped with more than one chiller, provisions
shall be made so that the chilled water flow in the chiller plant can be automatically reduced
when a chiller is shut down. Chillers must provide adequate chilled water flow and cooling capacity
that the AHUs and fan coils require.
Usually, a fairly constant-volume flow in the evaporator of the water chiller is preferable to
avoid an extremely high temperature drop in the chiller and to prevent water from freezing at a reduced
flow during part-load operation. A constant flow of chilled water in chillers is also beneficial
to the capacity control of multiple chillers.
WATER SYSTEMS 7.39
FIGURE 7.17 Relationship between fraction of design volume flow rate , coil load Qcs, and water temperature
rise Tw,c. (a) Qcs versus ; (b) Tw,c versus . V?w,d V?w
V?
w,d
Variable Flow for Saving Energy
ASHRAE/IESNA Standard 90.1-1999 specifies that water systems having a total pump system
power exceeding 10 hp (7.5 kW) shall include control valves to modulate or step open and close as
a function of load and shall be designed for variable flow for energy savings. Water systems should
be able to reduce system flow to 50 percent of design flow or less. Individual pumps serving variable
flow systems having a pump head exceeding 100 ft (30 m) and motors exceeding 50 hp (37
kW) shall have controls and devices (such as variable speed control) that will reduce pump motor
demand of no more than 30 percent of design wattage at 50 percent of design water flow.
Water Systems in Commercial Buildings
The following types of water systems are currently used in commercial buildings in the United States:
 Plant-through-building loop: bypass throttling flow
 Plant-through-building loop: distributed pumping
 Plant-building loop
 Plant-distributed pumping loop
 Plant-distribution-building loop
 Plant-distributed building loop
 Multiple sources-distributed building loop
Since the 1960s, one of the old, energy-inefficient water systems, a constant-flow system using
three-way valves which has constant flow in its chiller/boiler and its supply and return mains, is
gradually being replaced by a water system that is equipped with cheaper and more effective twoway
valves and uses energy-efficient variable flow for distribution. Constant-flow systems using
three-way valves are not discussed here.
7.10 PLANT-THROUGH-BUILDING LOOP
In a plant-through-building loop system, water is transported only by plant (chiller/boiler) pump(s)
or by distributed pump(s). Plant-through-building loops can be classified into three categories: bypass
throttling flow, distributed pumping, and variable flow.
Bypass Throttling Flow
A plant-through-building loop water system using bypass throttling flow is one of the older
chilled /hot water systems that has been adopted in commercial buildings since the use of two-way
control valves. For each chiller/ boiler, a corresponding plant constant-speed water pump is
equipped as shown in Fig. 7.9a. The chilled or hot water is supplied to the coils and terminals
through the supply and return mains and branches, and is then returned to the chiller/boiler. There is
a crossover bridge that connects the supply and return mains at junctions P and Q. A bypass twoway
control valve is often installed on the crossover bridge. A pressure-differential transmitter and
a pressure relief valve are used to maintain a set pressure differential across the supply and return
mains by modulating the opening of the bypass two-way control valve when the system pressure
7.40 CHAPTER SEVEN
tends to increase during part-load operation. A portion of the water is throttled in the control valve
and flows through the bypass crossover. It is then combined with water from the return main and returns
to the chiller/boiler. A constant flow (or approximately constant flow) is maintained in the
chiller / boiler.
A plant-through-building loop using bypass throttling control cannot save pumping energy if the
set point of the pressure-differential transmitter is fixed at the value of pressure drop across the
crossover bridge at design load during part-load operation. In comparison with a plant-building
loop water system, it is simpler, and lower in first cost and needs a smaller pump room space.
However, a plant-building loop (primary-secondary loop, which is discussed in a later section)
water system saves far more pump energy than a plant-through-building loop using bypass throttling
control.
Plant-through-building loop using bypass throttling control still has applications in small projects
and especially in retrofit where space may not be available for a plant-building loop system.
Distributed Pumping
A plant-through-building loop water system using distributed pumping is often used for hot water
systems. Hot water supplied from a boiler is divided into several distributed-piping loops, as
shown in Fig. 8.7. For each of the distributed-piping loop in a control zone, there is either a variable-
speed on-line hot water circulating pump or a constant-speed on-line pump with a two-way
control valve, finned-tube baseboard heaters, supply and return pipes, accessories, and controls.
Hot water flows through the distributed-piping loops and then returns to the boiler for heating
again. The amount of hot water flowing through the boiler is reduced within a limit during partload
operation.
A space temperature sensor sends a signal to a direct digital control (DDC) unit controller and
modulates the amount of hot water extracted from the supply header by means of a variable-speed
on-line pump, or a two-way control valve and an on-line pump to maintain a preset zone temperature.
A temperature sensor is also mounted at the water outlet of the boiler to maintain a preset hot
water leaving temperature through the modulation of the burner’s capacity by a DDC unit
controller.
A plant-through-building loop with distributed-pumping water system is a simple, energy-effi-
cient system.
Variable Flow
A plant-through-building loop using variable flow, as shown in Fig. 7.18, is an ideal chilled water
system with variable flow in chiller, supply main, coils, return main, and variable-speed
pump(s) during part-load operation. For each chiller, there is often a corresponding variablespeed
pump. Chilled water is supplied to various control valves and coils in branches through
supply and return mains. It is then extracted by the pump(s) and returned to the chillers for cooling
again.
Three related direct digital controls are equipped for this system:
 A control of chilled water temperature leaving the chiller
 An air discharge temperature control by modulation of the control valve and thus the water flow
rate entering the coil
 A pressure-differential control using pressure-differential transmitter with a DDC unit controller
that modulates the variable-speed pump to maintain a preset pressure differential between supply
and return mains
WATER SYSTEMS 7.41
Today, although many chiller manufacturers allow a reduction of chilled water flow of 30 to 40 percent
of the design volume flow rate, multiple variable-speed pumps must run at approximately the
same speed at the same pump head. If Qcs /Qcs,d drops to 0.5, the system flow ratio needs to
reduce to only about 0.28 (refer to Table 7.8). This may cause operating troubles, control problems,
and system instability.
Until the middle of 1997, there hasn’t been a single project actually using plant-throughbuilding
loop with variable-flow chilled water system reported in HVAC&R publications in the
United States and operated successfully.
V?
w/V?w,d
7.42 CHAPTER SEVEN
TABLE 7.8 Magnitudes of bg/ bg,d and Calculated Results for Example 7.1
Qcs/Qcs,d
0.3 0.4 0.5 0.6 0.65 0.7 0.8 0.9 1.0
1 chiller 2 chillers 2 chillers 3 chillers
bg/ bg,d 0.17 0.17 0.22 0.275 0.34 0.38 0.38 0.45 0.60 0.8 1.0
Tw,c, °F 36 36 36 36 35.2 34.2 34.2 31.1 26.7 22.5 20
pt, gpm 350 700 700 700 700 700 1050 1050 1050 1050 1050
cn, gpm 180 530 480 425 360 320 670 600 450 250 0
Tee, °F 57.5 48.8 51.3 54.2 57.1 58.6 52.4 53.3 55.3 57.1 60
V? V?
V?
V?
V?
V?
2 1 1 2
Coil
Chiller
Coil
Coil
Coil
PD T
PD Pressure-differential transmitter Variable-speed drive
VSD
VSD
Chiller
VSD
Chiller
VSD
FIGURE 7.18 A chilled water system of plant-through-building loop using variable flow.
7.11 PLANT-BUILDING LOOP
System Description
Plant-building loop water systems, also called primary-secondary loop (or circuit) water systems,
are the widely adopted water systems for large new and retrofit commercial HVAC&R installations
in the United States today. A plant-building loop chilled water, hot water, or dual-temperature water
system consists of two piping loops:
 Plant loop (primary loop). In a plant loop, there are chiller(s)/boiler(s), circulating water pumps,
diaphragm expansion tank, corresponding pipes and fittings, and control systems, as shown by
loop ABFG in Fig. 7.19. A constant volume flow rate is maintained in the evaporator of each
chiller. For a refrigeration plant equipped with multiple chillers, the chilled water volume flow
rate in the plant loop will vary when a chiller and its associated chiller pump are turned on or off.
 Building loop (secondary loop). In a building loop, there are coils, terminals, probably variablespeed
water pumps, two-way control valves and control systems, and corresponding pipes, fittings,
and accessories, as shown by loop BCDECF in Fig. 7.19. The water flow in the building
loop is varied as the coil load is changed from the design load to part-load.
A short common pipe, sometimes also called a bypass, connects these two loops and combines
them into a plant-building loop.
Control Systems
For a plant-building loop water system, there are four related specific control systems:
Coil Discharge Air Temperature Control. A sensor and a DDC system or unit controller are used
for each coil to modulate the two-way control valve and the water flow into the coil. The discharge
temperature after the coil can be maintained within predetermined limits.
Water Leaving Chiller Temperature Control. In a chiller, chilled water temperature leaving the
chiller is always maintained at a preset value within a specified period by varying the refrigerant
flow in the chiller. In a boiler, the leaving temperature of hot water is maintained at a predetermined
value by varying the fuel flow to the burner.
During part-load, the chilled water temperature leaving the chiller is reset to a higher value,
such as between 3 and 10°F (1.7 and 5.6°C) according to system loads or outdoor temperature
both to reduce the pressure lift between evaporating and condensing pressure and to save compressing
power. ASHRAE/IESNA Standard 90.1-1999 specifies that a chilled water system with a
design capacity exceeding 300,000 Btu/h (87,900 W) supplying chilled water to comfort air conditioning
systems shall be equipped with controls that automatically reset supply chilled water
temperature according to building loads (including return chilled water temperature) or outdoor
temperature.
Staging Control. Chillers are turned on and off in sequence depending on the required system
cooling capacity Qsc Btu/h (W), or the sum of the coils’ loads. The required system cooling capacity
can be found by measuring the product of the temperature difference across the supply and
return mains, as shown by temperature sensors T8 and T7 and the water volume flow rate by the
flowmeter F2 in Fig. 7.19. If the produced refrigeration capacity Qrf , Btu /h (W), measured by the
product of chilled water supply and return temperature differential (T6 and T5) and the flowmeter
(F1) is less than Qsc, a DDC system controller turns on a chiller. If Qrf Qsc is greater than the refrigeration
capacity of a chiller Qrfc, the system controller turns off a chiller. Chillers should not be
staged on or off based on the chilled water volume flow rate flowing through the common pipe.
WATER SYSTEMS 7.43
7.44
6 5 3 3
7 3 3
3 3 1 4 2
5
4 3 1
6
6
Boiler
Evaporator
Condenser
Plant
hot water
pump
Building
hot water
pumps
Chiller
pump
Condenser
pump
Chilled water
return main
Hot water
return main
Temperature sensor
Hot water
supply main
Chiller water
supply main
Coils
Building variablespeed
pumps
Common pipe
T2
T1
T
Pressure-differential
transmitter
PD
Flow meter F
Flow switch FS
Adjustable frequency
drives
VSD
T5
T7
T1 PD1
T9
T6 T8 F2
FS1
A
F
B C
C
D
E
G FS2
VSD
VSD
FIGURE 7.19 A dual-temperature water system with plant-building loop.
Pressure-Differential Control. These controls are used to maintain the minimum required pressure
differential between the supply and return mains at a specific location, as shown by PD1 in
Fig. 7.19. If only one differential-pressure transmitter is installed for chilled or hot water supply
and return mains, it is usually located at the end of the supply main farthest from the building pump
discharge. If multiple differential-pressure transmitters are installed, they are often located at places
remote from the building pump discharge, with a low signal selector to ensure that any coil in the
building loop has an adequate pressure differential between the supply and return mains.
The set point of the differential-pressure transmitter should be equal to or slightly greater than
the sum of the pressure drops of the control valve, coil, pipe fittings, and piping friction of the
branch circuit between the supply and return mains. A low set point cannot ensure adequate water
flow through the coils. A high set point consumes more pump power at a reduced flow. A set point
of 15 to 25 ft (4.5 to 7.5 m) of head loss may be suitable.
System Characteristics
For a plant-building loop chilled water system, when the volume flow rate of the chilled water in the
building loop is at its design value , the volume flow rate in the plant loop is equal to that in
the building loop theoretically, all in gpm (m3/min). In actual practice, is slightly (less than 3
percent) higher than to guarantee a sufficient chilled water supply to the building loop.
At design load, chilled water leaving the chiller(s) at point A flows through the junction of the
common pipe, plant loop, and building loop (point B), is extracted by the variable-speed building
pump; and is supplied to the coils. From the coils, chilled water returns through another junction of
the building loop, common pipe, and plant loop (point F). There is only a very small amount of
bypass chilled water in the common pipe flows in the direction from point B to F. The chilled water
return from the coils is then combined with the bypass water from the common pipe and is
extracted by the chiller pump(s) and enters the chiller(s) for cooling again.
When the coils’ load drops during part-load operation, and the water volume flow rate bg
reduces in the building loop because the control valves have been partially closed, is now
considerably greater than bg. Chilled water then divides into two flows at junction B: water at the
reduced volume flow rate is extracted by the variable-speed building pump in the building loop and
is supplied to the coils; the remaining water bypasses the building loop by flowing through the common
pipe, is extracted by the chiller pump(s), and returns to the chiller(s).
For a water system that includes a plant-building loop with a common pipe between the two
loops, Carlson (1968) states the following rule: One pumped circuit affects the operation of the
other to a degree dependent on the flow and pressure drop in piping common to both circuits. The
lower the pressure drop in the common pipe, the greater the degree of isolation between the plant
and building loops. The head-volume flow characteristics of these loops act as two separate
systems.
A plant-building loop has the following characteristics:
 It provides variable flow at the building loop with separate building pump(s) and constant flow
through the evaporator of the chiller and thus saves pumping power during periods of reduced
flow in the building loop. According to Rishel (1983), the annual pump energy consumption of a
plant-building loop with variable flow in a building loop that uses a variable-speed building pump
is about 35 percent that of a plant-through-building loop constant-flow system using three-way
control valves.
 It separates the building loop from the plant loop and makes the design, operation, and control of
both loops simpler and more stable.
 Based on the principles of continuity of mass and energy balance, if differences in the density
of chilled water are ignored at junctions B and F, the sum of the volume flow rates of chilled
water entering the junction must be equal to the sum of volume flow rates of water leaving that
junction. Also, for an adiabatic mixing process, the total enthalpy of chilled water entering the
V?
V?
pt
V?
V?
bg,d
V?
pt V?bg,d
V?
pt V?bg,d
WATER SYSTEMS 7.45
junction must be equal to the total enthalpy of water leaving the junction. At junction B or F,
chilled water has the same water pressure and temperature.
Sequence of Operations
Consider a chilled water system in a dual-temperature water system that is in a plant-building loop,
as shown in Fig. 7.19. There are three chillers in the plant loop, each of which is equipped with a
constant-speed chiller pump. In the building loop, there are two variable-speed building pumps connected
in parallel. One is a standby pump. Chilled water is forced through the water cooling coils in
AHUs that serve various zones in the building. For simplicity, assume that the latent coil load
remains constant when the coil load varies. Based on the data and information from Ellis and
McKew (1996), for such a chilled water system, the sequence of operations of the DDC system is
as follows:
1. When the system controller of the water system is in the off position, the chiller pump is off,
condenser pump is off, building pump is off, and the cooling tower fan is off.
2. If the system controller is turned on, then the chiller’s on/off switch in the unit controller is
placed in the on position; and interlock signals are sent to three chiller pumps, one variable-speed
building pump, and three condenser pumps and start all these pumps. The variable-speed building
pump is always started from zero speed and increases gradually for safety and energy saving. As
the chilled water flow swiches confirm that all the pumps are delivering sufficient water flow, the
compressor of the leading chiller (first chiller) is turned on.
3. Temperature sensor T2 tends to maintain the set point of the chilled water leaving chiller temperature
often at 45°F (7.2°C) by means of refrigerant flow control through multiple on/off compressors,
modulation of inlet vanes, or variable-speed compressor motor (details are discussed in
later chapters). Temperature sensors T7 and T8 and flowmeter F2 measure the required system
cooling capacity Qsc; and sensors T5 and T6 and flowmeter F1 measure the produced refrigeration
capacity Qrf. If Qsc  Qrf, chiller is staged on in sequence, until Qrf  Qsc.
4. Condenser water temperature sensor T measures the water temperature entering the condenser
so that it will not be lower than a limit recommended by the manufacturer for normal operation.
5. When the coils’ control valves in AHUs close, the chilled water flow drops below the design
flow. As the pressure-differential transmittter DP1 senses that the pressure differential between
chilled water supply and return mains increases to a value which exceeds the set point, such as 15 ft
(4.5 m), the system controller modulates the variable-speed drive (VSD) and reduces the speed of
the variable-speed pump to maintain a 15-ft (4.5-m) pressure differential.
6. At the design system load, three chillers shall provide nearly their maximum cooling capacity,
and the veriable-speed building pump shall provide maximum flow through pump speed control.
All two-way valves shall be nearly opened fully. A constant chilled water flow is maintained in the
evaporator of each chiller. Cooling tower fan shall be continuously operated at full speed.
7. During part-load operation as the sum of the coils load (system load) decreases, the two-way
valves close their openings to reduce the chilled water flowing through the coils. At a specific fraction
of design sensible coil load Qcs /Qcs,d, there is a corresponding water volume flow rate in the
building loop, expressed as a fraction of design flow , that offsets this coil load. The building
variable-speed pump should operate at this building volume flow rate [gpm (m3/min)] with a
head sufficient to overcome the head loss in the building loop through the modulation of the variable-
speed pump.
The supply and return temperature differential of the building loop, or the mean chilled water
temperature rise across the coils Twc [°F°C] depends on the fraction of the design sensible coil load
Qcs /Qcs,d and the fraction of the design volume flow rate through the coils . The smaller the
value of , the greater the temperature rise Tw,c. At part load (Qcs /Qcs,d  1), Tw,c is always
greater than that at the design load, as shown in Fig. 7.20d.
V?
bg/V?bg,d
V?
bg/V?bg,d
V?
bg
V?
bg / V?bg,d
7.46 CHAPTER SEVEN
WATER SYSTEMS 7.47
FIGURE 7.20 System performance curves for plant-building loop. (a) Schematic diagram; (b) head of building
variable-speed pump at various volume flow rates; (c) versus Qcs /Qcs, d and versus Qcs /Qcs,d;
(d) Tw,c versus Qcs /Qcs,d.
V?
bg / V?bg, d V?bg
Coil
Chiller
Chiller
Chiller
Coil
Coil
Coil
pFG
pCK
A
E D C B F
H J K L G
(a)
M
8. During part-load operation, temperature sensors T5, T6, T7, and T8 and flowmeters F1 and F2
measure the readings which give the produced cooling capacity Qrf and required system cooling
capacity Qsc. If Qrf  Qsc  Qrf,c (one chiller’s cooling capacity, Btu/ h), none of the chillers is staging
off. When Qrf  Qs  Qrf,c, one of the chillers is then turned off until Qrf  Qsc  Qsc,c. To turn
off a chiller, the compressor is turned off first, then the chiller pump, condenser pump, and cooling
tower fans corresponding to that chiller.
9. During part-load operation, a constant flow of chilled water is still maintained in the evaporator
of each turned-on chiller. However, the volume flow rate in the plant loop depends on the
number of operating chillers and their associated chiller pumps. The staging on or off of the chillers
and their associated pump causes a variation of chilled water volume flow rate in the plant loop.
The difference between the volume flow rate of chilled water in the plant loop and the volume
flow rate in the building loop gives the volume flow rate of chilled water in the com- V?pt  V?bg
V?
pt
V?
pt
7.48 CHAPTER SEVEN
FIGURE 7.20 (Continued)
mon pipe , that is, . At part-load operation, there is always a bypass flow of
chilled water from the plant loop returning to the chiller via the common pipe.
10. During part-load operation, the set point of the chilled water leaving temperature is often reset
to a higher value according to either the outdoor temperature or the reduction of system load.
11. During part-load operation, as the two-way control valves close, the chilled water pressure in
the supply main of the building loop tends to increase. The pressure-differential transmitter PD1
senses this increase and reduces the speed of the variable-speed building pump by means of a
variable-speed drive to maintain a constant 15-ft (4.5-m) pressure differential between supply and
return mains. When the system load increases, the two-way control valves open wider and PD1
senses the drop of the pressure differential, increases the speed of the pump, and still maintains a
required 15-ft (4.5-m) pressure differential.
12. When the water system is shut down, the system controller should be in the off position.
First, the compressor(s) are turned off, then the variable-speed building pump is off, condenser
pump(s) are off, chiller pump(s) are off, and cooling tower fan(s) are off. The speed of the variablespeed
pump is gradually reduced to zero first, and then the pump is turned off.
Low T between Chilled Water Supply and Return Temperatures
Many chilled water systems suffer a lower actual T between chilled water supply and return temperatures
compared to the design value. There is also argument that a primary-secondary control
scheme that depends on system flow to gauge system load is virtually blind to load variation.
First, from Eq. (7.1), T  Qsc /(500 ), a lower T is due to an overestimated coil load (system
load) Qsc, or an underestimated water volume flow rate , or both. Second, a plant-building
(primary-secondary) loop should measure system load Qsc, which is the product T and is not
the only water flow to stage on and off the chillers. Third, the space load and coil load of many
projects are often overestimated, the equipment is oversized, and pump head is often calculated on
the safe side. All these result in a far greater actual flow and cause a low T. Fourth, the cleanliness
of the coils including the air-side cleanliness has an influential effect on low T. Finally, for a
chiller plant equipped with three chillers, if the design T  20°F (11.1°C), when one of the
chillers is operated at 50 percent of the design load, T will be lower, to about 16°F (8.9°C) only. A
chilled water system using a plant-building loop is not an essential factor causing a low T between
the supply and return mains.
Variable-Speed Pumps Connected in Parallel
In a chilled water system, if two or more variable-speed pumps are connected in parallel, all the
variable-speed pumps must generate the same head. The purpose of variable-speed pumps connected
in parallel is to increase the volume flow rate. Their flow is additive.
For identical variable-speed pumps connected in parallel, the best overall efficiency is often
obtained if the pumps are operated at identical speeds. Parallel-connected variable-speed pumps
should be reduced or increased to approximately the same speed. If two parallel-connected identical
variable-speed pumps are operated at different speeds, or a large pump with higher head is connected
to a small pump with lower head, then it is possible that the lower-speed pump or pump with
lower head sometimes may contribute negative effects since they must operate at the same head.
Different speed pumps or different sized pumps are hardly operated at higher efficiency at the same
time. The performance of two or more variable-speed pump-piping systems connected in parallel is
further complicated in that only variable-speed pumps are connected in parallel.
V?
gal
V?
gal
V?
gal
V?
cn  V?pt  V?bg V?cn
WATER SYSTEMS 7.49
Use of Balancing Valves
Equal-percentage two-way control valves are widely used to modulate the flow rate of chilled water
flowing through the coils during part-load operation. Because of the lower installation cost and
since often there is only limited space available inside the ceiling plenum, the direct-return piping
system is often the best choice for the chilled water system in a multistory building. For a variable-
flow building loop using a direct-return piping arrangement, the argument concerns whether a balance
valve is necessary for each branch pipe to balance the water flow according to its requirement,
such as for branches CK, DJ, EH, and FG, as shown in Fig. 7.20a. If there are no balancing valves
installed in the branch pipes, after the control valve balances the water flow at design load, can it
still effectively adjust the amount of chilled water entering the coil as required in part-load operation.
This depends mainly on the type of control valve, the control mode adopted, the variation in
pressure drop between various branches, and the difference in main pipe pressure drop between the
farthest and the nearest branch regardless of whether equal-percentage two-way control valve with
modulation control (such as proportional plus integral control ) or two-way control valve with twoposition
on/off control (including pulse-width-adjusted two-position control) is used.
For an equal-percentage two-way control valve with modulation control for many AHUs, consider
a plant-building loop in a chilled water system, as shown in Fig. 7.20a. At the design load, the
chilled water volume flow rate through branch FG is 80 gpm (0.30 m3/min), the corresponding
pressure drop of its fully opened equal-percentage two-way valve is 7.5 ft WC (3.3 psi or 2.3 m
WC), and the pressure drop across the farthest branch FG is 20 ft WC (8.7 psi or 6.1 m WC). From
Eq. (5.8), the flow coefficient
Usually, the difference between the pressure drop across the farthest branch from the building pump
FG and the pressure drop across the nearest branch CK is often within 60 ft WC (26 psi or 18 m
WC). At design load, for a fully opened two-way control valve in branch CK, even if all this 26 psi
(18 m head loss) has been added, the chilled water volume flow rate is then
From Fig. 5.16, for a typical equal-percentage two-way valve, when the percentage of the valve
stem travel lies between 80 and 100 percent, the relationship between the percentage of full-range
travel of valve stem z and the percentage of water flow rate when the valve is fully opened is
 Kekz  0.004e5.5 (7.17)
where k  proportional constant
K  flow parameter affected by size of valve
Since 80/238  0.336, substituting into Eq. (7.17) gives
0.336  0.004e5.5z
And the percentage of full-range travel of valve stem z  0.805. That is, an equal-percentage twoway
control valve in branch CK will close its opening from 100 percent fully open to 80.5 percent
for water flow balance at design load. Avery et al. (1990) emphasized that “If properly selected
valves (those with equal percentage ports and with the correct actuators) are used, 20 percent or less
of the stroke will be used to balance the flow. The rest of the stroke will still be available to modulate
the flow within the design limits.” Rishel (1997) also stated that manual balance valves and automatic
balance valves should not be used on variable-volume, direct-return, modulating type, coil
control valve, chilled water systems.
Therefore, an equal-percentage, two-way control valve, direct-return VAV system can balance
the water flow in a direct-return chilled water system, and at least 80 percent of its stroke is still
V?
v
V?
v
V?
 Cv?pvv  44?3.3 	 26  238 gpm (15 L / s or 0.90 m3 / min)
Cv 
V?
?pvv

80
?3.3
 44
7.50 CHAPTER SEVEN
available to perform the control actions at part-load if the variations in pressure drop of various
branches are small and the difference in main pipe pressure drop between the farthest and nearest
branches is within 60 ft WC (18 m WC).
Common Pipe and Thermal Contamination
For a plant-building loop in a chilled water system, if there is a backflow of a portion of the return
chilled water from the building loop that enters the common pipe at junction L and is mixed with
the supply chilled water from the plant loop at junction B, as shown in Fig. 7.20a, then the thermal
contamination of building return chilled water occurs. Lizardos (1995) suggested that the length
of the common pipe (expressed as the number of diameters of the common pipe) should be as
follows:
The diameter of the common pipe should be equal to or greater than the diameter of the return main
of the building loop.
Example 7.1. Consider a chilled water system using a plant-building loop. The design chilled water
flow rate is 1000 gpm (63.1 L/s) with a chilled water temperature rise across the coils of
Tw,c  20°F (11.1°C). There are three chillers in the central plant, each equipped with a constantspeed
chiller pump that provides 350 gpm (22 L/ s) at 40-ft (12-m) total head. In the building loop,
there are two variable-speed building pumps, each with a volume flow rate of 1000 gpm (63.1 L/ s)
at 60-ft (18-m) head. One of these building pumps is a standby pump. At design conditions, chilled
water leaves the chiller at a temperature Tel  40°F (4.4°C) and returns to the chiller at 60°F
(15.6°C). Chilled water leaving the chiller is controlled at 40°F (4.4°C) for both design and partload
operation. Once the fouling and inefficiency of the coils have been taken into account, the fractions
of design volume flow rate required to absorb the coil load at various fractions of the
sensible load Qcs /Qcs,d are listed in Table 7.8.
When the system load drops, plant chiller 1 will turn off when Qcs /Qcs.d equals 0.65 and chiller 2
will turn off when Qcs /Qcs,d equals 0.30. When the system load increases, chiller 2 turns on at
Qcs /Qcs,d  0.35, and chiller 1 turns on at Qcs /Qcs,d  0.7. Plant chiller 3 operates continuously.
Calculate the following based on the chillers’ on/off schedule at various fractions of the sensible
coil load:
1. Mean chilled water temperature rise across the coil
2. Water flow in the common pipe
3. Temperature of water returning to the water chiller
Solution
1. From the given information in Table 7.8, and Eq. (7.1), the mean water temperature rise across
the coil for Qs,c /Qsc,d  0.9 is
Values of Tw,c at other values Qsc /Qsc,d can be similarly calculated and are listed in Table 7.8.
Tw,c  20QcsV?bg,d / Qcs,dV?bg 
20  0.9
0.8
 22.5F (12.5C)
V?
bg / V?bg,d
Chilled water velocity in return main, ft / s (m/ s) Minimum length of common pipe
 5 (1.5) 3 diameters, or  2 ft (0.6 m)
 5 (1.5) 10 diameters
WATER SYSTEMS 7.51
2. For Qcs /Qcs,d  0.9, because all three chillers are operating, the water volume flow rate in the
plant loop  3(350)  1050 gpm. The water volume flow rate in the building loop is
Then, the water volume flow rate in the common pipe is
When Qcs /Qcs,d  0.65, chiller 1 is turned off. Just before the chiller is turned off, the volume
flow rate in the common pipe is
 1050  0.38(1000)  670 gpm (42.3 L/ s)
Immediately after chiller 1 is turned off
 2(350)  0.38(1000)  320 gpm (20.2 L/ s)
Values of for other values of Qcs /Qcs,d can be similarly calculated and are listed in Table 7.8.
3. For Qcs /Qcs,d  0.9, after the adiabatic mixing of water from the building loop and common
pipe, the temperature of chilled water returning to the water chiller can be calculated as
For Qcs /Qcs,d  0.65, just before chiller 1 is turned off,
Immediately after chiller 1 is turned off,
Other chilled water temperatures upon entering the chiller can be similarly calculated and are
listed in Table 7.8.
7.12 PLANT-DISTRIBUTED PUMPING
A water system using plant-distributed pumping consists of two loops: a plant loop and a distributed
pumping loop connected by a bypass (common pipe), as shown in Fig. 7.21. As in the plantbuilding
loop, the plant loop comprises chiller(s) / boiler(s), constant-speed plant pumps, piping, and
controls. A constant flow is maintained in the evaporator of each chiller / boiler.
In each of the distributed pumping loops connected to an associated air-handling unit, there is a
corresponding variable-speed distributed pump, a coil, two isolating valves, a drain valve, and other
accessories. The discharge air temperature of the AHU is controlled by the DDC unit through the
modulation of the water flow rate by a variable-speed drive (VSD) and the asssociated variablespeed
pump. There is no two-way control valve, and no pressure-differential transmitter is equipped
to maintain a fixed pressure differential between the supply and return mains at the farthest branch.
In summer, as the AHU is operated at part-load operation for cooling, if a temperature sensor
senses the discharge air temperature drops below the preset limit, a DDC unit controller modulates
Tee 
350
700 (40) 	
0.35(1000)(40 	 37.1)
700
 58.6F (14.8C)
Tee 
700
1050 (40) 	
0.35(1000)(40 	 34.2)
1050
 51.4F (10.8C)
Tee 
250
1050 (40) 	
0.8(1000)(40 	 22.5)
1050
 57.1F (13.9C)
V?
cn
V?
cn
V?
cn
V?
cn  V?pt  V?bg  1050  800  250 gpm (15.8 L / s)
V?
bg 
1000V?bg
V?
bg,d
 1000(0.8)  800 gpm (50.5 L / s)
V?
pt
7.52 CHAPTER SEVEN
the VSD and the associated variable-speed pump to reduce the amount of chilled water flowing
through the coil of the AHU, to maintain an approximately constant discharge air temperature. In
the plant-distributed pumping loop, a portion of chilled water supply from the plant loop will return
to the chiller by means of the bypass. During part-load operation for winter heating, the discharge
air temperature increases, and the controller modulates the VSD and the variable-speed pump to reduce
the amount of hot water flowing through the coil of the AHU.
Compared to a plant-building loop, a plant-distributed pumping loop has the following advantages:
 A variable-speed pump replaces the two-way control valve and balancing valve.
 No pressure-differential transmitter and control is required.
 When a distributed pumping loop is nearer to the plant (chiller/boiler), less distributing energy to
transport water in the mains is needed.
The disadvantages include higher first cost and that more maintenance is required.
Distributed pumping is suitable to apply for the water system that serves large coils in AHUs
and where the AHU must be installed inside a fan room to avoid inconvenient maintenance and
noise.
7.13 CAMPUS-TYPE WATER SYSTEMS
Chilled water or chilled and hot water is often supplied to many buildings separated from a central
plant in universities, medical centers, and airports. The benefits of using a campus-type central
plant chilled water system instead of individual building installations are cost savings, minimal
environmental impact (e.g., from cooling towers), effective operation and maintenance, and
reliability.
WATER SYSTEMS 7.53
Constantspeed
pumps
Chiller
Boiler
C
pCK
pbranch
F
K
K
G
1
1
Chiller
Boiler
Variable-speed
pumps Supply
main
Return
main
VSD VSD VSD
AHUn
VSD
AHU3
AHU2
AHU1
Bypass
FIGURE 7.21 Schematic diagram of a plant-distributed pumping loop .
The following are three types of currently used campus-type water systems: plant-distributionbuilding
loop, plant-distributed building loop, and multiple sources-distributed building loop.
Plant-Distribution-Building Loop
System Description. Many recently developed campus-type central plant chilled water systems
use a plant-distribution-building loop, as shown in Fig. 7.22a. As in a plant-building loop, constant
flow is maintained in the evaporator of each chiller in the plant loop. Each chiller also has its own
constant-speed chiller pump. Chilled water leaves the chiller at a temperature of 40 to 42°F (4.4 to
5.6°C). It is then extracted by the distribution pumps and forced to the supply main of the distribution
loop.
Chilled water in the supply and return mains of the distribution loop operates under variable
flow. Multiple variable-speed pumps are often used to transport chilled water at a volume flow rate
slightly higher than the sum of the volume flow rates required in the building loops. At each building
entrance, the variable-speed building pump extracts the chilled water and supplies it to the coils
in AHUs and terminals in various zones by means of building supply mains. Chilled water is then
returned to the water chillers through building return mains, a distribution-loop return main, and
chiller pumps.
The system performances of the plant loop and building loop are similar to those in a plantbuilding
loop.
Pressure Gradient of Distribution Loop. A campus-type chilled water central plant may transport
several thousand gallons of water per minute to the farthest building at a distance that may be several
thousand feet away from the plant. The pressure gradient of the distribution supply and return
mains due to the pipe friction and fitting losses causes uneven pressure differentials among the supply
and return mains [Hs,r (ft WC or m WC)] of buildings along the distribution loop, as shown in
Fig. 7.22b. Buildings nearer to the central plant have a greater Hs,r than buildings farther from the
plant. Along the distribution loop, a smaller pressure gradient results in a lower pumping power but
a larger diameter of chilled water pipe. A more even Hs,r does not impose excessive pressure drop
across the control valves of coils.
Using a lower pressure drop Hf is an effective means of reducing the pressure gradient and
pressure differential Hs,r and saves energy. For a distribution loop, a value of Hf between 0.5 and
1 ft /100 ft (0.5 and 1 m/100 m) pipe length, sometimes even lower, may be used. Low values of
Hs,r can be offset at the coil control valves without affecting the coil’s proper operation when there
is no building pump in the building loop. A life-cycle cost analysis should be conducted to determine
the optimum Hf .
Using two-way distribution from the central plant, with two supply and return distribution loops,
may reduce the pipe distance and the diameter of the supply and return mains. Such a distribution
loop layout depends on the location of the central plant and the air conditioned buildings as well as
the cost analyses of various alternatives.
Using a reverse-return piping arrangement instead of a direct-return one does even the pressure
differentials Hs,r along the supply and return mains of the distribution loop. However, having an
additional pipe length equal to that of the return main significantly increases the piping investment.
A simpler and cheaper way is to install a pressure throttling valve at each building entrance to offset
the excess Hs,r. Usually, direct return is used for a distribution loop.
Variable-Speed Building Pumps. The function of a variable-speed building pump is (1) to provide
variable-flow and corresponding head to overcome the pressure drop of the building loop at
design and reduced coil loads and (2) to provide different magnitudes of head for the building loop
according to the needs of various types of buildings. When only variable-speed distribution
pump(s) are used instead of both variable-speed distribution pump(s) and variable-speed building
pump(s), it must provide sufficient pump head to overcome the pressure drop of the building
7.54 CHAPTER SEVEN
WATER SYSTEMS 7.55
FIGURE 7.22 Chilled water system using plant-distribution-building loop. (a) Schematic diagram; (b) pressure gradient for distribution
loop; (c) crossover bridge with temperature control valve.
loops. However, the pressure characteristics of the supply and return mains of the distribution loop at
reduced flows make it difficult to satisfy various load profiles in different buildings during part-load
operation. Having a variable-speed building pump for each building also saves more pump energy.
Therefore, the use of variable-speed pumps for both distribution and building loops is preferable.
Control of Variable-Speed Distribution Pump. Two types of controls can be used to modulate a
variable-speed distribution pump to transport the required chilled water volume flow to various
buildings:
 A differential-pressure transmitter may be located near the farthest end of the distribution supply
main, as shown in Fig. 7.22a. Theoretically, the head of the building pump should extract the exact
required amount of chilled water corresponding to the sum of the coil loads in the building
loop, force it through the coils, and discharge it to the distribution return main. Therefore, a set
point for the pressure differential of about 5 ft (1.5 m) may be appropriate. This type of control is
widely used.
 A DDC system measures the total water flow that returns from each building by means of
flowmeters and modulates the variable-speed distribution pump to supply exactly the required
amount to various building loops. This type of control is more precise but more expensive.
Building Entrance. Chilled water is usually supplied directly from the distribution supply main
to the building supply main. A pressure throttling valve may be used to offset the excess pressure
differential Hs,r along the distribution loop.
Although using a heat exchanger at the building entrance entirely isolates the chilled water in
the distribution loop from the chilled water in the building loop, a temperature increase of about 3
to 7°F (1.7 to 3.9°C) is required for a chilled water heat exchanger. Because chilled water has a supply
and return temperature differential of only about 15 to 20°F (8.3 to 11.1°C), a heat exchanger is
seldom used at a building or zone entrance for a chilled water system. Because a hot water system
has a greater supply and return temperature differential, a heat exchanger is sometimes used at the
building entrance for a hot water system.
A chilled water building loop may be divided into various zones based on different height levels
within the building. In such an arrangement, the coils in the lower floors of the building loop will
not suffer a high static pressure because the low-level water loop is often isolated from the highlevel
water loop by means of a heat exchanger at the zone entrance.
If a building requires a chilled water supply temperature higher than that given by the distribution
supply main, a crossover bridge with a temperature control valve can be arranged for this purpose,
as shown in Fig. 7.22c.
The return temperature from the coils in a building loop is affected by the cleanliness of the coil,
including air-side cleanliness, and the control system in the building loop. The building’s variablespeed
pump is often controlled by the pressure-differential transmitter located at the end of the
building supply main, as shown in Fig. 7.22a and c.
Plant-Distributed Building Loop
A plant-distributed building loop water system has nearly the same configuration as a plant-distribution-
building loop system except that there is no distribution pump in the distribution loop. Constant-
speed chiller /boiler pumps in the plant loop supply water to the beginning of the supply main
of the distribution loop, point S, and extract water from the end of the return main of the distribution
loop, point R. Various variable-speed building pumps in the building loops extract water from
the distribution main point S. They also overcome the pressure loss of the distribution supply main
piping up to the building entrance, the pressure loss of the building loop pbg including building
supply and return mains, coils, two-way control valve, and fittings; and the pressure loss of the distribution
return main piping from the building outlet to point R.
7.56 CHAPTER SEVEN
Compared to a plant-distribution-building loop, a plant-distributed building loop has the following
advantages:
 It saves the installation cost of variable-speed distribution pumps, related controls, and pump
room space.
 If the pressure differential between the distribution supply and return mains at the building entrance
Hs,r is greater than the pressure loss of the building loop pbg either at design load or at
part-load, then a plant-distributed building loop can save more pumping energy.
On the other hand, a plant-distributed building loop requires variable-speed building pumps of
higher head which means more attention to pump noise control in the building’s mechanical room.
Multiple Sources-Distributed Building Loop
Many conversions and retrofits of campus-type chilled water systems require existing chilled water
plants in addition to the developed central plant. In these cases, there are buildings with chilled water
sources (chillers); buildings with chilled water coils and loads; and buildings with chilled water
sources and loads connected to the same plant-distributed building loop, as shown in Fig. 7.23a. A
WATER SYSTEMS 7.57
FIGURE 7.23 A multiple sources-distributed building loop. (a) Schematic diagram; (b) building with sources and loads.
DDC system controller is used for each type of building. There is a central microprocessor for the
whole chilled water system. The optimization program may proceed as follows:
1. Measure the chilled water temperature across each load and source, as well as its rate of water
flow.
2. Add all the loads.
3. Turn on the most efficient source first, including the demand and downtime, according to available
sources. A chiller’s running capacity should match the required loads. There should be a
time delay to start or stop the chiller.
4. Use trending and expert system control strategies to predict load changes according to past experience
and outdoor conditions.
Chilled and Hot Water Distribution Pipes
Chilled and hot water distribution pipes are large pipes mounted in underground accessible tunnels
or trenches. They are well insulated, although underground return chilled water mains for which return
temperature Tret  60°F (15.6°C) may not be insulated, depending on a detailed cost analysis.
Factory-made conduits consist of inner steel pipe, insulation, airspace, and outer conduit; or steel
pipe, with and without insulation, and outer casing. Expansion loops or couplings should be included,
and a good drainage system is important to protect the insulating quality. Please refer to
ASHRAE Handbook 1996, HVAC Systems and Equipment, chapter 11, for details on design and installation.
7.14 COMPUTER-AIDED PIPING DESIGN AND DRAFTING
General Information
According to “Selecting Piping System Software” by Amistadi (1994), three currently widely used
piping design and drafting computer programs were reviewed: University of Kentucky’s
KYCAD/KYPIPE, Trane’s Water Piping Design, and Softdesk’s Piping. The Trane and Softdesk
computer programs require AutoCAD as a base product whereas the University of Kentucky software
has its own integrated CAD system.
The University of Kentucky programs have extensive hydraulic modeling capacities including
transient analysis. They are intended for mechanical engineers and are used to design and analyze
large, complex water systems. Trane’s and Softdesk’s software programs are limited to steady-state
imcompressible fluids. They are intended for contractors, mechanical engineers, and drafters. Kentucky’s
and Softdesk’s programs support both inch-pound and metric units, while Trane’s package
supports only inch-pound units. The University of Kentucky software requires 2.0 Mbytes of disk
space and 2.0 Mbytes RAM. Softdesk’s software requires 10 Mbytes disk space and 8 Mbytes
RAM. Trane’s software requires the most disk space—10 Mbytes and 12 Mbytes RAM. The
University of Kentucky program is MS-DOS applications, whereas Trane’s and Softdesk’s are
available for Windows platforms.
Computer-Aided Drafting Capabilities
Trane’s piping software allows designers to create a schematic two-dimensional (2D) piping system
in AutoCAD and to link to the computer programs of piping size calculations for use in piping system
design. Softdesk’s software is intended for drafting in 2D or 3D graphics. Design information
databases and engineering analysis software programs are linked to the drawing for support. The
7.58 CHAPTER SEVEN
University of Kentucky program is aimed for the design and diagnosis of large, complicated water
systems.
Softdesk’s software provides all the graphical components, such as pipes, valves, fittings, equipment,
pumps, tanks, controls, and structural components to form a piping system. Software by
Trane and the University of Kentucky only provides components that are needed for piping sizing
or hydraulic analysis. The University of Kentucky software provides annotation options for all hydraulic
parameters. Softdesk does not offer all, and Trane provides only selections of annotation.
Trane and the University of Kentucky only offer schematic layout and symbols to represent the
piping system and components and provide plan view engineering drawings as the standard option.
Softdesk offers schematic, double-line, and 3D model with wire-frame 3D version of its components.
Softdesk is the only computer program to translate automatically from single-line to 3D
model. The translation is bidirectional. It also allows the translation to go from 3D wire-frame to
double- or single-line. Softdesk offers a full range of engineering drawings including plan, section,
elevation, isometric, and perspective views.
Softdesk’s and University of Kentucky’s software recognize most of the types of layouts, such as series,
parallel, branching, and network. The check of continuity is the basis of the hydraulic calculation,
which begins with the system continuity. The Softdesk software provides the most complete checking
of graphical elements of the piping system including gaps, overlaps, pump location, and size compatibility
of adjacent sections, followed by the University of Kentucky and Trane programs.
Computer-Aided Design Capabilities
System Size. The University of Kentucky software is for large piping systems and supports a system
up to 1000 legs. The Trane and Softdesk software support 400 legs.
Pipe Sizing. Trane’s and Softdesk’s computer programs allow the designer many options based on
maximum head loss, such as 2.5 ft/100 ft (2.5 m/100 m) or velocity for pipe sizing. The University
of Kentucky computer program has extensive constraint capabilities that are linked to meet the
pressure at given node(s). A node is the junction of pipes and is the place where the flow rate
changes.
Pump and System Operations. Trane’s and Softdesk’s programs allow the designer to first set the
system water flow rate and then calculate the pressure drop and the flow of the system components.
The University of Kentucky software is able to determine system operating points for series, parallel,
plant-building loop (primary and secondary), and variable-speed pumping applications by
means of pump and system curves. The University of Kentucky software also has the capabilities to
set control and pressure-regulating valves, or to locate check valves which affect the system hydraulic
calculations.
Pressure Losses and Network Technique. All three programs use the Darcy-Weisbach equation
with empirical fits to the Moody diagram to calculate pipe frictional losses. Unique pipe roughness
is used in Trane’s and Softdesk’s calculations, but it can be varied for each pipe in the University of
Kentucky software. Trane’s and Softdesk’s software uses equivalent length and Cv to calculate dynamic
losses for pipe fittings and control valves, whereas the University of Kentucky software
adopts the local loss coefficient k method. Trane’s and Softdesk’s software read the drawing and automatically
places the node at points where there is a change in flow. The University of Kentucky
software expects the designer to input the nodes as fittings. All automatically number the nodes and
edit them if necessary.
Regarding piping network technique, Trane’s and Softdesk’s software programs adopt sequential
method, a once-through stepwise approach, and use arithmetic sum of flow in parallel circuits to
determine the combined flow and the component head loss. The University of Kentucky program
uses the simultaneous method by solving simultaneous algebraic-equations through successive
approximation.
WATER SYSTEMS 7.59
Input Data and Reports. Trane’s software relies on the AutoCAD attribute functions. Its data are
stored with the drawing data. Softdesk’s software uses extended entity capabilities, and data are
stored in external AutoCAD’s drawing exchange format (DXF) files. The University of Kentucky
software has a dedicated CAD system. Trane’s and Softdesk’s software programs let the designer
select a kind of fluid and temperature, and the computer program calculates the fluid properties. The
University of Kentucky software requires the designer to input the data each time. Trane’s and
Softdesk’s databases include size, cost, and hydraulic data, whereas the University of Kentucky
only includes hydraulic data.
All three computer programs provide tabular reports of pipe diameter, length, flow, velocity, and
head. Trane’s software identifies the critical path, and Trane’s and Softdesk’s software programs
offer quantity and cost bill of material part, whereas the University of Kentucky software offers
multiple design condition reports, such as cavitation and metering reports.
REFERENCES
Ahmed, O., Life-Cycle Cost Analysis of Variable-Speed Pumping for Coils Application, ASHRAE Transactions,
1988, Part I, pp. 194–211.
Amistadi, H., Selecting Piping System Software, Engineered Systems, no. 6, 1994, pp. 57–62.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
Avery, G., Microprocessor Control for Large Chilled Water Distribution Systems, Heating/Piping/Air Conditioning,
October 1987, pp. 59–61.
Avery, G., Stethem,W. C., Coad,W. J., Hegberg, R. A., Brown, F. L., and Petitjean, R., The Pros and Cons of
Balancing a Variable Flow Water System, ASHRAE Journal, no.10, 1990, pp. 30–55.
Ball, E. F., and Webster, C. J. D., Some Measurements of Water Flow Noise in Copper and ABS Pipes with
Various Flow Velocities, The Building Services Engineer, May 1976, pp. 33–40.
Binkowski, R. O.,Water Treatment for HVAC Systems,/Heating/Piping/Air Conditioning, October 1989, pp.
131–133.
Braun, J. E., Klein, S. A., Mitcell, J. W., and Beckman,W. A., Applications of Optimal Control to Chilled Water
Systems without Storage, ASHRAE Transactions, 1989, Part I, pp. 663–675.
Burr, G. C., and Pate, M. E., Conversion of Campus Central Plant from Constant Flow to Variable Flow at
University of West Florida, ASHRAE Transactions, 1984, Part I B, pp. 891–901.
Carlson, G. F., Hydronic Systems: Analysis and Evaluation—Part I, ASHRAE Journal, October 1968, pp.
2–11.
Carlson, G. F., Central Plant Chilled Water Systems—Pumping and Flow Balance Part I, ASHRAE Journal,
February 1972, pp. 27–34.
Coad,W. J., Centrifugal Pumps: Construction and Application, Heating/Piping/Air Conditioning, September
1981, pp. 124–129.
Ellis, R., and McKew, Howard, Back to Basics: Test 9—Chilled Water System Using Centrifugal Chiller
Advanced Energy Efficient Design, Engineered Systems, no. 11, 1996, p. 11.
Eppelheimer, D. M., Variable Flow—The Quest for System Energy Efficiency, ASHRAE Transactions, 1996,
Part II, pp. 673–678.
Griffith, D., Distribution Problems in Central Plant Systems, Heating/Piping/Air Conditioning, November 1987,
pp. 59–76.
Haines, R. W., Bahnfleth, D. R., Luther, K. R., Landman,W. J., and Kirsner,W., Open for Discussion: Primary-
Secondary Pumping, HPAC, no. 3, 1997, pp. 67–73.
Hansen, E. G., Parallel Operation of Variable-Speed Pumps in Chilled Water Systems, ASHRAE Journal, no. 10,
1995, pp. 34–38.
Hull, R. F., Effect of Air on Hydraulic Performance of the HVAC System, ASHRAE Transactions, 1981, Part I,
pp. 1301–1325.
7.60 CHAPTER SEVEN
Kelly, D. W., and Chan, T., Optimizing Chilled Water Plants, HPAC, no. 1, 1999, pp. 145–147.
Lizardos, E. J., Engineering Primary-Secondary Chilled-Water Systems, Engineered Systems, no. 8, 1995, pp.
30–34.
MacDonald, K. T., Valves: An Introduction, Heating/Piping/Air Conditioning, October 1988, pp. 109–117.
Mannion, G. F., High Temperature Rise Piping Design for Variable Volume Systems: Key to Chiller Energy
Management, ASHRAE Transactions, 1988, Part II, pp. 1427–1443.
Miller, R. H., Valves: Selection, Specification, and Application, Heating/Piping/Air Conditioning, October
1983, pp. 99–118.
Ocejo, J., Program Estimates Expansion Tank Requirements, Heating/Piping/Air Conditioning, November
1986, pp. 89–93.
Peterson, P. A., Medical Center Expands Utilities Distribution, Heating/Piping/Air Conditioning, May 1985,
pp. 84–96.
Pompei, F., Air in Hydronic Systems: How Henry’s Law Tells Us What Happens, ASHRAE Transactions, 1981,
Part I, pp. 1326–1342.
Prescher, R., Hydronic System Design Guidelines, Heating/Piping/Air Conditioning, May 1986, pp. 132–134.
Redden, G. H., Effect of Variable Flow on Centrifugal Chiller Performance, ASHRAE Transactions , 1996, Part
II, pp. 684–687.
Rishel, J. B., Energy Conservation in Hot and Chilled Water Systems, ASHRAE Transactions, 1983, Part II B,
pp. 352–367.
Rishel, J. B., Twenty Years’ Experience with Variable Speed Pumps on Hot and Chilled Water Systems,
ASHRAE Transactions, 1988, Part I, pp. 1444–1457.
Rishel, J. B., Distributed Pumping for Chilled- and Hot-Water Systems, ASHRAE Transactions, 1994, Part I,
pp. 1521–1527.
Rishel, J. B., Use of Balance Valves in Chilled Water Systems, ASHRAE Journal, no. 6, 1997, pp. 45–51.
Scientific Computing, Software Review: Up for Review (Again), Engineered Systems, no. 1, 1998, pp. 76–84.
Solden, H. M., and Siegel, E. J., The Trend toward Increased Velocities in Central Station Steam and Water
Piping, Proceedings of the American Power Conference, vol. 26, 1964.
Stewart,W. E., Jr., and Dona, C. L.,Water Flow Rate Limitations, ASHRAE Transactions, 1987, Part II, pp.
811–825.
Uglietto, S. R., District Heating and Cooling Conversion of Buildings, ASHRAE Transactions, 1987, Part II,
pp. 2096–2106.
Utesch, A. L., Variable Speed CW Booster Pumping, Heating/Piping/Air Conditioning, May 1989, pp. 49–58.
Waller, B., Piping—From the Beginning, Heating/Piping/Air Conditioning, October 1990, pp. 51–71.
Wilkins, C., NPSH and Pump Selection:Two Practical Examples, Heating/Piping/Air Conditioning, October
1988, pp. 55–58.
Zell, B. P., Design and Evaluation of Variable Speed Pumping Systems, ASHRAE Transactions, 1985, Part I B,
pp. 214–223.
WATER SYSTEMS 7.61
CHAPTER 8
HEATING SYSTEMS,
FURNACES, AND BOILERS
8.1
8.1 HEATING SYSTEMS 8.1
Selection of a Heating System 8.2
8.2 WARM AIR FURNACES 8.3
Types of Warm Air Furnace 8.3
Upflow Gas-Fired Furnace 8.3
Horizontal Gas-Fired Furnace 8.6
Furnace Performance Factors 8.6
Saving Energy 8.7
Control and Operation 8.8
8.3 HOT WATER BOILERS 8.9
Selection of Fuel 8.9
Types of Hot Water Boiler 8.10
Fire-Tube Boiler 8.10
Scotch Marine Packaged Boiler 8.10
Cast-Iron Sectional Boiler 8.12
Gas and Oil Burners 8.13
Condensing and Noncondensing
Boilers 8.13
Boiler Efficiency 8.13
Chimney, or Stack 8.14
Operation and Safety Controls 8.14
8.4 ELECTRIC FURNACES, HEATERS, AND
BOILERS 8.15
Electric Heating Fundamentals 8.15
Electric Furnaces, Electric Heaters, and
Duct Heaters 8.16
Electric Hot Water Boilers 8.17
8.5 LOW-PRESSURE DUCTED WARM AIR
HEATING SYSTEMS 8.17
System Characteristics 8.17
Types of Low-Pressure Ducted Warm Air
Heating System 8.18
Heat Supplied to Conditioned
Space 8.18
Duct Efficiency and System Efficiency for
Heating 8.20
Location of Furnace and Duct
Insulation 8.20
Duct Leakage 8.20
Thermal Stratification 8.21
Part-Load Operation and Control 8.21
8.6 DUCTED WARM AIR HEATING
SYSTEMS 8.22
8.7 HOT WATER HEATING SYSTEMS
USING FINNED-TUBE HEATERS 8.23
Types of Hot Water Heating System 8.23
Two-Pipe Individual Loop System 8.23
Finned-Tube Heaters 8.24
Design Considerations 8.25
Part-Load Operation and Control 8.26
8.8 HYDRONIC RADIANT FLOOR-PANEL
HEATING SYSTEMS 8.27
System Description 8.27
Radiant Floor Panel 8.27
Thermal Characteristics of
Floor Panel 8.28
Design Nomograph 8.30
Design Considerations 8.30
Control and Operations of Multizone
Hydronic Radiant Floor-Panel
Heating System 8.30
System Characteristics and
Applications 8.31
8.9 INFRARED HEATING 8.31
Basics 8.31
Gas Infrared Heaters 8.32
Electric Infrared Heaters 8.32
Design and Layout 8.33
REFERENCES 8.35
8.1 HEATING SYSTEMS
According to the data in DOE/EIA Commercial Buildings Consumption and Expenditures 1995 and
Household Energy Consumption and Expenditures 1993 for the 54.3 billion ft2 (5.05 billion m2)
of heated commercial buildings and the 96.6 million households in the United States, the percentage
of use of various types of heating systems may be estimated as follows:
In commercial buildings, the use of a boiler as a primary heat source had a share of about 28 percent
to supply hot water or steam to the heating coils, radiators, baseboard heaters, and heating panels.
Heating systems in new constructed commercial projects that use boilers are mainly hot water
systems because these are energy-efficient, simpler to operate, and easier to maintain than steam
heating systems. In new constructed residential buildings, the heat sources are mainly gas-fired warm
air furnaces and heat pumps.
Heating systems that use finned coils to heat the air, including water heating coils, electric heating
coils, steam coils, or condensing heating coils of heat pumps in air-handling units and packaged
units, are convective or warm air heating systems. Heating systems that use high-temperature
infrared heaters, radiant panels, or radiators in which emited radiant heat exceeds the released convective
heat are radiant heating systems.
Selection of a Heating System
Several factors must be considered prior to selection of a suitable heating system. They include the
following:
 Whether it is a separate heating system or part of the air conditioning system
 An open or enclosed space, or a space with high infiltration
 Size of heating system (small, medium, or large)
 Available existing heat source, such as hot water or steam
 Cost of gas, oil, or electricity
 Design criteria and local customs
In both commercial and residential buildings, a warm air furnace is generally linked with
unitary packaged air conditioning systems and, therefore, packaged units. In buildings with
rooftop packaged units, a warm air heating system using a warm air furnace is often the
most direct, economical, and suitable choice. In locations where the outdoor climate is mild
in winter, the heat pump is often a convenient and energy-efficient heatsource. For buildings
with central hydronic air conditioning systems, a hot water heating system using a boiler is
often a suitable choice because the central primary plant and boilers are usually far away from
the air-handling units and conditioned space. According to DOE/EIA 0318 (1998), heat
pumps consumed the least annual energy use, district heating and boilers the most, the
“others in between. ASHRAE/IESNA Standard 90.1-1999 specifies that radiant heating shall
be used when heating is required for unenclosed spaces except loading docks equipped with air
curtains.
Commercial, percent Residential, percent
Heating systems using boilers 28 33
Warm air furnaces and packaged heating units 24 37
Heat pumps 10 30
Individual space heaters including electric, gas,
and radiant heaters 28
District heating 10
8.2 CHAPTER EIGHT
8.2 WARM AIR FURNACES
Types of Warm Air Furnace
A warm air furnace is a combustion and heating device in which gas or oil is directly fired to heat
the air through a heat exchanger, or air is directly heated by electric resistance elements in order to
supply warm air to the conditioned space. Warm air furnaces can be classified into various categories
according to
 Fuel types: natural gas, liquefied petroleum gas (LPG), oil, electric energy, or wood
 Airflow directions: upflow, horizontal, or downflow
 Applications: residential, commercial, or industrial
Natural gas is the primary fuel used in warm air furnaces. Warm air furnaces used for residences
usually have a heating capacity up to 175,000 Btu/h (51.3 kW). Upflow models are the most popular
models in residences. For commercial applications, heating capacities are usually greater than
150,000 Btu/h (44 kW). Horizontal models mounted inside a rooftop packaged unit or packaged
heat pump for supplementary heating are widely used.
Upflow Gas-Fired Furnace
An upflow natural gas-fired warm air furnace consists of: a one or more gas burners, a heat exchanger,
a forced-draft circulating fan or blower, a venting system, a filter, and an outer casing, as
shown in Figs. 8.1a and b.
Gas Burners. Gas burners in small furnaces in residences are often atmospheric burners or fanassisted
burners. An atmospheric burner consists of an air shutter, a gas orifice, and outlet ports and
is usually die-formed and made of aluminum painted, heavy-gauge steel or aluminized steel (or
sometimes stainless steel). Atmospheric burners are either in-shot or upshot and are installed in
single or multiple ports. In-shot burners are installed horizontally and are also used for Scotch marine
boilers. Upshot burners are placed vertically and are suitable for vertical fire-tube boilers. Both
Scotch marine and fire-tube boilers are discussed in later sections. Atmospheric burners are simple,
require only a minimal draft of air, and maintain sufficient gas pressure for normal functioning. A
fan-assisted gas burner uses a small fan to induce the combustion air through the heat exchanger
and forces it to the outdoors via an air vent.
A power burner that uses a fan to supply and control combustion air is often employed for large
furnaces. Conversion burners are integrated with furnaces for safety and efficiency. Older gas furnaces
often use conversion burners.
Ignition. The ignition device is often a standing pilot ignition. These pilots are small. The pilot’s
flame is monitored by a sensor that shuts off the gas supply if the flame is extinguished. Another
type of ignition is called spark ignition. It ignites intermittently only when required. Spark ignition
saves more gas fuel than a standing pilot if the furnace is not operating.
Heat Exchangers. They are in the shape of clamshell, bent tube, or crossflow plate type. The
combustion air is burned and flows inside clamshells, or tubes. Air to be heated flows over the outer
surface of the heat exchanger. A heat exchanger is usually made of aluminum painted heavy-gauge
steel, aluminized steel, or stainless steel.
Circulating Fan or Blower. A circulating fan or blower is always installed in a warm air furnace
to force the air to flow over the heat exchanger, to distribute it to the conditioned space (except
HEATING SYSTEMS, FURNACES, AND BOILERS 8.3
when the warm air furnace is a part of an air-handling unit or packaged unit in which a supply
fan is always provided), and to extract warm space air for heating again. A centrifugal fan with
forward-curved blades and a double inlet intake is usually used.
Filter. A disposable filter, which is discussed in Chap. 17, is used to remove dust from the recirculating
air. It is often located upstream of the fan.
Venting Arrangements. In a natural-vent warm air furnace, a draft hood is employed to connect
the flue gas exit at the top of the heat exchanger section to a vent pipe or chimney. A relief air opening
is also used to guarantee that the pressure at the flue gas exit is always atmospheric. A draft
hood diverts the backdraft from the chimney, bypassing the burner without affecting the combustion
operation. A direct-vent warm air furnace does not have a draft hood. If the vent pipe or chimney is
blocked, a control system shuts down the warm air furnace. A power vent including a fan-assisted
combustion furnace uses a fan to force or induce the combustion products or flue gas flowing
through the heat exchanger and air vent.
Casing. The outer casing is usually made of heavy-gauge steel with removable access panels.
Furnace Operation. Gas is generally brought from the main to the pressure regulator. The regulator
reduces the gas pressure to about 3.5 in. WG (870 Pa). Gas then flows through a gas valve controlled
by a room thermostat. When a solenoid gas valve opens, the gas flows to the burners and
mixes with the necessary outside primary air for combustion. The primary air–gas mixture then
8.4 CHAPTER EIGHT
FIGURE 8.1 Upflow gas-fired furnaces. (a) Natural vent; (b) fan-assisted combustion.
discharges from the port slots, mixes with ambient secondary air, and is burned. For a natural-vent
furnace, the combustion products flow through the heat exchanger and the draft diverter and
discharge outdoors through the vent pipe or chimney. For a fan-assisted combustion furnace, the
induced-draft fan extracts the combustion products and the flue gas flowing through the heat
exchanger, and forces them to discharge outdoors through the vent pipe or chimney.
A mixture of space recirculating air and outdoor air is pulled by the circulating fan from the
return ducts and enters the bottom inlet of the warm air furnace. This air is then forced through
the heat exchanger for heating again. For a heating-only furnace, the air temperature is raised by
50 to 80°F (28 to 45°C). Warm air discharges from the top outlet and is distributed to various
conditioned spaces through the supply duct and outlets. An upflow gas-fired furnace is usually
installed indoors and is often installed within a vented closet or vented basement in residential
buildings.
Most natural gas furnaces can use liquefied petroleum gas. The main difference is their gas
pressure. Natural gas usually needs a pressure of 3 to 4 in. WG (746 to 994 Pa) at the manifold,
whereas LPG needs a higher pressure of about 10 in. WG (2486 Pa) and more primary air for gas
burners.
HEATING SYSTEMS, FURNACES, AND BOILERS 8.5
Vent
(b)
Induceddraft
assisting
fan
Heat
exchanger
Gas
burner
Circulating fan
Warm air
supply plenum
Vent
FIGURE 8.1 (Continued).
Horizontal Gas-Fired Furnace
A typical horizontal gas-fired furnace in a rooftop packaged unit is shown in Fig. 8.2. It mainly
consists of multiple gas burners, a heat exchanger, a combustion blower, and an ignition device. The
supply fan used to force air to flow through the gas-fired furnace is the same supply fan used to
force air through the filters and DX coil in the rooftop unit.
A power burner is often used in a horizontal gas-fired furnace. This type of burner provides
better combustion and higher efficiency than atmospheric burners. A power draft centrifugal blower
may be added to extract the combustion products and discharge them to a vent or chimney. The gas
supply to the burner is controlled by a pressure regulator and a gas valve for the purpose of controlling
the firing rate. In a premix power burner, gas and primary air are mixed first; then the mixture
is forced to mix with secondary air in the combustion zone. A power burner usually has a higher gas
pressure than the atmospheric burner used in residences.
The heat exchanger usually has a tubular two-pass arrangement, typically with 16-gauge
(1.5-mm-thickness) stainless steel for primary surfaces and 18-gauge (1.2-mm-thickness) stainless
steel for secondary surfaces, as shown in Fig. 8.2. The primary surface is the heat-transfer surface
of the combustion chamber. The secondary surface is the surface of the tubes through which the
flue gas flows after the combustion chamber. A cone-shaped flame is injected into a tubular or
drum-shaped combustion chamber. A centrifugal blower is used to provide secondary air for forced
combustion. Another small centrifugal blower may be used to induce the flue gas at the exit of the
heat exchanger, the power vent, to maintain a negative pressure at the heat exchanger section so as
to prevent the mixing of any leaked flue gas with the heated air. The mixture of outdoor and recirculating
air is forced by the circulating blower and is heated when it flows over the primary and
secondary surfaces.
Furnace Performance Factors
The performance of a gas-fired furnace is usually measured by the following parameters:
 Thermal efficiency Et, in percent, is the ratio of the energy output of the fluid (air or water) to the
fuel input energy. Input and output energy should be expressed in the same units. The value of Et
can be calculated as
(8.1) Et 
100 (fluid energy output)
fuel energy input
8.6 CHAPTER EIGHT
FIGURE 8.2 Horizontal gas-fired furnace in a rooftop packaged unit.
 Annual fuel utilization efficiency (AFUE) is the ratio of annual output energy from air or water to
the annual input energy, expressed in the same units:
(8.2)
AFUE also includes nonheating-season standing-pilot input energy loss. AFUE is similar to thermal
efficiency Et, except that AFUE is the ratio of annual energy output to energy input, whereas Et is
the ratio of energy output to energy input at specific test periods and conditions.
 The steady-state efficiency (SSE) is the efficiency of a given furnace according to an ANSI test
procedure and is calculated as
(8.3)
The steady-state efficiency of gas-fired furnaces varies from 65 to 95 percent.
Test data in Jakob et al. (1986) for ASHRAE Special Project SP43, based on a nighttime setback
period of 8 h and a setback temperature of 10°F (5.6°C), gave the following performance factors for
gas-fired furnaces of two test houses with different construction characteristics:
Saving Energy
Based on the above results, factors that affect the energy saving of the fuel and improve the furnace
performance are as follows:
 Condensing or noncondensing. When the water vapor in the flue gas is condensed by indirect
contact with recirculating air, part of the latent heat released during condensation is absorbed by
the recirculating air, which increases furnace efficiency and saves fuel. The difference in the
values of AFUE between condensing and noncondensing power vent furnaces may be around
10 percent. Corrosion-resistant materials such as stainless steel must be used for condensing heat
exchangers due to the presence of chloride compounds in the condensate.
 Power vent or natural vent. Power vent guarantees the air supply for combustion and also forces
the combustion product flows through heat exchangers including condensing heat exchangers
with higher flow resistance.
 Preheated outdoor combustion air. If combustion air is taken from outdoors and preheated
through an annular pipe that surrounds the flue pipe, part of the heat in the flue gas will be saved.
 Automatic vent damper. An automatic thermal or electric vent damper located in the vent pipe
closes the vent when the furnace is not in operation. This decreases exfiltration from the house
and restricts the amount of heat escaping from the heat exchanger.
Construction characteristics AFUE, percent SSE, percent
Natural vent
Pilot ignition 64.5 77
Intermittent ignition 69 77
Intermittent ignition  venting damp 78 77
Power vent
Noncondensing 81.5 82.5
Condensing 92.5 93
SSE 
100(fuel input  fuel loss)
fuel input
AFUE 
100(annual output energy)
annual input energy
HEATING SYSTEMS, FURNACES, AND BOILERS 8.7
 Intermittent ignition. Intermittent ignition saves more energy than standing-pilot ignition, especially
during times when the furnace is not operating.
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency of warm air furnaces,
warm air duct furnaces, and warm air unit heaters for sizes as follows:
In 1987, the National Appliance Energy Conservation Act (NAECA) set minimum efficiency
standards for gas furnaces at an AFUE of 78 percent, effective January 1, 1992. According to the
venting systems and the steady-state efficiency, ANSI/AGA Standard Z21.47 classifies the central
gas furnaces into the following four categories:
Here Tflue represents the vent gas (flue gas) temperature, and Tf, dew the dew-point temperature of the
flue gas, both in °F.
Paul et al. (1993) reported that a category I fan-assisted gas furnace typically operated at an SSE
of 80 to 83 percent, resulting in AFUEs greater than 78 percent (a 78 percent AFUE corresponds to
an SSE of about 80.5 percent). Prior to 1987, the majority of gas furnaces installed in the United
States for residences were draft-hood-equipped models. Because of the effect of the NAECA, for
new installations and replacements, gas furnaces using natural vent draft hood systems will be gradually
phased out in favor of the fan-assisted power vent furnaces.
Control and Operation
Capacity. The heating capacity of a gas-fired furnace is controlled by a gas valve and ignition device.
For small furnaces the gas valve is usually controlled by a room thermostat that has on/off
control. Either standing-pilot or intermittent ignition may be used. For large furnaces a two-stage
gas valve, controlled by a two-stage thermostat, may be operated with a full gas supply, or at a reduced
rate when the outdoor weather is mild. Energy or fuel savings may not be significant unless
both the gas and the combustion air supply are controlled.
Venting system Tflue  Tf, dew,°F Approximately SSE, percent
Category I Nonpositive (natural vent)  140 (noncondensing )  83
Category II Nonpositive (natural vent)  140 (condensing) 	 83
Category III Positive (power vent)  140 (noncondensing)  83
Category IV Positive (power vent)  140 (condensing) 	 83
Efficiency as of
Equipment type Size, Btu/h Rating conditions Minimum efficiency 10/29/2001
Gas-fired furnaces  225,000 78% AFUE or 80% Et 78% AFUE or 80% Et
 225,000 Maximum capacity 80% Et 80% Ec
Oil-fired furnaces  225,000 78% AFUE or 80% Et 78% AFUE or 80% Et
 225,000 Maximum capacity 81% Et 81% Et
Gas-fired duct furnaces All sizes Maximum capacity 78% Et 80% Ec
Minimum capacity 75% Et
Gas-fired unit heaters All sizes Maximum capacity 78% Et 80% Ec
Minimum capacity 74% Et
Oil-fired unit heaters All sizes Maximum capacity 81% Et 80% Ec
Minimum capacity 81% Et
8.8 CHAPTER EIGHT
Nighttime Setback. ASHRAE research project SP43 found that a nighttime thermostat setback
of 10°F (5.6°C) lower for 8 h improved Et slightly (only a 0.4 percent increase). However, there is a
10 to 16 percent annual savings in energy input compared to those furnaces without nighttime
setbacks.
Oversizing the furnace shortens the morning pickup time caused by nighttime setback. Use of an
oversized furnace has a significant effect on the swing of space air temperature when an on/off control
is used for the gas valve. Project SP43’s results showed that when the furnace size corresponded
to 1.4 times the design heating load (DHL), the furnace had a space temperature swing of
4.9°F (2.7°C) and a morning pickup time of about 1 h. If the furnace size was based on 1.7 times
the DHL, the space temperature swing increased to 5.9°F (3.3°C), and the pickup time reduced to
about 0.5 h. A furnace size based on 1.4 times the DHL is more suitable for a nighttime setback
period of 8 h and a setback temperature of 10°F (5.6°C).
Fan Operation. In the past, continuous fan operation in small upflow gas-fired burners was said to
offer the benefits of better air circulation, reduced noise (because the fan did not start and stop), and
an even temperature distribution. Project SP43 showed that continuous fan operation resulted in a
higher furnace efficiency. However, continuous operation consumes more electricity than intermittent
operation in which the fan shuts off and the supply temperature drops below 90°F (32.2°C). In
many locations with a high electricity-to-fossil-fuel cost ratio, an energy cost analysis based on
AFUE may determine whether continuous or intermittent operation is more efficient.
The fan often starts about 1 min after the burner starts. Such a delay allows the heat exchanger
to warm up and prevents a flow of cold air. The fan will shut down 2 to 3 min after the burner is
shut off. The supply of residual heat from the heat exchanger also improves the performance of the
furnace.
Safety. For safety, the vicinity of the furnace should be free of combustible gas, vapor, and material.
Any passage to provide combustion air must be carefully planned. Gas and vent pipes should be
installed according to local and federal codes.
8.3 HOT WATER BOILERS
A hot water boiler for space heating is an enclosed pressure vessel in which water is heated to a
required temperature and pressure without evaporation. Hot water boilers are manufactured according
to the American Society of Mechanical Engineers (ASME) boiler and pressure vessel codes.
Boilers are usually rated according to their gross output heat capacity, i.e., the rate of heat delivered
at the hot water outlet of the boiler, in MBtu/ h, or thousands of Btu/h (kW). Hot water boilers are
available in standard sizes up to 50,000 Mbtu/h (14,650 kW).
Selection of Fuel
Natural gas, oil, coal, and electricity are energy sources that can be used in hot water boilers. It is
necessary to provide for an adequate supply during normal and emergency conditions and to take
into account the limitations imposed by any building and boiler codes for certain types of equipment
due to safety and environmental concerns. In addition, storage facilities and cost should be
considered before a fuel is selected.
Whereas natural gas and electricity are supplied by a utility, LPG, oil, and coal all need space
for storage within and outside the boiler plant.
Cost includes energy cost, initial cost, and maintenance cost. A gas-fired boiler plant requires
the lowest initial costs and maintenance costs, oil-fired boiler plants are moderately higher, and
coal-fired boiler plants are significantly higher (although their energy cost is lowest). An electric
boiler is simple to operate and maintain. In addition, it does not require a combustion process,
HEATING SYSTEMS, FURNACES, AND BOILERS 8.9
chimney, or fuel storage. In locations where electricity costs are low, electric boilers become
increasingly more attractive.
According to the data in the Commercial Building Characteristics 1992 by the EIA, the percentages
of floor area in all commercial buildings served by different kinds of primary energy source in
hot water and steam boilers in 1992 in the United States are as follows:
Gas-fired boilers 71 percent
Oil-fired boilers 15 percent
Electric boilers 11 percent
Others 2 percent
Types of Hot Water Boiler
According to their working temperature and pressure, hot water boilers can be classified as follows:
1. Low-pressure boilers. These hot water boilers are limited to a working pressure of 160 psig
(1103 kPag) and a working temperature of 250°F (121°C).
2. Medium- and high-pressure boilers. These boilers are designed to operate at a working pressure
above 160 psig (1103 kPag) and a temperature above 250°F (121°C).
A low-pressure hot water boiler is generally used for a low-temperature water (LTW) heating
system in a single building, regardless of the building’s size. Medium- and high-pressure boilers are
often used in medium-temperature water (MTW) and high-temperature water (HTW) heating systems
for campuses or building complexes in which hot water temperature may range from 300 to
400°F (150 to 205°C).
Based on their construction and materials, hot water boilers can also be classified as fire-tube
boilers, water-tube boilers, cast-iron sectional boilers, and electric boilers. Water-tube boilers,
mainly used for steam at higher pressure and temperature, are not discussed here. Electric boilers
are discussed in the next section.
Fire-Tube Boiler
A fire-tube boiler’s combustion chamber and flue gas passages are in tubes, which are all enclosed
in a shell filled with water. Heat released from the combustion process and the flue gases is absorbed
by the surrounding water, the temperature of which is increased to a required value.
Many kinds of fire-tube boilers have been developed. One of the more recently developed models
is known as the modified Scotch marine boiler, which is a compact and efficient design originally
used on ships. The Scotch marine boiler is probably the most popular hot water boiler manufactured
today.
Scotch Marine Packaged Boiler
A Scotch marine packaged boiler consists mainly of a gas, oil, or gas/ oil burner; a mechanical draft
system; a combustion chamber; fire tubes; and a flue vent. A packaged boiler is a one-piece, integrated,
factory-assembled boiler that includes a burner, outer steel shell, fire tubes, draft system, external
insulation, controls, interconnecting piping, and wiring. A schematic diagram of a typical
Scotch marine packaged boiler is shown in Fig. 8.3a and a cutaway illustration in Fig. 8.3b.
Flow Processes. Gas or oil and air are taken into the burner in measured quantities. The burner
then injects a combustible air-fuel mixture containing the necessary combustion air into the
8.10 CHAPTER EIGHT
HEATING SYSTEMS, FURNACES, AND BOILERS 8.11
FIGURE 8.3 Scotch marine packaged boiler. (a) Schematic diagram; (b) a cutaway photograph of a typical product.
(Source: Cleaver-Brooks. Reprinted with permission.)
combustion chamber, where it is initially ignited by an ignition device. The mixture burns, and the
combustion process sustains itself once the heat it generates is greater than the heat it transfers to
the water, i.e., once a high enough temperature in the combustion chamber is attained. The injected
air-fuel mixture is then burned spontaneously.
The direct radiation from the flame and the high temperature in the combustion chamber both
conduct heat through the wall of the chamber—the primary surface adjacent to the water that fills
the shell. The combustion product from the combustion chamber—flue gas—is directed into fire
tubes by headers. As the flue gas at a higher temperature flows through the fire tubes, it transfers
heat to the surrounding water through the pipewall of the tubes—the secondary surface. The flue
gas temperature drops, and its volume contracts. Because the number of fire tubes continuously decreases
in the second, third, and fourth passes of fire tubes, which matches the volume contraction
of flue gas, the gas velocity is maintained in a more uniform manner. Flue gas leaves after the
fourth pass and is vented into the stack.
Return water enters the side of the boiler, sinks to the bottom, is heated, and rises again. Hot
water is finally supplied at the top outlet. Such an arrangement prevents cold return water from
surrounding the combustion chamber and producing thermal shock. It also promotes good water
circulation.
Construction Characteristics. The constructional characteristics for Scotch marine packaged
boilers are as follows:
 Four flue gas passes. Each pass means a horizontal run of the fluid flow passage. A four-pass flow
arrangement for the flue gas, with a gradual reduction in gas flow area as the flue gas becomes
colder, helps maintain a higher gas velocity and, therefore, a clean surface and a higher rate of
heat transfer.
 Sufficient heat-transfer surface area. In a packaged Scotch marine boiler, the combustion chamber,
fire tubes, and tube sheets at both ends are all heating surfaces. The amount of heating surface
area directly affects a boiler’s output. And 3 to 5 ft2 (0.9 to 1.5 m2) of heating surface area for
each boiler horsepower, or 33,475 Btu/h (9.8 kW), of output, is the key value for providing suffi-
cient capacity and a long-lasting boiler. A smaller heating surface area for each boiler horsepower
of output often results in a higher heat flux, probably a higher surface temperature, and therefore a
shorter life for the boiler.
 Forced-draft arrangement. In a forced-draft arrangement, the fan is located adjacent to the burner.
The fan forces air into the combustion chamber. It also forces the flue gas to flow through the fire
tubes and to discharge from the vent or chimney. A forced-draft fan supplies a controllable quantity
of combustion air to the combustion chamber. The forced-draft fan is located upstream of the
combustion chamber, so it handles dense, clean boiler room air at a comparatively lower temperature
and lower volume flow rather than the hot, dirty, high-temperature, expanded flue gas downstream.
The temperature fluctuation for the room air is smaller than that for the exhausted flue
gas, so more accurate control of combustion air is possible than with an induced-draft design.
Cast-Iron Sectional Boiler
A cast-iron sectional boiler consists of many vertical cast-iron hollow sections in the shape of an
inverted U filled with water. When the sections are linked together by bolts, the center part of the
inverted U forms the furnace or the combustion chamber.
Because of its thick, heavy sections, a cast-iron sectional boiler has a large heat storage capacity
and thus is slow to heat up. This reduces temperature swings when heat demand varies. Thick,
heavy sections are also helpful in extending the boiler’s service life by preventing corrosion.
The heating capacity of a cast-iron sectional boiler depends on the number of sections connected.
Because of its relatively large combustion chamber and lower flow resistance, such a boiler
is able to use atmospheric gas burners with a lower chimney height. Cast-iron sectional boilers are
low-pressure boilers used in residences and in small and medium-size commercial buildings. They
can be field-assembled and fitted in existing buildings.
8.12 CHAPTER EIGHT
Gas and Oil Burners
When natural gas is used as the fuel in boilers, as in gas-fired furnaces, atmospheric and power gas
burners are usually used. Oil burners used in boilers differ from gas burners in that oil requires atomization
and vaporization prior to combustion. Combustion air is supplied by either natural drafts
or forced drafts. Ignition is usually provided by an electric spark or a standing pilot using gas or oil.
Oil burners used in residences are usually pressure atomizing-gun burners. This type of burner
either directly atomizes oil at 100 to 300 psig (690 to 2069 kPag) high pressure or uses combustion
air to atomize the oil. Combustion air is supplied by a blower, and electric spark ignition is
generally adopted. Oil burners for boilers in commercial buildings inject oil and atomize it into fine
sprays. Burners also force combustion air to mix with atomized oil and ignite the mixture, often
with an electric spark. Complete combustion can be sustained with as little as 20 percent excess air.
Sometimes a boiler can be designed to use either gas or oil. A combination of a ring-type gas
burner with an oil burner at the center is often used. Combination gas-oil burners are used more
often than single-fuel burners in large boilers.
Condensing and Noncondensing Boilers
In condensing boilers, the water vapor in the flue gas is condensed and drained by means of a heat
exchanger, as in gas furnaces. Thus, the latent heat of condensation can be recovered. In hot water
boilers, the cooling medium is usually the return water from the conditioned space. The lower the
temperature of the return water, the higher the amount of heat recovered. If the dew point of the flue
gas is 130°F (54°C), return water or service water at a temperature below 125°F (51.7°C) may be
used as the condensing cooling medium. Corrosion in the heat exchanger and flue gas passage
caused by the condensate should be avoided. Noncondensing boilers have no way to condense the
water vapor contained in the flue gas.
Boiler Efficiency
Boilers can be assessed by two efficiency values. The combustion efficiency Ec, in percent, is the
ratio of heat output from the hot water or steam Qout to the heat content rate of the fuel consumed
Qfuel, that is,
(8.4)
Both Qout and Qfuel are expressed in Btu/h (kW). Annual fuel utilization efficiency (AFUE) is also
an efficiency value used for hot water boilers. For noncondensing boilers, Ec varies from 80 to 85
percent. For condensing boilers, Ec varies from 85 to 90 percent.
ASHRAE/IESNA Standard 90.1-1999 mandates that the minimum efficiency requirements of
gas- and oil-fired hot water boilers are as follows:
Efficiency as of
Equipment type Size, Btu/h Rating condition Minimum efficiency 10/29/2001
Gas-fired boilers  300,000 Hot water 80% AFUE 80% AFUE
 300,000 and Maximum capacity 80% Ec 75% Et
	 2,500,000
	 2,500,000 Hot water 80% Ec 80% Ec
Oil-fired boilers  300,000 80% AFUE 80% AFUE
 300,000 and Maximum capacity 83% Ec 78% Et
	 2,500,000
	 2,500,000 Hot water 83% Ec 83% Ec
(Continued)
Ec 
100Qout
Qfuel
HEATING SYSTEMS, FURNACES, AND BOILERS 8.13
Chimney, or Stack
The chimney, or stack, is the vertical pipe or structure used to discharge flue gas. The breeching, or
lateral, is the part of the horizontal duct that connects the flue gas outlet of the boiler and the vertical
chimney. Breeching is commonly fabricated from 10-gauge steel (3.4-mm thickness) covered
with a high-temperature insulation layer. Flue gas discharge from hot water boilers that burn gas or
oil usually has a temperature rise of 300 to 400°F (167 to 222°C) over the ambient temperature.
Chimneys for burning gas or oil should be extended to a certain height above adjacent buildings according
to the local codes and topographical conditions. In high-rise buildings, locating the boiler
room at the basement level may be very expensive because of the space occupied by the vertical
stack and insulation inside the building.
Operation and Safety Controls
Heating Capacity Control. For hot water boilers, heating capacity control during periods of reducing
heat demand is achieved by sensing the return water temperature and controlling the firing
rate of the gas and oil burners. The firing rate of burners can be controlled in on /off, off / low/high,
and modulating modes. An on/off control mode is usually used for small boilers in which hot water
temperature control is not critical. A solenoid valve is used to open or close the fuel supply line to
the burner. An off / low/high control mode offers better output control than an on /off mode. If the
demand for heat is low, the boiler starts to fire at approximately one-half of its full capacity (the low
fire setting). Upon further increase in demand, the burner fires at its maximum capacity (the high
fire setting). A two-stage control valve and damper are used to control the supply of fuel and combustion
air to the burners.
With modulating control, the boiler starts at a low fire of about one-third of the full-load capacity.
As the demand for heat increases, an increasing amount of fuel and combustion air is supplied
to the burner through the modulation of either a butterfly valve in a gas burner or an orifice of variable
size in an oil burner. Both types of burner are linked to a damper to vary the supply of combustion
air. For gas burners, two pressure sensors are often provided to maintain the gas pressure
within a narrow range for proper operation. If one of the sensors detects an improper pressure, the
gas supply is cut off and the boiler shuts down.
To maintain an optimum air-fuel ratio, the amount of O2 in the stack flue gas can be maintained
at 1 to 5 percent by modulating the air damper of combustion air to provide higher combustion effi-
ciency. Usually, CO2 is measured and monitored in small boilers, and O2 is measured and monitored
in large boilers.
Safety Control. One or more pressure or temperature relief valves should be equipped for each
boiler. These are mechanical devices that open when the boiler pressure or temperature exceeds the
rated value. An additional limit control is installed to open the switch and shut down the boiler as
soon as the sensed water pressure or temperature exceeds the predetermined limit. When the pressure
or temperature falls below the limit, the switch closes and the firing starts again.
A flame detector is used to monitor the flame. When the flame is extinguished, the controller
closes the fuel valve and shuts down the boiler. An airflow sensor is also used to verify a continuous
supply of combustion air. Once the combustion air is not present, the fuel valve closes before a dangerous
fuel-air ratio can form.
Efficiency as of
Equipment type Size, Btu/h Rating condition Minimum efficiency 10/29/2001
Oil-fired (residual)  300,000 and Maximum capacity 83% Ec 78% Ec
	 2,500,000
	 2,500,000 Hot water 83% Ec 83% Ec
8.14 CHAPTER EIGHT
Modern packaged boilers often include a totally enclosed and factory-assembled controlled
cabinet that offers the latest microprocessor-based programmable direct digital controls, including
flame safeguard and control systems.
8.4 ELECTRIC FURNACES, HEATERS, AND BOILERS
Electric Heating Fundamentals
Except for hot water electric boilers using electrodes, most electric furnaces, electric heaters, and
electric heating coils used in HVAC&R installations are of the resistance type. When an electric
current flows through a resistor under electric potential, heat is released to the ambient air or water.
Electric resistance heaters of small wattage (e.g., an electric furnace in a residence with a heating
capacity of less than 8 kW) usually use single-phase 120/240-V supply. Large-capacity electric
resistance heaters for commercial and industrial applications usually use a three-phase 240/480-V
supply.
For a single-phase electric heater, the electric power input P, in W, which is equal to the heat
released from the resistor, can be calculated as
P  EI  I 2R (8.5)
where E  electric voltage, V
I  electric current, A
R  electric resistance, 
For a three-phase electric heater, the electric power input is given as
P  1.73El Il 1.73Il
2R (8.6)
where El  line voltage, V, and Il  line current, A. When a three-phase electric heater is in a delta
connection, as shown in Fig. 8.4b, then
and (8.7)
where Ep  phase voltage, V, and Ip  phase current, A. When a three-phase electric heater is in a
wye connection, then
and (8.8)
Several kinds of overload protection devices are used for electric heaters to open the circuit if
the electric current reaches a value that will cause a dangerous temperature:
 Fuses are alloys in the form of links to be inserted in the electric circuit. A fuse melts at a preset
temperature, opening the circuit, when an overload current passes through.
 Circuit breakers are mechanical devices that open their contacts, by means of a warp action of a
bimetallic strip or disk, when an excessive load passes through and heats them.
There are also thermal overload current-sensing devices and magnetic overload relays, both of
which open a circuit when the electric circuit is overloaded.
National Electrical Codes (NEC) and local codes must be followed during the design, installation,
and operation of electric heaters.
Ip  Il Ep 
El
1.73
Ip 
Il
1.73
Ep  El
HEATING SYSTEMS, FURNACES, AND BOILERS 8.15
Electric Furnaces, Electric Heaters, and Duct Heaters
Electric furnaces installed in rooftop packaged units, electric heaters equipped in air-handling units
and packaged units, electric duct heaters mounted in air systems, and electric unit heaters installed
directly above the conditioned space are all electric resistance warm air heaters. Air is forced
through the electric heating element by the supply fan in a rooftop packaged unit or in an airhandling
unit, or by the separately installed propeller fan in an electric unit heater.
Because the mixture of outdoor air and recirculating air or the outdoor air alone is clean and
nonhazardous after passing through the air filter, most electric resistance heaters use heating
8.16 CHAPTER EIGHT
FIGURE 8.4 Typical electric heater wiring diagram. (a) Single-phase line voltage control with thermal cutout and fan interlock;
(b) three-phase line voltage control with fan interlock and backup contactor.
elements made of bare wire of various alloys, such as 80 percent nickel and 20 percent chromium
wire. Bare wire is supported by bushes and brackets of insulating materials such as ceramics.
Sheathed elements are also used for better protection.
The face velocity of air flowing through warm air electric heaters in rooftop packaged units is
approximately the same as the face velocity of air at the cooling coil in AHUs. For a face velocity
of 500 to 600 fpm (2.5 to 3 m/ s), the temperature rise of warm air is often between 30 and 60°F
(16.7 and 33.4°C).
The total pressure loss of the warm air flowing through an electric resistance heater in AHU or
packaged units is usually less than 0.3 in. WC (75 Pa). For electric duct heaters, the air velocity and
pressure loss may be higher. Figure 8.4a shows a typical wiring diagram for an electric heater using
a single-phase electric power supply. Figure 8.4b shows a wiring diagram for a three-phase electric
power supply. In the design and installation of an electric duct heater, the following requirements
should be fulfilled:
 An electric duct heater should be installed at least 4 ft (1.2 m) downstream of the fan outlet,
elbow, or other obstructions. Otherwise, devices should be installed to provide for an even distribution
of airflow over the face of the heater.
 Do not install two electric duct heaters in series. Install electric duct heaters at least 4 ft (1.2 m)
from the heat pumps and air conditioners. Following these suggestions prevents excessively high
temperatures and the accumulation of moisture in the heater.
 A fan interlock circuit must be provided so that the electric heater does not operate unless the fan
is on.
 Automatic-reset overload cutout safety control should be placed in series with the thermostat. A
manual-reset overload cutout safety control is used in the backup safety control circuit. Limit
controls may be used instead of a manual-reset cutout control. Disconnecting devices should be
installed within sight of the electric heater.
 On each electric circuit, the maximum load of the electric heater is 48 A. The maximum fuse size
is 60 A. Divide a larger electric heater into smaller heaters with smaller loads. The loading of an
electric heater is often divided into stages for step control. Small heaters may be divided into two
or three stages; large electric heaters may have as many as 16 stages.
Electric Hot Water Boilers
There are two types of electric hot water boiler: resistance and electrode. Resistance electric boilers
use metal-sheathed heating elements, such as thin nickel chromium alloy wire, submerged in hot
water. Such safety valves as a relief valve, thermal overload cutout, and high-limit switch should be
provided. As with electric heaters, the electric load is divided into stages. The heating capacity is
controlled in steps by sensing the return water temperature. Resistance boilers are more commonly
used than electrode boilers. Resistance boilers are available in sizes with outputs up to about
3360 kW. Electrode boilers use water as a resistor. They are generally found in very large installations
and require high-quality water and may operate at high voltages.
8.5 LOW-PRESSURE DUCTED WARM AIR HEATING SYSTEMS
System Characteristics
A low-pressure ducted warm air (forced-air) heating system has the following characteristics:
 It is often equipped with an upflow gas-fired furnace, oil-fired furnace, or electric heater as well
as a centrifugal fan to force air through the furnace or electric heater.
HEATING SYSTEMS, FURNACES, AND BOILERS 8.17
 The external static pressure loss for the supply and return duct system is usually no greater than
0.5 in. WC (125 Pa).
 It uses a furnace heating capacity Qf to airflow ratio ranging from 50 to 70 Btu/hcfm (31 to
43 W s /L) and a temperature rise immediately after the furnace between 50 and 70°F (28 and
40°C).
 Because of the heat storage capacity of the supply duct system, the supply temperature differential
Ts  Tr is often 20 to 35°F (11.1 to 19.4°C). Here Ts represents the supply air temperature at
the supply register, and Tr indicates the space temperature, both in °F (°C).
 The heating system is often integrated with the cooling system to form a heating/cooling air
conditioning system.
 Most low-pressure warm air heating systems have a capacity no greater than 100,000 Btu/h
(29.3 kW).
Low-pressure ducted warm air heating systems are usually used in residences and sometimes in
small commercial buildings.
Types of Low-Pressure Ducted Warm Air Heating System
Low-pressure ducted warm air heating systems can be divided into two categories according to
their duct systems:
 Supply and return duct system
 Supply duct and return plenum
Figure 8.5a shows a typical low-pressure ducted warm air heating system with a supply and return
duct system for a single-family house. Recirculating air from the living room, dinning room, and
bedrooms flows through return ducts and enters the lower part of the upflow gas furnace. It mixes
with outdoor air. The mixture is forced through a gas furnace and heated to a required supply temperature
at the supply plenum. Warm air is then supplied to various rooms through supply ducts and
registers.
Figure 8.5b shows a typical low-pressure ducted warm air heating system with supply ducts and
a return plenum. There are no return ducts. Warm air is supplied to various rooms through supply
ducts. Recirculating air flows back to the return plenum by means of undercut or louvered doorways
and corridors. Once an interior door between rooms is closed, air squeezes through the door
undercuts and louvers to the return plenum as well as other cracks and gaps to the attic or outdoors.
The room is therefore maintained at a positive pressure. This causes a reduction in the supply flow
rate and an increase in the infiltration rate of the whole house.
Field tests performed by Cummings and Tooley (1989) on five Florida houses show that, for a
low-pressure ducted warm air heating system with supply ducts and a return plenum, the average
whole-house infiltration was 0.31 ach (air change per hour) when all the interior doors between
rooms were open. If the interior doors were closed, the average whole house infiltration increased to
0.91 ach. The typical pressure difference between rooms separated by a closed door was 0.032 in.
WC (8 Pa), for a door undercut of 0.5 in. (13 mm).
Heat Supplied to Conditioned Space
Heat supplied to the conditioned space by a low-pressure ducted warm air heating system, denoted
by Qsh, Btu /h (W), can be calculated as
(8.9) Qsh  60V?s 
scpa(Ts  Tr)
V?
a
8.18 CHAPTER EIGHT
where volume flow rate of supply air, cfm [m3 / (60 s)]
Ts, Tr  temperature of supply and space air, °F (°C)
In Eq. (8.9), 
s represents the density of supply air. For a supply air temperature of 100°F (38°C)
and a relative humidity around 20 percent, 
s can be taken as 0.07 lb / ft3 (1.12 kg/m3). The term cpa
indicates the specific heat of moist air. For simplicity, its value can still be taken as 0.243 Btu/ lb °F
(1017 J/kg°C).
V?
s 
HEATING SYSTEMS, FURNACES, AND BOILERS 8.19
FIGURE 8.5 Typical low-pressure ducted warm air heating system. (a) Supply and return duct; (b) supply
duct and return plenum.
Duct Efficiency and System Efficiency for Heating
Duct efficiency for heating du, h, in percent, can be calculated by dividing the heat energy output
from the supply and return duct system by the heat energy input to it, i.e.,
(8.10)
where qho, s, qho, r  output heat energy from supply ducts and return ducts, Btu/h (W)
qhi, s, qhi, r  input heat energy to supply ducts and return ducts, Btu/h (W)
qhl, s, qhl, r  heat loss from supply ducts and return ducts, Btu/h (W)
and the supply duct efficiency for heating sd, h , in percent, is
(8.11)
The system efficiency for heating sy, h, in percent, can be calculated as
(8.12)
where Qf, h  jacket loss and equipment losses from furnace and released to conditioned space,
Btu/h (W)
Qf, in  total energy input to furnace, including auxiliary energy input, Btu/h (W)
Location of Furnace and Duct Insulation
The location of the furnace has a significant effect on the system efficiency of heating. In Jakob et al.
(1986), if the gas furnace is installed in a closet, the supply duct is mounted inside the conditioned
space, and the equipment losses of the furnace become the direct heat gains of the conditioned space,
then sy, h might be 20 percent higher than for those installations where the furnace and supply ducts
are in the attic or basement.
Jakob et al. (1986) also showed that if ducts had an exterior or interior insulation of R5, that
is, an insulation layer with an R value of 5 h  ft2  °F/Btu (0.9 m2  °C/W), the duct efficiency increased
from 61 percent (without duct insulation) to about 78 percent. The system efficiency saw
a smaller increase because a portion of the heat loss from the duct without insulation had been
used to heat the attic or basement, which in turn reduced the heat loss from the conditioned
space.
Duct Leakage
ASHRAE Standard 90.1-1999 mandates the minimum duct seal level as discussed in Sec. 17.8.
Field tests in many houses have showed that actual duct leakage in many low-pressure ducted warm
air heating systems is considerably higher.
Lambert and Robinson (1989) analyzed the duct leakage, whole-house leakage, and heat energy
use of 800 electric-heated houses built since 1980 in the Pacific northwest. Tested houses were
 100 Et du, h 
Qf, h
Qf, in 
hsy, h  100
qho, s  Qf, h
Qf, in
sd, h  100
qho, s
qhi, s
 100
1  qhl, s
qhi, s
 1001 
qhl, s  qhl, r
qhi, s  qhi, r 
du, h  100
qho, s  qho, r
qhi, s  qhi, r
8.20 CHAPTER EIGHT
divided into two groups: highly energy-efficient Model Conservation Standard (MCS) houses and a
control group built to current (regional) practice (CP) or standards. Ducted heating systems in CP
houses had 26 percent more air leakage than unducted systems and used 40 percent more heating
energy. Ducted heating systems in MCS houses had 22 percent more air leakage and used 13 percent
more heating energy than unducted MCS houses.
Gammage et al. (1986) studied the ducted warm air systems of 31 Tennessee houses. They
found that the air infiltration rate was 0.44 ach when the forced-air systems were off and 0.78 ach
when the forced-air systems were on.
Field tests have also shown that leakages are greater for the return duct than for the supply
duct because of the greater importance of the supply air. Also, the return plenum is often not
carefully sealed. In five Florida homes with low-pressure ducted warm air systems with supply
and return plenums, Cummings and Tooley (1989) found that when repairs were made to seal
the return plenums, the infiltration in these five homes dropped from an average of 1.42 ach to
0.31 ach.
When a ducted warm air heating system is operating, supply duct leakage in such nonconditioned
space as the attic or basement raises the space pressure to a positive value and promotes
exfiltration. Return duct leakage extracts the space air, lowers the space pressure to a negative
value, and promotes infiltration. Both types of leakage increase the whole-house infiltration.
Suggested remedies to reduce duct leakage and energy use are as follows:
 Externally seal the duct with tape; seal the ducts internally if possible.
 Seal the return plenum and equipment if there is leakage. Seal the duct, pipes, and cable penetrations
through the structures.
 Avoid locating ducts in unconditioned spaces. Provide insulation for ducts running in unconditioned
spaces.
 For a low-pressure ducted warm air system without return ducts, an adequate door undercut or
door louver should be provided.
Thermal Stratification
A low-pressure ducted warm air heating system with a gas-fired furnace always has a high supply
temperature differential Ts. If Ts exceeds 30°F (16.7°C), or if there is a high ceiling, thermal
stratification may form in the conditioned space. The vertical temperature difference may be greater
than 5°F (2.8°C) and cause discomfort, as mentioned in Sec. 4.8. In addition, a higher temperature
near the ceiling may increase heat transfer through the ceiling, attic, and roof.
A greater supply air volume flow rate, lower Ts, higher downward air jet velocity, and suitable
location for the supply grille are remedial measures that reduce thermal stratification and vertical
temperature differences.
Part-Load Operation and Control
For low-pressure ducted warm air heating systems, a thermostat or a temperature sensor is usually
installed in a representative space to control the gas valve of a furnace operating under an on/off or
two-stage step control mode to maintain the required space temperature. The proportion of on and
off time in each operating on/off cycle varies to meet the varying space heating load. Figure 8.6
shows a typical on/off control for a gas furnace, resulting in a temperature variation for a point
20 ft (6.1 m) downstream of the supply plenum and at the supply register.
When the space temperature increases above the upper limit as the space heating load falls, the
controller shuts off the gas valve so that the heat supply to the space is cut off. When the space temperature
falls below the lower limit, the gas valve will open again and raise the space temperature.
HEATING SYSTEMS, FURNACES, AND BOILERS 8.21
The time period for an on/off operating cycle is generally between 5 and 15 min. Too short a cycle
may result in unstable operation—a condition known as hunting.
8.6 DUCTED WARM AIR HEATING SYSTEMS
Ducted warm air heating systems are heating systems that are part of the air-handling unit, packaged
units, packaged heat pumps, or fan-coil units that use water heating coils, gas furnaces, electric
heaters, or indoor coils in heat pumps to heat the air. Warm air is then supplied to the conditioned
space through the duct system.
A ducted warm air system has the following characteristics:
 It is often integrated with a cooling system, forming a heating/cooling system.
 Its external pressure loss for a duct system is usually between 0.5 and 2 in. WC (125 and 500 Pa),
except for fan-coil units.
 The heating capacity is controlled by modulating the water flow rate to the water heating coil, by
modulating the gas valve, or by controlling the refrigeration flow in an on/off or step control
mode in the heat pumps.
 Heating systems often have a capacity greater than 100,000 Btu/h (29 kW).
Because the perimeter zones of office buildings need strict control of the vertical temperature
difference in their conditioned space, a smaller supply air temperature differential, such as Ts 	
8.22 CHAPTER EIGHT
FIGURE 8.6 Typical on/off control of a gas furnace and the variation of temperature of supply plenum and supply register.
15°F (8.3°C) is usually used. The performance and design of these types of heating and cooling
systems will be discussed in later chapters.
8.7 HOT WATER HEATING SYSTEMS USING
FINNED-TUBE HEATERS
Types of Hot Water Heating System
Like hot water boilers, hot water heating systems can be classified according to their operating temperature
into two groups:
 Low-temperature water systems. These operate at a temperature not exceeding 250°F (120°C),
typically 190°F (88°C) supply and 150°F (65°C) return, and a maximum working pressure not
exceeding 150 psig (1034 kPa g), usually less than 30 psig (207 kPa g).
 Medium- and high-temperature water systems. In medium-temperature water systems, the operating
temperature is 350°F (177°C) or less, and the operating pressure is 150 psig (1034 kPa g) or
less. In high-temperature water systems, the maximum operating temperature can be from 400 to
450°F (205 to 232°C), and the maximum operating pressure is 300 psig (2070 kPag). In both
medium- and high-temperature systems, the hot water supplied to and returned from the heating
coils and space heaters is typically 190 and 150°F (88°C and 65°C), respectively.
Low-temperature hot water heating systems are widely used for space heating in residential and
commercial buildings. Medium- and high-temperature hot water heating systems are often used in
central heating plants for university campuses or groups of buildings, or in industrial applications
for process heating. Only low-temperature hot water systems are discussed in this section.
In low-temperature hot water heating systems, the two-pipe individual loop and radiant floorpanel
heating systems are currently used in residential, commercial, and industrial buildings.
Two-Pipe Individual Loop System
Most current low-temperature hot water heating systems that use finned-tube heaters are equipped
with zone control facilities. Without such a control system, rooms that face south may be overheated
and rooms that face north may be underheated in northern latitudes because of the effects
of solar radiation. The typical piping arrangement for a low-temperature hot water heating system
is the two-pipe individual-loop system, as shown in Fig. 8.7. On a cold winter day, water in a typical
individual-loop system returns from various finned-tube baseboard heaters through different
individual return loops at a temperature between 150 and 155°F (65 and 68°C). It is heated at the
hot water boiler to a temperature of 190°F (88°C), extracted by the on-line circulating pump on
each individual loop, and distributed to the finned-tube baseboard heaters by means of supply and
branch pipes. Because of the higher average temperature of hot water in the baseboard heaters,
heat is released to the space air to offset the heating load by means of radiation and convection in
various control zones, such as north, south, east, and west zones. The temperature of hot water
then drops to between 150 and 155°F (65 and 68°C) and returns to the hot water boiler again via a
return main.
In Fig. 8.7, there are several individual loops, or control zones. In each individual loop, several
finned-tube baseboard heaters in a large room can be connected in series. Finned-tube baseboard
heaters in small rooms belonging to the same individual loop can be connected in a reverse-return
arrangement. In such an arrangement, the length of pipeline that hot water travels is nearly the same
for each baseboard heater for a better system balance.
If a hot water heating system using finned-tube heaters is part of an air conditoning system, then
the outdoor ventilation air during winter heating will be provided for the occupants in the heating
HEATING SYSTEMS, FURNACES, AND BOILERS 8.23
space by the air conditioning system that cools the conditioned space in summer. In locations
of cold climate, if only winter heating is required, then a separate outdoor ventilation system
sometimes is required instead of opening the windows manually. Outdoor ventilating systems are
discussed in Chaps. 23 and 24.
The design procedure for a hot water heating system involves the following:
 Calculation of the space heating load, as discussed in Chap. 6
 Selection of suitable finned-tube heaters
 Division of the heating space into various control zones, or individual loops
 Planning of a piping layout, containing branches and necessary piping fittings
 Location of the boiler, circulating pumps, air separator, and expansion tank in the mechanical
equipment room, and determination of their capacities
 Specification of the control functions of the capacity and safety control systems, including the
sequence of operation
Finned-Tube Heaters
A finned-tube heater is a terminal unit installed directly inside the conditioned space to deliver heat
energy to the space by means of radiation and convection. A finned-tube heater consists of a finnedtube
element and an outer enclosure, as shown in Fig. 8.8. The tubes are often made of copper and
steel. Copper tubes are usually of 0.75-, 1-, and 1.25-in. (19-, 25-, and 32-mm) diameters, and steel
tubes of 1.25- and 2-in. (32- and 50-mm) diameters. For copper tubes, fins are often made of aluminum.
For steel tubes, the fins are made of steel. Fin spacing varies from 24 to 60 fins per foot (79
to 197 fins per meter). A finned-tube heater may have a length up to 12 ft (3.6 m).
8.24 CHAPTER EIGHT
FIGURE 8.7 A two-pipe individual-loop low-temperature hot water heating system for a factory.
Although various configurations of the outer enclosure have been designed and manufactured,
each enclosure must have a bottom inlet and a top outlet for better convection. The enclosure is usually
made of 18-gauge (1.2-mm-thickness) steel with corrosion-resistant coating to provide protection
and improve appearance. To allow a higher heating capacity, two finned-tube elements can be
set in a two-tier (two-row) arrangement. The most widely used finned-tube heater is the baseboard
heater, which is often installed at a level of 7 to 10 in. (175 to 250 mm) above the floor. It is usually
3 in. (75 mm) deep and has only one pier of finned-tube elements. Baseboard heaters are usually
mounted on cold walls and release heat nearly at the floor level. They also interfere less with indoor
decor than other heaters.
A wall finned-tube heater has a greater height and is available in various shapes to meet the
architectural interior design. A convector has a cabinet-type enclosure with one or two tiers of
finned-tube heating elements installed under the window sill. The rated heating capacity of a
finned-tube heater depends mainly on its length, the fin spacing, the average water temperature
flowing through the heating element, and the temperature of air entering the heater. Table 8.1 lists
the rated heating capacity of typical finned-tube heaters for entering air at 65°F (18.3°C) and different
average hot water temperatures.
Design Considerations
In the older low-temperature hot water heating system design, a water temperature drop of Tw 
20°F (11.1°C) at the finned-tube element was usually used. The current trend is to use a greater
temperature drop, such as 20 to 50°F (11.1 to 27.8°C). A detailed cost analysis should be performed
to determine the optimum temperature drop in a finned-tube heater.
The sizing of low-temperature hot water pipes is usually based on a pressure drop of 0.5 to
1.5 psi (1 to 3 ft head loss per 100 ft or 1 to 3 m head loss per 100 m of pipe length). Friction charts
for hot water steel pipes and the equivalent length of pipe fittings are discussed in Chap. 7.
HEATING SYSTEMS, FURNACES, AND BOILERS 8.25
FIGURE 8.8 Baseboard finned-tube heater.
TABLE 8.1 Heating Capacity of Finned-Tube Elements for Entering Air Temperature of 65°F, Btu/h ft
Average water temperature, °F
Type of finned-tube heater No. of rows 220 210 200 190 180 170
Steel tube: 1.25-in. dia., steel fin 1 1260 1140 1030 940 830 730
2 2050 1850 1680 1520 1350 1190
Copper tube: 1-in. dia., aluminum fin 1 1000 900 820 740 660 580
2 1480 1340 1210 1100 970 860
Source: Adapted with permission from Handbook of HVAC Design 1990.
For small low-temperature hot water heating systems, an open expansion tank is usually used.
For medium-size and large systems, a diaphragm tank may be more suitable. Circulating water
pumps are often on-line pumps (circulators) with low pump head.
Part-Load Operation and Control
In a low-temperature hot water heating system, one of the basic part-load controls is the variation of
the hot water supply temperature from the boiler in response to a variation in outdoor temperature.
For instance, in the midwest, a low-temperature hot water heating system has a winter design
outdoor temperature To  0°F (17.8°C). At the winter design outdoor temperature, the hot water
supply temperature Tws is 190°F (88°C). When the outdoor temperature drops, Tws is reset as follows:
A hot water temperature sensor located at the hot water exit of the boiler, whose set point is
reset by an outdoor temperature sensor, is used to control the firing rate of the boiler by means of a
DDC unit. Zone control can be performed better by sensing the hot water temperature that returns
from each individual loop and then modulating the control valve to vary the mass flow rate of hot
water supplied to that zone through an on-line circulating pump (see Fig. 8.7).
ASHRAE/IESNA Standard 90.1-1999 specifies that a hot water system with a design heating
capacity exceeding 300,000 Btu/ h (88 kW) shall include controls to automatically reset to a lower
hot water supply temperature according to building loads or outdoor temperature during part-load
for energy saving except for systems that use variable flow to reduce pumping energy.
For a low-temperature hot water heating system installed with multiple boilers, the control strategy
is to decide when to turn a boiler on or off. This optimum operation control can be accomplished
by using a microprocessor-based controller that fires a standby boiler according to a preprogrammed
software instruction. This strategy depends on not only the increase or reduction in heating demand
but also how the operating cost of such a hot water heating system can be minimized.
ASHRAE/IESNA Standard 90.1-1999 also specifies that for a boiler plant equipped with more than
one boiler, provisions shall be made so that the hot water flow in the boiler plant can be automatically
reduced when a boiler is shut down.
Example 8.1. A two-pipe individual-loop low-temperature hot water heating system is used to
heat a factory that has a layout shown in Fig. 8.7. At winter design conditions, hot water is supplied
to the heated space at a temperature of 190°F (88°C) and returns from the baseboard finned-tube
heaters at a temperature of 150°F (65°C).
1. If the space heating load for the largest room (facing north) is 90,000 Btu/h (26.4 kW) and for
the northwest corner room is 11,000 Btu/h (3.2 kW), and if steel tubes and fins are used, determine
the number of feet of finned tubing required for each of these two rooms.
2. If a pressure drop of 1 ft/100 ft (1 m/100 m) of pipe is used and the hot water system is
equipped with an open expansion tank, determine the diameter of the hot water supply main for
these two rooms.
3. Divide this hot water system into appropriate control zones, or individual loops.
Solution
1. For the largest room, if a two-row finned-tube heater is used, then from Table 8.1, for an average
hot water temperature of (190  150)/2  170°F (77°C), the heat output of each foot of two-row
Outdoor temperature To, °F (°C) Supply temperature Tws, °F (°C)
0 (17.8) 190 (88)
32 (0) 135 (57)
60 (16) 85 (29)
8.26 CHAPTER EIGHT
finned tube is 1190 Btu/h (350 W). The number of feet required is therefore
For the northwest corner room, a single-row finned-tube heater is used. The number of feet required
is
Along the north side’s external wall, an 8-ft (2.5-m) finned tube is mounted. On the west side’s
external wall, a 7-ft (2-m) finned tube is mounted.
2. Because the density of water at 170°F is 60.8 lb / ft3, from Eq. (7.1), the flow rate of hot water
supplied to the largest room and the northwest corner room can be calculated as
Then from the water friction chart for water in steel pipes (Fig. 7.2 in Chap. 7), for a pressure
drop of 1 ft/ 100 ft, the pipe diameter for the hot water supply main to the largest room and
northwest corner room is 1.25 in. (32 mm).
3. This two-pipe individual-loop low-temperature hot water heating system should be divided into
four zones: north, south, east, and west. Each zone, or individual loop, includes the rooms
whose outer walls face that direction (except the south zone, which also includes interior rooms
without external walls). Such a setup can offset the variation in solar radiation and the operating
conditions for each zone to prevent overheating of rooms facing south and underheating of
rooms facing north. For each individual loop, or control zone, a thermostat senses the return hot
water temperature and modulates the control valve and the flow rate of hot water supplied to that
loop by a DDC unit in order to meet the variations in zone heating load.
8.8 HYDRONIC RADIANT FLOOR-PANEL HEATING SYSTEMS
System Description
A hydronic (hot water) radiant floor-panel heating system provides heating and thermal comfort to
the occupants in an enclosed indoor environment primarily by means of the radiant heat transfer
from the floor panels. As with the hot water systems using finned-tube heaters, a hydronic radiant
floor-panel heating system consists of a hot water boiler, several individual-loop pipings with radiant
floor panels, an air separator, an expansion tank, circulating hot water pumps, control valves,
and accessories, as shown in Fig. 8.9a and b. Instead of finned-tube heaters, radiant floor panels are
used. For each individual loop that forms a control zone, there may be only one large radiant floor
panel, or several small panels connected in series.
Radiant Floor Panel
Tubes made of plastic, rubber, steel, or copper of diameters from 0.3875 to 1.0 in. (10 to 25 mm)
are often embedded in a floor slab with a tube spacing from 6 to 15 in. (150 to 375 mm) to form a
heating floor panel, as shown in Fig. 8.9c. Tubes are laid in the shape of serpentine to ensure an
even panel surface temperature. Floor slabs are often made of concrete with a thickness between

90,000  11,000
500(190  150)
 5.05 gpm
Vgal 
Qw
500(Twe  Twl)
11,000
730
 15.06 ft (15 ft, or 4.5 m)
90,000
1190
 75.63 ft (76 ft, or 23 m)
HEATING SYSTEMS, FURNACES, AND BOILERS 8.27
2 and 6 in. (50 and 150 mm). The thickness of the floor slab has a significant effect on its thermal
storage and slow response to heating load changes. Floor coverings like carpet and wood affect the
overall thermal resistance of the floor panel. The water temperature, therefore, should be increased
accordingly in order to provide the required amount of heat that is released from the floor panel. An
insulating layer and a moisture barrier under the concrete slab are often needed to reduce the heat
losses, as shown in Fig. 8.9c.
Thermal Characteristics of Floor Panel
According to ASHRAE Handbook 1996, HVAC Systems and Equipment, the heat flux of the radiant
heat transfer irradiated from a heating floor panel qr, Btu/h ft2 (W/m2), can be calculated as
(8.13)
where TRp  absolute temperature of effective panel surface,°R (K)
TR,AUST  absolute temperature of area-weighted average temperature of unheated surfaces exposed
to panels (AUST),°R (K)
The heat flux due to natural convection flows upward from a heated floor panel qc, Btu/h ft2 (W/m2), is
(8.14) qc  0.31(Tp  Ta)0.31(Tp  Ta)
qr  0.15  108(T 4
Rp  T 4
R,AUST )
8.28 CHAPTER EIGHT
(a)
(c)
(b)
DDC
unit controller
Outdoor
Supply
sensor
Supply main
Return main Mixing
valve
Boiler
Return
sensor
Plastic
tube
Concrete
Boiler
Floor
panel
(zone 1)
Pump
Thermostat
Floor
panel
(zone 2)
TO
TS
T1
TR
Control
value
T2
T1 T2
Insulation
layer
FIGURE 8.9 Hydronic radiant floor heating systems. (a) Zone control valves; (b) outdoor reset plus zone control valves; (c) floor panel;
(d) design nomograph. (Source: ASHRAE Transactions 1995 Part I, p. 1204, Kilkis and Coley. Reprinted by permission.)
where Tp  effective panel surface temperature, °F (°C)
Ta  space air temperature that surrounds floor panel, °F (°C)
Then the heat flux from the radiant heating floor panel to the space air due to the combined effect of
radiation and convection qu, Btu / h ft2 (W/m2), can be calculated as
qu  qr  qc (8.15)
If the tubes are embedded in the concrete slab with a tube spacing of 1 ft (12 in. or 0.3 m), the characteristic
panel R value ru, h ft2 °F/Btu (m2 °C/W), can be calculated as
(8.16)
where rt  thermal resistance of tube wall per unit tube spacing, h ft2 °F/Btu (m2°C/W)
M  tube spacing, ft (m)
rp  R value of floor panel, h ft2 °F/Btu (m2  °C/W)
rc  R value of floor-panel covering, h  ft2  °F/Btu (m2  °C/W)
ru  rt M  rp  rc
HEATING SYSTEMS, FURNACES, AND BOILERS 8.29
(AUST-Ta) F
8
0.1 0.2 0.3 0.4 0.5
0.6
0.7
0.8
0.9
1.0
1.2
1.4
1.6
1.8
2.0
2.5
3.0
4.0
6 4 2 0 2 Ceiling heating
Floor heating
P S
R
Use when
rc /rp ? 4
Floor
heating
0F
AUST 10
20
30 40 50
50
60
70
80
90
100 110
60
70 80 90 100 110
Ceiling
M
ru  rprcrtrs
Panel resistance line
ru  0.05
xp
xp
qu
Do DI
Ta
Tp
qd
Floor
Temperature difference (TpTa), F
Heating flux qu, Btu/h•ft2
40
35
30
25
20
15
10
5
0 0 0
10
20
30
40
50
60
70
6
9
M12
15
18
50
45
40
35
30
20
25
15
10
5
(d)
FIGURE 8.9 (Continued)
Design Nomograph
Kilkis and Coley (1995) noted that the mean hot water temperature needed to maintain a required
heat flux from the heating panel depends primarily upon the heating panel’s effective surface temperature,
its tube spacing, and its thermal resistance. Kilkis and Coley also recommended a design
nomograph for the heating floor panel to determine the required design parameters, as shown in
Fig. 8.9d, which comprises the following parameters:
 Mean hot water temperature Tw, m,  (AUST  Tw, add), °F (°C); here, Tw, add indicates the additional
water temperature,°F (°C).
 Heat flux released from the floor panel qu  Qrh /Ap, Btu/h ft2 /(W/m2). Here Qrh represents the
space heating load, Btu /h (W), and Ap the area of the floor panel(s), ft2 (m2).
 Tube spacing M, in. (mm).
 Temperature differential AUST  Ta , °F (°C).
 Panel temperature differential Tp  Ta , °F (°C).
During the design of a hydronic radiant floor heating system, for a given indoor design space air
temperature Ta, first AUST  Ta must be predicted. For a heating space having small outdoor exposures
like windows, skylights, external walls, and roof, AUST  Ta. For a room equipped with a normal
size of external window and outdoor exposures, AUST  Ta can be predicted to be 2°F (1.1°C).
For a given room, its winter indoor design temperature Ta is 68°F (20°C), the heat flux released
from the floor panel qu is 25 Btu/h ft2 (W/m2), tube spacing M  12 in. (300 mm), ru is 0.8
h ft2 °F/Btu (0.14 m2 °C/W), and rc /rp  4. Then at the intersection of horizontal line qu  25
Btu/h ft2 (79 W/m2) and incline line ru  0.8 h ft2 °F/Btu (0.14 m2 °C/W), point P (Fig. 8.9d),
draw a vertical line downward that intersects the tube spacing line M  12 in. (300 mm) at point R.
The reading of Tw, add is then 45°F (7.2°C). The required mean hot water temperature is
Tw, m  AUST  Tw, add  68  2  45  111°F (43.9°C)
To determine the effective panel surface temperature Tp (floor temperature), extend the horizontal
line of qu  25 Btu/h ft2 (79 W/m2) until it intersects the vertical line AUST  Ta  2°F at point
S. From Fig. 8.9d, Tp  Ta  13°F (7.2°C). Therefore, Tp  Ta  16  68  13  81°F (27.2°C).
Design Considerations
 Because of a higher space mean radiant temperature when a hydronic radiant floor heating system
is used, a lower winter indoor design temperature Ta, such as 68°F (20.0°C), is often adopted.
 For occupant comfort, the surface temperature of the floor heating panels Tp in occupied space
should be less than 84°F (28.9°C).
 In an interior heating space if only a hydronic floor heating system is installed instead of a
cooling/heating (air conditioning) system, an outdoor ventilating system is required to provide
outdoor air for the occupants.
 Setback of the space temperature during nighttime unoccupied periods produces less satisfactory
results for heavy floor heating panels. The surface temperature of the floor panels cannot drop
quickly when the temperature is set back and cannot pick up quickly either during the warm-up
period the next morning.
Control and Operations of Multizone Hydronic Radiant
Floor-Panel Heating System
Gibbs (1994) simulated and compared two different control strategies for hydronic radiant floor
heating systems:
8.30 CHAPTER EIGHT
 Pulse-width-modulated zone control. In this control, hot water is supplied from the boiler at a
constant temperature, as shown in Fig. 8.9a. A room thermostat with a proportional control mode
has a 3°F (1.7°C) proportional band and pulse-width-modulates the zone control valves 4 times
per hour. If the zone air temperature is at a set point Tset of 68°F (20°C), the on time of the zone
control valve is 7.5 min and the off time is also 7.5 min.
During a cold day in winter, hot water at a constant temperature of 120°F (48.9°C) is supplied to
various zones and returned at a temperature of 100°F (37.8°C). The surface of the floor panel is
maintained at a mean temperature of 82°F (27.8°C), and the space air temperature of zone 1 is
Tr1  68°F (20°C). As soon as the space heating load reduces, Tr1 increases to 68.3°F (20.2°C),
the thermostat in zone 1 pulse-width-modulates the control valve so that it is now on 6 min and off
9 min.
Pulse-width-modulated zone control is simple, effective, and lower in first cost. However, the
boiler is not protected from cold return water.
 Outdoor reset plus pulse-width-modulated zone control. In this control, a four-way mixing valve
is added to pulse-width-modulaed zone control, and a proportional-integral (PI) control is used
instead of proportional control, as shown in Fig. 8.9b. The hot water supply temperature Tws from
the boiler is no longer constant; Tws is reset according to outdoor temperature To. The zone
requires that the hottest supply temperature provide the indoor temperature feedback to the outdoor
reset control. When the zone air temperature is at the set point Trl  Tset, the control valve is
on (opened) all the time and cycles off only when the Trl 	 Tset.
To prevent excessive thermal stress in the boiler, when the return temperature is too cold to enter
the boiler and is sensed by the return temperature sensor TR as shown in Fig. 8.9b, part of the
return water is mixed with the recirculated hot water supply, until the return water temperature is
raised above a safe limit.
Outdoor reset plus pulse-width-modulated zone control is safer and more flexible and has more
continuously distributed hot water. On the other hand, it is more expensive.
System Characteristics and Applications
Compared with a warm air heating system, a hydronic radiant floor-panel heating system has the
following characteristics:
 It provides a better thermal comfort for occupants, because of less cold draft, and a higher mean
radiant temperature and an even temperature distribution.
 There is a lower natural infiltration rate and a lower winter indoor design temperature, and therefore,
it provides greater potential for saving energy.
 The peak load may be shaved (leveled) due to the thermal energy stored in the floor panels.
 Response to the load changes and heating capacity modulation are slower.
 It is higher in first cost if a separate outdoor ventilating system is installed.
Because of the use of plastic and rubber tubes for floor panels in recent years, radiant floor-panel
heating systems were used by many residences in Europe since 1980s, as well as for buildings in
locations where winter heating is dominant.
8.9 INFRARED HEATING
Basics
Infrared heating uses radiant heat transfer from a gas-fired or electrically heated high-temperature tube
or panel to provide heating to a localized area for the comfort of the occupants or the maintenance of a
HEATING SYSTEMS, FURNACES, AND BOILERS 8.31
suitable environment for a manufacturing process. Heat radiates from an infrared heater in the form of
electromagnetic waves in all directions. Most infrared heaters have reflectors to focus the radiation on
a localized target; hence, they are often known as beam radiant heaters. Infrared heating is widely
used in factories, warehouses, garages, gymnasiums, skating rinks, outdoor loading docks, and racetrack
stands.
In environments with a low indoor ambient air temperature during the winter, the warmth that an
occupant feels depends on the radiant energy absorbed by the occupant from all sources that have
temperatures higher than that of the indoor ambient air. This warmth is closely related to the effective
radiant flux Irad, or ERF Btu / h ft2/(W/m2). And ERF indicates the effective radiant flux that
must be added to an unheated space, Btu /h ft2/(W/m2). ERF and ERF can be calculated as
ERF  hr(Trad  Ta)
 h (To  Ta) (8.17a)
ERF  h(To  Tuo) (8.17b)
where hr  radiative heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
h  convective and radiative combined heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
Trad  mean radiant temperature, °F (°C)
Ta  indoor ambient air temperature, °F (°C)
To  operative temperature as defined in Eq. (4.6), °F (°C)
Tuo  operative temperature of unheated space, °F (°C)
In Eq. (8.17), Trad  Ta indicates the amount Trad is above the indoor ambient air Ta; it is important
because it influences the infrared radiant heat transfer to the occupant. There are two types of infrared
heaters: gas infrared heaters and electric infrared heaters.
Gas Infrared Heaters
Gas infrared heaters can be divided into indirect infrared heaters and porous matrix infrared heaters.
An indirect infrared heater consists of a burner, a radiating tube, and a reflector. Combustion takes
place within the radiating tube at a temperature up to 1200°F (650°C). In a porous matrix infrared
radiation heater, a gas-air mixture is supplied to an enclosure and distributed evenly through a
porous ceramic, stainless-steel, or metallic screen that is exposed at the other end. Combustion
takes place at the exposed surface and has a maximum temperature of about 1600°F (870°C).
Gas infrared heaters are usually vented and have a small conversion efficiency. Only 10 to 20
percent of the energy output of an open combustion gas infrared heater is in the form of infrared
radiant energy. Gas infrared heaters should not be operated under conditions in which the ambient
air contains ignitable gas or materials that may decompose to hazardous or toxic gases or
vapors.
Adequate combustion air must be provided. Venting is preferred in order to prevent a buildup of
combustion products. Usually, 4 cfm (1.9 L/s) of makeup air is required for 1000 Btu/h (293 W)
of gas input. If unvented infrared heaters are used, humidity and condensation control should be
provided to account for the accumulation of water vapor that forms during combustion.
A thermostat usually controls the supply of gas by means of a gas valve in on/off mode. For
standing-pilot ignition, a sensing element and controller are also used to cut off the gas supply
when the pilot flame extinguishes. If the combustion air is blocked, the gas supply is also cut off for
safety.
Electric Infrared Heaters
Electric infrared heaters are usually made of nickel-chromium wire or tungsten filament inside an
electrically insulated metal tube or quartz tube with or without an inert gas. The heaters also contain
8.32 CHAPTER EIGHT
a reflector, which directs the radiant beam to the area that needs heating. Nickel-chromium wires
often operate at a temperature between 1200 and 1800°F (650 and 980°C). The tungsten filament
can stand a temperature as high as 4000°F (2200°C) while the outer envelope is at a temperature of
about 1000°F (540°C).
Electric infrared heaters also use a thermostat to switch on or cut off the electric current to control
the input to the nickel-chromium electric heater. Some input controllers can preset the on/off
time period. For a quartz tube, the output can be controlled by varying its voltage supply. As with
gas infrared heaters, electric infrared heaters should not be used where there is a danger of igniting
flammable dust or vapors or decomposing contaminated matter into toxic gases.
Electric infrared heaters have a far higher infrared radiant energy conversion efficiency than gas
infrared heaters. They are clean and more easily managed. Although the cost of electric energy is
several times higher than that of natural gas, a comprehensive analysis should be conducted before
selecting an electric or gas infrared heater.
Design and Layout
Determination of Watt Density. According to tests, an acceptable temperature increase Trad  Ta
for normal clothed occupants in an indoor environment using infrared heating is often between 20
and 25°F (11.1 and 13.9°C). The effective radiant flux, or required watt density, provided by the
infrared heaters has been determined from field experiments, with the data listed in Table 8.2 for
different operating conditions.
Single or Multiple Heaters. Based on the required coverage area and the width and length of the
floor area that can be covered with the specified watt density listed in Table 8.3, it is possible to
determine whether a single heater or multiple heaters should be installed. It is also necessary to
select the coverage pattern: three-fourths overlap or full overlap. If there are many occupants within
the covering area or if the occupant will stay for a long period, a full-overlap pattern is preferable.
Mounting Height and Clearances. The mounting height of infrared heaters should not be below
10 ft (3 m); otherwise, the occupant may feel discomfort from the radiant beam overhead. Because
of the spread of the radiant beam, a higher mounting height results in a smaller watt density and a
greater coverage, and allows a greater spacing between heaters. An optimum mounting height
should be selected based on the ceiling height of the building or the outdoor structures.
Adequate clearance (recommended by the manufacturer) between the infrared heater and any
combustible material, especially between the heater and the roof, must be maintained to prevent fire.
Other Considerations. When infrared heaters are used for total indoor space heating, the following
must be taken into account:
HEATING SYSTEMS, FURNACES, AND BOILERS 8.33
TABLE 8.2 Required Watt Density,W/ ft2
Temperature Tight Drafty indoors Outdoor shielded, Outdoor unshielded,
increase uninsulated or large glass Loading area, less than 5 mph less than 10 mph
TradTa, qc building area one end open wind speed wind speed
20 30 40 50 55 60
25 37 50 62 70 75
30 45 60 75 85 90
40 60 80 100 115 120
50 75 100 125 145 150
Source: Abridged with permission from Handbook of HVAC Design 1990, Chapter 7 (Lehr Associates).
8.34 CHAPTER EIGHT
TABLE 8.3 Watt Density and Coverage
Two asymmetric heaters
Watt Wattt
Mounting density, density,
height, ft W/ft2 W  L, ft Spacing s, ft W/ft2 W  L, ft Spacing s, ft
10 33 13  12 8.5 38 11  12 6
11 27 14  13 9 33 12  13 6.5
12 24 15  14 10 29 12  14 7
13 21 16  15 10.5 25 13  15 7.5
14 18 17  16 11 22 14  16 8
15 16 18  17 12 20 15  17 8.5
16 14 20  18 12.5 18 16  18 9
18 11 22  20 14 14 18  20 10
20 9.6 24  22 15.5 11 20  22 11
22 8.0 26  24 17 9.6 22  24 12
25 6.4 30  27 19 8 24  27 13.5
Single heater
Watt Wat Wattt
Mounting density, density, density,
height, ft W/ft2 W  L, ft W/ft2 W  L, ft W/ft2 W  L, ft
10 18 12  12 25 8.5  12 19 11  12
11 15 13  13 22 9  13 16 12  13
12 13 14  14 18 10  14 14 12  14
13 11 15  15 16 10  15 13 13  15
14 10 16  16 14 11  16 11 14  16
15 9 17  17 13 12  17 9.6 15  17
16 8 18  18 11 12  18 8.8 16  18
18 6.4 20  20 8.8 14  20 7.2 18  20
20 5.2 22  22 7.6 15  22 5.8 20  22
22 4.4 24  24 6.2 17  24 4.8 24  24
25 3.5 27  27 5.0 19  27 3.9 24  27
W  width, L  length, both in ft.
Source: Abridged with permission from Handbook of HVAC Design 1990, Chapter 7 (Lehr Associates).
 The space heating load can be lower because of a lower space temperature.
 The area near the external walls needs a greater intensity of infrared heating.
 Heaters are often mounted around the perimeter at a suitable height and are usually arranged 15
to 20 ft (4.5 to 6 m) from the corners of the building.
 The key to successful design is to supply a proper amount of heat evenly to the occupying zone.
REFERENCES
Adams, C. W., Performance Results of a Pulse-Combustion Furnace Field Trial, ASHRAE Transactions, 1983,
Part I B, pp. 693–699.
The American Boiler Manufacturer Association, Why Packaged Firetube Boilers? Heating/Piping/Air Conditioning,
November 1990, pp. 79–83.
Andrews, J. W., Impact of Reduced Firing Rate on Furnace and Boiler Efficiency, ASHRAE Transactions, 1986,
Part I A, pp. 246–262.
ASHRAE, ASHRAE Handbook 1995, HVAC Applications, ASHRAE Inc., Atlanta, GA, 1995.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, Atlanta, GA, 1996.
Axtman,W. H., Boiler Types and General Selection Requirements, Heating/Piping/Air Conditioning, November
1987, pp. ABMA 5–9.
Census Bureau, Heating and Cooling Unit in Home 1991, AC Heating & Refrigeration News, August 28, 1995
pp. 23–24.
Cummings, J. B., and Tooley, J. J., Infiltration and Pressure Differences Induced by Forced Air Systems in
Florida Residences, ASHRAE Transactions, 1989, Part II, pp. 551–560.
DOE/EIA, 1998 Nonresidential Buildings Energy Consumption Survey: Commercial Buildings Consumption
and Expenditures 1995, DOE/EIA-0318(95).
Fischer, R. D., Jacob, F. E., Flanigan, L. J., and Locklin, D. W., Dynamic Performance of Residential Warm-Air
Heating Systems—Status of ASHRAE Project SP43, ASHRAE Transactions, 1984, Part II B, pp. 573–590.
Gammage, R. B., Hawthorne, A. R., and White, D. A., Parameters Affecting Air Infiltration and Air Tightness in
Thirty-one East Tennessee Homes, Measured Air Leakage in Buildings, ASTM STP 904, American Society of
Testing Materials, Philadelphia, 1986.
Gibbs, D. R., Control of Multizone Hydronic Radiant Floor Heating Systems, ASHRAE Transactions, 1994,
Part I, pp. 1003–1010.
Grimm, N. R., and Rosaler, R. C., Handbook of HVAC Design, McGraw-Hill, New York, 1990.
Int-Hout, D., Analysis of Three Perimeter Heating Systems by Air-Diffusion Methods, ASHRAE Transactions,
1983, Part I B, pp. 101–112.
Jakob, F. E., Fischer, R. D., and Flanigan, L. J., ASHRAE Transactions, 1987, Part I, pp. 1499–1514.
Jakob, F. E., Fischer, R. D., Flanigan, L. J., and Locklin, D. W., Validation of the ASHRAE SP43 Dynamic Simulation
Model for Residential Forced-Warm-Air Systems, ASHRAE Transactions, 1986, Part II B, pp.
623–643.
Kesselring, J. P., Blatt, M. H., and Hough, R. E., New Option in Commercial Heating, Heating/Piping/Air Conditioning,
July 1990, pp. 41–43.
Kilkis, B., and Coley, M., Development of a Complete Design Software for Hydronic Floor Heating of Buildings,
ASHRAE Transactions, 1995, Part I, pp. 1201–1213.
Lambert, L. A., and Robinson, D. H., Effects of Ducted Forced-Air Heating Systems on Residential Air Leakage
and Heating Energy Use, ASHRAE Transactions, 1989, Part II, pp. 534–541.
Modera, M. P., Residential Duct System Leakage: Magnitude, Impacts, and Potential for Reduction, ASHRAE
Transactions, 1989, Part II, pp. 561–569.
Palm, Jr., R. B., Pulse Combustion: A New Approach, Heating/Piping/Air Conditioning, January 1989,
pp. 148–150.
Parker, D. S., Evidence of Increased Levels of Space Heat Consumption and Air Leakage Associated with
Forced Air Heating Systems in Houses in the Pacific Northwest, ASHRAE Transactions, 1989, Part II,
pp. 527–533.
HEATING SYSTEMS, FURNACES, AND BOILERS 8.35
Patani, A., and Bonne, U., Operating Cost of Gas-Fired Furnace Heating Systems with Add-on Heat Pumps,
ASHRAE Transactions, 1983, Part I B, pp. 319–329.
Paul, D. D., Whitacre, G. R., Fischer, R. D., Rutz, A. L., Dewerth, D. W, Borgeson, R. A., and Leslie, N. P.,
Development of Vent Capacity Tables for Category I Gas Appliances with Fan Assisted Combustion Systems,
ASHRAE Transactions, 1993, Part I, pp. 1163–1179.
Robison, D. H., and Lambert, L. A., Field Investigation of Residential Infiltration and Heating Duct Leakage,
ASHRAE Transactions, 1989, Part II, pp. 542–550.
Slattery, L. T., A Look at Packaged Boilers, Heating/Piping/Air Conditioning, no. 12, 1995, pp. 65–72.
Spolek, G. A., Herriott, D. W., and Low, D. M., Air Flow in Rooms with Baseboard Heat: Flow Visualization
Studies, ASHRAE Transactions, 1986, Part II A, pp. 528–536.
Tao,W., Modern Boiler Plant Design, Heating/Piping/Air Conditioning, November 1984, pp. 69–82.
The Trane Company, Boilers for Steam and Hot Water, La Crosse, WI, 1971.
The Trane Company, Basics of Heating with Electricity, La Crosse, WI, 1973.
Trehan, A. K., Fortmann, R. C., Koontz, M. D., and Nagda, N. L., Effect of Furnace Size on Morning Picking
Time, ASHRAE Transactions, 1989, Part I, pp. 1125–1129.
Trewin, R. R., Langdon, F. M., Nelson, R. M., and Pate, M. B., An Experimental Study of A Multipurpose Commercial
Building with Three Different Heating Systems, ASHRAE Transactions, 1987, Part I, pp. 467–481.
8.36 CHAPTER EIGHT
CHAPTER 9
REFRIGERANTS,
REFRIGERATION CYCLES,
AND REFRIGERATION SYSTEMS
9.1
9.1 REFRIGERATION AND REFRIGERATION
SYSTEMS 9.2
9.2 REFRIGERANTS 9.3
Refrigerants, Cooling Media, and Liquid
Absorbents 9.3
Azeotropic, Near Azeotropic, and
Zeotropic 9.3
Numbering of Refrigerants 9.4
9.3 PROPERTIES AND CHARACTERISTICS
OF REFRIGERANTS 9.5
Safety Requirements 9.5
Effectiveness of Refrigeration Cycle 9.5
Evaporating and Condensing
Pressures 9.6
Oil Miscibility 9.6
Inertness 9.6
Thermal Conductivity 9.6
Refrigeration Capacity 9.6
Physical Properties 9.6
Operating Characteristics 9.6
9.4 PHASEOUT OF OZONE DEPLETION
REFRIGERANTS 9.7
Refrigerant Use 9.7
Ozone Depletion and Global Warming
Potentials 9.7
Phaseout of CFCs, Halons, and
HCFCs 9.10
Montreal Protocol and Clean Air Act 9.10
Action and Measures 9.11
Status of CFC Replacements 9.13
9.5 CLASSIFICATION OF
REFRIGERANTS 9.13
Hydrofluorocarbons 9.13
Azeotropic HFC 9.14
Near-Azeotropic HFC 9.14
Zeotropic HFC 9.15
HCFCs and Their Zeotropes 9.15
Inorganic Compounds 9.16
CFCs, Halons, and Their Zeotropes 9.16
9.6 REFRIGERATION PROCESSES AND
REFRIGERATION CYCLES 9.16
Refrigeration Processes 9.16
Refrigeration Cycles 9.17
Unit of Refrigeration 9.17
9.7 GRAPHICAL AND ANALYTICAL
EVALUATION OF REFRIGERATION 9.17
Pressure-Enthalpy Diagram 9.17
Temperature-Entropy Diagram 9.18
Analytical Evaluation of Cycle
Performance 9.19
9.8 CARNOT REFRIGERATION CYCLE 9.19
Performance of Carnot Refrigeration
Cycle 9.19
9.9 COEFFICIENT OF PERFORMANCE OF
REFRIGERATION CYCLE 9.21
9.10 SINGLE-STAGE IDEAL VAPOR
COMPRESSION CYCLE 9.22
Flow Processes 9.22
Cycle Performance 9.22
Determination of Enthalpy by
Polynomials 9.24
Refrigeration Effect, Refrigerating Load,
and Refrigerating Capacity 9.25
9.11 SUBCOOLING AND
SUPERHEATING 9.26
Subcooling 9.26
Superheating 9.26
9.12 MULTISTAGE VAPOR COMPRESSION
SYSTEMS 9.29
Compound Systems 9.29
Interstage Pressure 9.30
Flash Cooler and Intercooler 9.31
9.13 TWO-STAGE COMPOUND SYSTEM
WITH A FLASH COOLER 9.31
Flow Processes 9.31
Fraction of Evaporated Refrigerant in
Flash Cooler 9.31
Enthalpy of Vapor Mixture Entering
Second-Stage Impeller 9.32
Coefficient of Performance 9.33
Characteristics of Two-Stage Compound
System with Flash Cooler 9.33
9.14 THREE-STAGE COMPOUND SYSTEM
WITH A TWO-STAGE FLASH
COOLER 9.35
Flow Processes 9.35
Fraction of Refrigerant Vaporized in
Flash Cooler 9.35
Coefficient of Performance of
Three-Stage System 9.38
9.15 TWO-STAGE COMPOUND
SYSTEM WITH A VERTICAL
INTERCOOLER 9.38
Comparison between Flash Cooler and
Vertical Coil Intercooler 9.40
9.1 REFRIGERATION AND REFRIGERATION SYSTEMS
Refrigeration is defined as the process of extracting heat from a lower-temperature heat source,
substance, or cooling medium and transferring it to a higher-temperature heat sink. Refrigeration
maintains the temperature of the heat source below that of its surroundings while transferring
the extracted heat, and any required energy input, to a heat sink, atmospheric air, or surface
water.
A refrigeration system is a combination of components and equipment connected in a sequential
order to produce the refrigeration effect. The refrigeration systems commonly used for air
conditioning can be classified by the type of input energy and the refrigeration process as
follows:
1. Vapor compression systems. In vapor compression systems, compressors activate the refrigerant
by compressing it to a higher pressure and higher temperature level after it has produced
its refrigeration effect. The compressed refrigerant transfers its heat to the sink and is
condensed to liquid form. This liquid refrigerant is then throttled to a low-pressure, lowtemperature
vapor to produce refrigerating effect during evaporation. Vapor compression
systems are the most widely adopted refrigeration systems in both comfort and process air
conditioning.
2. Absorption systems. In an absorption system, the refrigeration effect is produced by thermal
energy input. After absorbing heat from the cooling medium during evaporation, the vapor
refrigerant is absorbed by an absorbent medium. This solution is then heated by direct-fired
furnace, waste heat, hot water, or steam. The refrigerant is again vaporized and then condensed
to liquid to begin the refrigeration cycle again.
3. Air or gas expansion systems. In an air or gas expansion system, air or gas is compressed
to a high pressure by mechanical energy. It is then cooled and expanded to a low pressure.
Because the temperature of air or gas drops during expansion, a refrigeration effect is
produced.
9.2 CHAPTER NINE
9.16 CASCADE SYSTEMS 9.40
Advantages and Disadvantages 9.40
Performance of Cascade System 9.42
9.17 AIR EXPANSION REFRIGERATION
CYCLES 9.45
Thermodynamic Principle 9.45
Flow Processes of a Basic Air Expansion
System for Aircraft 9.47
Air Expansion Cycle 9.48
9.18 REFRIGERATION SYSTEMS—
CLASSIFICATIONS AND
DEVELOPMENTS 9.49
Classifications 9.49
Recent Developments 9.51
9.19 REFRIGERATION
COMPRESSORS 9.51
Positive Displacement and Non–Positive
Displacement Compressors 9.51
Hermetic, Semihermetic, and Open
Compressors 9.53
Direct Drive, Belt Drive, and
Gear Drive 9.53
9.20 PERFORMANCE OF
COMPRESSORS 9.53
Volumetric Efficiency 9.53
Motor, Mechanical, and Compression
Efficiency 9.54
Isentropic and Polytropic Analysis 9.54
Energy Use Index 9.55
9.21 SAFETY REQUIREMENTS AND
MACHINERY ROOM 9.56
Refrigerant Safety 9.56
Application Rules for High-Probability
Systems 9.56
Application Rules for Low-Probability
Systems 9.57
Refrigerating Systems of 100 hp
(74.6 kW) or less 9.58
Refrigerating Machinery Room 9.58
Storage of Refrigerants 9.59
REFERENCES 9.59
9.2 REFRIGERANTS
Refrigerants, Cooling Media, and Liquid Absorbents
Refrigerants. A refrigerant is the primary working fluid used for absorbing and transmitting heat
in a refrigeration system. Refrigerants absorb heat at a low temperature and low pressure and
release heat at a higher temperature and pressure. Most refrigerants undergo phase changes during
heat absorption—evaporation—and heat releasing—condensation.
Cooling Media. A cooling medium is the working fluid cooled by the refrigerant to transport the
cooling effect between a central plant and remote cooling units and terminals. In a large, centralized
system, it is often more economical to use a coolant medium that can be pumped to remote locations
where cooling is required. Chilled water, brine, and glycol are used as cooling media in many
refrigeration systems. The cooling medium is often called a secondary refrigerant, because it obviates
extensive circulation of the primary refrigerant.
Liquid Absorbents. A solution known as liquid absorbent is often used to absorb the vaporized
refrigerant (water vapor) after its evaporation in an absorption refrigeration system. This solution,
containing the absorbed vapor, is then heated at high pressure. The refrigerant vaporizes, and the
solution is restored to its original concentration for reuse.
Lithium bromide and ammonia, both in a water solution, are the liquid absorbents used most
often in absorption refrigerating systems.
Azeotropic, Near Azeotropic, and Zeotropic
A refrigerant can either be a single chemical compound or a mixture (blend) of multiple compounds.
Azeotropic. These are blends of multiple components of volatilities (refrigerants) that evaporate
and condense as a single substance and do not change their volumetric composition or saturation
temperature when they evaporate or condense at a constant pressure. Components in a mixture of
azeotropes cannot be separated from their constituents by distillation. Properties of azeotropic
refrigerants are entirely different from those of their components and may be conveniently treated
as a single chemical compound.
Near Azeotropic. Near-azeotropic refrigerants are blends whose characteristics are near to
azeotropic. Although properties of near-azeotropic refrigerants are nearer to azeotropic than to nonazeotropic
(zeotropic), near-azeotropic refrigerants are defined as zeotropic or nonazeotropic.
Zeotropic. These are blends of multiple components of volatilities (refrigerants) that evaporate
and condense as a single substance and do change volumetric composition or saturation temperature
when they evaporate or condense at a constant pressure.
Blends. Mixtures of refrigerants of two or more chemical compounds are blends. The advantage
of a blend of multiple chemical compounds compared to a single compound is that the required
properties of the blend can possibly be achieved by varying the fractional composition of the
components.
Glide. Zeotropic mixtures, including near-azeotropic blends, show changes in composition because
of the leaks, the difference between liquid and vapor phases, or the difference between the
charge and circulation, or their combined effect. The shift in composition causes the change in
evaporating and condensing temperature and pressure. The difference in dew point and bubble point
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.3
in the temperature-concentration diagram of a zeotropic refrigerant during evaporation and condensation
is called glide, expressed in °F (°C). A near-azeotropic refrigerant has a smaller glide than
a zeotropic one. The midpoint between the dew point and bubble point is usually taken as the
evaporating or condensing temperature for a nonazeotropic and near-azeotropic refrigerant. Hwang
et al. (1997) showed that temperature drops during condensation and temperature increases during
evaporation. Ideal or perfect azeotropic refrigerants are uncommon, whereas near-azeotropic ones
are fairly common.
Numbering of Refrigerants
Before the invention of chlorofluorocarbons (CFCs), refrigerants were called by their chemical names.
Because of the complexity of these names, especially the CFCs, the fully halogenated CFCs, and
hydrochlorofluorocarbons (HCFCs), the not fully halogenated HCFCs (see Fig. 9.1), a numbering system
was developed for hydrocarbons and halocarbons, and is used widely in the refrigeration industry.
According to ANSI/ASHRAE Standard 34-1997, the first digit from the right is the number of fluorine
atoms in the compound. The second digit from the right is one more than the number of hydrogen
atoms in the compound. The third digit from the right is one less than the number of the carbon atoms
in the compound. If the digit is zero, it is omitted from the number. The fourth digit from the right is
the number of unsaturated carbon-carbon bonds in the compound. If the digit is zero it is also omitted
from the number. For example, the chemical formula of HCFC-123 is CHCl2CF3:
There are 3 fluorine atoms, first digit from the right is 3
There is 1 hydrogen atom, second digit from the right is 1  1  2
There are 2 carbon atoms, third digit from the right is 2  1  1
No unsaturated C9C bonds, the fourth digit from the right is 0
9.4 CHAPTER NINE
FIGURE 9.1 Fully halogenated CFCs and not fully halogenated HCFCs.
9.3 PROPERTIES AND CHARACTERISTICS OF REFRIGERANTS
Today, the preservation of the ozone layer is the first priority of refrigerant selection. In addition,
the global warming effect and the following factors should be considered.
Safety Requirements
Refrigerant may leak from pipe joints, seals, or component parts during installation, operation, or
accident. Therefore, refrigerants must be acceptably safe for humans and manufacturing processes,
with little or no toxicity or flammability.
In ANSI/ASHRAE Standard 34-1997, the toxicity of refrigerants is classified as class A or B.
Class A refrigerants are of lower toxicity. A class A refrigerant is one whose toxicity has not been
identified when its concentration is less than or equal to 400 ppm, based on threshold limit
value–time-weighted average (TLV-TWA) or equivalent indices. The TLV-TWA concentration is a
concentration to which workers can be exposed over an 8-h workday and a 40-h workweek without
suffering adverse effect. Concentration ppm means parts per million by mass.
Class B refrigerants are of high toxicity. A class B refrigerant produces evidence of toxicity
when workers are exposed to a concentration below 400 ppm based on a TLV-TWA concentration.
Flammable refrigerants explode when ignited. If a flammable refrigerant is leaked in the area of a
fire, the result is an immediate explosion. Soldering and welding for installation or repair cannot be
performed near such gases.
ANSI/ASHRAE Standard 34-1997 classifies the flammability of refrigerants into classes 1, 2,
and 3. Class 1 refrigerants show no flame propagation when tested in air at a pressure of 14.7 psia
(101 kPa) at 65°F (18.3°C). Class 2 refrigerants have a lower flammability limit (LFL) of more than
0.00625 lb/ ft3 (0.1 kg/m3) at 70°F (21.1°C) and 14.7 psia (101 kPa abs.), and a heat of combustion
less than 8174 Btu/lb (19,000 kJ/kg). Class 3 refrigerants are highly flammable, with an LFL less
than or equal to 0.00625 lb/ ft3 (0.1 kg/m3) at 70°F (21.1°C) and 14.7 psia (101 kPa abs.) or a heat
of combustion greater than or equal to 8174 Btu /lb (19,000 kJ/ kg).
A refrigerant’s safety classification is its combination of toxicity and flammability. According to
ANSI/ASHRAE Standard 34-1997, safety groups are classified as follows:
 A1 lower toxicity and no flame propagation
 A2 lower toxicity and lower flammability
 A3 lower toxicity and higher flammability
 B1 higher toxicity and no flame propagation
 B2 higher toxicity and lower flammability
 B3 higher toxicity and higher flammability
For zeotropic blends whose flammability and toxicity may change as their composition changes, a
dual safety classification should be determined. The first classification denotes the classification of
the formulated composition of the blend. The second classification lists the classification of the
blend composition at the worst case of fractionation.
Effectiveness of Refrigeration Cycle
The effectiveness of refrigeration cycles, or coefficient of performance (COP), is one parameter that
affects the efficiency and energy consumption of the refrigeration system. It will be clearly defined
in a later section. The COP of a refrigeration cycle using a specific refrigerant depends mainly upon
the isentropic work input to the compressor at a given condensing and evaporating pressure differential,
as well as the refrigeration effect produced.
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.5
Evaporating and Condensing Pressures
It is best to use a refrigerant whose evaporating pressure is higher than that of the atmosphere so
that air and other noncondensable gases will not leak into the system and increase the condensing
pressure. The condensing pressure should be low because high condensing pressure necessitates
heavier construction of the compressor, piping, condenser, and other components. In addition, a
high-speed centrifugal compressor may be required to produce a high condensing pressure.
Oil Miscibility
When a small amount of oil is mixed with refrigerant, the mixture helps to lubricate the moving
parts of a compressor. Oil should be returned to the compressor from the condenser, evaporator, accessories,
and piping, in order to provide continuous lubrication. On the other hand, refrigerant can
dilute oil, weakening its lubricating effect; and when the oil adheres to the tubes in the evaporator
or condenser, it forms film that reduces the rate of heat transfer.
Inertness
An inert refrigerant does not react chemically with other materials, thus avoiding corrosion, erosion,
or damage to the components in the refrigerant circuit.
Thermal Conductivity
The thermal conductivity of a refrigerant is closely related to the efficiency of heat transfer in the
evaporator and condenser of a refrigeration system. Refrigerant always has a lower thermal conductivity
in its vapor state than in its liquid state. High thermal conductivity results in higher heat transfer
in heat exchangers.
Refrigeration Capacity
The cubic feet per minute (cfm) suction vapor of refrigerant required to produce 1 ton of refrigeration
(liters per second to produce 1 kW of refrigeration) depends mainly on the latent heat of vaporization
of the refrigerant and the specific volume at the suction pressure. It directly affects the size
and compactness of the compressor and is one of the criteria for refrigerant selection.
Physical Properties
Discharge Temperature. A discharge temperature lower than 212°F (100°C) is preferable because
temperatures higher than 300°F (150°C) may carbonize lubricating oil or damage some of the
components.
Dielectric Properties. Dielectric properties are important for those refrigerants that will be in direct
contact with the windings of the motor (such as refrigerants used to cool the motor windings in
a hermetically sealed compressor and motor assembly).
Operating Characteristics
Leakage Detection. Refrigerant leakage should be easily detected. If it is not, gradual capacity reduction
and eventual failure to provide the required cooling will result. Most of the currently used
refrigerants are colorless and odorless. Leakage of refrigerant from the refrigeration system is often
detected by the following methods:
9.6 CHAPTER NINE
 Halide torch. This method is simple and fast. When air flows over a copper element heated by a
methyl alcohol flame, the vapor of halogenated refrigerant decomposes and changes the color of
the flame (green for a small leak, bluish with a reddish top for a large leak).
 Electronic detector. This type of detector reveals a variation of electric current due to ionization
of decomposed refrigerant between two oppositely charged electrodes. It is sensitive, but cannot
be used where the ambient air contains explosive or flammable vapors.
 Bubble method. A solution of soup or detergent is brushed over the seals and joints where leakage
is suspected, producing bubbles that can be easily detected.
9.4 PHASEOUT OF OZONE DEPLETION REFRIGERANTS
Refrigerant Use
The use of CFCs and HCFCs is a global concern. Approximately two-thirds of all fully halogenated
CFCs were used outside the United States in the mid-1980s. In 1985, the total use of halocarbons
in the United States was 611 million lb (0.28 million ton). These halocarbons were used in foam insulation,
automotive air conditioners, new systems of Air Conditioning and Refrigeration Institute
(ARI) members, and other products. Foam insulation blown by CFCs was the largest user. Automotive
air conditioners made up 19 percent of the total and CFCs purchased by ARI members for new
systems made up 5 percent of the total use. Of the CFCs and HCFCs purchased by ARI members,
HCFC-22 made up 77 percent, while CFC-11 and CFC-12 each made up about 10 percent.
Ozone Depletion and Global Warming Potentials
To compare the relative ozone depletion caused by various refrigerants, an index called the ozone
depletion potential (ODP) has been proposed. ODP is the ratio of the rate of ozone depletion of 1 lb
of any halocarbon to that of 1 pound of CFC-11. The ODP of CFC-11 is assigned a value of 1. The
following are the ODP values for various refrigerants:
Similar to the ODP, the halocarbon global warming potential (HGWP) is the ratio of calculated
warming for each unit mass of gas emitted to the calculated warming for a unit mass of reference
gas CFC-11. The HGWPs of various refrigerants are listed in Table 9.1. In addition to the HGWP,
another global warming index uses CO2 as a reference gas. For example, 1 lb of HCFC-22 has the
Refrigerant Chemical formula ODP value
CFC-11 CCl3F 1.0
CFC-12 CCl2F2 1.0
CFC-13B1 CBrF3 10
CFC-113 CCl2FCClF2 0.8
CFC-114 CClF2CClF2 1.0
CFC-115 CClF2CF3 0.6
CFC/HFC-500 CFC-12 (73.8%)/HFC-152a (26.2%) 0.74
CFC/HCFC-502 HCFC-22 (48.8%)/CFC-115 (51.2%) 0.33
HCFC-22 CHClF2 0.05
HCFC-123 CHCl2CF3 0.02
HCFC-124 CHClFCF3 0.02
HCFC-142b CH3CClF2 0.06
HFC-125 CHF2CF3 0
HFC-134a CF3CH2F 0
HFC-152a CH3CHF2 0
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.7
9.8
TABLE 9.1 Properties of Commonly Used Refrigerants at 40°F Evaporating and 100°F Condensing
Halocarbon
Ozone global
depletion warming Evaporating Condensing
Chemical Molecular potential potential pressure, pressure, Compression Refrigeration
formula mass (ODP) (HGWP) psia psia ratio effect, Btu/ lb
Hydrofluorocarbons (HFCs)
HFC-32 Difluoromethane CH2F2 52.02 0 0.11 135.6 340.2 2.51
HFC-125 Pentafluoroethane CHF2CF3 120.02 0 0.84 112.4 276.8 2.46 36.4
HFC-134a Tetrafluoroethane CF3CH2F 102.3 0 0.28 49.8 138.9 2.79
HFC-143a Trifluoroethane CH3CF3 84.0 0 1.1
HFC-152a Difluoroethane CHF2CH3 66.05 0 0.03 44.8 124.3 2.77
HFC-245ca Pentafluoropropane CF3CF2CH3 134.1 0 0.09
Azeotropic HFC
HFC-507 HFC-125/HFC-143a(45/55) 0 0.98
HFC-507A HFC-125/HFC-143a(50/50) 0 104.6 257.6 2.46
Near-azeotropic HFC
HFC-404A HFC-125/HFC-143a/
HFC-134a (44/52/4) 0 0.95 109.16 251.18 2.30
HFC-410A HFC-32/HFC-125(50/50) 0 0.48 132.90 332.33 2.50
HFC Zeotropic
HFC-407A HFC-32/HFC-125/
HFC-134a (20/40/40) 0 0.47
HFC 407C HFC-32/HFC-125/
HFC-134a (23/25/52) 0 0.38 86.13 225.22 2.61
Hydrochlorofluorocarbons (HCFCs) and their zeotropes
HCFC-22 Chlorodifluoromethane CHClF2 86.48 0.05 0.34 82.09 201.5 2.46 69.0
HCFC-123 Dichlorotrifluoroethane CHCl2CF3 153.0 0.016 0.02 5.8 20.8 3.59 62.9
HCFC-124 Chlorotetrafluoroethane CHFClCF3 136.47 0.02 0.10 27.9 80.92 2.90 52.1
Near-azeotropic HCFC
HCFC-402A HCFC-22/HFC-125/
PRO-290 (38/60/2) 0.02
Zeotropic HCFC
HCFC-401A HCFC-22/HCFC-124/
HFC-152a (53/34/13) 0.037 0.22
HCFC-401B HCFC-22/HCFC-124/
HFC-152a (61/28/11) 0.04 0.24
Inorganic compounds
R-717 Ammonia NH3 17.03 0 0 71.95 206.81 2.87 467.4
R-718 Water H2O 18.02 0
R-729 Air 28.97 0
Chlorofluorocarbons (CFCs), halons, and their azeotropes
CFC-11 Trichlorofluoromethane CCl3F 137.38 1.00 1.00 6.92 23.06 3.33 68.5
CFC-12 Dichlorodifluoromethane CCl2F2 120.93 1.00 3.1 50.98 129.19 2.53 50.5
BFC-13B1 Bromotrifluoromethane CBrF3 148.93 10
CFC-113 Trichlorotrifluoroethane CCl2FCClF2 187.39 0.80 1.4 2.64 10.21 3.87 54.1
CFC-114 Dichlorotetrafluoroethane CCl2FCF3 170.94 1.00 3.9 14.88 45.11 3.03 42.5
CFC-500 CFC-12/HFC-152a (73.8/26.2) 99.31 0.74 59.87 152.77 2.55 60.5
CFC-502 HCFC-22/CFC-115 (48.8/51.2) 111.63 0.22 3.7
Source: Adapted with permission from ASHRAE Handbook 1997, Fundamentals, and ANSI/ASHRAE Standard 34-1992 and Addenda 1997.
9.9
Specific
volume of Power Critical Discharge
Trade suction Compressor, consumption, temperature, temperature,
Alternatives name vapor, ft3/ lb cfm/ton hp/ton °F °F Flammability Safety
Hydrofluorocarbons (HFCs)
HFC-32 0.63 173.1 Lower flammability A2
HFC-125 0.33 150.9 103 Nonflammable A1
HFC-134a CFC-12, HCFC-22 0.95 213.9 Nonflammable A1
HFC-143a Lower flammability A2
HFC-152a 1.64 236.3 Lower flammability A2
HFC-245ca CFC-11, HCFC-123 353.1
Azeotropic HFC
HFC-507 CFC-502, CFC-12 Genetron AZ-50 Nonflammable A1
HFC-507A 0.43 Nonflammable A1
Near-azeotropic HFC
HFC-404A CFC-502, CFC-12 SUVA HP-62 0.44 A1/A1
HFC-410A HCFC-22, CFC-502 AZ-20 0.45 A1/A1
Zeotropic HFC A1/A1
HFC-407A CFC-502, CFC-12 KLEA60, AC-9000 A1/A1
HFC-407C HCFC-22, CFC-502 KLEA66 0.63
Hydrochlorofluorocarbons (HFCs) and their zeotropes
HCFC-22 0.66 1.91 0.696 204.8 127 Nonflammable A1
HCFC-123 CFC-11 5.88 18.87 0.663 362.6 Nonflammable B1
HCFC-124 1.30 5.06 0.698 252.5 Nonflammable A1
Near-azeotropic HCFC
HCFC-402A CFC-502 SUVAHP-80 A1/A1
Zeotropic HCFC
HCFC-401A CFC-12 MP-39 A1/A1
HCFC-401B CFC-12 MP-66 A1/A1
Inorganic compounds
R-717 3.98 1.70 0.653 271.4 207 Lower flammability B2
R-718 Nonflammable A1
R-729 Nonflammable
Chlorofluorocarbons (CFCs), halons, and their azeotropes
CFC-11 5.43 15.86 0.636 388.4 104 Nonflammable A1
CFC-12 0.79 3.08 0.689 233.6 100 Nonflammable A1
BFC-13B1 0.21 152.6 103 Nonflammable A1
CFC-113 10.71 39.55 0.71 417.4 86 Nonflammable A1
CFC-114 2.03 9.57 0.738 294.3 86 Nonflammable A1
CFC-500 CFC-12/HFC152a (73.8/26.2) 0.79 3.62 0.692 221.9 105 Nonflammable A1
CFC-502 HCFC-22/CFC115 (48.8/51.2) 98 Nonflammable A1
same effect on global warming as 4100 lb (1860 kg) of CO2 in the first 20 years after it is released
into the atmosphere. Its impact drops to 1500 lb (680 kg) at 100 years.
Phaseout of CFCs, Halons, and HCFCs
The theory of depletion of the ozone layer was proposed in 1974 by Rowland and Molina. (The
1995 Nobel Prize was awarded to F. Sherwood Rowland, Mario Molina, and Paul Crutzen for their
work in atmospheric chemistry and theory of ozone depletion.) Network station in Halley Bay,
Antarctica, established a baseline trend of ozone levels that helped scientists to discover the ozone
hole in 1985. National Aeronautics and Space Administration (NASA) flights into the stratosphere
over the arctic and antarctic circles found CFC residue where the ozone layer was damaged. Approximately
the same ozone depletion over the antarctic circle was found in 1987, 1989, 1990, and
1991. By 1988, antarctic ozone levels were 30 percent below those of the mid-1970s. The most severe
ozone loss over the antarctic was observed in 1992. Ground monitoring at various locations
worldwide in the 1980s has showed a 5 to 10 percent increase in ultraviolet radiation. Although
there is controversy about the theory of ozone layer depletion among scientists, as discussed in
Rowland (1992), action must be taken immediately before it is too late.
Montreal Protocol and Clean Air Act
In 1978, the Environmental Protection Agency (EPA) and the Food and Drug Administration (FDA)
of the United States issued regulations to phase out the use of fully halogenated CFCs in nonessential
aerosol propellants, one of the major uses at that time. On September 16, 1987, the European
Economic Community and 24 nations, including the United States, signed the Montreal Protocol.
This document is an agreement to phase out the production of CFCs and halons by the year 2000.
The Montreal Protocol had been ratified by 157 parties.
The Clean Air Act Amendments, signed into law in the United States on November 15, 1990,
governed two important issues: the phaseout of CFCs and a ban (effective July 1, 1992) on the deliberate
venting of CFCs and HCFCs. Deliberate venting of CFCs and HCFCs must follow the regulations
and guidelines of the EPA. In February 1992, then-President Bush called for an accelerated
phaseout of the CFCs in the United States. Production of CFCs must cease from January 1, 1996,
with limited exceptions for service to certain existing equipment.
In late November 1992, representatives of 93 nations meeting in Copenhagen also agreed to the
complete cessation of CFC production beginning January 1, 1996, and of halons by January 1,
1994, except continued use from existing (reclaimed or recycled) stock in developed nations. In
addition, the 1992 Copenhagen amendments and later a 1995 Vienna meeting revision agreed to
restrict the production of HCFCs relative to a 1989 level beginning from 2004 in developed nations
according to the following schedule:
Consumption indicates the production plus imports minus exports and feedstocks. The value of
2.8 percent cap is the revised value of the Vienna meeting in 1995 to replace the original value of
Date Production limit
January 1, 1996 100 percent cap
Cap  2.8 percent of ODP of 1989 CFC consumption
plus total ODP of 1989 HCFC consumption
January 1, 2004 65 percent cap
January 1, 2010 35 percent cap
January 1, 2015 10 percent cap
January 1, 2020 0.5 percent cap
January 1, 2030 Complete cessation of production
9.10 CHAPTER NINE
3.1 percent in the Copenhagen amendments. The Copenhagen amendments had been ratified by 58
parties.
Action and Measures
The impact of CFCs on the ozone layer poses a serious threat to human survival. The following
measures are essential:
Conversions and Replacements. Use alternative refrigerants (substitutes) to replace the CFCs in
existing chillers and direct-expansion (DX) systems. During the conversion of the CFC to
non–ozone depletion alternative refrigerants, careful analysis should be conducted of capacity, ef-
ficiency, oil miscibility, and compatibility with existing materials after conversion. For many refrigeration
systems that already have a service life of more than 15 years, it may be cost-effective to
buy a new one using non-CFC refrigerant to replace the existing refrigeration package.
 HFC-134a and HCFC-22 are alternative refrigerants to replace CFC-12.
 HCFC-123, and HFC-245ca are alternative refrigerants to replace CFC-11 in large chillers.
It is important to realize that HCFC-123 and HCFC-22 themselves are interim refrigerants and will
be restricted in consumption beginning in 2004. HCFC-123 has a very low global warming potential and
is widely used in centrifugal chillers. HCFC-22 is widely used in small and medium-size DX systems.
 HFC-134a, HFC-407C, and HFC-410A are alternative refrigerants to replace HCFC-22. HFC-
407C is a near-azeotropic refrigerant of HFC-32/HFC-125/HFC-134a (23/25/52) [means
(23%/25%/525)], and HFC-410A also a near-azeotropic refrigerant of HFC-32/HFC-125 (50/50).
 HFC-245ca or another new HFC possibly developed before 2004 will be the hopeful alternative
to replace HCFC-123.
In supermarkets, CFC-502 is a blend of HCFC-22/CFC-115 (48.8/ 51.2).
 HFC-404A, HFC-507, and HFC-410A are alternative refrigerants to replace CFC-502. HFC-
404A is a near-azeotropic refrigerant of HFC-125/HFC-143a/HFC-134a(44/52/4); and HFC-
507 is an azeotropic refrigerant of HFC-125/HFC-143a (45/ 55).
Reducing Leakage and Preventing Deliberate Venting. To reduce the leakage of refrigerant from
joints and rupture of the refrigeration system, one must detect the possible leakage, tighten the
chillers, improve the quality of sealing material, and implement preventive maintenance.
Prevent the deliberate venting of CFCs and HCFCs and other refrigerants during manufacturing,
installation, operation, service, and disposal of the products using refrigerants.
Avoid CFC and HCFC emissions through recovery, recycle, and reclaiming. According to
ASHRAE Guideline 3-1990, recovery is the removal of refrigerant from a system and storage in an
external container. Recycle involves cleaning the refrigerant for reuse by means of an oil separator
and filter dryer. In reclamation, refrigerant is reprocessed for new product specifications.
To avoid the venting of CFCs and HCFCs and other refrigerants, an important step is to use an
ARI-certified, portable refrigerant recovering/ recycling unit to recover all the liquid and remaining
vapor from a chiller or other refrigeration system. An outside recovery/ reclaiming service firm may
also be employed. A typical refrigerant recovery unit is shown in Fig. 9.2. It includes a recovery
cylinder, a vacuum pump or compressor, a water-cooled condenser, a sight glass, a shutoff float
switch, necessary accessories, pipes, and hoses.
To recover refrigerant from a chiller that has been shut down involves two phases: liquid recovery
and vapor recovery. Liquid recovery is shown in Fig. 9.2a. The vacuum pump or compressor
in the recovery unit creates a low pressure in the recovery cylinder. Liquid refrigerant is then
extracted from the bottom of the chiller into the recovery cylinder. If the recovery cylinder is not
large enough, the shutoff float switch ceases to operate the vacuum pump or compressor when the
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.11
recovery cylinder is 80 percent full. Another empty recovery cylinder is used to replace the filled
cylinder. If the vapor enters the sight glass, which means that the liquid refrigerant is all extracted,
then the vacuum pump or compressor is stopped and the vapor recovery phase begins.
Vapor recovery is shown in Fig. 9.2b. The vacuum pump or compressor extracts the refrigerant
vapor from the top of the chiller. Extracted refrigerant vapor is then condensed to liquid form that
flows through the water-cooled condenser and is stored in the recovery cylinder. Noncondensable
gases are purged into the atmosphere from the recovery unit. Water at a temperature between 40
and 85°F (4.4 and 29.4°C) is often used as the condensing cooling medium. The recovered refrigerant
can be recycled or reclaimed as required.
In addition to the recovery of refrigerants from the chiller or other refrigeration system, refrigerant
vapor detectors should be installed at locations where refrigerant from a leak is likely to
concentrate. These detectors can set off an alarm to notify the operator to seal the leak.
9.12 CHAPTER NINE
FIGURE 9.2 A typical refrigerant recovery unit: (a) liquid recovery; (b) vapor recovery.
Because of the worldwide effort to phase out CFCs, the latest result of a study conducted by the
National Oceanic and Atmospheric Administration (NOAA) and published in the journal Science
was reported by UPI science writer Susan Milius “. . . scientists are cheering a 1 percent reduction
during 1995 in the chemicals that slowly carry chlorine and bromine aloft to the stratosphere.”
(Air Conditioning, Heating and Refrigeration News, June 10, 1996, p. 2).
Status of CFC Replacements
Dooley (1997) showed that according to the ARI survey, in the United States there were about
80,000 CFC large-tonnage chillers in 1992, and most used CFC-11 as refrigerant. At the beginning
of 1997, some 18,981 chillers, or 24 percent of the total 80,000, had been phased-out CFC refrigerants.
Of these, 4813 chillers were converted to non-CFC refrigerant, and 14,168 chillers were
replaced by new chillers which used non-CFC refrigerant. The ratio of new replacements to conversions
was about 3 to 1. It is often cost-effective to replace an old chiller instead of to convert the
CFC refrigerant to alternatives in an existing chiller.
The ARI estimates that 53 percent of the 80,000 chillers in 1992 will remain in service on
January 1, 2000. This shows that the actual phaseout process was slower than expected.
For CFC-12, automotive cooling and supermarkets acount for more than 90 percent of their servicing
requirements. Due to the vibrations and unsteady operating conditions, automotive cooling
required greater servicing losses and a faster phaseout schedule than large-tonnage chillers. The annual
amount of CFCs required to compensate for servicing losses came from stockpiles of virgin
CFCs and reclaimed CFCs. Because of the slower phaseout of CFC chillers, servicing demands
were greater than could be supplied from the reclaimed CFCs. The using up of the stockpiles of
virgin CFCs and thus a shortage of CFC supply may occur at the begining of the twenty-first century
in the United States. The slower phaseout of CFCs indicated that there is a possibility of a
considerably longer period of servicing of HCFC equipment than called for in the production
phaseout schedule in the twenty-first century.
9.5 CLASSIFICATION OF REFRIGERANTS
Before the introduction of chlorofluorocarbons in the 1930s, the most commmonly used refrigerants
were air, ammonia, sulfur dioxide, carbon dioxide, and methyl chloride. Until 1986, nontoxic
and nonflammable halogenated hydrocarbons with various ozone depletion potentials were used
almost exclusively in vapor compression refrigeration systems for air conditioning. The impact of
ozone depletion of CFCs, halons, and HCFCs since the 1980s caused a worldwide decision to
phase out these refrigerants. A new classification of refrigerants into six groups based mainly on
ozone depletion will be helpful for the selection of non–ozone depletion refrigerants as well as
replacement of CFCs by alternative refrigerants (Table 9.1).
Hydrofluorocarbons
HFCs contain only hydrogen, fluorine, and carbon atoms. They contain no chlorine atoms, therefore
are environmentally safe, and cause no ozone depletion. They are designated by the prefix HFC.
HFC-134a is an attractive, long-term alternative to replace CFC-12 in reciprocating, scroll,
screw, and centrifugal compressors; and a long-term alternative for HCFC-22. It has a low 0.28
HGWP. HFC-134a is nonflammable, has an extremely low toxicity, and is classified as AI in
ANSI/ASHRAE Standard 34-1997 safety rating.
HFC-134a has a molecular mass of 102.3 instead of CFC-12’s molecular mass of 120.93. At a
condensing temperature of 100°F (37.8°C), HFC-134a’s condensing pressure is 138.83 psia (957
kPa abs.), whereas CFC-12’s is 131.65 psia (908 kPa abs.). A larger impeller of higher speed is
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.13
needed for a centrifugal chiller to provide the same cooling capacity. Parsnow (1993) reported a
capacity loss of direct conversion from CFC-12 to HFC-134a of 8 to 10 percent, and an efficiency
loss of 1 to 2 percent. In Lowe and Ares (1995), in the conversion from CFC-12 to HFC-134a in the
Sears Tower centrifugal chillers, the compressor’s speed increased about 8.5 percent, there was a
cooling capacity loss of 12 to 24 percent, and efficiency was 12 to 16 percent worse.
HFC-134a has a poor mutual solubility with mineral oil because of a higher interfacial tension
between them. Polyolester-based synthetic lubricants should be used. Polyolester-based synthetic
oils are hygroscopic, so monitoring of the moisture content of the refrigerant is important. Halocarbons,
including HFC-134a, are compatible with containment materials. Concerning nonmetallic or
elastomer (such as gaskets) compatibility, Corr et al. (1993) reported that HFC-134a, an ester-based
synthetic oil mixture, has a smaller volume change of elastomer than CFC-12 and mineral oil.
HFC-134a may become one of the most widely used single-chemical-compound refrigerants during
the first half of the twenty-first century.
HFC-245ca also does not contain chlorine and bromine atoms, and its ODP is 0. Compared to
CFC-11, its efficiency will be 3 to 4 percent lower. Synthetic polyolester lubricant oil will be used.
Except for neoprene at high moisture content, common materials used in the refrigeration system
were shown to be compatible with HFC-245ca in tests. Smith et al. (1993) reported that mixtures of
HFC-245ca in air with a relative humidity of 43 percent and a HFC-245ca concentration range of
7 to 14.4 percent were observed to be flammable in tests.
Because of the higher isentropic work required by HFC-245ca compared to CFC-11 and HCFC-
123, for direct-drive centrifugal chillers, a large impeller is required during the conversion from
CFC-11 to HFC-245ca or from HCFC-123 to HFC-245ca.
HFC-245ca is a possible long-term alternative to CFC-11 and HCFC-123 in large centrifugal
chillers in the future. In the HFC group, HFC-32, HFC-125, HFC-143a, and HFC-152a all are seldom
used as a refrigerant of single compound only.
Azeotropic HFC
HFC-507 is an azeotrope of refrigerant blends of HFC-125/HFC-143a (45/55) of zero ozone
depletion and an HGWP of 0.96. It is a long-term alternative refrigerant to replace CFC-502 and
CFC-12 in low-temperature refrigeration systems whose evaporating temperatures are below10°F
(12.2°C). HFC-507 needs synthetic lubricant oil. According to ANSI/ASHRAE Standard 34-
1997, HFC-507 is allowed alternative designations for HFC-507A, a refrigeration blend of HFC-
125/HFC-143a (50/ 50).
The Linton et al. (1995) test results showed that compared to CFC-502, the cooling capacity of
HFC-507 was between 0.95 and 1.05. HFC-507 had an energy efficiency of 0.87 to 0.97 compared
to CFC-502.
Near-Azeotropic HFC
Near-azeotropic HFC is a refrigerant blend of zero ozone depletion and having rather small changes
in volumetric composition or saturation temperature, a small glide, when it evaporates or condenses
at a constant pressure. Near-azeotropic HFC-404A and HFC-410A require synthetic lubricant oil
instead of mineral oil and are nontoxic and nonflammable with a safety classification of A1/A1.
HFC-404A is a blend of HFC-125/HFC-143a/HFC-134a (44/52/4) of zero ozone depletion
and an HGWP of 0.94. It is a long-term alternative refrigerant for CFC-502 and CFC-12 both in
low-temperature refrigeration systems. HFC-404A has a temperature glide of 0.9°F (0.5°C) during
evaporation and a temperature glide of 0.6°F (0.33°C) during condensation. Snelson et al. (1995)
compared HFC-404A with CFC-502 from their test results. HFC-404A had the same, slightly
higher, or lower evaporating capacity at various condensing and evaporating temperatures. The energy
efficiency ratio of 0.89 to 0.99 was found at different evaporating temperatures Tev. The lower
the Tev, the lower the energy efficiency ratio, because of the higher compressor pressure ratio.
9.14 CHAPTER NINE
HFC-410A is a blend of HFC-32/HFC-125 (50/50) of zero ozone depletion and an HGWP of
0.43. It is a long-term alternative refrigerant to replace HCFC-22 and CFC-502. HFC-410A has a
temperature glide of 0.2°F (0.11°C) during evaporation and condensation. Hickman (1994) showed
that the compressor displacement, cfm/ton (L/ s kW), for HFC-410A is about 50 percent smaller
than that for HCFC-22; and the discharge pressure for 130°F (54.4°C) condensing is about 490 psia
(3379 kPa abs.), which is much higher than that for HCFC-22. It is often necessary to change the
original reciprocating compressor to a scroll compressor. A higher energy efficiency was reported
by a refrigerant manufacturer.
Zeotropic HFC
Zeotropic (nonazeotropic) HFCs are refrigerant blends of zero ozone depletion that have greater
temperature glide during evaporation and condensation. Zeotropic HFC-407A and HFC-407C also
require synthetic lubrication oil, instead of mineral oil; and both are nontoxic and nonflammable
with a safety classification of A1/A1.
HFC-407A is a blend of HFC-32/HFC-125/HFC-134a (20/40/40) of zero ozone depletion
with an HGWP of 0.49. It is a long-term alternative refrigerant for CFC-502 and CFC-12 in lowtemperature
refrigeration sysems. HFC-407A showed a reduction in heat transfer in the evaporator
of a low-temperature system during tests. The system performance of HFC-407A was the lowest
compared to HFC-404A and HFC-507.
HFC-407C is a blend of HFC-32/HFC-125/134a (23/25/52) of zero ozone depletion with an
HGWP of 0.38. It is a long-term alternative refrigerant to replace HCFC-22 and CFC-502. Bivens
et al. (1994) compared HFC-407C to HCFC-22 during tests. For cooling and heating, the capacity
ratio ranged from 0.93 to 1.06, and the energy ratio ranged from 0.94 to 0.97. In-tube heat-transfer
coefficients during evaporation and condensation were 85 to 95 percent of HCFC-22 values.
HCFCs and Their Zeotropes
HCFCs contain hydrogen, chlorine, fluorine, and carbon atoms and are not fully halogenated.
HCFCs have a much shorter atmospheric life than CFCs and cause far less ozone depletion (0.02 to
0.1 ODP). They are designated by the prefix HCFC. Their consumption is scheduled to be reduced
gradually starting from 2004 and will be completely phased out in 2030 in developed nations, except
for a limited amount for service, as discussed previously.
HCFC-22 has an ODP of 0.05 and an HGWP of 0.40. It is nonflammable with a safety classifi-
cation of A1. HCFC-22 is partially miscible with mineral oil. At 40°F (4.4°C), its evaporating pressure
is 82.09 psia (566 kPa abs.), and at 100°F (37.8°C) its condensing pressure is 201.5 psia (1389
kPa abs.), the highest of currently used HCFC and CFC refrigerants. HCFC-22 has a smaller compressor
displacement among the HCFCs and CFCs. All these factors make it an interim alternative
to replace CFC-12. HCFC-22 was the most widely used refrigerant in reciprocating and scroll compressors
in small and medium-size packaged units in the 1990s in the United States.
HCFC-123 is an interim alternative to replace CFC-11 in low-pressure centrifugal chillers. It has
an ODP of 0.02 and a very low HGWP of 0.02. HCFC-123 is nonflammable and of lower toxicity
with a safety classification of B1. In 1997, DuPont raised the allowable exposure limit (AEL) of
HCFC-123 to 50 ppm. Smithhart and Crawford (1993) reported that for chillers with direct conversion
from CFC-11 to HCFC-123, there was about a decrease of 0 to 5 percent in capacity and a 2 to
4 percent decrease in efficiency. A conversion of refrigerant from CFC-11 to HCFC-123 in an existing
chiller may require changing its lubricants, seals, and motor windings of hermetic compressors.
Because HCFC-123 has a low ODP, a very low HGWP, and a 15 percent higher energy effi-
ciency in centrifugal chillers than HFC-134a does, if no acceptable alternative can be found, the use
of HCFC-123 in centrifugal chillers may be considered longer than the cap specified in the Vienna
meeting in 1995, as listed in Sec. 9.4, in the twenty-first century.
HCFC-124 has an ODP of 0.02. It is nonflammable and has a safety classification of A1. HCFC-
124 is an interim alternative refrigerant to replace CFC-114.
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.15
Near-azeotropic HCFC-402A is a blend of HCFC-22/HFC-125/PRO-290 (38/60/2) with an
ODP of 0.02 and an HGWP of 0.63. Here PRO-290 represents propane, which is a more highly
flammable refrigerant with a safety classification of A3. HCFC-402A is nonflammable and has a
safety classification of A1/A1. It needs polyolester or alkyl-benzene-based lubricant oil. HCFC-
402A is an interim alternative refrigerant to replace CFC-502.
Zeotropic HCFC-401A is a blend of HCFC-22/HCFC-124/HFC-152a (53/34/13) with an ODP
of 0.037 and an HGWP of 0.22; and HCFC-401B is a blend of HCFC-22/HCFC-124/HFC-152a
(61/28/11) with an ODP of 0.04 and an HGWP of 0.24. Both HCFC-401A and HCFC-401B are
nonflammable and have a safety classification of A1/A1. They both need alkyl-benzene-based lubricant
oil. HCFC-401A is an interim alternative refrigerant to replace CFC-12, and HCFC-401B is an
interim alternative refrigerant to replace CFC-12 in low-temperature refrigeration systems.
Inorganic Compounds
These compounds include ammonia (NH3), water (H2O), and gases used in the gas expansion systems.
As refrigerants, they were used far earlier than the halocarbons. Air is a mixture of nitrogen,
oxygen, argon, rare gases, and water vapor. Air has zero ozone depletion and is a zeotropic blend that
has a temperaure glide of 5.5°F (3°C) at atmospheric pressure. Ammonia also has zero ozone
depletion. It has a high operating pressure at 40°F (4.4°C) evaporating and 100°F (37.8°C) condensing.
Ammonia compressors show a smaller cfm/ton displacement and higher energy efficiency than
HCFC-22 compressors. Leakage of ammonia is easily detected due to its objectionable odor.
Ammonia attacks copper even in the presence of a small amount of moisture. It is of higher toxicity.
An ammonia-air mixture is flammable if the concentration of NH3 by volume is within 16 to
25 percent. The mixture may explode if the ignition source is above 1200°F (650°C). Because the
safety classification of ammonia is B2—lower flammability and higher toxicity—it is not allowed
to be used in comfort air conditioning in the United States.
Water has a zero ODP and is readily available. At 40°F (4.4°C) evporating and 100°F (37.8°C)
condensing, water’s evaporating and condensing pressures are both below atmospheric. Air and
other noncondensable gases must be purged out of the refrigeration system periodically.
CFCs, Halons, and Their Zeotropes
CFCs including CFC-11, CFC-12, CFC-113, and CFC-114, have an ODP from 0.8 to 1.0. Halons
including BFC-13B1 have an ODP of 10. Their azeotropics CFC-500 and CFC-502 have ODPs of
0.74 and 0.22, respectively. Production of all these CFCs, halons, and their azeotropes ceased in developed
nations since January 1, 1996. However, a limited amount of these refrigerants, used to service
the refrigeration systems that have not been converted or replaced by non–ozone depletion refrigerants,
may be extended to the beginning of the twenty-first century.
9.6 REFRIGERATION PROCESSES AND REFRIGERATION CYCLES
Refrigeration Processes
A refrigeration process indicates the change of thermodynamic properties of the refrigerant and the
energy transfer between the refrigerant and the surroundings. The following refrigeration processes
occur during the operation of a vapor compression refrigerating system:
 Evaporation. In this process, the refrigerant evaporates at a lower temperature than that of its
surroundings, absorbing its latent heat of vaporization.
 Superheating. Saturated refrigerant vapor is usually superheated to ensure that liquid refrigerant
does not flow into the compressor.
9.16 CHAPTER NINE
 Compression. Refrigerant is compressed to a higher pressure and temperature for condensation.
 Condensation. Gaseous refrigerant is condensed to liquid form by being desuperheated, then condensed,
and finally subcooled, transferring its latent heat of condensation to a coolant.
 Throttling and expansion. The higher-pressure liquid refrigerant is throttled to the lower evaporating
pressure and is ready for evaporation.
The following refrigeration processes occur during the operation of an air or gas expansion
refrigeration system:
 Compression. Air or gas is compressed to a higher pressure and temperature.
 Heat release. Heat is released to the surroundings at constant pressure in order to reduce the temperature
of the air or gas.
 Throttling and expansion. Air or gas is throttled and expanded so that its temperature is lowered.
 Heat absorption. Heat is absorbed from the surroundings because of the lower air or gas temperature.
Refrigeration Cycles
Most refrigerants undergo a series of evaporation, compression, condensation, throttling, and expansion
processes, absorbing heat from a lower-temperature reservoir and releasing it to a highertemperature
reservoir in such a way that the final state is equal in all respects to the initial state. It is
said to have undergone a closed refrigeration cycle. When air or gas undergoes a series of compression,
heat release, throttling, expansion, and heat absorption processes, and its final state is not
equal to its initial state, it is said to have undergone an open refrigeration cycle.
Both vapor compression and air or gas expansion refrigeration cycles are discussed in this chapter.
Absorption refrigeration cycles are covered in Chap. 14.
Unit of Refrigeration
In inch-pound (I-P) units, refrigeration is expressed in British thermal units per hour, or simply Btu/h.
A British thermal unit is defined as the amount of heat energy required to raise the temperature of one
pound of water one degree Fahrenheit from 59°F to 60°F; and 1 Btu/h  0.293 watt (W).
Another unit of refrigeration widely used in the HVAC&R industry is ton of refrigeration, or
simply ton. As mentioned before, 1 ton  12,000 Btu/h of heat removed. This equals the heat
absorbed by 1 ton (2000 lb) of ice melting at a temperature of 32°F over 24 h.
Because the heat of fusion of ice at 32°F is 144 Btu / lb,
also 1 ton  3.516 kW
9.7 GRAPHICAL AND ANALYTICAL EVALUATION
OF REFRIGERATION
Pressure-Enthalpy Diagram
The pressure-enthalpy p-h diagram is the most common graphical tool for analysis and calculation
of the heat and work transfer and performance of a refrigeration cycle. A single-stage refrigeration
1 ton 
1  2000  144
24
 12,000 Btu / h
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.17
cycle consists of two regions: the high-pressure region, or high side, and the low-pressure region, or
low side. The change in pressure can be clearly illustrated on the p-h diagram. Also, both heat and
work transfer of various processes can be calculated as the change of enthalpy and are easily shown
on the p-h diagram.
Figure 9.3 is a skeleton p-h diagram for refrigerant HCFC-22. Enthalpy h (in Btu/ lb) is the
abscissa, and absolute pressure (psia) or gauge pressure (psig), both expressed in logarithmic scale,
is the ordinate. The saturated liquid line separates the subcooled liquid from the two-phase region in
which vapor and liquid refrigerants coexist. The saturated vapor line separates this two-phase
region from the superheated vapor. In the two-phase region, the mixture of vapor and liquid is
subdivided by the constant-dryness-fraction quality line.
The constant-temperature lines are nearly vertical in the subcooled liquid region. At higher temperatures,
they are curves near the saturated liquid line. In the two-phase region, the constant-temperature
lines are horizontal. In the superheated region, the constant-temperature lines curve down
sharply. Because the constant-temperature lines and constant-pressure lines in the two-phase region
are horizontal, they are closely related. The specific pressure of a refrigerant in the two-phase
region determines its temperature, and vice versa.
Also in the superheated region, the constant-entropy lines incline sharply upward, and constantvolume
lines are flatter. Both are slightly curved.
Temperature-Entropy Diagram
The temperature-entropy T-s diagram is often used to analyze the irreversibilities in a refrigeration
cycle, as well as in the system, in order to select optimum operating parameters and improve performance
of the system. In a temperature-entropy T-s diagram, entropy s, Btu / lb°R, is the abscissa of
9.18 CHAPTER NINE
FIGURE 9.3 Skeleton of pressure-enthalpy p-h diagram for HCFC-22.
the diagram and temperature T, °R, is the ordinate. A T-s diagram is more suitable for evaluating
the effectiveness of an air expansion refrigeration cycle.
Analytical Evaluation of Cycle Performance
Swers et al. (1972) proposed a thermodynamic analysis of degradation of available energy and irreversibility
in a refrigerating system, and Tan and Yin (1986) recommended a method of exergy
analysis. The exergy of a working substance e, Btu / lb (kJ / kg), is defined as
e  h  ha  TRa (s  sa) (9.1)
where h, ha  enthalpy of working substance and ambient state, Btu / lb (kJ /kg)
TRa  absolute temperature of ambient state, °R (K)
s, sa  entropy of working substance and ambient state, Btu / lb°R (kJ/kgK)
Both analyses are effective tools in the selection of optimum design and operating parameters by
means of complicated analysis. They require extensive supporting data and information.
For most analyses of refrigeration cycle performance and design and operation of refrigeration
systems in actual applications, satisfactory results can be obtained by using the steady flow energy
equation, heat and work transfer, and energy balance principle. If a more precise and elaborate
analysis is needed in research or for detailed improvements of refrigeration systems, the references
at the end of this chapter can be consulted.
9.8 CARNOT REFRIGERATION CYCLE
The Carnot refrigeration cycle is a reverse engine cycle. All processes in a Carnot refrigeration
cycle are reversible, so it is the most efficient refrigeration cycle.
Figure 9.4a is a schematic diagram of a Carnot cycle refrigerating system, and Fig. 9.4b shows
the Carnot refrigeration cycle using gas as the working substance. This Carnot cycle is composed of
four reversible processes:
1. An isothermal process 4-1 in which heat q#1 is extracted at constant temperature TR1 per lb (kg)
of working substance
2. An isentropic compression process 1-2
3. An isothermal process 2-3 in which q#2 is rejected at constant temperature TR2 per lb (kg) of
working substance
4. An isentropic expansion process 3-4
Figure 9.4c shows the Carnot refrigeration cycle using vapor as the working substance. Wet vapor
is the only working substance where heat supply and heat rejection processes can occur easily
at constant temperature. This is because the temperatures of wet vapor remain constant when latent
heat is supplied or rejected.
As in the gas cycle, there are two isothermal processes 4-1 and 2-3 absorbing heat at temperature
TR1 and rejecting heat at TR2, respectively, and two isentropic processes, one for compression
1-2 and another for expansion 3-4.
Performance of Carnot Refrigeration Cycle
According to the first law of thermodynamics, often called the law of conservation of energy, when
a system undergoes a thermodynamic cycle, the net heat supplied to the system is equal to the net
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.19
work done, or
Heat supply  heat rejected  net work done (9.2)
Referring to Fig. 9.4a, in a Carnot refrigeration cycle,
q#1  q#2   W
or q#1  q#2  W (9.3)
q#2  q#1  W
where q#1  heat supplied from surroundings per lb (kg) of working substance at temperature T1;
sign of q#1 is positive
q#2  heat rejected to sink per lb (kg) of working substance at temperature T2; sign of q#2
is negative
W  net work done by system; sign is positive, or if it is a work input to system, sign is
negative
9.20 CHAPTER NINE
FIGURE 9.4 Carnot refrigeration cycle: (a) schematic diagram; (b) gas cycle; (c) vapor cycle.
The heat extracted from the source at temperature TR1 by the working substance, i.e., the refrigerating
effect per lb (kg) of working substance, is
q#1  TR1(s1  s4) (9.4)
where s1, s4  entropy at state points 1 and 4, respectively, Btu / lb°R (kJ/kgK). Heat rejected to
the heat sink at temperature TR2 can be calculated as
q#2   TR2(s3  s2)  TR2 (s2  s3) (9.5)
where s2, s3  entropy at state points 2 and 3, respectively, Btu / lb°R (kJ/kgK). Because in the
isentropic process 1-2, s1  s2, and in isentropic process 3-4, s3  s4,
q#2  TR2(s1  s4) (9.6)
9.9 COEFFICIENT OF PERFORMANCE
OF REFRIGERATION CYCLE
The coefficient of performance is an index of performance of a thermodynamic cycle or a thermal
system. Because the COP can be greater than 1, COP is used instead of thermal efficiency. The
coefficient of performance can be used for the analysis of the following:
 A refrigerator that is used to produce a refrigeration effect only, that is, COPref
 A heat pump in which the heating effect is produced by rejected heat COPhp
 A heat recovery system in which both the refrigeration effect and the heating effect are used at the
same time, COPhr
For a refrigerator, COP is defined as the ratio of the refrigeration effect q#1 to the work input Win,
both in Btu/ lb (kJ / kg), that is,
COPref  Refrigeration effect /Work input
 q#1 /Win (9.7)
For the Carnot refrigeration cycle, from Eq. (9.3),
(9.8)
With a heat pump, the useful effect is the heating effect because of the rejected heat q#2, so COPhp is
the ratio of heat rejection to the work input, or
(9.9)
For a heat recovery system, the useful effect is q#1 and q#2. In such a condition, COPhr is defined as
the ratio of the sum of the absolute values of refrigerating effect and heat rejection to the absolute
value of the work input, i.e.,
(9.10) COPhr 
q#1   q#2 
Win
COPhp 
q#2
Win

TR1(s1  s4)
(TR2  TR1)(s1  s4)

TR1
TR2  TR1
COPref 
q#1
q#2  q#1
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.21
9.10 SINGLE-STAGE IDEAL VAPOR COMPRESSION CYCLE
The Carnot cycle cannot be achieved for the vapor cycle in actual practice because liquid slugging
would occur during compression of the two-phase refrigerant. In addition, the mixture, mostly liquid,
does very little work when it expands after condensation in the heat engine. Therefore, a single-
stage ideal vapor compression cycle is used instead of the Carnot cycle.
Figure 9.5 shows an ideal single-stage vapor compression cycle in which compression occurs in
the superheated region. A throttling device, such as an expansion valve, is used instead of the heat
engine. Single-stage means that there is only one stage of compression. An ideal cycle is one in
which the compression process is isentropic and the pressure losses in the pipeline, valves, and
other components are negligible. All the refrigeration cycles covered in this chapter are ideal cycles
except the air expansion refrigeration cycle.
Vapor compression means that the vapor refrigerant is compressed to a higher pressure, and then
the condensed liquid is throttled to a lower pressure to produce the refrigerating effect by evaporation.
It is different from the absorption or air expansion refrigeration cycle.
Flow Processes
Figure 9.5b and c shows the refrigeration cycle on p-h and T-s diagrams. The refrigerant evaporates
entirely in the evaporator and produces the refrigerating effect. It is then extracted by the compressor
at state point 1, compressor suction, and is compressed isentropically from state point 1 to 2. It
is next condensed to liquid in the condenser, and the latent heat of condensation is rejected to the
heat sink. The liquid refrigerant, at state point 3, flows through an expansion valve, which reduces it
to the evaporating pressure. In the ideal vapor compression cycle, the throttling process at the expansion
valve is the only irreversible process, usually indicated by a dotted line. Some of the liquid
flashes into vapor and enters the evaporator at state point 4. The remaining liquid portion evaporates
at the evaporating temperature, thus completing the cycle.
Cycle Performance
For the evaporating process between points 4 and 1, according to the steady flow energy equation,
(9.11)
where h1, h4  enthalpy of refrigerant at points 1 and 4, respectively, Btu / lb (J /kg)
v1, v4  velocity of refrigerant at points 1 and 4, respectively, ft / s (m/ s)
q#  heat supplied per lb (kg) of working substance during evaporation process, Btu/ lb
(J /kg)
gc  dimensional conversion factor, 32 lbmft / lbf  s2
Because no work is done during evaporation, the change of kinetic energy between points 4 and 1 is
small compared with other terms in Eq. (9.11), and it is usually ignored. Then
h4  q#  h1  0
The refrigerating effect qrf, Btu / lb (J / kg), is
qrf  q#  h1  h4 (9.12)
h4  v4 2
2gc  778
 q# 
h1  v1 2
2gc  778
 W
9.22 CHAPTER NINE
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.23
FIGURE 9.5 Single-stage ideal vapor compression cycle: (a)
schematic diagram; (b) p-h diagram; (c) T-s diagram.
For isentropic compression between points 1 and 2, applying the steady flow energy equation and
ignoring the change of kinetic energy, we have
h1  0  h2  W
 W  h2  h1
Work input to the compressor Win, Btu / lb (kJ / kg), is given as
Win  h2  h1 (9.13)
Similarly, for condensation between points 2 and 3,
h2  q#  h3  0
The heat released by the refrigerant in the condenserq#, Btu / lb (kJ / kg), is
q#  h2  h3 (9.14)
For the throttling process between points 3 and 4, assuming that the heat loss is negligible,
h3  0  h4  0
or h3  h4 (9.15)
The COPref of the single-stage ideal vapor compression cycle is
(9.16)
The mass flow rate of refrigerant , lb/h (kg/ s), flowing through the evaporator is
(9.17)
where Qrc  refrigerating capacity, Btu /h (W). From Eq. (9.16), the smaller the difference between
the condensing and evaporating pressures, or between condensing and evaporating temperatures,
the lower the Win input to the compressor at a specific Qrc and, therefore, the higher the COP.
A higher evaporating pressure pev and evaporating temperature Tev or a lower condensing pressure
pcon and condensing temperature Tcon will always conserve energy.
Determination of Enthalpy by Polynomials
During the performance analysis of a refrigeration cycle, the enthalpies h of the refrigerant at
various points must be determined in order to calculate the refrigeration effect, work input, and
COP. The enthalpy of refrigerant at saturated liquid and saturated vapor state is a function of
saturated temperature or pressure. In other words, saturated temperature Ts and saturated pressure
pss of the refrigerant are dependent upon each other. Therefore, it is more convenient to evaluate
the enthalpy of refrigerant in terms of saturated temperature Ts within a certain temperature
range
h  f (Ts) (9.18)
m?r 
Qrc
qrf
m?r

qrf
Win

h1  h4
h2  h1
COPref 
refrigerating effect
work input
9.24 CHAPTER NINE
The enthalpy differential along the constant-entropy line within a narrower temperature range can
be calculated as
h2  h1  F(Ts2  Ts1) (9.19)
where h1, h2  enthalpy of refrigerant on constant-entropy line at points 1 and 2, Btu/ lb (kJ /kg)
Ts1, Ts2  temperature of saturated refrigerant at points 1and 2,°F (°C)
From the refrigerant tables published by ASHRAE, the following polynomial can be used to calculate
the enthalpy of saturated liquid refrigerant hlr, Btu / lb (kJ / kg), from its temperature Ts1 at a saturated
temperature from 20 to 120°F ( 7 to 50°C) with acceptable accuracy:
hlr  a1  a2Ts1  a3Ts1
2  a4Ts1
3 (9.20)
where a1, a2, a3, a4  coefficients. For HCFC-22,
a1  10.409 a2  0.26851 a3  0.00014794 a4  5.3429  107
Similarly, the polynomial that determines the enthalpy of saturated vapor refrigerant hvr, Btu / lb
(kJ/ kg), from its temperature Tsv in the same temperature range is
hvr  b1  b2Tsv  b3Tsv
2  b4Tsv
3 (9.21)
where b1, b2, b3, b4  coefficients. For HCFC-22,
b1  104.465 b2  0.098445 b3  0.0001226 b4  9.861  107
The polynomial that determines the enthalpy changes of refrigerant along the constant-entropy line
for an isentropic compression process between initial state 1 and final state 2 is
(9.22)
where c1, c2, c3, c4  coefficients
Ts1, Ts2  saturated temperature of vapor refrigerant corresponding to its pressure at initial
state 1 and final state 2,°F (°C)
For HCFC-22 within a saturated temperature range of 20 to 100°F:
c1  0.18165 c2  0.21502 c3  0.0012405 c4  8.198  106
Computer programs are available that calculate the coefficients based on ASHRAE’s refrigerant tables
and charts.
Refrigeration Effect, Refrigerating Load, and Refrigerating Capacity
The refrigeration effect qrf , Btu / lb (J /kg or kJ/ kg), is the heat extracted by a unit mass of refrigerant
during the evaporating process in the evaporator. It can be calculated as
qrf  hlv  hen (9.23)
where hen, hlv  enthalpy of refrigerant entering and leaving evaporator, Btu / lb (J / kg). Refrigerating
load Qrl, Btu /h (W), is the required rate of heat extraction by the refrigerant in the evaporator. It
h2  h1  c1  c2(Ts2  Ts1)  c3(Ts2  Ts1)2  c4(Ts2  Ts1)3
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.25
can be calculated as
(9.24)
where mass flow rate of refrigerant flowing through evaporator, lb/h (kg/ s).
Refrigerating capacity, or cooling capacity, Qrc, Btu /h (W), is the actual rate of heat extracted by
the refrigerant in the evaporator. In practice, the refrigeration capacity of the equipment selected is
often slightly higher than the refrigerating load. This is because the manufacturer’s specifications
are a series of fixed capacities. Occasionally, equipment can be selected so that its capacity is just
equal to the refrigeration load required. Refrigeration capacity Qrc can be calculated as
(9.25)
where hren, hrlv  enthalpy of refrigerant actually entering and leaving evaporator, Btu / lb (J /kg)
9.11 SUBCOOLING AND SUPERHEATING
Subcooling
Condensed liquid refrigerant is usually subcooled to a temperature lower than the saturated temperature
corresponding to the condensing pressure of the refrigerant, shown in Fig. 9.6a as point 3	.
This is done to increase the refrigerating effect. The degree of subcooling depends mainly on the
temperature of the coolant (e.g., atmospheric air, surface water, or well water) during condensation,
and the construction and capacity of the condenser.
The enthalpy of subcooled liquid refrigerant hsc, Btu / lb (J / kg), can be calculated as
hsc  hs,con  cpr(Ts,con  Tsc) (9.26)
where hs,con  enthalpy of saturated liquid refrigerant at condensing temperature, Btu/ lb (J /kg)
cpr  specific heat of liquid refrigerant at constant pressure, Btu/ lb °F (J /kg°C)
Ts,con  saturated temperature of liquid refrigerant at condensing pressure,°F (°C)
Tsc  temperature of subcooled liquid refrigerant,°F (°C)
Enthalpy hsc is also approximately equal to the enthalpy of the saturated liquid refrigerant at subcooled
temperature.
Superheating
As mentioned before, the purpose of superheating is to avoid compressor slugging damage. Superheating
is shown in Fig. 9.6b. The degree of superheat depends mainly on the type of refrigerant
feed and compressor as well as the construction of the evaporator. These are covered in detail in
Chap. 11.
Example 9.1. A 500-ton (1760-kW) single-stage centrifugal vapor compression system uses
HCFC-22 as refrigerant. The vapor refrigerant enters the compressor at dry saturated state. The
compression process is assumed to be isentropic. Hot gas is discharged to the condenser and
condensed at a temperature of 95°F (35°C). The saturated liquid refrigerant then flows through a
throttling device and evaporates at a temperature of 35°F (1.7°C). Calculate:
1. The refrigeration effect
2. The work input to the compressor
Qrc  m?r (hrlv  hren)
m?r 
Qrl  m?r (hlv  hen)
9.26 CHAPTER NINE
3. The coefficient of performance of this refrigeration cycle
4. The mass flow rate of the refrigerant
Recalculate the COP and the energy saved in work input if the refrigerant is subcooled to a temperature
of 90°F (32.2°C).
Solution
1. From Eq. (9.20), the enthalpy of the saturated liquid refrigerant at a temperature of 95°F (35°C),
point 3 as shown in Fig. 9.5a, is
 10.409  25.508  1.335  0.458  37.71 Btu / lb
 10.409  0.26851(95)  0.0001479(95)2  5.3429  107(95)3
h3  h4  10.409  0.26851Ts  0.0001479T s
2  5.3429  107T s
3
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.27
FIGURE 9.6 (a) Subcooling and (b) superheating.
From Eq. (9.21), the enthalpy of saturated vapor refrigerant at a temperature of 35°F (1.7°C),
point 1, is
Then the refrigeration effect is calculated as
qrf  h1  h4  107.72  37.71  70.01 Btu/ lb (162.8 kJ /kg)
2. From Eq. (9.22), the enthalpy differential h2  h1 on the constant-entropy line corresponding to
a saturated temperature differential Ts2  Ts1  95 35  60°F in the two-phase region is
That is, the work input Win  h2  h1  10.024 Btu/ lb (23.73 kJ / kg).
3. According to Eq. (9.16), COPref of the refrigerating system is calculated as
4. From Eq. (9.17), the mass flow rate of the refrigerant can be calculated as
If the liquid refrigerant is subcooled to a temperature of 90°F, (32.2°C), then
h3  h4  10.409  0.26851(90)  0.0001479(90)2  5.3429  107(90)3
 10.409  24.166  1.198  0.389  36.16 Btu/ lb
Refrigeration effect is then increased to
qrf  107.72  36.16  71.56 Btu/ lb (166 kJ/kg)
Also COPref is increased to
Since 1 ton  200 Btu/min and 1 hp  42.41 Btu/min, electric power input to the compressor Pin
without subcooling is
Pin 
500  200
42.41  6.98
 337.8 hp (252 kW)
COPref 
71.56
10.024
 7.14
m? r 
Qrc
qrf

500  12,000
70.01
 85,702 lb / h (38,874 kg / h)
COPref 
qrf
Win

70.01
10.024
 6.98
 0.182  12.901  4.466  1.771  10.02 Btu / lb
  0.18165  0.21502(60)  0.0012405(60)2  8.1982  106(60)3
8.1982  106(Ts2  Ts1)3
h2  h1  0.18165  0.21502(Ts2  Ts1)  0.0012405(Ts2  Ts1)2
 104.47  3.445  0.150  0.042  107.72 Btu / lb
 104.47  0.098445(35)  0.0001226(35)2  9.861  107(35)3
h1  104.465  0.098445Ts  0.0001226T s
2  9.861  107T s
3
9.28 CHAPTER NINE
With subcooling,
Savings in electric energy are calculated as
9.12 MULTISTAGE VAPOR COMPRESSION SYSTEMS
When a refrigeration system uses more than single-stage compression process, it is called a multistage
system (as shown in Fig. 9.7), and may include the following:
1. A high-stage compressor and a low-stage compressor
2. Several compressors connected in series
3. Two or more impellers connected internally in series and driven by the same motor or prime
mover, as shown in Fig. 9.7
4. A combination of two separate refrigeration systems
The discharge pressure of the low-stage compressor, which is equal to the suction pressure of the
high-stage compressor, is called the interstage pressure.
The reasons for using a multistage vapor compression system instead of a single-stage system
are as follows:
1. The compression ratio Rcom of each stage in a multistage system is smaller than that in a singlestage
unit, so compressor efficiency is increased. Compression ratio Rcom is defined as the ratio
of the compressor’s discharge pressure pdis, psia (kPa abs.), to the suction pressure at the compressor’s
inlet psuc, psia (kPa abs.), or
(9.27)
2. Liquid refrigerant enters the evaporator at a lower enthalpy and increases the refrigeration effect.
3. Discharge gas from the low-stage compressor can be desuperheated at the interstage pressure.
This results in a lower discharge temperature from the high-stage compressor than would be
produced by a single-stage system at the same pressure differential between condensing and
evaporating pressures.
4. Two or three compressors in a multistage system provide much greater flexibility to accommodate
the variation of refrigeration loads at various evaporating temperatures during part-load operation.
The drawbacks of the multistage system are higher initial cost and a more complicated system
than that for a single-stage system.
Compound Systems
Multistage vapor compression systems are classified as compound systems or cascade systems.
Cascade systems are discussed in a later section.
A compound system consists of two or more compression stages connected in series. For
reciprocating, scroll, or screw compressors, each compression stage usually requires a separate
Rcom 
pdis
psuc
337.8  330.2
337.8
 0.022, or 2.2%
Pin, s 
500  200
42.41  7.14
 330.2 hp (246 kW)
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.29
compressor. In multistage centrifugal compressors, two or more stages may be internally compounded
by means of several impellers connected in series.
Interstage Pressure
Interstage pressure is usually set so that the compression ratio at each stage is nearly the same for
higher COPs. For a two-stage compound system, interstage pressure pi, psia (kPa abs.), can be
9.30 CHAPTER NINE
FIGURE 9.7 Two-stage compound system with a flash cooler: (a)
schematic diagram; (b) refrigeration cycle.
calculated as
(9.28)
where pcon  condensing pressure, psia (kPa abs.)
pev  evaporating pressure, psia (kPa abs.)
For a multistage vapor compression system with z stages, the compression ratio Rcom for each stage
can be calculated as
(9.29)
Flash Cooler and Intercooler
In compound systems, flash coolers are used to subcool liquid refrigerant to the saturated temperature
corresponding to the interstage pressure by vaporizing part of the liquid refrigerant. Intercoolers
are used to desuperheat the discharge gas from the low-stage compressor and, more often, to
subcool also the liquid refrigerant before it enters the evaporator.
9.13 TWO-STAGE COMPOUND SYSTEM WITH
A FLASH COOLER
Flow Processes
Figure 9.7a is a schematic diagram of a two-stage compound system with a flash cooler, and Fig. 9.7b
shows the refrigeration cycle of this system. Vapor refrigerant at point 1 enters the first-stage impeller
of the centrifugal compressor at the dry saturated state. It is compressed to the interstage pressure pi at
point 2 and mixes with evaporated vapor refrigerant from the flash cooler, often called an economizer.
The mixture then enters the second-stage impeller at point 3. Hot gas, compressed to condensing pressure
pcon, leaves the second-stage impeller at point 4. It is then discharged to the condenser, in which
the hot gas is desuperheated, condensed, and often subcooled to liquid state at point 5	. After the condensing
process, the subcooled liquid refrigerant flows through a throttling device, such as a float
valve, at the high-pressure side. A small portion of the liquid refrigerant flashes into vapor in the flash
cooler at point 7, and this latent heat of vaporization cools the remaining liquid refrigerant to the
saturation temperature corresponding to the interstage pressure at point 8. Inside the flash cooler, the
mixture of vapor and liquid refrigerant is at point 6. Liquid refrigerant then flows through another
throttling device, a small portion is flashed at point 9, and the liquid-vapor mixture enters the evaporator.
The remaining liquid refrigerant is vaporized at point 1 in the evaporator. The vapor then flows to
the inlet of the first-stage impeller of the centrifugal compressor and completes the cycle.
Fraction of Evaporated Refrigerant in Flash Cooler
In the flash cooler, out of 1 lb of refrigerant flowing through the condenser, x lb of it cools down the
remaining portion of liquid refrigerant (1  x) lb to saturated temperature T8 at interstage pressure
pi. Because h5	 is the enthalpy of the subcooled liquid refrigerant entering the flash cooler, h6 is the
enthalpy of the mixture of vapor and liquid refrigerant after the throttling device, for a throttling
process, h5	  h6. Enthalpies h7 and h8 are the enthalpies of the saturated vapor and saturated
liquid, respectively, at the interstage pressure, and h9 is the enthalpy of the mixture of vapor and
liquid refrigerant leaving the flash cooler after the low-pressure side throttling device. Again, for a
throttling process, h8  h9.
Rcom   pcon
pev 1/z
pi  ?pcon pev
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.31
If the heat loss from the insulated flash cooler to the ambient air is small, it can be ignored. Heat
balance of the refrigerants entering and leaving the flash cooler, as shown in Fig. 9.8a, gives
Sum of heat energy of refrigerant entering flash cooler
 Sum of heat energy of refrigerant leaving flash cooler
that is,
h5	  xh7  (1  x)h8
The fraction of liquid refrigerant evaporated in the flash cooler x is given as
(9.30)
The fraction x also indicates the quality, or dryness fraction, of the vapor and liquid mixture in the
flash cooler at the interstage pressure.
Enthalpy of Vapor Mixture Entering Second-Stage Impeller
Ignoring the heat loss from mixing point 3 to the surroundings, we see that the mixing of the
gaseous refrigerant discharged from the first-stage impeller at point 2 and the vaporized refrigerant
from the flash cooler at point 7 is an adiabatic process. The heat balance at the mixing point before
the second-stage impeller, as shown in Fig. 9.7b, is given as
h3  (1  x)h2  xh7 (9.31)
where h2  enthalpy of gaseous refrigerant discharged from first-stage impeller, Btu/ lb (kJ /kg)
h3  enthalpy of mixture at point 3, Btu/ lb (kJ /kg)
h7  enthalpy of saturated vapor refrigerant from flash cooler at point 7, Btu/ lb (kJ /kg)
x 
h5	  h8
h7  h8
9.32 CHAPTER NINE
FIGURE 9.8 Heat balance of entering and leaving refrigerants in a
flash cooler and at the mixing point: (a) in the flash cooler; (b) at the
mixing point 3 before entering the second-stage impeller.
Coefficient of Performance
For 1 lb (kg) of refrigerant flowing through the condenser, the amount of refrigerant flowing
through the evaporator is (1  x) lb (kg). The refrigeration effect qrf per lb (kg) of refrigerant flowing
through the condenser, Btu / lb, (kJ/ kg), can be expressed as
qrf  (1  x)(h1  h9) (9.32)
where h1  enthalpy of saturated vapor leaving evaporator, Btu / lb (kJ /kg)
h9  enthalpy of refrigerant entering evaporator, Btu / lb (kJ /kg)
Total work input to the compressor (including the first- and second-stage impeller) Win per lb (kg)
of refrigerant flowing through the condenser, Btu / lb (kJ / kg), is
Win  (1  x)(h2  h1)  h4  h3 (9.33)
where h4  enthalpy of the hot gas discharged from the second-stage impeller, Btu/ lb (kJ / kg). The
coefficient of performance of the two-stage compound system with a flash cooler COPref is
(9.34)
The mass flow rate of refrigerant at the condenser , lb/h (kg/ s), is
(9.35)
where Qrc  refrigeration capacity, Btu /h (W).
Characteristics of Two-Stage Compound System with Flash Cooler
In a two-stage compound system with a flash cooler, a portion of the liquid refrigerant is flashed
into vapor going directly to the second-stage suction inlet, so less refrigerant is compressed in
the first-stage impeller. Furthermore, the remaining liquid refrigerant is cooled to the saturated
temperature corresponding to the interstage pressure, which is far lower than the subcooled liquid
temperature in a single-stage system. The increase in refrigeration effect and the drop in compression
work input lead to a higher COPref than in a single-stage system.
Although the initial cost of a two-stage compound system is higher than that for a single-stage
system, the two-stage compound system with a flash cooler is widely used in large central hydronic
air conditioning systems because of the high COPref.
Example 9.2. For the same 500-ton (1758-kW) centrifugal vapor compression system as in
Example 9.1, a two-stage compound system with a flash cooler is used instead of a single-stage
centrifugal compressor. Vapor refrigerant enters the first-stage impeller at a dry saturated state, and
the subcooled liquid refrigerant leaves the condenser at a temperature of 90°F (32.2°C). Both
compression processes at the first-stage impeller and the second-stage impeller are assumed to be
isentropic. Evaporating pressure is 76.17 psia (525kPa abs.), and the condensing pressure is 196.5
psia (1355 kPa abs.). Other conditions remain the same as in Example 9.1. Calculate
1. The fraction of liquid refrigerant vaporized in the flash cooler
2. The refrigeration effect per lb (kg) of refrigerant flowing through the condenser
3. The total work input to the compressor
m? r 
Qrc
qrf

Qrc
(1  x)(h1  h9)
m? r

(1  x)(h1  h9)
(1  x)(h2  h1)  (h4  h3)
COPref 
qrf
Win
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.33
4. The coefficient of performance of this refrigerating system
5. The mass flow rate of refrigerant flowing through the condenser
6. The percentage of saving in energy consumption compared with the single-stage vapor compression
system
Solution
1. Based on the data calculated in Example 9.1, enthalpy of vapor refrigerant entering the firststage
impeller h1  107.72 Btu/lb and the enthalpy of the subcooled liquid refrigerant leaving
the condenser h5	  36.162 Btu/ lb, as shown in Fig. 9.6b. From Eq. (9.28) and the given data,
the interstage pressure can be calculated as
From the Table of Properties of Saturated Liquid and Vapor for HCFC-22 in ASHRAE Handbook
1989, Fundamentals, for pi  122.34 psia, the corresponding interstage saturated temperature Ti in
the two-phase region is 63.17°F.
From Eq. (9.20), the enthalpy of liquid refrigerant at saturated temperature 63.17°F is
h8  h9  10.409  0.26851(63.17)  0.00014794(63.17)2  5.3429(63.17)3
 10.409  16.961  0.59  0.135  28.095 Btu/ lb
Also, from Eq. (9.21), the enthalpy of the saturated vapor refrigerant at a temperature of 63.17°F is
h7  104.465  0.098445(63.17)  0.0001226(63.17)2  9.861  107(63.17)3
 104.465  6.219  0.489  0.249  109.946 Btu/ lb
Then, from Eq. (9.30), the fraction of vaporized refrigerant in the flash cooler is
2. From Eq. (9.32), the refrigeration effect is
qrf  (1  x)(h1  h9)  (1  0.09856)(107.72  28.095)
 71.78 Btu/ lb (167 kJ/kg)
3. From Eq. (9.22), the enthalpy differential h2  h1 corresponding to a saturated temperature differential
Ts2  Ts1  63.17  35  28.17°F in the two-phase region is
h2  h1  0.18165  0.21502(28.17)  0.0012405(28.17)2  8.1982  106(28.17)3
 0.182  6.057  0.984  0.183  5.074 Btu/ lb
Similarly, from Eq. (9.22), the enthalpy differential h4  h3 corresponding to a saturated temperature
differential Ts  Ti  95  63.17  31.83°F is
h4  h3  0.020  0.16352(31.83)  0.00035106(31.83)2  1.9177  106(31.83)3
 0.020  5.205  0.356  0.062  4.891 Btu/ lb

36.162  28.095
109.946  28.095
 0.09856
x 
h5	  h8
h7  h8
pi  ?pcon pev  ?196.5  76.17  122.34 psia
9.34 CHAPTER NINE
Then, from Eq. (9.33), the total work input to the compressor is calculated as
Win  (1  x)(h2  h1)  h4  h3
 (1  0.09856)(5.074)  4.891  9.465 Btu/ lb (22.0 kJ /kg)
4. The coefficient of performance of this two-stage compound system is
5. The mass flow rate of refrigerant flowing through the condenser can be evaluated as
6. From the results in Example 9.1, the power input to the single-stage system is 330.2 hp. The
power input to the two-stage compound system is
Energy saving compared with the single-stage system is calculated as
9.14 THREE-STAGE COMPOUND SYSTEM WITH
A TWO-STAGE FLASH COOLER
To reduce the energy consumption of refrigeration systems in air conditioning, the three-stage compound
system with a two-stage flash cooler became a standard product in the 1980s. Fig. 9.9 shows
the schematic diagram and refrigeration cycle of this system.
Flow Processes
In Fig. 9.9, vapor refrigerant enters the first-stage impeller of the centrifugal compressor at a dry
saturated state, point 1. After the first-stage compression process, at point 2, it mixes with the vaporized
refrigerant coming from the low-pressure flash cooler at point 12. At point 3, the mixture
enters the second-stage impeller. After the second-stage compression process at point 4, it mixes
again with vaporized refrigerant from the high-pressure flash cooler at point 9. The mixture, at
point 5, is then compressed to the condensing pressure in the third-stage impeller. At point 6, the
hot gas enters the condenser, condenses to liquid, and subcools to a temperature below the condensing
temperature. Subcooled liquid refrigerant leaves the condenser at point 7	 and flows through a
two-stage flash cooler and the associated throttling devices, in which a small portion of liquid refrigerant
is successively flashed into vapor at interstage pressure pi1, point 9, and interstage pressure
pi2, point 12. The liquid-vapor mixture enters the evaporator at point 14, and the remaining liquid
refrigerant evaporates into vapor completely in the evaporator.
Fraction of Refrigerant Vaporized in Flash Cooler
Based on the heat balance of the refrigerants entering and leaving the high-pressure flash cooler, as
shown in Fig. 9.10a, the fraction of liquid refrigerant vaporized in the high-pressure flash cooler
330.2  311.1
330.2
 0.058, or 5.8%
Pin 
500  200
42.41  7.58
 311.1 hp (232 kW)
m? r 
Qrc
qrf

500  12,000
71.78
 83,588 lb / h (37,916 kg / h)
COPref 
qrf
Win

71.78
9.465
 7.58
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.35
9.36
FIGURE 9.9 Three-stage compound system with a two-stage flash
cooler: (a) schematic diagram; (b) refrigeration cycle.
x1 at interstage pressure pi1 can be calculated as
h7	  (1  x1)h10  x1h9
Therefore,
(9.36)
where h7	, h9, h10  enthalpies of refrigerants at points 7	, 9, and 10, respectively, Btu / lb (kJ / kg).
In the same manner, the heat balance of the refrigerant entering and leaving the low-pressure flash
cooler, as shown in Fig. 9.10b, may be expressed as
(1  x1)h10  x2h12  (1  x1  x2)h13
where h12, h13  enthalpies of the refrigerant at points 12, and 13, respectively, Btu / lb (kJ / kg). The
fraction of liquid refrigerant vaporized in the low-pressure flash cooler x2 at an interstage pressure
pi2 can be evaluated as
(9.37) x2 
(1  x1)(h10  h13)
h12  h13
x1 
h7	  h10
h9  h10
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.37
FIGURE 9.10 Heat balance of refrigerants entering and leaving the high-pressure and low-pressure flash
cooler and mixing points. (a) High-pressure flash cooler; (b) low-pressure flash cooler; (c) at the mixing point
before entering second-stage impeller; (d) at the mixing point before entering third-stage impeller.
Coefficient of Performance of Three-Stage System
From a heat balance of the refrigerants entering and leaving the mixing point before the inlet of the
second-stage impeller, as shown in Fig. 9.9c, the enthalpy of the mixture at point 3, h3 (Btu/ lb or
kJ/ kg) is given as
(9.38)
where h2  enthalpy of the gaseous refrigerant discharged from the first-stage impeller at point 2,
Btu/ lb (kJ / kg). As shown in Fig. 9.9d, the enthalpy of the mixture of vapor refrigerants at point 5,
h5, Btu / lb (kJ / kg), can also be evaluated as
h5  (1  x1)h4  x1h9 (9.39)
where h4  enthalpy of the gaseous refrigerant discharged from the second-stage impeller at point
4, Btu/ lb (kJ / kg). The refrigeration effect qrf in Btu/ lb of refrigerant flowing through the condenser
is given as
qrf  (1  x1  x2)(h1  h14) (9.40)
where h1, h14  enthalpies of the refrigerants leaving the evaporator at point 1 and entering the
evaporator at point 14, respectively, Btu / lb (kJ / kg). Total work input to the three-stage compressor
Win, Btu / lb (kJ / kg) of refrigerant flowing through the condenser, is given by
Win  (1  x1  x2)(h2  h1)  (1  x1)(h4  h3)  h6  h5 (9.41)
where h6  enthalpy of the hot gas discharged from the third-stage impeller at point 6, Btu/ lb
(kJ/ kg).
From Eqs. (9.40) and (9.41), the coefficient of performance of this three-stage system with a
two-stage flash cooler is
(9.42)
A three-stage compound system with a two-stage flash cooler often has a further energy saving of
about 2 to 5 percent compared to a two-stage compound system with a flash cooler.
The mass flow rate of the refrigerant flowing through the condenser , lb/ h (kg/h), can be calculated
as
(9.43)
9.15 TWO-STAGE COMPOUND SYSTEM WITH
A VERTICAL INTERCOOLER
When an evaporating temperature in the range of 10 to 50°F (23.3 to 45.6°C) is required, a
two-stage compound system using reciprocating or screw compressors is usually applied. Figure 9.11
shows the schematic diagram and the refrigeration cycle of a two-stage compound system with a vertical
coil intercooler. The subcooled liquid refrigerant from the receiver at point 5	 is divided into two
streams. One stream enters the coil inside the intercooler. The other stream enters its shell after throttling
to point 6, the interstage pressure.
m? r 
Qrc
qrf

qrc
(1  x1  x2)(h1  h14)
m?r

(1  x1  x2)(h1  h14)
(1  x1  x2)(h2  h1)  (1  x1)(h4  h3)  (h6  h5)
COPref 
qrf
Win
h3 
(1  x1  x2)h2  x2 h12
1  x1
9.38 CHAPTER NINE
In the intercooler shell, some of the liquid refrigerant vaporizes to saturated vapor at point 7,
drawing latent heat from the liquid in the coil at point 5	, further subcooling it to point 10. This subcooled
liquid is throttled by the expansion valve at point 10 and then evaporates to saturated vapor
at point 1 in the evaporator. Vapor refrigerant from the evaporator at point 1 enters the low-stage
compressor. The compressed hot gas at point 2 discharges into the intercooler, mixing with the
liquid from the receiver at the interstage pressure. The liquid level in the intercooler is controlled by
the saturated temperature at the interstage pressure in the intercooler. The saturated vapor from
the vertical coil intercooler at point 7 enters the high-stage compressor. At point 4, hot gas is
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.39
FIGURE 9.11 Two-stage compound system with a vertical coil intercooler:
(a) schematic diagram; (b) refrigeration cycle.
discharged from the high-stage compressor and then condensed and subcooled to point 5	 in the
condenser.
In this system, x is the fraction of liquid refrigerant vaporized in the intercooler, and h10 is the
enthalpy of the liquid refrigerant that has been subcooled in the vertical coil. Based on the heat balance
of the refrigerants entering and leaving the intercooler, as shown in Fig. 9.11a,
(1  x)h5	  xh5  (1  x)h2  h7  (1  x)h10
Then
(9.44)
This type of system is often used in low-temperature refrigerated cold storage and other industrial
applications. Ammonia is often used as the refrigerant.
Comparison between Flash Cooler and Vertical Coil Intercooler
Hot gas discharged from the low-stage compressor is always desuperheated to a nearly saturated vapor
state at the interstage pressure in the vertical coil intercooler. This process is more appropriate
for a refrigerant like ammonia, which has a high discharge temperature. In flash coolers, desuperheating
is caused by the mixing of flashed vapor and hot gas, and will not result in a dry saturated
state. Therefore, flash coolers are usually used in refrigeration systems using HCFCs or HFCs.
The liquid refrigerant flowing inside the coils of a vertical coil can be maintained at a slightly
lower pressure than condensing pressure, whereas the pressure of liquid refrigerant in the flash
cooler is decreased to the interstage pressure. Some refrigerant may be preflashed before the throttling
device, causing a waste of refrigerating capacity. For a flash cooler, the available pressure drop
in the throttling device is lower.
9.16 CASCADE SYSTEMS
A cascade system consists of two separate single-stage refrigeration systems: a lower system that
can better maintain lower evaporating temperatures and a higher system that performs better at
higher evaporating temperatures. These two systems are connected by a cascade condenser in
which the condenser of the lower system becomes the evaporator of the higher system as the higher
system’s evaporator takes on the heat released from the lower system’s condenser.
It is often desirable to have a heat exchanger between the liquid refrigerant from the cascade
condenser and the vapor refrigerant leaving the evaporator of the lower system. The liquid refrigerant
can be subcooled to a lower temperature before entering the evaporator of the lower system, as
shown in Fig. 9.12a. Because the evaporating temperature is low, there is no danger of too high a
discharge temperature after the compression process of the lower system.
When a cascade system is shut down while the temperature of the ambient air is 80°F (26.7°C),
the saturated vapor pressure of the refrigerant increases. For a lower system using HFC-125 as the
refrigerant, this saturated pressure may increase to 208.91 psia (1440 kPa abs.). For safety reasons, a
relief valve at the cascade condenser connects to an expansion tank, designed to store the refrigerant
from the lower system in case of shutdown. For extremely low evaporating temperatures, a multistage
compression system may be used in either the lower or higher system of a cascade system.
Advantages and Disadvantages
The main advantage of a cascade system is that different refrigerants, equipment, and oils can
be used for the higher and the lower systems. This is especially helpful when the evaporating
x 
h2  h7  h5	  h10
h2  h10
9.40 CHAPTER NINE
temperature required in the lower system is less than 100°F (73°C). The specific volume of the
suction vapor vsuc is extremely important in such temperature applications. A greater vsuc always requires
a large compressor and more space. For instance, at 80°F (62.2°C), saturated vapor of
HCFC-22 has a vsuc of 9.70 ft3 / lb (0.605 m3 / kg), whereas for HFC-125 vsuc is only 4.67 ft3 / lb
(0.291 m3 / kg).
One disadvantage of a cascade system is the overlap of the condensing temperature of the lower
system and the evaporating temperature of the higher system for heat transfer in the condenser. The
overlap results in higher energy consumption. Also a cascade system is more complicated than a
compound system.
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.41
FIGURE 9.12 Cascade system: (a) schematic diagram; (b) p-h diagram; (c) T-s diagram.
Performance of Cascade System
The performance of the cascade system can be measured in terms of 1 lb (kg) of refrigerant in the
lower system, for the sake of convenience. In Fig. 9.12b and c, which shows the refrigeration cycles
on p-h and T-s diagrams, the lower system is indicated by points 1, 2, 3	, and 4, and the higher system
by points 5, 6, 7, and 8. The desired low-temperature refrigeration effect of the cascade system
can be calculated as
qrf  h1  h4 (9.45)
If the heat transfer between the cascade condenser and the surroundings is ignored, then the heat released
by the condenser of the lower system is equal to the refrigerating load on the
evaporator of the higher system , that is
where mass flow rate of refrigerants in lower and higher systems respectively, lb/h
(kg/ s). Therefore, the ratio of the mass flow rate of refrigerant in the higher system to the mass flow
rate in the lower system is
(9.46)
The mass flow rate of refrigerant in the lower system , lb/h (kg/ s), is
(9.47) m?l 
qrc, l
qrf,l
m?l
m? h
m? l

h2  h3
h5  h8
m?l, m? h 
m? l(h2  h3)  m? h (h5  h8)
m? h (h5  h8)
m?l (h2  h3)
9.42 CHAPTER NINE
FIGURE 9.12 (Continued)
where qrc,l  refrigeration capacity of lower system, Btu/h (W)
qrf,l  refrigeration effect of lower system, Btu/ lb (J / kg)
Total work input to the compressors in both higher and lower systems Win, Btu / lb (J / kg), of refrigerant
in the lower system can be evaluated as
(9.48)
where h1	  enthalpy of vapor refrigerant leaving heat exchanger, Btu/ lb (J/kg). The coefficient of
performance of the cascade system is given as
(9.49)
Example 9.3. A cascade system for an environmental chamber is designed to operate at the following
conditions during summer:
Vapor refrigerant enters the compressor of the higher system at a dry saturated state, and liquid refrigerant
leaves the condenser of the higher system without subcooling. Ignore the pressure losses
in the pipelines and valves, and assume that the compression processes for both the higher and
lower systems are isentropic.
Based on the data from the Properties of Saturated Liquid and Saturated Vapor of HFC-125
listed in ASHRAE Handbook 1997, Fundamentals, the following formulas can be used to calculate
the enthalpies of the refrigerants:
Enthalpy of refrigerant HFC-125 at saturated liquid state at temperature Tsl between 0 and 50°F
(17.8 and 10°C) hsl, Btu / lb (kJ / kg), is
hsl  11.04  0.300Tsl
Enthalpy of refrigerant HFC-125 at saturated vapor state at temperature Tsv between 80 and
30°F ( 62.2 and 34.4°C) hsv, Btu / lb (kJ / kg), is
hsv  63.30  0.14[Tsv(80)]
Enthalpy difference of vapor refrigerant HFC-125 along the constant-entropy line when Tcon 
30°F (1.1°C) and Tev > 80°F (26.7°C) is
	hsv  0.175(Tcon  Tev)
Here Tcon represents the condensing temperature and Tev the evaporating temperature, both in
°F (°C). The specific heat of saturated vapor of HFC-125 cpv, in the temperature range of 80 to
40°F (62.2 to 40°C) is 0.14 Btu/ lb °F (586 J/kg°C). The specific heat of saturated liquid
Lower system Higher system
Refrigerant HFC-125 HCFC-22
Evaporating temperature,°F 60 10
Evaporating pressure, psia 12.75 47.46
Condensing temperature,°F 20 100
Condensing pressure, psia 78.5 210.6
Suction temperature,°F 20
Subcooling after heat exchanger,°F 1
Refrigerating load, Btu /h 100,000
COPref 
qrf
Win

m? l(h1  h4)
m? l(h2  h1	)  m? h(h6  h5)
Win  (h2  h1	) 
m? h(h6  h5)
m? l
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.43
of HFC-125 cpl in the temperature range of 0 to 50°F (17.8 to 10°C) is 0.300 Btu/ lb °F (1256
J/kg°C).
Calculate:
1. The refrigeration effect per pound (kilogram) of refrigerant in the lower system
2. Total work input to the compressors
3. Coefficient of performance of this cascade system
4. Mass flow rates of the refrigerants in the higher and lower systems
Solution
1. From the given values, the enthalpy of the saturated vapor refrigerant at 60°F is
h1  63.30  0.14[Tsv  (80)]
 63.30  0.14[ 60  (80)]  66.10 Btu/ lb
Enthalpy of saturated liquid refrigerant at a condensing temperature of 20°F is
h3  11.04  0.30Tsl
 11.04  0.30(20)  17.04 Btu/ lb
and from Eq. (9.26)
h1	  h4  h3  cpl (T3  T3	)
 17.04  0.30(20  1)  11.34 Btu/ lb
Then the refrigeration effect of the lower system is
qrf,l  h1  h4  66.12  11.34  54.78 Btu/ lb (127.4 kJ /kg)
2. Enthalpy of superheated vapor refrigerant after the heat exchanger in the lower system is
h1	  h1  cpv(degree of superheat)
 66.12  0.14[ 20  (60)]  71.72 Btu/ lb
Enthalpy difference (h2  h1	) along the constant-enthalpy line of the lower system is
h2  h1	  0.175 (Tcon  Tlv)
 0.175[20  (60)]  14.0 Btu/ lb
Enthalpy of the hot gas discharged from the compressor of the lower system is
h2  h1	  (h2  h1	)
 71.72  14.0  85.72 Btu/ lb
From Eq. (9.20), the enthalpy of saturated vapor refrigerant HCFC-22 in the higher system is
h5  104.465  0.098445Ts  0.0001226Ts
2  9.861  107Ts
3
 104.465  0.098445(10)  0.0001226(10)2  9.861  107(10)3
 104.465  0.984  0.012  0.001  105.436 Btu/ lb
9.44 CHAPTER NINE
From Eq. (9.22), the enthalpy difference along the constant-entropy line of the higher system is
h6  h5  0.18165  0.21502(Ts2  Ts1)  0.0012405(Ts2  Ts1)2  8.1982  10 6
(Ts2  Ts1)3
 0.18165  0.21502(100  10)  0.0012405(100  10)2  8.1982  106
(100  10)3
 15.09 Btu/ lb
According to Eq. (9.20), the enthalpy of liquid refrigerant HCFC-22 at condensing temperature
100°F in the higher system is
From Eq. (9.46), the ratio of the mass flow rate of refrigerant in the higher system to that of the
lower system is
Then, from Eq. (9.48), the total work input to the compressors is
3. Coefficient of performance of the cascade system is
4. Mass flow rate of refrigerant in the lower system is
The mass flow rate of refrigerant in the higher system is
9.17 AIR EXPANSION REFRIGERATION CYCLES
Thermodynamic Principle
According to the steady flow energy equation
For an adiabatic process, q#  0. Assuming that the difference between the kinetic energy v1
2/(2gc)
and v2
2/(2gc) is small compared with the enthalpy difference h1  h2, it can be ignored. For air
expanding in a turbine and producing work, the above equation can be rewritten as
W  h1  h2
h1 
v1 2
2gc  778
 q#  h2 
v2 2
2gc  778
 W
m? h  1.04 m? l  1.04(30.42)  1898 lb / h (861 kg / h)
m? l 
Qrc, l
qrf, l

100,000
54.78
 1825 lb / h (828 kg / h)
COPref 
qrf
Win

54.78
29.69
 1.85
 14.0  1.04(15.09)  29.69 Btu / lb (69.06 kJ / kg)
Win  (h2  h1	) 
m?h(h6  h5)
m? l
m?h
m? l

h2  h3
h5  h8

85.72  17.04
105.44  39.27
 1.04
 10.409  0.26851(100)  0.00014794(1002)  5.3429  107(1003)  39.27 Btu / lb
h7  h8  10.409  0.26851Ts  0.00014794 (T s
2)  5.3429  107 T s
3
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.45
Usually air can be considered a perfect gas, so
	h  cp	TR (9.50)
where cp  specific heat of air at constant pressure, Btu/ lb °R (kJ/kgK)
TR  absolute temperature, °R (K)
and
W  cp(TR1  TR2) (9.51)
When the expansion process is a reversible adiabatic process, the temperature-pressure relationship
can be expressed as
(9.52)
TR1
TR2
  p1
p2 ( 1) / 
9.46 CHAPTER NINE
FIGURE 9.13 Air expansion refrigeration system and refrigeration cycle. (a) refrigeration system;
(b) air expansion refrigeration cycle.
where p1, p2  air pressure at inlet and outlet of turbine, psia (kPa abs.)
 ratio of specific heat at constant pressure to specific heat at constant volume
Equations (9.51) and (9.52) demonstrate that when air expands in a turbine and produces work, its
absolute temperature TR1 drops to TR2. The decrease in absolute temperature is proportional to the
pressure ratio (p1 /p2) ( 1) / and the work produced W. The greater the pressure drop in the turbine,
the lower the temperature at the outlet of the turbine. This type of refrigeration has been used for air
conditioning systems in many aircraft.
Flow Processes of a Basic Air Expansion System for Aircraft
Figure 9.13a is a schematic diagram of a basic air expansion refrigerating system, and Fig. 9.13b
depicts the air expansion refrigeration cycle. In Fig. 9.13, ambient air that surrounds an aircraft
flying at subsonic speed and at high altitude, at point 0 with pressure p0, is rammed into the engine
scoop. Before entering the engine, the pressure of the air at point 1	 is increased to p1. When ram air
enters the engine compressor, it is compressed to pressure p2 and bleeds to a heat exchanger at point
2	. In the heat exchanger, the bleed air is cooled at constant pressure to point 3 by another stream of
ram air extracted by a fan. Bleed air is then expanded in a turbine and cooled to point 4	. This cold
air is supplied to the cabin and flight deck for air conditioning. Cabin air is then discharged to the
atmosphere after absorbing the sensible and latent heat from these conditioned spaces. Air expansion
refrigeration cycles for aircraft are open cycles because there is no recirculated air.
Another alternative is to add an evaporative cooler between section A-A and B-B, as shown in
Fig. 9.13, to cool the supply air a few degrees more. An evaporant such as water or alcohol can be
used to extract heat from the supply air by indirect contact.
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.47
FIGURE 9.13 (Continued)
The stream of ram air, extracted by a fan driven by the cooling turbine to cool the bleed air, is
also discharged to the atmosphere at the cooling air exit. The work output of the turbine is mainly
used to drive the fan.
Air Expansion Cycle
The air expansion refrigeration cycle is different from a vapor compression cycle because the refrigerant,
air, remains in gaseous form, does not evaporate, and condenses. Air is also not recirculated
and is used directly as the supply air for cooling the conditioned spaces.
Work input to the engine compressor Wc, Btu / lb (kJ / kg), is given by
(9.53)
where TR1	, TR2	 absolute temperature of ram air and air after compression process, respectively,
°R (K). The rate of heat transfer in the heat exchanger between the bleed air and the ambient air for
cooling Qex, Btu /h (W), can be calculated as
(9.54)
where TR3  absolute temperature of air after heat exchanger, °R (K)
mass flow rate of bleed air, lb/h (kg/ s)
Work output from the turbine Wo during the expansion process, Btu/ lb (J / kg), can be calculated as
(9.55)
where TR4	  absolute temperature of cold supply air at exit of turbine, °R (K). The amount of sensible
heat Qsen absorbed by the cold supply air in the air cabin or in the flight deck, Btu/h (W), is
(9.56)
where TRr  absolute temperature of air cabin, °R (K). The relationship between temperature and
pressure for an isentropic compression process is
(9.57)
For an isentropic expansion process, the relationship between temperature and pressure is
(9.58)
where TR2, TR4  absolute temperature of air after isentropic compression and isentropic expansion
processes, °R (K).
The isentropic efficiency for the compressor 
c is
(9.59)
Similarly, isentropic efficiency for the turbine 
t is
(9.60)

t 
TR3  TR4	
TR3  TR4

TR2  TR1	
TR2	  TR1	

c 
cp(TR2  TR1	 )
cp(TR2	  TR1	 )
TR3
TR4
  p3
p4 ( 1) / 
TR2
TR1	
  p2
p1 ( 1) / 
Qsen  m? bcp(TRr  TR4	 )
Wo  cp(TR3  TR4	 )
m? b 
Qex  m? bcp(TR2	  TR3)
Wc  cp(TR2	 TR1	)
9.48 CHAPTER NINE
If 
c, 
t, ,TR3, and pressure ratios p2 /p1 and p3 /p4 are known values, , and , can be determined
by using Eqs. (9.57) to (9.60).
The effectiveness of the heat exchanger can be evaluated as
(9.61)
where , TR3  absolute temperature of bleed air entering and leaving heat exchanger, °R (K)
TRt1  absolute total temperature of cooling ram air entering heat exchanger,
measured at outer heat-transfer surface at rest, °R (K)
From Eq. (9.61),  depends mainly on the construction of the heat exchanger and the ratio , if
the difference of cp between bleed and cooling ram air is ignored. Here indicates the mass flow
rate of cooling ram air, lb/h (kg/ s).
If air at a velocity v with an absolute temperature TR finally reaches a total temperature TRt when
brought to rest adiabatically, then, based on the steady flow energy equation;
or
(9.62)
When the air velocity v  200 ft/ s (61 m/ s), the term v2 / [(2cp)(778gc)] is small compared with TR
and can be ignored. Hence, only TRt is used for cooling ram air.
Net work input to the air expansion refrigeration cycle is calculated as
(9.63)
If Wt is used to drive the fan to extract the cooling air, then the coefficient of performance of the
basic air expansion refrigeration cycle can be calculated as
(9.64)
The main advantage of an air expansion refrigeration system for an aircraft is its simple construction
and light weight. The main drawbacks are the lower COPref and the need for an additional air
conditioning and refrigeration system when the aircraft is on the ground.
9.18 REFRIGERATION SYSTEMS—CLASSIFICATIONS
AND DEVELOPMENTS
Classifications
As discussed in Sec. 9.1, based on the type of energy input and refrigeration process, refrigeration
systems can be classified as vapor compression, absorption, or air / gas expansion systems. In the
1990s, the absorption system has a share of only less than 8 percent in new installations compared

TRr  TR4	
TR2	  TR1	
COPref 
cp(TRr  TR4	)
cp(TR2	  TR1	)
Wnet  Wc  Wt  cp(TR2	  TR1	)  cp(TR3  TR4	)
TRt  TR 
v2
2cp(778gc)
cpTR 
v2
2gc  778
 cpTRt
m?c
m? b / m?c
TR2	

TR2	  TR3
TR2	  TRt1
 
cp m? b(TR2	  TR3)
cp m? b (TR2	  TRt1)
TR4	 TR2	 TR1	
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.49
with centrifugal chillers in the United States. Applications of air or gas expansion refrigeration systems
are limited in aircraft and cryogenics. Vapor compression systems dominate in both comfort
and processing air conditioning.
Vapor compression refrigeration systems can also be classified according to
 The type of compressor used, such as reciprocating, rotary, scroll, screw, or centrifugal systems.
 The type of evaporator: direct-expansion (DX) cooler or liquid chillers. In a DX cooler system,
space air is directly cooled and dehumidified by the expanded and evaporated refrigerant in a DX
coil. A liquid chiller is a factory-assembled refrigeration package in which chilled water is
produced in its evaporator. Air is then cooled and dehumidified by chilled water in air-handling
units (AHUs) or in fan coils.
 The size (tonnage in capacity ) of the refrigeration system, such as small ( 2.5 ton, or 8.8 kW),
medium, and large ( 75 tons or 264 kW).
For a refrigeration system, the type of compressor, system size, and whether a DX coil or a water
chiller is used are often interrelated. A centrifugal chiller is often a large-tonnage refrigeration system.
A small refrigeration system with a cooling capacity of less than 2.5 tons (8.8 kW) is often an
air-cooled DX cooler using a reciprocating, rotary, or scroll compressor. Refrigeration systems can
therefore be primarily classified as follows:
 DX (direct-expansion) cooler systems:
Air-cooled, reciprocating DX coolers
Air-cooled, scroll DX coolers
Air-cooled rotary DX coolers
 Liquid chillers
Centrifugal chillers
Screw chillers
Scroll chillers
Reciprocating chillers
 Liquid overfeed systems
 Multistage systems
 Absorption systems
 Air or gas expansion systems
Among these, DX coolers have a share of about 70 percent based on the air conditioned floor
area in commercial buildings in 1992. For air conditioned homes, more than 90 percent use
DX coolers. Chillers have a share of about 28 percent of floor space in air conditioned commercial
buildings. According to the U.S. domestic market research 1997 estimate provided by
BSRIA/Ducker in Air Conditioning, Heating and Refrigeration News, September 22, 1997, for a
total number of nearly 18,460 chillers sold, the large-tonnage centrifugal and screw chillers with a
capacity  75 tons (264 kW) had a share of 28 percent in units and a value of about $505 million.
Reciprocating, screw, and scroll chillers with a capacity  75 tons (264 kW) using step unloading
had a share of 61 percent in units and a value of about $380 million. Absorption and gas-engine
chillers had a value of about $66 million.
Refrigeration systems are also classified according to their evaporating temperatures Tev as
9.50 CHAPTER NINE
 40°F  Tev  0°F (40°C  Tev 18°C)
0°F Tev  32°F (18°C  Tev  0°C)
Tev  32°F (Tev  0°C)
Low-temperature systems
Medium-temperature systems
High-temperature systems
ASHRAE Standard 15-1994 classifies refrigeration systems as direct systems and indirect systems
based on the method employed for extracting or delivering heat. A direct refrigeration system is one
in which the evaporator or condenser of the refrigeration system is in direct contact with air or other
media to be cooled or heated. An indirect refrigeration system uses a secondary coolant to cool the
air or other substance after the coolant is cooled by the refrigeration system.
According to the probability that leakage of refrigerant will enter the occupied space, ASHRAE
Standard 15-1994 divides refrigeration systems into the following two categories:
High-probability system. Any refrigeration system in which a leakage of refrigerant from a
failed connection, seal, or component could enter the occupied space (the space is frequently
occupied by people) is a high-probability system. A direct system is a high-probability system.
Low-probability system. This is a refrigeration system in which leakage of refrigerant from a
failed connection, seal, or component cannot enter the occupied space. An indirect system that
uses secondary coolants, such as chilled water, brines, or glycols, is a low-probability system.
Recent Developments
In the United States in the 1990s, refrigeration systems underwent the following major developments:
 The production of CFC and halons in developed countries ceased from January 1, 1996. However,
the conversion of CFCs to non–ozone depletion HFCs and HCFCs was slower than expected.
 For small and medium-size refrigeration systems, the direct-expansion systems using reciprocating,
rotary, scroll, and screw compressors dominate in the individual and packaged air conditioning
systems. For large-size refrigeration systems, centrifugal and screw chillers are widely adopted.
 Energy-efficient rotary, scroll, and screw compressors are gradually replacing reciprocating compressors.
According to Air Conditioning, Heating and Refrigerating News, May 6, 1996, of the
compressors in new packaged units and room air conditioners 35 percent are scroll compressors.
Based on BSRIA/Ducker analyses for the U.S. domestic market in 1996, rotary compressors
dominated the room air conditioner market of less than 2 tons (7 kW), scroll compressors dominated
the small chillers, screw compressors had a two-thirds share at the top of medium-size units
(3 to 75 tons, or 10 to 264 kW), and reciprocating chillers showed a share of 70 percent of the
units with a cooling capacity around 5 tons (18 kW).
 The energy use of new centrifugal compressors dropped considerably from an average of 0.85
kW/ton (4.14 COP) in the late 1970s in the United States to 0.60 kW/ ton (5.86 COP) in 1996. In
1997, the lowest-energy-use new centrifugal chiller is 0.49 kW /ton (7.17 COP).
The rotary, scroll, screw, and centrifugal compressors will form a series of rotary refrigeration
products to meet mainly the requirements of new and retrofit, small, medium, and large refrigeration
systems in the twenty-first century. Air conditioning becomes the largest user of refrigeration.
9.19 REFRIGERATION COMPRESSORS
A refrigeration compressor is the heart of a vapor compression refrigeration system. Its function is
to raise the pressure of the refrigerant and provide the primary force to circulate the refrigerant. The
refrigerant thus produces the refrigeration effect in the evaporator, condenses into liquid form in the
condenser, and throttles to a lower pressure through the throttling device.
Positive Displacement and Non–Positive Displacement Compressors
According to the characteristics of the compression process, currently used refrigeration compressors
can be classified as positive displacement compressors or non–positive displacement
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.51
compressors. A positive displacement compressor increases the pressure of the vapor refrigerant by
reducing the internal volume of the compression chamber through mechanical force applied to the
compressor. This type of refrigeration compressor mainly includes reciprocating, scroll, screw, and
rotary compressors.
9.52 CHAPTER NINE
FIGURE 9.14 Hermetic and open-type compressors:
(a) hermetic; (b) semihermetic; and (c) open.
The only type of non–positive displacement refrigeration compressor widely used in refrigeration
systems is the centrifugal compressor. In a centrifugal compressor, the increase of the pressure
of the vapor refrigerant depends mainly on the conversion of dynamic pressure to static pressure.
Hermetic, Semihermetic, and Open Compressors
In a hermetic compressor, the motor and the compressor are “sealed” or “welded” in the same housing,
as shown in Fig. 9.14a. Hermetic compressors have two advantages: They minimize leakage of
refrigerant, and the motor can be cooled by the suction vapor flowing through the motor windings,
which results in a small and cheaper compressor-motor assembly.
Motor windings in hermetic compressors must be compatible with the refrigerant and lubrication
oil, resist the abrasive effect of the suction vapor, and have high dielectric strength. Welded
compressors are usually used for small installations from  1 hp to 24 hp ( 0.7 kW to 18 kW).
Semihermetic compressors are also known as accessible hermetic compressors (see Fig. 9.14b).
The main advantage that semihermetic compressors have over hermetic compressors is accessibility
for repair during a compressor failure or for regular maintenance. Other features are similar to those
of the hermetic compressor. Most of the medium compressors are semihermetic.
In an open compressor, the compressor and the motor are enclosed in two separate housings, as
shown in Fig. 9.14c. An open compressor needs shaft seals to minimize refrigerant leakage. In most
cases, an enclosed fan is used to cool the motor windings by using ambient air. An open compressor
does not need to evaporate the liquid refrigerant to cool the hermetic motor windings. Compared
with hermetic compressors, open compressors may save 2 to 4 percent of the total power input.
Many very large refrigeration compressors are open compressors.
Direct Drive, Belt Drive, and Gear Drive
Hermetic compressors are driven by motor directly or driven by gear trains. Both semihermetic and
open compressors can be driven directly, driven by gear trains, or driven by motor through V belts.
The purpose of a gear train is to increase the speed of the compressor. Gear drive is compact in size
and rotates without slippage. Like belt drive, gear drive needs about 3 percent more power input
than direct-drive compressors. Some large open compressors may be driven by steam turbine, gas
turbine, or diesel engine instead of electric motor.
9.20 PERFORMANCE OF COMPRESSORS
Volumetric Efficiency
The volumetric efficiency 
v of a refrigeration compressor is defined as
(9.65)
where Va,v  actual induced volume of suction vapor at suction pressure, ft3 (m3)
Vdis  theoretical displacement of compressor, ft3 (m3)
Factors that influence the 
v of the compressor include:
 Clearance volume and compression ratio Rcom. Both factors affect the volume of reexpansion gas
trapped in the clearance volume.

v 
Va,v
Vdis
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.53
 Heating effect. When vapor refrigerant enters the compressor, heat absorbed by the vapor results
in a heating effect that increases the specific volume of the refrigerant and, therefore, the Va,v
value.
 Leakage. Refrigerant leaks through the gap and the clearance across the high- and low-pressure
sides of the compressor, such as the clearance between the piston ring and the cylinder in a reciprocating
compressor.
Motor, Mechanical, and Compression Efficiency
Motor efficiency 
mo is defined as
(9.66)
where Pcom, Pmo  power input to shaft of compressor and to motor, respectively, hp (kW). Mechanical
efficiency 
mec is defined as
(9.67)
where Wv, Wcom  work delivered to vapor refrigerant and to compressor shaft, Btu/lb (kJ/ kg).
Compression efficiency 
cp is defined as
(9.68)
where Wisen  work required for isentropic compression, Btu/lb (kJ/kg). Here, Wisen  h2  h1,
where h1 and h2 represent the enthalpy of intake vapor refrigerant and of discharged hot gas, respectively,
in an isentropic compression process, in Btu/ lb (kJ / kg). The difference between work input
on the compressor shaft and work delivered to the vapor refrigerant Wcom  Wv is mainly caused by
the friction loss and turbulent loss of refrigerant flow.
Compressor efficiency 
com is the product of 
cp and 
mec, that is,

com  
cp 
mec (9.69)
Isentropic and Polytropic Analysis
The isentropic efficiency 
isen of a compressor is defined as
(9.70)
where h2	  enthalpy of discharged hot gas if compression process is not isentropic, Btu / lb
(kJ/ kg). The difference between h2	 and h2 implies, first, the deviation of a reversible polytropic
process from an isentropic process and, second, the deviation of an irreversible polytropic process
from a reversible process. Isentropic efficiency is equal to compressor efficiency, or 
isen  
com.
The actual power input to the compressor Pcom, hp, can be calculated as
(9.71) Pcom 
m? r (h2	  h1	)
42.41

m? r (h2  h1)
42.41 
isen

m? r (h2  h1)
42.41 
com

isen 
h2  h1
h2	  h1

cp 
Wisen
Wv

mec 
Wv
Wcom

mo 
Pcom
Pmo
9.54 CHAPTER NINE
In Eq. (9.71), represents the mass flow rate of refrigerant, lb /min (kg/min). For reciprocating
compressors, can be calculated as
(9.72)
where piston displacement, cfm (or m3/min, to be covered in Chap. 11)
suc  density of suction vapor, lb/ ft3 (kg/m3)
Although the actual compression processes for most compressors are irreversible polytropic
processes, for simplicity, an isentropic analysis is often used. In other words, 
isen is a widely used
efficiency index for refrigeration compressors, and actual power input to the compressor Pcom is
usually calculated by Eq. (9.71).
Energy Use Index
According to ASHRAE/IESNA Standard 90.1-1999, the following are the current energy use indices
for refrigeration compressors, packaged units, heat pumps, and chillers:
 Energy efficiency ratio (EER) is defined as the ratio of the net cooling capacity of a refrigeration
compressor, a packaged unit, or other device, in Btu/ h, to the electric power input to that device,
in W, under designated operating conditions.
 SEER indicates the seasonal energy efficiency ratio. This is the total cooling capacity of a device
during its normal annual usage period, in Btu, divided by the total electric energy input during the
same period, in watt-hours.
 IPLV indicates the integrated part-load value. This is a single index of merit that is based on partload
EER or COP. It expresses part-load efficiency for refrigeration compressors, packaged units,
and heat pumps based on the weighted operation at various load capacities.
 HSPF indicates the heating season performance factor. This is the total heating output of a heat
pump during its normal annual usage period for heating, in Btu, divided by the total electric energy
input to the equipment during the same period, in watt-hours.
 kW/ ton indicates the electric power consumption of a refrigerating compressor per ton of refrigeration
output. It is a clear and widely used energy index.
From Eq. (9.71) and because , the kW/ton for a hermetic refrigeration compressor
can be calculated as
(9.73)
where Win  work input to compressor (including single-, two-, or three-stage) per lb of refrigerant
flowing through condenser, Btu / lb (kJ /kg)
qrf  refrigeration effect per lb of refrigerant flowing through condenser, Btu / lb (kJ /kg)

isen  isentropic efficiency of compressor

mo  motor efficiency
From Eq. (9.73), primary factors that affect the energy use of the kW/ton index for a refrigeration
compressor are as follows:
 Difference between the condensing and evaporating temperature Tcon  Tev
 Isentropic efficiency 
isen or compressor efficiency 
com
 Refrigeration effect of the specific refrigerant

3.516
COPref
kW / ton 
3.516 Win
qrf 
isen 
mo
m? r  Qrc /qrf
V?
p V?p 
m? r  V?p 
v suc
m? r
m? r
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.55
The following are the EERs indicating the energy use of a hermetic refrigeration compressor
and motor, and the equivalent COP and kW/ton values:
EER is widely used for reciprocating and scroll compressors in air-cooled, direct-expansion refrigeration
systems. The coefficient of performance and kW/ton are usuallly used for water-cooled
centrifugal chillers. They are often rated at different operating conditions. Water-cooled chillers
always show a more energy-efficient index than air-cooled DX systems.
9.21 SAFETY REQUIREMENTS AND MACHINERY ROOM
Refrigerant Safety
Refrigerant hazards stemming from leaks in the pipe joints, the rupture of system components, and the
burning of escaping refrigerant depend on the type of refrigerant, the occupancy classification, and the
refrigeration system. As discussed in Sec. 9.3, refrigerants can be classified into six safety groups that
range from lower toxicity and no flame propagation (safety group A1) to higher toxicity and higher
flammability (safety group B3). The type of occupancy may be one of the following five categories:
 Institutional or health care, such as hospitals and nursing homes
 Public assembly, such as auditoriums and department stores
 Residential, including hotels and apartments
 Commercial, such as offices, restaurants, and markets
 Industrial, such as factories and warehouses
Application Rules for High-Probability Systems
A tightly sealed refrigerant system is always necessary to reduce refrigerant leaks that may produce
refrigerant hazards. In addition, limiting the quantity of refrigerant in a refrigeration system per occupied
space, thereby reducing the possible leaks from joints and seals in a high-probability system,
is an effective means of reducing the hazards of refrigerants for the safety of people and property.
ASHRAE Standard 15-1994 specifies a number of rules and requirements for various refrigerants
in high-probability refrigeration systems. These rules are described in the following paragraphs.
1. Any refrigeration system in a room air conditioner or packaged terminal air conditioner
(PTAC), or any small packaged unit for which the refrigerant charge does not exceed 6.6 lb (3 kg),
is considered to meet the system refrigerant safety application requirements.
2. Refrigerants that belong to the A1 safety group (HCFC-22, HFC-134a, etc.) and are used in
high-probability refrigeration systems have the following restrictions:
EER COP kW/ton
6 1.758 2.0
8 2.344 1.5
10 2.930 1.2
12 3.516 1.0
15 4.395 0.8
20 5.860 0.6
30 8.790 0.4
9.56 CHAPTER NINE
For institutional or health care occupancies (except in kitchens, laboratories, or mortuaries), the
following conditions apply:
 The quantity of refrigerant HCFC-22 in the largest refrigeration system mHCFC-22 is limited to
4.7 lb /1000 ft3 (2.13 kg/28.3 m3) of occupied space.
 The quantity of refrigerant HFC-134a, or mHFC-134a, is limited to 8 lb /1000 ft3 (3.6 kg/28.3 m3)
of occupied space. A flame-sustaining device in an occupied space must be provided with a
hood to exhaust combustion products to open air if the refrigeration system contains more than
1 lb (0.45 kg) of A1 refrigerant (except R-744 carbon dioxide).
For public assembly, residential, and commercial occupancies, the following apply:
 The value of mHCFH-22 in a refrigeration system is limited to 9.4 lb /1000 ft3 (4.3 kg /28.3 m3)
of occupied space.
 The value of mHFC-134a in a refrigeration system is limited to 16 lb /1000 ft3 ( 7.3 kg /28.3 m3)
of occupied space.
In industrial applications, refer to ANSI/ASHRAE Standard 15-1994 for details.
3. Refrigerants in the B1 safety group, such as HCFC-123, that are used in high-probability refrigeration
systems should not be allowed for applications involving air conditioning for human
comfort. The quantity of refrigerants in a refrigeration system is limited to under 0.4 lb /1000 ft3
(0.18 kg/28.3 m3) occupied space for commercial occupancy and 0.2 lb /1000 ft3 (0.09 kg/28.3 m3)
occupied space for institutional occupancy.
4. Refrigerants in the B2 safety group, such as R-717 ammonia, that are used in high-probability
refrigeration systems are not allowed to be used for institutional occupancy as well as comfort air
conditioning in other occupancies. The following rules apply: When R-717 is used in high-probability
refrigeration systems in process air conditioning (including food storage), mR-717, lb (kg), the
maximum permissible quantity, is limited to the following values:
Institutional Public Residential Commercial
Sealed absorption system:
In public hallways or lobbies 0 0 3.3 (1.5) 3.3 (1.5)
In adjacent outdoor locations 0 0 22 (10) 22 (10)
Other than hallways or lobbies 0 6.6 (3) 6.6 (3) 22 (10)
Individual and packaged systems:
Other than hallways or lobbies 0 0 6.6 (3) 22 (10)
Application Rules for Low-Probability Systems
ASHRAE Standard 15-1994 states the requirements for low-probability refrigeration systems that
have complete, factory-assembled chillers tested by an approved and nationally recognized laboratory
and that are used in health care, public assembly, residential, commercial, and industrial buildings.
The following paragraphs describe these requirements.
1. For A1 safety group refrigerants, if mHCFC-22 in the largest refrigeration system exceeds 9.4
lb /1000 ft3 (4.3 kg/28.3 m3) of occupied space or mHFC-134a exceeds 16 lb/1000 ft3 (7.3 kg/28.3 m3)
of occupied space, all refrigerant-containing parts, except pipes and parts located outdoors, must be
installed in a machinery room that meets general safety requirements.
2. For B1 safety group refrigerants, if mHCFC-123 in the largest refrigeration system exceeds 0.4
lb /1000 ft3 (0.18 kg/28.3 m3) of occupied space, all refrigerant-containing parts, except pipes and
parts located outdoors, must be installed in a machinery room that meets general safety requirements.
3. For B2 safety group refrigerants such as R-717 ammonia, the requirements are as follows:
When mR-717 in the largest refrigeration system exceeds 0.022 lb/1000 ft3 (0.01 kg/28.3 m3) of
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.57
occupied space, all refrigerant-containing parts, except pipes and those parts outdoors, must be
installed in a machinery room that meets special requirements; and mR-717 in the refrigeration
system must not exceed 550 lb (250 kg) for institutional occupancy. For details, refer to ASHRAE
Standard 15-1994 and local codes.
Refrigerating Systems of 100 hp (74.6 kW) or Less
Separated fan rooms with light construction and with controlled access by authorized personnel
may contain refrigeration machinery. Refer to ASHRAE Standard 15-1994 and Section 16.6.
Refrigerating Machinery Room
A refrigerating machinery room is an enclosure with tightly fitted doors to safely house compressors,
refrigeration components, and other types of mechanical equipment. A refrigerating machinery
room must be designed so that it is easily accessible, with adequate space for proper servicing,
maintenance, and operation. A refrigerating machinery room must have doors that open outward
and are self-closing if they open into the building; there must be an adequate number of doors to
allow easy escape in case of emergency.
According to ASHRAE Standard 15-1994, a refrigerating machinery room must meet the following
requirements:
 It must be ventilated to the outdoors by means of mechanical ventilation using power-driven fans
or multiple-speed fans. The minimum volume flow rate of mechanical ventilation Vm,v, cfm (L/ s),
required to exhaust the refrigerant from leaks or ruptured component is
(9.74)
where mref is the mass of refrigerant in the largest system, in pounds. When a refrigeration system
is located outdoors, its location is more than 20 ft (6 m) from any building opening, and it is enclosed
by a penthouse or other open structure, then both mechanical and natural ventilation can
be used. The minimum free-aperture section Afa, ft2 (m2), required for natural ventilation of the
machinery room is
(9.75)
Provisions for venting catastrophic leaks and component ruptures should be considered.
 For safety group A1 refrigerants such as HCFC-22 and HFC-134a, the machinery room must
include an oxygen sensor to warn the operator when oxygen levels drop below 19.5 percent by
volume. For other refrigerants, a vapor detector should be installed to actuate the alarm and the
mechanical ventilation system.
 There must be no open flames that use combustion air from the machinery room except matches,
lighters, leak detectors, and similar devices.
A refrigerating machinery room of special requirements must meet the following specifications
in addition to the general requirements:
 Inside the room, there must be no flame-producing device or hot surface continuously operated at
a surface temperature exceeding 800°F (427°C).
 There must be an exit door that opens directly to the outdoors or to a similar facility.
 Mechanical ventilation for ammonia must be either continuously operating or equipped with a vapor
detector that actuates a mechanical ventilation system automatically at a detection level not
exceeding 4 percent by volume.
Afa  mref
0.5
V?
m,v  100 mref
0.5
9.58 CHAPTER NINE
 It must be provided with remote pilot control panel immediately outside the machinery room to
control and shut down the mechanical equipment in case of emergency.
Because the refrigerating machinery room itself is a fire compartment, the building structure and
its material (including the door, wall, ceiling, and floor) should meet National Fire Protection
Association (NFPA) fire codes. Refrigerating machinery rooms are different from fan rooms where
air-handling units with fans are mounted inside. In a refrigerating machinery room, there are general
and special requirements. The installation of mechanical ventilation and an oxygen or vapor
sensor is mandatory. A central refrigeration plant must be a kind of machinery room.
For more details, refer to ASHRAE Standard 15-1994.
Storage of Refrigerants
Refrigerants are usually stored in cylinders during transport and while on site. During storage, the
pressure of the liquid refrigerant must be periodically checked and adjusted. Excessive pressure
may cause an explosion. According to Interstate Commerce Commission (ICC) regulations, liquid
refrigerants must not be stored above 130°F (54.4°C), although the containers are designed to withstand
up to 3 times the saturated pressure at 130°F (54.4 °C). If a container bursts, liquid refrigerant
flashes into vapor. Such a sudden expansion in volume could cause a violent explosion inside a
building, blasting out windows, walls, and roofs.
Containers should never be located near heat sources without sufficient ventilation. They must
also not be put in a car or truck in direct sunlight. The valve of the container is attached by thread
only. If the threads are damaged, the force of escaping vapor could propel the container like a
rocket.
According to ASHRAE Standard 15-1994, in addition to the refrigerant charge in the system
and receiver, refrigerant stored in a machinery room shall not exceed 330 lb (150 kg). The receiver
is a vessel used to store refrigerant after the condenser when necessary.
REFERENCES
ARI, Switch from CFC Chillers Is Slower Than Projected, ASHRAE Journal, no.6, 1997, pp. 11–12.
ASHRAE, ASHRAE Handbook 1998, Refrigeration, ASHRAE Inc., Atlanta, GA, 1994.
ASHRAE, ASHRAE Standard 15-1994, Safety Code for Mechanical Refrigeration, Atlanta, GA, 1998.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
ASHRAE, ANSI/ASHRAE Standard 34-1997, Number Designation and Safety Classification of Refrigerants,
Atlanta, GA, 1997.
Bivens, D. B, Allgood, C. C., Rizzo, J. J., Shiflett, M. B., Patron, D. M., Shealy, G. S.,Yokozeki, A.,Wells,W.
D., and Geiger, K. A., HCFC-22 Alternative for Air Conditioners and Heat Pumps, ASHRAE Transactions,
1994, Part II, pp. 566–572.
Corr, S., Dekleva, T. W., and Savage, A. L., Retrofitting Large Refrigeration Systems with R-134a, ASHRAE
Journal, no. 2, 1993, pp. 29–33.
Cox, J. E., and Miro, C. R., Industry Helping U.S. Comply with Laws Governing CFCs, ASHRAE Journal,
no. 12, 1996, pp. 26–27.
Denny, R. J., The CFC Foot Print, ASHRAE Journal, November 1987, pp. 24–28.
Dooley, E. W., Industry Status of CFC Chiller Retrofits and Replacements, HPAC, no. 1, 1997, pp. 131–135.
Dossat, R. J., Principles of Refrigeration,Wiley, 2d ed., New York, 1978.
Gilkey, H. T., The Coming Refrigerant Shortage, Heating/Piping/Air Conditioning, April 1991, pp. 41–46.
Hickman, K. E., Redesigning Equipment for R-22 and R-502 Alternatives, ASHRAE Journal, no. 1, 1994,
pp. 43–47.
REFRIGERANTS, REFRIGERATION CYCLES, AND REFRIGERATION SYSTEMS 9.59
Hwang,Y., Judge, J., and Radermacher, R., Experience with Refrigerant Mixtures, ASHRAE Transactions, 1997,
Part I, pp. 765–776.
King, G. R., Modern Refrigeration Practice, 1st ed., McGraw-Hill, New York, 1971.
Lavelle, J., HFCs Will Find Greater Use as CFCs Diminish, Air Conditioning, Heating & Refrigeration News,
Feb. 17, 1997, p. 25.
Linton, J. W., Snelson,W. K., Triebe, A. R., and Hearty, P. F., System Performance Comparison of R-507 with
R-502, ASHRAE Transactions, 1995, Part II, pp. 502–510.
Lowe, R., and Ares, R., From CFC-12 to HFC-134a: An Analysis of a Refrigerant Retrofit Project,
Heating/Piping/Air Conditioning, no. 1, 1995, pp. 81–90.
McGuire, J., An Industry at Risk, User Concerns, ASHRAE Journal, August. 1988, pp. 48–49.
Parker, R.W., Reclaiming Refrigerant in OEM Plants, ASHRAE Transactions, 1988, Part II, pp. 2139–2144.
Parsnow, J. R., The Long Term Alternative: R-134a in Positive Pressure Chillers, ASHRAE Journal, no. 5, 1993,
pp. 54–56.
Powell, P., Recips on Decline, Gas-Electric Power Ploys, Secondary-Loop Hoopla: Top Mechanical Focus, Air
Conditioning, Heating & Refrigeration News, June 16, 1997, pp. 22–26.
Rowland, S., The CFC Controversy: Issues and Answers, ASHRAE Journal, December 1992, pp. 20–27.
Smith, N. D., Ratanaphruks, K., Tufts, M. W., and Ng, A. S., R-245ca: A Potential Far-Term Alternative for
R-11, ASHRAE Journal, no. 2, 1993, pp. 19–23.
Smithart, E. L., and Crawford, J. G., R-123: A Balanced Selection for Low Pressure Systems, ASHRAE Journal,
no. 5, 1993, pp. 66–69.
Snelson,W. K., Linton, J. W., Triebe, A. R., and Hearty, P. F., System Drop-In Tests of Refrigerant Blend
R-125/R-143a/R-134a (44%/52%/4%) Compared to R-502, ASHRAE Transactions, 1995, Part I, pp. 17–24.
Stamm, R.H., The CFC Problem: Bigger Than You Think, Heating, Piping and Air Conditioning, April 1989,
pp. 51–54.
Swers, R., Patel,Y. P., and Stewart, R. B., Thermodynamics Analysis of Compression Refrigeration Systems,
ASHRAE, New Orleans, LA, January 1972.
Tan, L. C., and Yin, J. M., The Exergy-Enthalpy Diagram and Table of R-22 and Their Application, ASHRAE
Transactions, 1986, Part I A, pp. 220–230.
Tuck, A., The Chemistry of Stratospheric Ozone Depletion, HPAC, no. 1, 1997, pp. 111–116.
Wang, S. K., Principles of Refrigeration, vol. 1, 1st ed., Hong Kong Polytechnic, 1984.
Wilson, D. P., and Basu, R. S., Thermodynamics Properties of a New Stratospherically Safe Working Fluid—
Refrigerant 134A, ASHRAE Transactions, 1988, Part II, pp. 2095–2118.
9.60 CHAPTER NINE
CHAPTER 10
REFRIGERATION SYSTEMS:
COMPONENTS
10.1
10.1 EVAPORATORS 10.2
10.2 DX COILS—WET COILS 10.3
Construction and Installation 10.3
Two-Region Model 10.4
Simultaneous Heat and Mass
Transfer 10.5
DX Coil Effectiveness 10.6
Face Velocity and Air-Side Pressure Drop
of Wet Coil 10.7
DX Coils at Part-Load Operation 10.8
Selection of DX Coils 10.10
10.3 FLOODED LIQUID COOLER 10.12
Construction 10.12
Heat Transfer in a Flooded Liquid
Cooler 10.14
Fouling Factor 10.14
Pool Boiling and Forced-Convection
Model 10.15
Parameters That Influence Performance
of Flooded Liquid Cooler 10.16
Part-Load Operation of Flooded Liquid
Cooler 10.17
10.4 DIRECT-EXPANSION LIQUID
COOLER AND LIQUID
OVERFEED COOLER 10.18
Direct-Expansion Liquid Cooler 10.18
Liquid Overfeed Cooler 10.18
10.5 CONDENSATION AND
CONDENSERS 10.20
Condensation Process 10.20
Total Heat Rejection 10.21
Types of Condenser 10.22
10.6 WATER-COOLED CONDENSERS 10.22
Types of Water-Cooled Condenser 10.22
Shell-and-Tube Condensers 10.22
Heat Transfer in Shell-and-Tube
Condensers 10.24
Parameters That Influence the
Performance of Shell-and-Tube
Condensers 10.25
Part-Load Operation 10.26
10.7 AIR-COOLED CONDENSERS 10.26
Construction of Air-Cooled
Condenser 10.26
Heat-Transfer Processes
and Temperature Curves 10.26
Cooling Air Temperature Rise
and Volume Flow 10.28
Condenser Temperature Difference 10.28
Condensing Temperature, Subcooling,
and Oil Effect 10.29
Clearance, Warm Air Recirculation,
and Dirt Clogging 10.29
Low Ambient Air Control 10.29
Selection of Air-Cooled Condenser 10.30
10.8 EVAPORATIVE CONDENSERS 10.30
Condensation Process 10.30
Heat Transfer in Evaporative
Condensation 10.30
Water Spraying 10.32
Cooling Air 10.32
Site Location and Low Ambient
Air Control 10.32
Selection and Installation 10.33
10.9 COOLING TOWERS 10.34
Counterflow Induced-Draft
Cooling Towers 10.34
Crossflow Induced-Draft Cooling
Towers 10.34
Counterflow Forced-Draft Cooling
Towers 10.35
10.10 THERMAL ANALYSIS OF COOLING
TOWERS 10.36
Technological Terms 10.36
Tower Coefficient 10.36
Heat- and Mass-Transfer Process for
Counterflow Cooling Tower 10.37
Tower Capacity, Size, and
Coefficient 10.37
10.11 COOLING TOWER
PERFORMANCE 10.40
Factors That Affect Cooling Tower
Performance 10.40
Range and Water Circulating Rate 10.40
Tower Coefficient and Water-Air
Ratio 10.41
Approach 10.41
Outdoor Wet-Bulb Temperature 10.41
Fill Configuration and Water
Distribution 10.42
Construction Materials 10.43
10.12 PART-LOAD OPERATION AND
OPTIMUM CONTROL 10.43
Part-Load Operation 10.43
Optimum Control 10.43
10.13 OPERATING CONSIDERATIONS FOR
COOLING TOWERS 10.46
Recirculation and Interference 10.46
Freeze Protection and Fogging 10.46
Blowdown 10.47
Legionnaires’ Disease 10.47
Maintenance of Cooling Towers 10.47
10.1 EVAPORATORS
The evaporator is one of the main components of a refrigeration system, in which refrigerant evaporates
for the purpose of extracting heat from the surrounding air, chilled water, or other substances.
In vapor compression refrigeration systems, the evaporator is also an indirect-contact heat exchanger.
Evaporators can be classified into three categories, depending on the medium or substance to be
cooled:
 An air cooler is an evaporator that cools the air directly in a refrigerated space or piece of equipment
(such as a packaged unit). Conditioned air is then distributed through air distribution
systems. In an air cooler, the refrigerant flows inside the metal tubes or finned tubes while air
flows over them.
 In a liquid cooler, chilled water is cooled to a lower temperature and is pumped to remote airhandling
units, fan coils, or other terminals for air conditioning or other applications.
 An evaporator can be used to produce ice directly, such as an ice maker in an ice harvester of an
ice storage system. These are discussed in Chap. 31.
A liquid cooler is different from a chiller. A liquid cooler is an evaporator, a component of a refrigeration
system, whereas a chiller is a refrigeration package to produce chilled water.
The refrigerant feed for air and liquid coolers and ice makers can be classified mainly into the
following three types:
 Dry-expansion or direct-expansion (DX). In evaporators with dry- or direct-expansion refrigerant
feed, liquid refrigerant is fed through an expansion valve and a distributer, flows inside the tubes
in a finned coil, the evaporator, and is completely vaporized and superheated to a certain degree
before reaching the exit of the evaporator.
 Flooded refrigerant feed. In evaporators with flooded refrigerant feeds, liquid refrigerant is fed
through a throttling device and vaporizes outside the tubes within a shell. The refrigerant-side surface
area is always wetted by the liquid refrigerant, which results in a higher surface heat-transfer
coefficient.
 Liquid overfeed. In liquid overfeed evaporators, liquid refrigerant is fed by a mechanical or gas
pump and is then overfed to each evaporator. The inner surface in an overfeed evaporator is also
wetted by liquid refrigerant.
The various combinations of cooler types and refrigerant feeds make up the following evaporator
groups:
 DX cooler, which is a DX coil, acts as an air cooler with direct-expansion refrigerant feed
 Flooded shell-and-tube liquid cooler, or simply flooded liquid cooler
10.2 CHAPTER TEN
Thermostatic Expansion Valves 10.51
Liquid Charge in the Sensing Bulb 10.53
Electric Expansion Valves 10.55
Float Valves 10.56
Capillary Tube 10.57
REFERENCES 10.58
10.14 HEAT REJECTION SYSTEMS 10.48
Types of Heat Rejection System 10.48
Comparison between Various
Systems 10.48
ASHRAE/IESNA Standard
90.1-1999 10.50
10.15 REFRIGERANT FLOW CONTROL
DEVICES 10.51
 Shell-and-tube liquid cooler with direct-expansion refrigerant feed, or simply direct-expansion
liquid cooler
 Liquid overfeed cooler
 Direct-expansion ice maker
Among these, DX coils are most widely used. Flooded liquid coolers are widely used in large
chillers, and direct-expansion liquid coolers are often used in reciprocating or screw chillers. Liquid
overfeed coolers are sometimes adopted for processing air conditioning systems. Ice makers are
usually used in thermal storage systems and for industrial applications.
10.2 DX COILS—WET COILS
In a direct-expansion or dry-expansion DX coil for a comfort air conditioning system, the evaporating
temperature Tev, °F (°C), of refrigerant HCFC-22, HFC-134a, HFC-404A, HFC-410A, HFC-
407A, or HFC-407C inside the coil tubes is usually 37 to 52°F (2.8 to 11.1°C). In such a Tev, the
surface temperature of the coil is usually lower than the dew point of the air entering the coil. Condensation
occurs on the outside surface of the DX coil, so the coil becomes a wet coil. The air
conditioning process of a DX coil is always a cooling and dehumidifying process as well as a heatand
mass-transfer process.
Construction and Installation
Copper tubes and aluminum fins are widely used for DX coils using halocarbon refrigerants. The
diameter of the copper tubes is usually to in. (10 to 16 mm). Aluminum fins of a fin spacing
often from 12 to 18 fins / in. (1.4 to 2 fins / mm) are typically 0.006 in. (0.15 mm) thick. Corrugated
plate fins are often used for additional turbulence and increase in heat transfer. On the inner surface
of the copper tubes, microfins typically with a spacing of 60 fins / in. (0.42 fins / mm) and a height of
0.008 in. (0.2 mm) are widely used to enhance the boiling heat transfer. The Eckels and Pate (1991)
experiments showed that compared with smooth tubes, copper tubes with inner microfins had a
heat-transfer enhancement factor from 1.5 to 2.5 during in-tube evaporation or condensation for refrigerant
HFC-134a.
During the analysis and calculation of the performance of a water-cooling, water-heating, or DX
coil, Ao/Ai, fin spacing, and tube spacing are the coil construction parameters that determine coil
performance. Here Ao indicates the total outside surface area of the coil, and Ai the inner surface
area of the coil, ft2 (m2). The following coil construction parameters have been used by many manufacturers
to optimize the rate of heat transfer, air-side pressure drop, and manufacturing cost:
Longitudinal tube spacing SL 1.083 in. (27.5 mm)
Transverse tube spacing ST 1.25 in. (31.8 mm)
Outside nominal diameter of copper tube Do 0.5 in. (23 mm)
Aluminum fin thickness Ft 0.006 in. (0.15 mm)
Table 10.1 lists the coil construction parameters for smooth fin, based on the data given above. Parameters
are for a face area of 1 ft2 (0.305 m2) and a coil depth of one row, or 1.083 in. (27.5 mm).
If the edge effect, in both longitudinal and transverse directions, is included, then the values of
Ao/Ap, Ao/Ai, and Fs may be slightly higher. Here Fs indicates coil core surface area parameter, and
it is discussed in detail in Sec. 15.9. For corrugated fins, Ao/Ap, Ao/Ai, and Fs should be multiplied
by a corrugated factor Fc,a of 1.1 to 1.2.
The ratio of free flow (or minimum flow) area to face area Amin /Aa is usually 0.54 to 0.6. For a
smaller fin spacing, such as 12 to 15 fins / in. (1.7 to 2 fins /mm), Amin /Aa  0.58 to 0.60.
58
38
REFRIGERATION SYSTEMS: COMPONENTS 10.3
For even coolant distribution and to reduce pressure drop, flow paths of refrigerant in a DX coil
are always divided into a number of refrigerant circuits, according to the finned width of the coil.
Refrigerant is usually supplied through a thermostatic expansion valve. It then flows through a distributor,
which distributes the refrigerant evenly, as shown in Fig. 10.1. Venturi-type distributors are
usually used. The distributor is connected to various circuits through copper distributor tubes of
diameter typically between and in. (6.5 and 16 mm). Equal length of distributor tube and
approximately equal circuit lengths ensure even distribution of refrigerant.
To create a counterflow arrangement between air and refrigerant, the suction header should be
installed nearest to the air entering the side of the coil.
For a DX coil, copper finned tubes should be reasonably level to ensure proper air venting and
condensate drainage. When two or more DX coils are banked vertically, there must be an individual
drain pipe and drain pan for each coil.
Two-Region Model
Refrigerant enters the DX coil at a liquid-vapor two-phase state. It is cooled to evaporating
temperature Tev and gradually evaporated along the flow paths until it is completely vaporized.
Because of the pressure drop of the refrigerant as it flows through the coil, the evaporating temperature
Tev gradually drops along the refrigerant circuit as the evaporating pressure decreases, as
shown in Fig. 10.1.
For simplification, the flow path of the refrigerant and its corresponding outer surface in a DX
coil is divided into two regions.
Two-Phase Region. In the two-phase region, the quality of the refrigerant xr  1. The pressure
drop of the liquid refrigerant is negligible, and liquid refrigerant is evaporated at a uniform temperature.
In the two-phase region, the boiling heat-transfer coefficient hboil of refrigerant HCFC-22,
Btu/h ft2 °F (W/m2 °C), inside a DX coil is mainly a function of the mass flow rate and quality
of the refrigerant. Figure 10.2 shows the boiling heat-transfer coefficient of HCFC-22 in a tube with
an outside diameter of 0.665 in. (17 mm) at 40°F (4.4°C) when the quality of refrigerant is varied.
Schlager et al. (1989) found that naphthenic mineral oil with a viscosity of 150 SSU enhances
the surface heat-transfer coefficient of refrigerant HCFC-22 inside a microfin tube up to 25 percent
516
14
10.4 CHAPTER TEN
TABLE 10.1 Finned-Tube Coil Construction Parameters
Outside diameter of copper tube Do 0.528 in.
Inner diameter of copper tube Di 0.496 in.
Aluminum fin thickness Ft 0.006 in.
Longitudinal tube spacing 1.083 in.
Transverse tube spacing 1.25 in.
Fin spacing Sf
Fins/ in. in. Ao /Ap
† Ao /Ai Af /Ao
‡ Fs Sf /Ft
8 0.125 7.85 7.95 0.873 9.91 20.8
10 0.100 9.68 9.68 0.896 12.07 16.7
12 0.0833 11.54 11.40 0.913 14.21 13.9
14 0.0714 13.46 13.17 0.925 16.37 11.9
15 0.0667 14.47 14.03 0.928 17.48 11.1
†Ap is area of the primary surfaces (outside surface of the copper tubes).
‡Af is area of fins.
Note: For corrugated plate fins, Ao /Ap, Ao /Ai, and Fs should be multiplied by a factor of 1.10 to 1.20.
at an oil concentration of 3 percent. For 300-SSU oil, the surface heat-transfer coefficient inside the
microfin tube increases slightly and then decreases about 20 percent when the oil concentration
increases from 0 to 3 percent.
Superheated Region. In the superheated region, the quality of refrigerant xr  1. The refrigerantside
heat-transfer coefficient is low. Because of the relatively high outer surface temperature caused
by the superheated refrigerant, the outer surface of the DX coil corresponding to the superheated
region may be dry. The superheated region is small. For simplification, the rate of heat transfer
(including both sensible and latent heat) of the whole DX coil is often multiplied by a degrading
factor Fsuper to account for the existence of the superheated region.
Simultaneous Heat and Mass Transfer
In a DX coil, the driving potentials for the simultaneous heat and mass transfer during cooling and
dehumidifying are the enthalpy difference between the ambient air and the saturated air film at the
REFRIGERATION SYSTEMS: COMPONENTS 10.5
FIGURE 10.1 Temperature and pressure drop across a typical DX coil using HCFC-22 as refrigerant.
(All pressures are measured at gauge pressure.)
interface of condensate and air (in Btu / lb or J/kg) and the temperature difference between the air
and the evaporating refrigerant inside the tubes.
This heat and mass transfer Qc,wet, Btu /h (W), is more conveniently calculated as
(10.1)
where hae  enthalpy of entering air, Btu / lb (J / kg)
hs,r  enthalpy of saturated air film at coil surface corresponding to evaporating temperature
of refrigerant inside tubes, Btu/ lb (J / kg)
 air volume flow rate, cfm [m3 /(60 s)]
a  air density, lb/ ft3 (kg /m3)
DX Coil Effectiveness
The effectiveness of a heat exchanger  is defined as the ratio of the actual rate of heat transfer
between the hot and cold fluids to the maximum possible rate of heat transfer. Because the heat
energy of the condensate is small compared to that of air and water streams, it can be ignored.
DX coil effectiveness wet can be calculated as
wet 
hae 	 hal
hae 	 hs, r
V?
a
Qc, wet  wet 60V?a a (hae 	 hs, r)
10.6 CHAPTER TEN
FIGURE 10.2 Boiling heat-transfer coefficient of HCFC-22. (Source:
ASHRAE Handbook 1981, Fundamentals. Reprinted with permission.)
or
hal  hae 	 (hae 	 hs,r) (10.2)
where hal  enthalpy of air leaving the coil, Btu/ lb (kJ / kg). After hal is determined, the dry-bulb
and wet-bulb temperatures of the leaving air can be found from the cooling and dehumidifying
curve of the psychrometric chart with the same entering air condition.
For convenience, the refrigerant in a DX coil is assumed to evaporate at a constant temperature,
so Cmin /Cmax  0. Here Cmin and Cmax indicate the smaller and greater value of heat capacities between
air and water, Btu /h°F (W/ °C). The wet coil effectiveness is calculated as
wet  1 	 exp(-NTU) (10.3)
The number of transfer units NTU for a DX coil can also be calculated as
(10.4)
where Uo  overall heat-transfer coefficient based on outside surface of coil, Btu/h ft2 °F
(W/m2 °C)
Ao  total outside surface area of coil, ft2 (m2)
Ca  heat capacity rate of moist air, Btu /h°F (W/ °C)
and
(10.5)
where cpa  specific heat of moist air, Btu /h°F (J /kg°C)
a  density of moist air, lb/ ft3 (kg /m3 )
Also
(10.6)
where ho, hi  heat-transfer coefficient of outer surface of coil (including fins and tubes) and inner
surface of tubes, Btu/h ft2 °F (W/m2 °C)
Ai  inner surface area of coil, ft2 (m2)
In Eq. (10.6), s represents the finned surface efficiency and is given as
(10.7)
where Af  area of fins, ft2 (m2). The fin efficiency f can be calculated as
(10.8)
Face Velocity and Air-Side Pressure Drop of Wet Coil
Selection of face velocity for wet coils must consider the carryover of the condensate water
droplets, the influence of coil face velocity on heat-transfer coefficients, the pressure drop of the
coil, and the size of the coil. If air velocity distribution is even over the entire surface of the DX
coil, the carryover is usually not significant at face velocities lower than 500 fpm (2.5 m / s) for
f 
actual heat transfer
heat transfer when fins are at base temperature
s 
1 	 Af
Ao (1 	
f)
Uo Ao 
1
1/( s ho Ao) 	 1/(hi Ai)
Ca  60 V?a acpa
NTU 
Uo Ao
Ca
REFRIGERATION SYSTEMS: COMPONENTS 10.7
smooth plate fins or lower than 550 fpm (2.75 m / s) for corrugated fins [to prevent wet surfaces
formed in an air distribution system that nourishes the mold growth in fiberglass duct liners, a DX
coil face velocity of 500 fpm (2.5 m / s) or less is preferable]. When the supply fan is located upstream
from the DX coil (a blow-through fan and coil arrangement), there may be areas where the
face velocity is far higher than in the rest of the coil.
For wet DX coils, a rough estimate of air-side pressure drop for 15 fins / in. (1.7 fins / mm) corrugated
fins is as follows:
The air-side pressure drop for a three-row 15-fins / in.(1.7 fins / mm) corrugated fin coil from one
manufacturer may be 50 percent higher than that from another manufacturer, even at the same operating
conditions and same unit cooling capacity. It is important to carefully compare the coil performances
of different manufacturers.
The lower limit of coil face velocity at design load depends on initial and operating costs. For
DX coils in VAV systems, reduction of face velocity in part-load operation should also be considered.
Face velocities between 350 and 450 fpm (1.75 and 2.25 m / s) may be justified in many cases.
DX Coils at Part-Load Operation
The refrigerant circuits in a DX coil of capacity greater than 15 tons are usually divided into two or
more sections, each with its own expansion valve, distributor, and suction header. The refrigerant
circuits are controlled in three ways, as shown in Fig. 10.3.
Face Control. The refrigerant circuits are divided into upper and lower sections, as shown in
Fig. 10.3a. One section can be shut off by deenergizing its solenoid valve during capacity reduction.
Face control reduces the temperature of the conditioned air, and requires thorough mixing of
conditioned and bypass airstreams downstream from the DX coil.
Row Control. Row control is more effective when applied to a six-row DX coil, as shown in Fig.
10.3b. One distributor connects to the refrigerant circuits of two rows; another distributor connects
to the remaining four rows. The active refrigerant circuits can then be operated at 33, 66, or 100
percent capacity.
Intertwined Face Control. An intertwined face control (shown in Fig. 10.3c) always has a full
face active coil even when the refrigerant supply to one of the distributors is cut off by the solenoid
valve. When one distributor is inactive, the fin efficiency f and the heat capacitance ratio of wet
coil Cwet are changed, and the capacity of the DX coil is reduced to about 70 percent.
In the selection of a DX refrigeration system, it is important to check the minimum refrigerant
temperature in the DX coil during the part-load operation. The minimum saturated temperature of
the refrigerant Tr, min, °F (°C), for a DX coil at a face velocity about 500 fpm (2.5 m / s) must not
drop below 32°F (0°C). Otherwise, frost will form on the coil surface. Frost formation blocks the
air passage and severely reduces the rate of heat transfer. At part-load operation, the evaporating
temperature is lowered because less refrigerant is evaporated.
For DX coils using HCFC-22 or other oil-miscible refrigerants, the refrigrant pipe should be
designed so that the refrigerant-oil mixture can be returned in the pipe riser at minimum load operation.
This is covered in detail in Chap. 11.
Fin configuration
Face velocity, fpm (m/ s) Simple Corrugated
500 fpm (2.5 m / s) 0.15 in. WC (37 Pa) / row 0.23 in. WC (57 Pa) / row
600 fpm (3 m / s) 0.25 in. WC (62 Pa) / row 0.35 in. WC (87 Pa) / row
10.8 CHAPTER TEN
REFRIGERATION SYSTEMS: COMPONENTS 10.9
FIGURE 10.3 Control of DX coils at part-load operation. (a) Face control;
(b) row control; (c) intertwined face control.
When a DX coil is installed in a variable-air-volume system, the air volume flow rate and face
velocity are decreased during part-load operation. Figure 10.4 shows the relationship between the
cooling capacity of the DX coil at a specific face velocity and the cooling capacity at a face velocity
of 600 fpm (3 m / s), denoted by (Qc /Aa) / (Qc /Aa)600. Here, Qc indicates the coil capacity and Aa the
face area of the coil. The curve is plotted according to data from the catalogs of three manufacturers
at an air entering condition of 80°F (26.7°C) dry-bulb and 67°F (19.4°C) wet-bulb, and 45°F
(7.2°C) suction temperature, or 45°F (7.2°C) entering chilled water temperature and 55°F (12.8°C)
leaving. When the face velocity drops from 600 to 400 fpm (3 to 2 m / s), (Qc /Aa) / (Qc /Aa)600 drops
from 1.0 to about 0.8.
For a VAV packaged system using a DX coil, it is important to have effective multistep capacity
unloading and refrigeration circuit cutoff controls during part-load operation to maintain Tev above
32°F (0°C). This is discussed in Chap. 11.
Selection of DX Coils
The cooling capacity (refrigeration capacity) of a DX coil is actually the cooling capacity of the
corresponding DX refrigeration system as well as the cooling capacity of the packaged air conditioning
system. For comfort air conditioning, today a fixed pattern of DX coil, including number of
rows, fins per inch, and tubing spacing, is adopted for each model and size of packaged unit. The
trend is to use a denser fin spacing, such as 14 to 17 fins / in. (1.5 to 1.8 fins /mm). For the cooling
and dehumidifying of the mixture of outdoor and recirculating air in air conditioning, a two-, three-,
or four-row DX coil is often used. For a 100 percent outdoor air makeup unit, a four- or six-row DX
coil is often used. Check also the coil face velocity so that it does not exceed 550 fpm (2.75 m / s).
The selection of the equipment for the DX refrigeration system including the selection of the
DX coil is discussed in later chapters.
10.10 CHAPTER TEN
FIGURE 10.4 Variation of cooling capacity versus face velocity.
Example 10.1. A DX coil in a packaged conditioning system has a supply volume flow rate of
5500 cfm (2595 L/ s). Air enters the coil at a dry-bulb temperature of 80°F (26.7°C) and a wet-bulb
temperature of 67°F (19.4°C). Calculate the total cooling capacity of this DX coil at full load and
the conditions of the air leaving the coil, using the following data:
 A three-row, 15-fins / in. (1.7-fins / mm) coil with the same coil construction parameters shown in
Table 10.1
 Boiling heat-transfer coefficient for HCFC-22 inside the copper tubes of 700 Btu /h ft2 °F (3975
W/m2 °C) at Tev  45°F (7.2°C)
 Surface effectiveness for wet fins s  0.76
 Heat-transfer coefficient of the outer surface of the coil (corrugated fin) ho  12.90 Btu /h ft2 °F
(73.25 W /m2 °C)
Solution. For a DX coil using corrugated plate fins, select a face velocity va  550 fpm (2.75
m/ s) to provide an efficient heat transfer and a reasonable air-side pressure drop, and avoid condensate
carryover. Face area Aa for the DX coil can then be calculated as
For a 15-fins / in. (1.7-fins/mm) DX coil, from Table 10.1, the coil core surface area parameter Fs 
17.48. At full-load operation, for a face area of 10 ft2, and a multiplying factor of 1.15 for the corrugated
surface, the outer surface area of this DX coil is
Ao  FsNrAa  17.48 1.15 3 10  603 ft2 (56 m2)
For corrugated fins, Ao /Ai  14.03 1.15. The inner area is therefore
As given, for a wet surface s  0.76. Then, from Eq. (10.6),
The heat capacity rate for conditioned air is calculated as
The number of transfer units is
From Eq. (10.3), the effectiveness for the DX coil can be calculated as
wet  1 	 exp(-NTU)  1 	 e	0.831  1 	 0.436  0.564
Considering the influence of the superheated region, effectiveness of the DX coil should be multiplied
by a degrading factor Fsuper  0.95. Then
DX  0.0564(0.95)  0.536
NTU 
Uo Ao
Ca

4995
6014
 0.831
Ca  60V?a acpa  60 5500 0.075 0.243  6014 Btu / h  
F

1
1/(0.76 12.90 603) 	 1/(700 37.4)
 4995 Btu / h
F
Uo Ao 
1
1/( s ho Ao) 	 1/(hr Ai)
Ai 
603
14.03 1.15
 37.4 ft2 (3.5 m2)
Aa 
V?
s
va

5500
550
 10 ft2
REFRIGERATION SYSTEMS: COMPONENTS 10.11
From the psychrometric chart, hae  31.6 Btu / lb and the enthalpy of saturated air film at evaporating
temperature Tev  45°F is 17.65 Btu / lb. Then the total cooling capacity of the DX coil is
 60 5500 0.075 0.536(31.6 	 17.65)  185,060 Btu /h (54,223 W)
The cooling capacity is
MBtu/h ft2(58.36 kW/m2)
The enthalpy of the conditioned air leaving the DX coil hal is
hal  hae 	 h  hae 	(hae 	 hs,r)  31.6 	 0.536(31.6 	 17.65)  24.1 Btu / lb (56 kJ / kg)
Draw a cooling and dehumidifying curve from the air entering condition, 80°F (26.7°C) dry-bulb
and 67°F (19.4°C) wet-bulb temperature on the psychrometric chart. Determine the air leaving condition
on the curve since hal  24.1 Btu / lb (56 kJ / kg). From psychrometric chart, air leaves at
57.6°F (14.2°C) dry-bulb and 56.2°F (13.4°C) wet-bulb temperature with a relative humidity of 93
percent.
10.3 FLOODED LIQUID COOLER
Construction
Most medium-size and large liquid coolers are shell-and-tube flooded liquid coolers. In a flooded
liquid cooler, several straight tubes are aligned in a parallel staggered arrangement, usually held in
place at both ends by tube sheets, as shown in Fig. 10.5a. Chilled water circulates inside the tubes,
which are submerged in a refrigerant-filled shell.
Liquid-vapor refrigerant, usually at a quality x around 0.15 in air conditioning applications, is
fed into the bottom of the shell. It is evenly distributed over the entire length of the tubes. As the refrigerant
boils and bubbles rise, the upper part becomes increasingly bubbly. Vapor refrigerant is
discharged from the opening at the top of the cooler. A dropout area, or eliminator, is sometimes installed
to separate the liquid refrigerant from the vapor. The amount of refrigerant fed to the flooded
liquid cooler is controlled by a low-pressure-side float valve, or a multiple-orifice throttling device,
which is discussed in Chaps. 11 and 13.
When halocarbons are used as the refrigerants, copper tubes are always used because they provide
higher thermal conductivity and do not react with halocarbons. Tube diameters of and 1 in.
(19 and 25 mm) are often used, such as either internally enhanced copper tubing of 1-in.- (25-mm-)
diameter or smooth bore of -in. (19-mm) diameter, and the number of tubes inside the shell varies
from 50 to several thousand. Integral fins are extruded on the outer surface of the tubes to increase
the outer surface area. Other tube geometry on the outside surface is also used to enhance the boiling
heat-transfer coefficient, especially to coupling of a high-voltage and low current electric field
with the fluid field called electrohydrodynamics, and the boiling and condensing heat-transfer coef-
ficient may be increased in excess of tenfold. Fin spacing ranges from 19 to 40 fins / in. (0.6 to 1.3
fins /mm), typically 26 fins / in. (1 fin/ mm) at a height from 0.06 in. (1.5 mm). The outer-to-inner
surface ratio of integral finned tubes ranges from 2.5 to 3.5. Spiral grooves or other enhancements
may be added to the inner surface of the tubes to increase the inner surface heat-transfer coefficient.
By increasing the turbulence, dirt and suspended solids are prevented from settling on the inner surface
during operation.
Possible water flow arrangements in flooded shell-and-tube liquid coolers are shown in
Fig. 10.5b. A one-pass arrangement has the highest chilled water flow rate, two-pass has a lower
34
34
Qc
Aa

185,060
10 1000
 18.5
Qc  60V?s a (hae 	 hs,r)
10.12 CHAPTER TEN
rate, and three-pass has the lowest chilled water flow rate. The two-pass water flow arrangement,
with the water inlet and outlet on the same side, is the standard arrangement. Performance analyses
showed that there is no significant difference between a two-pass arrangement with the water inlet
and outlet located side by side and one with the inlet at the bottom and outlet at the top. Water
velocity inside the copper tubes is usually between 4 and 12 ft/ s (1.2 and 3.6 m / s); water velocity
exceeding 12 ft / s (3.6 m / s) may cause erosion. Water-side pressure drop is usually maintained at
or below 10 psi (70 kPa) to optimize pump energy consumption. Because of the greater flow passage
for refrigerant within the shell, the pressure loss on the refrigerant side is far lower than the
pressure loss in DX coils.
REFRIGERATION SYSTEMS: COMPONENTS 10.13
FIGURE 10.5 Flooded shell-and-tube liquid cooler: (a) schematic diagram; (b) passages of water flow.
Flooded liquid coolers provide a large surface area and higher heat-transfer coefficient, need
minimal space and headroom, and are easily maintained. They are most widely used in large refrigeration
plants.
Heat Transfer in a Flooded Liquid Cooler
In flooded liquid coolers, the refrigerant is assumed to evaporate at a uniform temperature. The rate
of heat transfer between the chilled water and the refrigerant Q, Btu /h (W); Uo and hi,
Btu/h ft2 °F (W/m2 °C); ; and NTU can be calculated from Eqs. (7.1), (9.25), (10.1) through
(10.4), and (10.6) and summarized as follows:
(10.9)
where hre, hrl  enthalpy of refrigerant entering and leaving liquid cooler, Btu / lb (kJ / kg or J / kg)
Tee, Tel  chilled water entering and leaving liquid cooler, °F (°C)
Tm  log-mean temperature difference between chilled water and refrigerant, °F (°C)
 mass flow rate of refrigerant, lb / min [kg / (60 s)]
ho  average heat-transfer coefficient of refrigerant-side surface, Btu /h ft2 °F (W/m2 °C)
hi  inner surface heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
Di  inner diameter of tube, ft (m)
Lt  thickness of tube, ft (m)
kw  thermal conductivity of water, Btu /h ft °F (W/m°C)
kt  thermal conductivity of metal tube, Btu /h ft °F (W/m°C)
Tev  evaporating temperature, °F (°C)
For the analysis of heat transfer in a flooded liquid cooler, both the log-mean temperature difference
and the NTU method can be used. However, NTU method is preferred because it clearly
describes the effectiveness and characteristics of the liquid cooler.
Fouling Factor
Fouling factor Rf, h ft2 °F/ Btu (m2 °C/W), is the additional resistance caused by the dirty film of
scale, rust, silt, or other deposits on the surface of the tube. The fouling factor may be defined as
(10.10) Rf 
1
Udirty
	
1
Uclean
m?r
Q  60m? wcpw (Tee 	 Tev)
NTU 
Uo Ao
cpw
  1 	 exp(-NTU)
hi Di
kw
 0.023 ReD
0.8 Pr n
Uo 
1
1/ho 	 Lt Ao /(Ai kt) 	 Ao Rf /Ai 	 Ao /(Ai hi)
 m?r (hrl 	 hre)
 Ao Uo Tm
 500V?gal(Tee 	 Tel)
Q  60m? wcpw (Tee 	 Tel)
10.14 CHAPTER TEN
where Udirty, Uclean  overall heat-transfer coefficient through tube wall with dirty or clean surface
respectively, Btu /h ft2 °F (W/m2 °C).
In 1982, ASHRAE Journal published research results on the fouling of heat-transfer surfaces
such as evaporators and condensers in air conditioning. The study showed that with a certain water
treatment and in the absence of biological growth and suspended particulates, long-term fouling did
not exceed 0.0002 h ft2 °F/Btu (0.000035 m2 °C/W) and short-term fouling did not exceed
0.0001 h ft2 °F/ Btu (0.000018 m2 °C/W).
The Air-Conditioning and Refrigeration Institute (ARI) Standard 550-88 specifies the following:
Field fouling allowance 0.00025 h ft2 °F/ Btu (0.000044 m2 °C/W)
ARI Rating Standard (new evaporators and
condensers) 0
The ARI Standard recommends specifying 0.00025 h ft2 °F/ Btu (0.000044 m2 °C/ W) total fouling
for closed-loop liquid coolers and for condensers served by well-maintained cooling towers.
Pool Boiling and Forced-Convection Model
If a surface is in contact with a liquid, and if the temperature of the surface is maintained above the
saturation temperature of the liquid, boiling may occur. The rate of heat transfer depends mainly on
the temperature difference between the surface giving heat and the saturation temperature of the
liquid at a specified pressure.
When the surface is submerged below the free surface of the liquid, the boiling process is known
as pool boiling. When a mixture of liquid and vapor is forced through a tube or a passage and the
surface surrounding the mixture is above the saturation temperature of the liquid, forced-convection
boiling may occur inside the tube or passage.
When the temperature of a surface that releases heat is a few degrees higher than the temperature
of the liquid, molecules within the thin layer of superheated liquid adjacent to the surface tend
to break away from the surrounding liquid molecules. They form nuclei, which grow into vapor
bubbles. As the number and size of these bubbles increase, the bubbles flow upward to the surface
of the liquid and form forced-convection boiling. This process of combined nucleate boiling and
forced convection occurs in flooded shell-and-tube liquid coolers.
Chen (1966) developed a forced-convection model that can be applied to flooded shell-and-tube
liquid coolers to calculate the boiling heat-transfer coefficient on the outer surface of the tube hb,
Btu/h ft2 °F (W/m2 °C), as follows:
hb  hnb 	 htp  Shnbl 	 Fhfc (10.11)
where hnbl  single-tube pool-boiling coefficient, Btu/h ft2 °F (W/m2 °C)
S  suppression factor
hfc  single-phase forced-convection coefficient for liquid phase flowing alone in tube
bundle, Btu /h ft2 °F (W/m2 °C)
F  forced-convection multiplier
The experimental results of Webb et al. (1990) showed that for CFC-11 and copper tubes with integral
fins, the forced-convection portion hnb  Shnbl is 1.6 times larger than the nucleate boiling portion
htp  Fhfc at the first row (bottom row) of the tube bundle and 5.3 times larger at the top row.
In tubes with high-flux outer boiling surfaces with porous coating, the capacity of the liquid cooler
is 36 percent higher than in tubes with integral fins.
Because the average heat-transfer coefficient of a tube’s outer surface ho, Btu/h ft2 °F
(W/m2 °C), is a function of the temperature difference between the tube wall and the refrigerant
Ts 	 Tr , also ho is a function of the heat flux at the tube’s outer surface Qev /Ao, Btu/h ft2 (W/m2).
Here Qev represents the rate of heat transfer through the tube surfaces of the liquid cooler or the
REFRIGERATION SYSTEMS: COMPONENTS 10.15
cooling capacity of a specific cooler, Btu /h (W), and Ao the outer surface area of the tubes, ft2 (m2).
Heat-transfer coefficient ho in a flooded liquid cooler can be calculated as
(10.12)
Constant C and exponent n are mainly determined by the outer surface conditions of the tubes and
the wettability of different refrigerants on tubes.
Parameters That Influence the Performance of Flooded Liquid Cooler
Evaporating Temperature Tev. In comfort air conditioning systems, chilled water usually leaves
the liquid cooler at a temperature Tel from 42 to 45°F (5.6 to 7.2°C) to provide the desirable air offcoil
temperature Tal. In ice storage and cold air distribution systems (which are discussed in
Chap. 31), brine, a mixture of chilled water and glycol, may leave the liquid cooler at a temperature
of 24 to 34°F (	4.4 to 1.1°C).
In a conventional chilled water system, if the value of Tel is fixed, the choice of optimum temperature
difference Tel 	 Tev determines the evaporating temperature in the flooded liquid cooler, as
shown in Fig. 10.6. A smaller Tel 	 Tev means a lower log-mean temperature difference Tm between
the chilled water and refrigerant. For a specific refrigeration capacity, a smaller Tm needs a
larger liquid cooler, a higher Tev, and a lower power input to the compressor. If a larger Tel 	 Tev is
chosen, the liquid cooler can be smaller, but the Tev value will be lower and compressor power input
will be higher. A compromise between energy cost and initial investment should be made through
life-cycle cost analysis.
Most liquid cooler manufacturers have adapted a value of Tel 	 Tev between 6 and 10°F (3.3 and
5.6°C). For energy-saving models, Tel 	 Tev may be as low as 4 to 7°F (2.2 to 3.9°C).
Temperature Difference Tee 	 Tel. For each ton (12,000 Btu /h, or 3.516 kW) of cooling capacity
of a liquid cooler, the water flow rate (gpm, or L / s) at various temperature differences of
chilled water entering and leaving Tee 	 Tel can be calculated from Eq. (10.9) as follows:
V?
gal
ho  C 
Qev
Ao n
10.16 CHAPTER TEN
FIGURE 10.6 Heat-transfer process in a flooded shell-and-tube cooler.
Temperature difference Tee 	 Tel has a direct impact on pump power. In addition, for a given Tel,
a greater Tee 	 Tel always means a greater log-mean temperature difference Tm and a greater
Tee 	 Tev and, therefore, a greater cooler capacity and higher off-coil temperature Tal, as shown in
Fig. 10.6. Here the log-mean temperature difference is calculated as
(10.13)
At full load, the chilled water temperature increase from pump power and piping heat gain is usually
negligible. Therefore, temperature difference Tee 	 Tel is often equal to the temperature difference
of chilled water leaving and entering the water cooling coils Twl 	 Twe at design full load, and
is closely related to the conditioned air temperature leaving the coil Tal. Many liquid coolers use a
Tee 	 Tel between 12 and 24°F (6.7 and 13.3°C). For Tee 	 Tel  20°F, the water flow rate
gpm/ ton (0.0215 L / s  kW) refrigeration at full load.
The minimum chilled water temperature leaving the liquid cooler should be 38°F (3.3°C), and
the mass flow rate of chilled water flowing through the flooded liquid cooler should be approximately
constant to prevent freezing. If the chilled water temperature is to be lower than 38°F
(3.3°C), brine should be used, as in ice storage systems.
Oil Effect. In flooded liquid coolers, vapor velocity above the tube bundle is usually insufficient
to return the oil to the compressor. Oil concentration in the refrigerant tends to increase during operation.
If oil concentration rises above 5 percent, it may degrade the rate of heat transfer. Generally,
oil-rich liquid is taken and returned to the compressor continuously.
Cooling Capacity. The cooling capacity of a flooded liquid cooler also represents the capacity of
the refrigeration system. From Eq. (10.10), for a liquid cooler with specified Tel, Tee, and Tev, the
cooling capacity Qev, Btu /h, depends on the following factors:
 Tubes’ outer surface area Ao and number of transfer units NTU
 Heat flux at the tubes’ outer surface Qev /Ao
 Heat-transfer coefficients ho and hi and overall heat-transfer coefficient Uo
 Heat capacity rate of chilled water
Recent developments in flooded shell-and-tube liquid coolers have emphasized high-flux and enhanced
boiling coefficient tubes to increase cooling capacity.
Part-Load Operation of Flooded Liquid Cooler
When a flooded liquid cooler in a comfort air conditioning system is at part-load operation, the
outer surface area Ao is the same as in full-load operation, so the reduction of its cooling capacity is
mainly caused by the following:
1. A decrease in heat flux and, therefore, a lower outer surface boiling heat-transfer coefficient ho
2. A smaller difference in Tee 	 Tev caused by the drop in Tee and increase in Tev
V?
gal  1.2
Tm 
Tee 	 Tel
ln[(Tee 	 Tev) / (Tel 	 Tev)]
REFRIGERATION SYSTEMS: COMPONENTS 10.17
Tee 	 Tel, °F , gpm
8 3
10 2.4
15 1.6
20 1.2
24 1.0
V?
gal
3. A smaller log-mean temperature difference resulting from a higher Tev
4. A decrease of effectiveness 
10.4 DIRECT-EXPANSION LIQUID COOLER AND LIQUID
OVERFEED COOLER
Direct-Expansion Liquid Cooler
In a direct-expansion liquid cooler (shown in Fig. 10.7), liquid refrigerant evaporates inside the
copper tubes while chilled water fills the shell. An expansion valve and sometimes a distributor are
used for each group of refrigerant circuits connected to the same suction header and compressor.
In a direct-expansion liquid cooler, various inner surface configurations and enhancements are
used to increase the boiling heat transfer. To provide optimum velocity and a higher rate of heat
transfer on the water side of the liquid cooler, baffle plates are used to guide the water flow in the
shell in multipass arrangements.
Direct-expansion liquid coolers are usually used for refrigerating systems equipped with multiple
compressors. In a direct-expansion liquid cooler, refrigerant circuits may be connected to a
single header and compressor or to two separate headers and multiple compressors.
Liquid Overfeed Cooler
A liquid overfeed cooler can be an air cooler or a liquid cooler. Liquid refrigerant is fed to multiple
evaporators by a mechanical pump or by a gas pump using the high-pressure gas discharged from
the compressor. Liquid refrigerant is fed at a mass flow rate 2 to 6 times greater than the actual
evaporating rate, which causes liquid recirculation, as shown in Fig. 10.8. Liquid recirculation pro-
10.18 CHAPTER TEN
FIGURE 10.7 Direct-expansion liquid cooler: (a) cut-away view of a typical DX liquid cooler. (Source: Trane Company.
Reprinted with permission.)
vides a sufficient volume of liquid in the tubes and ensures that the inner surface of the tubes is
fully wetted throughout its length.
Compared with DX coils, liquid overfeed air coolers have the following advantages:
 They have higher heat-transfer coefficient on the refrigerant side, eliminating the decline of the
heat-transfer coefficient at the dryout area when the quality of the refrigerant is raised to 0.85 to
0.90 in a DX coil.
 They allow nonuniform loading between refrigerant circuits or between evaporators, so that none
are starved, with insufficient refrigerant supply.
 Air-refrigerant temperature difference is lower, which is especially beneficial for low-temperature
systems.
 Evaporators can be conveniently defrosted.
Their main disadvantage is their higher initial cost.
REFRIGERATION SYSTEMS: COMPONENTS 10.19
FIGURE 10.7 (Continued) (b) Schematic diagram.
FIGURE 10.8 Liquid overfeed cooler using a mechanical pump.
During the design of a liquid overfeed cooler, the following choices should be carefully
considered:
Mechanical Pump or Gas Pump. Gas pumps generally require more energy than mechanical
pumps. According to Cole’s (1986) analysis, based on a typical plant of 275 tons (967 kW) of cooling
capacity and an evaporating temperature of 20°F (	6.7°C), the energy requirement of a gas
pumping system is about 7 percent of the total plant power, and that of mechanical pump is about
2 percent. However, the difference in energy consumption is a small percentage of the total plant
power, so in many circumstances, gas pumps are still used.
Among mechanical pump systems, centrifugal pumps are most popular. Rotary and gear pumps
may be suitable in circumstances where cavitation is critical.
Down-Feed or Up-Feed. In down-feed circuits, refrigerant is fed to the highest tube of the evaporating
coil. This arrangement improves oil return and eliminates static head. Up-feed circuits
simplify liquid distribution. The relative locations of the evaporators and receivers become less
important, and the system design and layout may be simplified.
Counterflow or Parallel Flow. Because of its lack of superheat, as in DX coils, a parallel flow
arrangement between air and refrigerant in a liquid overfeed air cooler has a small performance advantage
over the counterflow arrangement. Also, frosting occurs evenly in parallel flow, whereas
frosting is heavier in the first and second rows of the finned-tube coil in a counterflow arrangement.
Circulating Rate. Circulating rate or number n is defined as
(10.14)
where and mass flow rate of refrigerant, circulated and evaporated, respectively, lb/min
[kg / (60 s)]. According to ASHRAE Handbook 1998, Refrigeration, the minimum recommended
circulating rate is as follows:
Liquid overfeed air coolers are widely used in low-temperature refrigeration plants for food storage
and processing. Recently, overfeed liquid coolers have also been used in ice storage systems.
10.5 CONDENSATION AND CONDENSERS
Condensation Process
When saturated vapor comes into contact with a surface having a temperature below the saturation
temperature, condensation occurs on the surface. There are two types of condensation:
1. The condensed liquid, often called the condensate, wets the surface and forms a film covering
the entire surface. This type of condensation is called filmwise condensation.
2. The surface is not totally wetted by the saturated vapor, and the condensate forms liquid droplets
that fall from the surface. This type of condensation is called dropwise condensation.
Ammonia (R-717)
Down-feed (large-diameter tubes) 6 to 7
Up-feed (small-diameter tubes) 2 to 4
HCFC-22-Up-feed 3
HFC-134a 2
m?ev  m?cir
n 
m? cir
m? ev
10.20 CHAPTER TEN
Compared with filmwise condensation, dropwise condensation has a greater surface heat-transfer
coefficient because it has a greater surface area exposed to the saturation vapor. In practice, however,
the entire surface of the condenser tubes becomes wet during refrigerant condensation. Filmwise
condensate falls in droplets and often disturbs the condensate films in a tube bundle.
The heat-transfer process in a refrigerant condenser occurs in three stages:
1. Desuperheating of the hot gas
2. Condensing of the gas to liquid state and release of the latent heat
3. Subcooling of the liquid refrigerant
Although the surface heat-transfer coefficient is lower on the hot gas side during desuperheating,
there is a greater temperature difference between the hot gaseous refrigerant and the cooling medium
during desuperheating. Subcooling only occupies a small portion of the condenser’s surface area.
Therefore, for simplification, an average heat-transfer coefficient is used for the whole condenser’s
surface area, and the condensation of refrigerant is assumed to occur at the condensing temperature.
Total Heat Rejection
The capacity of a condenser is rated according to its total heat rejection Qrej, Btu /h (W). Total heat
rejection Qrej is defined as the total heat removed from the condenser during desuperheating, condensation,
and subcooling of the refrigerant in the condenser, and it is expressed as
(10.15)
where mass flow rate of refrigerant in condenser, lb/ min [kg / (60 s)]
h2, h3  enthalpy of hot gas entering condenser and enthalpy of subcooled liquid leaving condenser,
Btu / lb (J / kg)
In refrigerating systems using hermetic compressors, the heat released by the motor is absorbed by
the refrigerant. If the heat gains and losses from the ambient air at the evaporator, condenser, suction
line, and discharge line are ignored, then
(10.16a)
where Qrl  refrigeration load at evaporator, Btu /h
Pcom  power input to compressor, hp
mot  efficiency of hermetic compressor motor
For a refrigerating system equipped with open compressors, mot  1.
For water-cooled condensers used in refrigeration systems using chilled water as the cooling
medium, chilled water pump power heat gain and chilled water piping heat gain should be added to
the refrigeration load by multiplying by a water heat gain factor Fw, h. Its magnitude varies from a water
system with a small plant-building loop Fw,h  1.05 to a campus-type water system Fw, h  1.10.
Equation (10.16a) can then be modified to
(10.16b)
Heat rejection factor Frej is defined as the ratio of the total heat rejection to the refrigeration load
of the evaporator, and it can be calculated as
(10.17) Frej 
Qrej
Qrl
 Fw, h 	
2545 Pcom
Qrl mot
Qrej 
Fw, h Qrl 	 2545 Pcom
mot
Qrej 
Qrl 	 2545Pcom
mot
m? r 
Qrej  60m?r (h2 	 h3)
REFRIGERATION SYSTEMS: COMPONENTS 10.21
Here the power input to the compressor , and Win represents the work input. Therefore,
Frej depends mainly on the difference between the condensing and evaporating pressures,
hence Frej is always greater than 1.
Types of Condenser
A condenser is a major system component of a refrigeration system. It is also an indirect-contact
heat exchanger in which the total heat rejected from the refrigerant is removed by a cooling
medium, usually air or water. As a result, the gaseous refrigerant is cooled and condensed to liquid
at the condensing pressure. Liquid refrigerant is often subcooled to a temperature up to 15°F
(8.3°C) below the saturated temperature at condensing pressure to conserve energy.
Based on the cooling medium used, condensers used in refrigeration systems can be classified
into the following three categories:
1. Water-cooled condensers
2. Air-cooled condensers
3. Evaporative condensers
10.6 WATER-COOLED CONDENSERS
Types of Water-Cooled Condenser
In a water-cooled condenser, condenser water or cooling water is used to remove condensing heat
from the refrigerant. The condenser water is often the recirculating water from the cooling tower. It
also could be from a lake, river, or well near the refrigeration plant. Groundwater discharged from
the condenser may be treated and used for other purposes. If river or lake water is used as the condenser
water, an effective water filter system, or sometimes a heat exchanger, may be necessary to
prevent fouling of condenser tubes. Use of river, lake, or groundwater is a once-through method, as
it cannot be recirculated and used as condenser water again.
If a refrigeration plant is near the seashore, sometimes seawater can be used as condenser water
if it is more economical than water from the cooling tower. Seawater is corrosive to many metals.
Copper-nickel alloy, stainless steel, and titanium are often used for water tubes in seawater
condensers. An effective filter system is necessary to prevent dirt and impurities from blocking the
water passages. Water treatment is often required to prevent the growth of living organisms. The
whole seawater intake system may be very expensive. The use of seawater is also a once-through
method.
Two types of water-cooled condensers are widely used for air conditioning and refrigeration
purposes: shell-and-tube and double-tube condensers. A double-tube condenser consists of two
tubes, one inside the other, as shown in Fig. 10.9. Water is pumped through the inner tube while
refrigerant flows in the space between the inner and outer tubes in a counterflow arrangement.
Because double-tube condensers provide limited condensing area, they are only used in small
refrigeration systems such as indoor packaged units.
Shell-and-Tube Condensers
There are two types of shell-and-tube condensers: horizontal and vertical. Vertical shell-and-tube condensers
usually incorporate a one-pass water flow arrangement. They consume more condenser water
than the horizontal type and are often located outdoors because of their height. Vertical shell-and-tube
condensers are sometimes used for ammonia refrigeration systems in industrial applications.
Pcom  m?r Win
10.22 CHAPTER TEN
Horizontal shell-and-tube condensers are widely used in both comfort and process air conditioning
systems. A horizontal shell-and-tube condenser using halocarbon refrigerant has an outer shell
in which copper tubes are held in position by tube sheets, as shown in Fig. 10.10. As in the shelland-
tube liquid cooler, copper tubes with diameters of and 1 in. (19 and 25 mm) are often used.
Integral external fins with spacing between 19 and 40 fins/ in. (0.6 to 1.3 fins / mm) and at a height
of 0.035 to 0.06 in. (0.9 to 1.5 mm) are usually used. Surface coating is employed to enhance the
condensing coefficient of the tube’s outer surface on the refrigerant side. Spiral grooves or other
turbulators are added to the inner surface to increase the water-side heat-transfer coefficient.
Hot gas from the compressor enters at the top inlet and is distributed along the two sides of the
baffle plate to fill the shell. Condenser water enters the copper tubes at the bottom of the condenser
to provide effective subcooling. The water extracts heat from the hot gas and the liquid refrigerant
and is discharged near the top of the shell. Water flow arrangements are the same as in shell-andtube
liquid coolers: one-pass, two-pass, and three-pass, with two-pass as the standard equipment.
After heat has been removed, the hot gas is desuperheated, condensed to liquid, and subcooled
to a lower temperature. Because of the higher condensing pressure, the liquid refrigerant is forced
into the liquid line. In a shell-and-tube condenser, usually about one-sixth of the shell’s volume is
filled with liquid refrigerant for subcooling. This liquid level prevents gas bubbles from entering the
liquid line.
34
REFRIGERATION SYSTEMS: COMPONENTS 10.23
FIGURE 10.9 Double-tube condenser.
Heat Transfer in Shell-and-Tube Condensers
As in shell-and-tube liquid coolers, heat transfer in shell-and-tube condensers can be calculated by
using Eq. (10.9) with the following modifications:
 Heat is rejected from the refrigerant in the condenser, whereas heat is absorbed in the liquid
cooler.
 Evaporating temperature Tev should be replaced by condensing temperature Tcon, °F (°C).
 Tee 	 Tel should be replaced by Tcl 	 Tce. Here Tce and Tcl represent the condenser water temperature
entering and leaving the condenser. Also hrl 	 hre is replaced by hre 	 hrl, and Tee 	 Tev is
replaced by Tcon 	 Tce.
In 1915, Nusselt developed the correlation for laminar film condensation. Actual shell-side
condensing coefficient hcon, Btu/h ft2 °F (W/m2 °C), is needed to correct for vapor shear, inundation
due to tube bundle, surface effects, and enhancements. These factors are usually determined by
experiment.
10.24 CHAPTER TEN
FIGURE 10.10 Shell-and-tube condenser with two-pass water flow arrangement: (a) schematic diagram;
(b) heat transfer.
It is important to recognize that the shell-side condensing coefficient hcon is a function of
1 / (Tcon 	 Ts). Here Ts is the surface temperature of tube wall. Therefore, the condensing coeffi-
cient can be calculated as
(10.18)
where Ccon  constant. The difference between the boiling and condensing heat-transfer coefficient
is that the greater the heat flux Q/Ao, the higher the boiling heat-transfer coefficient hb and the
lower the condensing coefficient hcon.
Parameters That Influence the Performance of Shell-and-Tube Condensers
Temperature Difference Tcl 	 Tce. Temperature difference Tcl 	 Tce determines the condenser
water flow rate. When the condenser water is from the cooling tower, the temperature difference
between the condenser water leaving and entering the shell-and-tube condenser Tcl 	 Tce has a
direct impact on condensing temperature Tcon and the compressor’s power input, condenser water
pump power, and cooling tower fan power input. The performance of the cooling tower must also
be considered. A temperature difference Tcl 	 Tce  8°F (4.4°C) means a condenser water flow rate
of 3 gpm/ ton (0.054 L / s kW), and Tcl 	 Tce  12°F (6.7°C) means a flow rate of 2 gpm / ton
(0.036 L / s kW).
Kirsner (1996) showed that for a three-stage centrifugal chiller using HCFC-123 as refrigerant,
a 3 gpm / ton (0.054 L / s kW) condenser water flow rate had a compressor’s kW/ ton of 0.596 (5.9
COP) at Tev  37.6°F (3.1°C), whereas a 2 gpm/ ton (0.036 L / s kW) flow rate had a 0.644
kW/ ton (5.46 COP), an increase of compressor power input of about 8 percent. At the same time,
condenser water flow rate of 2 gpm / ton (0.036 L / s kW) showed a saving of condenser water
pump power of 3.4 percent and a cooling tower fan power saving of about 3 percent compared to
the flow rate of 3 gpm / ton (0.054 L / s kW). In addition, the initial cost of a condenser water flow
rate of 2 gpm / ton was lower than the flow rate of 3 gpm / ton.
When lake, river, well, or seawater is used, the scarcity of water and water temperature are factors
to be considered. A cost analysis is often required to determine the temperature difference Tcl 	 Tce.
Condensing Temperature Tcon. In a shell-and-tube condenser, Tcon is closely related to the
temperature of condenser water Tce available. If Tcl 	 Tce has been selected, then Tcl is a fixed value
and Tcon is determined by choosing an optimum value of Tcon 	 Tcl. This value is directly related to
the size of the shell-and-tube condenser, as shown in Fig. 10.10b. As in the shell-and-tube liquid
cooler, a smaller Tcon 	 Tcl means a lower Tcon and a large condensing surface area, whereas a
larger Tcon 	 Tcl indicates a higher Tcon and a smaller condenser. Condenser manufacturers usually
adopt a Tcon 	 Tcl value between 6 and 10°F (3.3 and 5.6°C), and it is 4 to 7°F (2.2 and 3.9°C) for
energy-efficient models.
Effect of Oil. If the lubrication oil is miscible with the refrigerants (like many halocarbon refrigerants),
experiments show that there is no significant effect on condensing coefficient hcon when oil
concentration is lower than 7 percent. In actual operation, oil concentration is usually lower than
3 percent.
Subcooling. Subcooling increases the cooling capacity and coefficient of performance. However,
subcooling also uses part of the surface area in the condenser to cool the liquid refrigerant at the
bottom of the condenser. The surface heat-transfer coefficient on the subcooled liquid refrigerant
side is smaller than the condensing coefficient hcon. Subcooling also depends on the temperature
of entering condenser water Tce. For a shell-and-tube condenser, it may vary from 2 to 8°F (1.1 to
4.4°C).
hcon  ho  Ccon 1
Qrej / Ao 1/3
REFRIGERATION SYSTEMS: COMPONENTS 10.25
Condenser Capacity. Parameters that influence the condenser capacity Qrej of a shell-and-tube condenser
are mainly Uo, Ao, Tm, and Qrej /Ao. Size and capacity of the condenser must be matched with
the evaporator and compressor. Condenser capacity is always rated at certain operating conditions.
Part-Load Operation
When the heat flux Qrej /Ao reduces at the condensing surface of the shell-and-tube condenser during
part-load operation, hcon increases. Therefore, the reduction of Qrej at part-load operation in a shelland-
tube condenser is mainly caused by the drop in condensing temperature Tcon and, therefore, a
smaller Tm between refrigerant and condenser water. At part-load operation, the degree of subcooling
is also reduced because of the drop in condensing temperature Tcon.
10.7 AIR-COOLED CONDENSERS
Construction of Air-Cooled Condenser
Air-cooled condensers use air to extract the latent heat of condensation released by the refrigerant
during condensation. An air-cooled condenser generally consists of a condenser coil in which there
are a main condensing coil section and a subcooling coil section connected in series, with several refrigerant
circuits to condense the gaseous refrigerant to liquid, and a subcooled refrigerant at a lower
level, as shown in Fig. 10.11. The condenser coil is usually equipped with copper tubes and aluminum
fins when halocarbon is used as refrigerant. The diameters of the tubes are usually between
and in. (6.5 and 19 mm), and the fin spacing is generally 12 to 20 fins / in. (1.3 to 2 fins /mm).
On the inner surface of the copper tubes, microfins (typically 60 fins / in. or a fin spacing of 0.4 mm)
with a height of 0.008 in. (0.2 mm) are used, as in DX coils. The condenser coil usually has two to
three rows of tubes because of the lower air-side pressure drop provided by the propeller fan.
Hot gas from the compressor enters the refrigerant circuits at the top. This arrangement provides
flexibility between the condensing and subcooling areas. As the condensate increases, part of the
condensing area can be used as subcooling for the storage of liquid refrigerant. A receiver is necessary
only when not all the liquid refrigerant can be stored in the condenser coil during the shutdown
period of the refrigerant plant in winter. Cooling air is usually forced through the coil by a propeller
fan, as shown in Fig. 10.11. A propeller fan has a lower fan total pressure and large volume flow
rate, which make it more suitable for air-cooled condensers. Fans are usually located downstream
from the coils in order to provide an even airstream through the coils. A damper may be installed
after or before the fan to modulate the air volume flow rate. In a small air-cooled condenser, the
coils, propeller fan, and damper may be installed in line horizontally. In large air-cooled condensers,
condensing and subcooling coils are usually located on two sides, and the propeller fans
and dampers are at the top of the unit.
Heat-Transfer Processes and Temperature Curves
Heat-transfer processes in a typical air-cooled condenser using HCFC-22 as refrigerant are divided
into three stages, as shown in Fig. 10.11b: desuperheating of hot gas, condensation of liquid refrigerant,
and subcooling. Desuperheating uses about 5 percent of the condensing surface area, condensation
uses 85 to 90 percent, and subcooling uses 5 to 10 percent.
For a condenser coil with a depth of two to three rows, airflow and refrigerant flow are usually in
a combined counterflow and cross-flow arrangement. Air enters the condenser coil at a temperature
equal to the summer outdoor design dry-bulb temperature To. In a year of average weather, To is the
temperature whose annual cummulative frequency of occurrence is about 0.01 8760  88 h. If
the air-cooled condenser is installed on the roof, an additional 3 to 5°F (1.7 to 2.8°C) should be
34
14
10.26 CHAPTER TEN
REFRIGERATION SYSTEMS: COMPONENTS 10.27
FIGURE 10.11 Air-cooled condenser: (a) typical construction; (b) temperature curves of refrigerant
HCFC-22 and cooling air.
added to the summer outdoor design dry-bulb temperature to account for the temperature increase of
the roof due to solar heat. Let To  90°F (32.2°C). For a typical condenser, entering air temperature
is gradually raised to Tal of about 102°F when it leaves the condenser coil. Hot HCFC-22 gas may
enter the top of the condenser coil at a temperature Tgas  170°F (76.7°C) and is desuperheated to a
saturated temperature of about 115°F (46.1°C). Gaseous refrigerant condenses to liquid and releases
the latent heat of condensation. As the refrigerant flows through the condensing coil, the condensing
pressure drops, typically by about 8 psi (55 kPa), before entering the subcooling coil. Because no
latent heat is released, the temperature of the subcooled liquid refrigerant drops sharply to 105°F
(40.6°C) when it leaves the air-cooled condenser.
Equations to calculate heat transfer between refrigerant and air in an air-cooled condenser are
similar to those for a water-cooled condenser, except that the temperature of condenser water
should be replaced by the temperature of cooling air, and the air-side heat-transfer coefficient ho is
discussed in Chap. 17. The condensing coefficient for horizontal tubes can be found in ASHRAE
Handbook 1997, Fundamentals.
Total heat rejection Qrej of an air-cooled condenser in a DX packaged system can still be calculated
by Eq. (10.16a). However, for air-cooled condensers, the refigeration load Qrl in Eq. (10.16a)
is equal to the DX coil load Qc, both in Btu / h (W), and Fw, h  1.
Cooling Air Temperature Rise and Volume Flow
For a specific Qrej, temperature rise between air entering and leaving the air-cooled condenser Tal 	
To depends mainly on the volume flow rate of cooling air per unit of total heat rejection .
Here Qrej is often expressed in tons of refrigeration capacity at the evaporator; therefore,
is expressed in cfm / ton (L / s kW). A smaller usually results in a lower air-side heattransfer
coefficient ho, a greater Tal 	 To, a higher Tcon, a lower condenser fan power, and a greater
log-mean temperature difference Tm between refrigerant and air. Conversely, a larger
and a smaller Tal 	 To mean a lower Tcon, a smaller Tm, greater fan power consumption, and probably
more noise from the propeller fans. Based on cost analysis, the optimum value of is
usually 600 to 1200 cfm / ton (80.5 to 161 L / s kW) refrigeration capacity at the evaporator. In
comfort air conditioning systems, if the heat rejection factor Frej  1.25, air volume flow may be
between 40 and 80 cfm per 1 MBtu / h total heat rejection. When  900 cfm / ton (120.8
L/skW) refrigeration capacity, Tal 	 To is around 13°F (7.2°C). Fan power consumption in aircooled
condensers is usually 0.1 to 0.2 hp / ton (0.075 to 0.15 kW / ton ) refrigeration capacity.
Condenser Temperature Difference
The condenser temperature difference (CTD) for an air-cooled condenser is defined as the difference
in saturated condensing temperature corresponding to the refrigerant pressure at the inlet and
the air intake dry-bulb temperature Tcon, i 	 To. The total heat rejection Qrej of an air-cooled condenser
is directly proportional to its CTD. The Qrej of an air-cooled condenser operated at a CTD of
30°F (16.7°C) is approximately 50 percent greater than if the same condenser is operated at a CTD
of 15°F (8.3°C). On the other hand, for a specific value of Qrej, an air-cooled condenser selected
with a CTD of 15°F (8.3°C) is larger than a condenser elected with a CTD of 30°F (16.7°C).
Air-cooled condensers are rated at a specific CTD related to the evaporating temperature Tev of the
refrigeration system in which the air-cooled condenser is installed. Typical CTD values are as follows:
Tev, °F CTD, °F
45, for air conditioning 20 to 30
20 15 to 20
	20 to 	40 10 to 15
V?
ca / Qu, rej
V?
ca / Qu, rej
V?
ca / Qu, rej
V?
ca / Qu, rej
V?
ca / Qu, rej
V?
ca / Qu, rej
10.28 CHAPTER TEN
A refrigeration system with a lower Tev is more economical to equip with a large condenser with a
small CTD.
Condensing Temperature, Subcooling, and Oil Effect
A high condensing temperature Tcon means a higher condensing pressure pcon and a higher discharge
temperature. High Tcon and pcon result in a higher compressor power input and, more often, unsafe
operating conditions. Both are undesirable. At a specific location, an air-cooled condenser always
has a higher Tcon than a water-cooled condenser.
A higher Tcon and pcon may be caused by an undersized air-cooled condenser or a low
value. It may be the result of a high To in a specific location or a high roof temperature. It may also
be caused by a lack of cooling air or warm air recirculation, which is discussed below.
In air-cooled condensers, the degree of subcooling usually ranges from 10 to 20°F (5.6 to
11.1°C). Based on the results of Schlager et al. (1989) for HCFC-22 in microfin tubes, the condensing
heat-transfer coefficient decreased by 15 to 20 percent with a 150-SSU (Saybolt seconds universal
viscosity) and 300-SSU oil concentration of 2.6 percent. Experiments also showed a larger
degradation of heat transfer with the more viscous oil.
Clearance, Warm Air Recirculation, and Dirt Clogging
Air-cooled condensers must provide clearance between the condenser coil and the wall, or adequate
space between two condenser coils, as specified in the manufacturer’s catalog, to provide sufficient
cooling air. Sufficient clearance can also prevent recirculation of warm air discharged from aircooled
condensers. Generally, the clearance should not be less than the width of the condenser coil.
Recirculation of discharged warm air from the air-cooled condenser raises the average air temperature
entering the condenser coil To and, therefore, increases Tcon and reduces the capacity of the
condenser. Recirculation of warm air may be caused by insufficient clearance, insufficient upward
discharge air velocity from the condenser, and wind effect.
When an air-cooled condenser is located in industrial areas or in locations where outdoor air has
a high concentration of dirt, solid particles, or sometimes pollens and seeds from trees and plants,
then dirts and particles may clog the condenser coil and considerably reduce the volume flow of
cooling air. In these circumstances, the surface of the condenser coil must be inspected and cleaned
regularly.
Low Ambient Air Control
Lower condensing pressure saves energy. However, when the ambient air or outdoor air temperature
is low, the condensing pressure may drop considerably. If the condensing pressure drops below a
certain value because of low outdoor temperature at part-load operation, a reciprocating vapor compression
refrigerating system using an expansion valve as the throttle device will not operate normally.
The supply of liquid refrigerant to the evaporator may be reduced. If refrigerant feed is insuf-
ficient, frost may form on the outer surface of the coil. The strategy is to lower the condensing
pressure to the minimum limit specified by the manufacturer for effective operation. Therefore, various
methods of pressure control for air-cooled condensers have been developed to prevent the condensing
pressure from falling below a specific value. These include the following:
 Duty cycling, or turning the fans on and off, so that air is not forced through part of the condensing
coils. This reduces the cooling air volume flow as necessary.
 Modulating the air dampers to decrease the volume flow of the cooling air.
 Reducing the fan speed so that less air passes through the condensing coil.
V?
ca / Qu, rej
REFRIGERATION SYSTEMS: COMPONENTS 10.29
Low ambient air control should follow the manufacturer’s instructions. One manufacturer speci-
fies the following low ambient air control scheme:
 Control energized at 65  2°F (18.3  1.1°C)
 Control deenergized at 70  2°F (21.1  1.1°C)
 Minimum entering outdoor temperature, 25°F (	3.9°C)
Selection of Air-Cooled Condenser
Total heat rejection Qrej of air-cooled condensers varies at different cooling air intake temperatures.
During the selection of an air-cooled condenser, the following steps should be taken:
 Calculate the total heat rejection of the refrigerating system.
 Compare the face area of the condenser coil at the same cooling air intake temperature.
 Choose an air-cooled condenser with the required Qrej at a specific cooling air intake temperature
with the appropriate Tev.
 Select a condenser for comfort air conditioning whose value is between 800 and 1000
cfm/ ton (107 and 134 L / s  kW) refrigeration capacity at the evaporator.
 Select an air-cooled condenser whose manufacturer specifies a lower low ambient air control
limit.
 Check the sound power level of the propeller fans. Noise attenuation is always expensive.
10.8 EVAPORATIVE CONDENSERS
Condensation Process
An evaporative condenser uses the evaporation of water spray to remove the latent heat of condensation
of the refrigerant during condensation. It is actually a simplified combination of a watercooled
condenser and cooling tower.
An evaporative condenser consists of a condensing coil, a water spray bank, a forced-draft or
induced-draft fan, an eliminator, a circulating pump, a water basin, and a casing, as shown in
Fig. 10.12. Water is sprayed over the outside surface of the condensing coil. Because of the evaporation
of water, heat is extracted through the wetted surface of the coil. The rest of the spray falls
and is collected in the water basin. Air enters from the inlet located just above the water basin. It
moves up through the condensing coil, spray nozzles, and water eliminator; is extracted by the fan;
and is finally discharged at the top outlet in a counterflow arrangement. Other airflow and water
flow arrangements have also been developed.
The condensing coils are usually made of bare pipes of copper, galvanized steel, or stainless
steel. The high heat-transfer coefficient of the wetted surface does not require fins to increase its
outer surface area. Bare pipes are not easy to foul and are easier to clear.
Heat Transfer in Evaporative Condensation
The heat-transfer process in an evaporative condenser consists also of desuperheating, condensation,
and subcooling. Heat transfer from the saturated air film to the ambient air, and from the
condensing refrigerant through the metal tube to the wetted surface, can be calculated as
hwetAo(Tos 	 To)  hi Ai (Tcon 	 Tis) (10.19)
V?
ca / Qu, rej
10.30 CHAPTER TEN
REFRIGERATION SYSTEMS: COMPONENTS 10.31
FIGURE 10.12 A typical evaporative condenser: (a) schematic diagram; (b) temperature curves.
where Tos, Tis  outer and inner surface temperature of tube,°F (°C)
To  ambient air temperature, °F (°C)
Tcon  condensing temperature, °F (°C)
Ao, Ai  outer and inner surface area of tube, ft2 (m2)
hi  inner surface heat-transfer coefficient, Btu/h ft2 °F (W/m2 °C)
In Eq. (10.19), hwet is the heat-transfer coefficient of the wet surface, Btu /h ft2 °F (W/m2 °C).
According to ASHRAE Handbook 1996, HVAC Systems and Equipment,
(10.20)
Here m is the slope of the air enthalpy saturation curve, m  dhs /dTs, where hs denotes the enthalpy
of saturated air film and Ts the saturated temperature; Rw represents the R value of the wet
film on copper tubes. And Rw is nearly equal to 1 /ho. For convenience,
(10.21)
Because of the higher Tos in an evaporative condenser, for instance, Tos  80°F (26.7°C), and the
ratio hwet /ho  m / cpa  1.085 / 0.245  4.4, so hwet is often 4 or 5 times greater than ho.
The number of transfer units, rate of heat transfer, coil effectiveness, and parameter UoAo can be
calculated from Eq. (10.9).
Water Spraying
The water circulation rate in an evaporative condenser is far less than in the cooling tower for
the same total heat rejection. Usually, it is around 1.6 to 2 gpm / ton refrigeration capacity at the
evaporator.
In evaporative condensers, drift carry over the water eliminator depends mainly on the eliminator
configuration and the face air velocity at the water spraying section. According to ASHRAE
Handbook 1996, HVAC Systems and Equipment, an efficient eliminator can reduce drift loss to a
range of 0.001 to 0.2 percent of the water circulation rate. The evaporated water is 1.6 to 2 gph
(gallon per hour) per ton (1.7 to 2.2 kg /h kW) of refrigeration. And the blowdown rate may be as
low as one-half of the evaporation rate, and the total water consumption is about 2.4 gph/ ton (2.6
L/hkW) for air conditioning and 3 gph/ton (3.3 L/hkW) for cold storage. According to Knebel
(1997), typical pump power requirement is 0.02 hp/ton (0.004 kW/kw).
Cooling Air
In evaporative condensers, air should not be used for sensible cooling of the outer surface of the
condensing coil. An air velocity of 400 to 700 fpm (2 to 3.5 m / s), typically 500 fpm (2.5 m / s),
should be maintained when it flows over the condensing coil to provide a higher hwet for the latent
heat transfer between the wetted surface and the ambient air. Sufficient air volume flow rate is necessary
to maintain a desirable enthalpy difference between the saturated film and air (hs 	 ha).
Both centrifugal and propeller fans are used in evaporative condensers to provide the airflow. According
to Knebel (1997), typical propeller fan power requirement is 0.06 hp/ton (0.013 kW/ kW).
Site Location and Low Ambient Air Control
Evaporative condensers are usually located outdoors. If they are located indoors, air is ducted outdoors
and centrifugal fans are generally used to overcome the greater fan total pressure required.
hwet 
mho
cpa
hwet 
m
cpa
l
Rw
10.32 CHAPTER TEN
The evaporative condenser should be located as near to the compressor as possible. If it is located at
a distance from the compressor, the discharge refrigerant pipe should be selected to have a pressure
drop of 2°F (1.1°C) of saturated temperature difference of the refrigerant between the compressor
and the condenser.
If an evaporative condenser is located outdoors and will be operated during winter, its water
basin and pump can be located indoors in a heated space. A water heater can be installed in the
water basin, and the water should be drained back to the water basin as soon as the pump ceases to
operate, to prevent freezing.
To prevent the condensing pressure from falling below a certain value at which the expansion
valve or other system components are difficult to operate, the capacity of the evaporative condenser
is reduced by means of fan cycling, dampers, and fan speed modulation at low ambient temperatures,
as in air-cooled condensers.
Selection and Installation
The procedure for selecting and installing an evaporative condenser is as follows:
1. Before selecting an evaporative condenser, calculate the total heat rejection of the refrigerating
system Qrej, MBtu/ h (kW).
2. According to the refrigerant used in the system, select the appropriate material for the condenser
coil. Copper tubes should be used for halocarbon refrigerants, and stainless steel or galvanized
steel should be used for ammonia.
3. Select a condenser coil that can provide a lower condensing temperature—below 95°F (35°C),
if possible. This will save a lot of energy and reduce scale accumulation on the outer surface of
the condenser coil significantly compared with Tcon  100 or 105°F (37.8 or 40.6°C).
4. Based on the location of the evaporative condenser and the sound control requirement, it is often
best to select a centrifugal fan. Check the volume flow rate and the fan total pressure that can be
provided by the fan to overcome the pressure drop of the condenser coil and eliminator or air
duct, if any.
5. Check the flow rate of the water spray of the circulating pump.
6. For outdoor installation, select a casing and other system components that are weatherproof and
corrosion-resistant.
7. Install the evaporative condenser in a location where there is ample space for outdoor air intake
and discharge. The influence of nearby exhaust systems should be avoided.
Evaporative condensers have the advantage of lower condensing temperature and lower installation
and energy costs. They are widely used in industrial applications and have been adopted more
and more frequently in commercial buildings recently. The following reasons may explain why they
were not previously popular in commercial and institutional applications:
 Monthly maintenance, inspection, and cleaning are required to prevent buildup of scale on the
condenser coil.
 Water basin and water spraying system must be regularly inspected and cleaned to prevent the
growth of Legionella bacteria and other microbiological growth. Water treatments with chemicals
are often required. Mazurkiewicz (1996) reported that between the end of September and the beginning
of October, 1996, there was a recent outbreak of Legionnaires’ disease in a food market’s
evaporative condenser, in Farmington, Michigan.
 Designers and operators have less experience with evaporative condensers.
 Heat rejected from the evaporative condenser is more difficult to use in winter heating for energy
conservation than that from a water-cooled condenser.
REFRIGERATION SYSTEMS: COMPONENTS 10.33
10.9 COOLING TOWERS
A cooling tower is a device in which recirculating condenser water from a condenser or cooling
coils is evaporatively cooled by contact with atmospheric air. Most cooling towers used in refrigerating
plants for commercial buildings or industrial applications for air conditioning are mechanical
draft cooling towers. A mechanical draft cooling tower uses fan(s) to extract atmospheric air. It consists
of a fan to extract intake air, a heat-transfer medium (or fill), a water basin, a water distribution
system, and an outer casing, as shown in Fig. 10.13.
According to the location of the fan corresponding to the fill and to the flow arrangements of air
and water, currently widely used mechanical draft cooling towers for HVAC&R can be classified
into the following categories:
 Counterflow induced-draft
 Crossflow induced-draft
 Counterflow forced-draft
Counterflow Induced-Draft Cooling Towers
In a counterflow induced-draft cooling tower, as shown in Fig. 10.13a, the fan is located downstream
from the fill at the air exit. Atmospheric air is drawn by the fan through the intake louver or,
more simply, an opening covered by wire mesh. Water from the condenser or recirculating water
from the coil, or a mixture of the two, is evenly sprayed or distributed over the fill and falls down
into the water basin.
Air is extracted across the fill and comes in direct contact with the water film. Because of the
evaporation of a small portion of the condenser water, usually about 1 percent of the water flow,
the temperature of the water gradually decreases as it falls down through the fill countercurrent to
the extracted air. Evaporated water vapor is absorbed by the airstream. Large water droplets
entrained in the airstream are collected by the drift eliminators. Finally, the airstream and drift are
discharged at the top exit. Drift, or carryover, is the minute water droplets entrained in the airstream
discharged out of the tower. The evaporatively cooled water falls into the water basin and flows to
the condenser.
In a counterflow induced-draft cooling tower, the driest air contacts the coldest water. Such a
counterflow arrangement shows a better tower performance than a crossflow arrangement. In
addition, air is drawn through the fill more evenly by the induced-draft fan and is discharged at
a higher velocity from the top fan outlet. Both higher exhaust air velocity and even velocity
distribution reduce the possibility of exhaust air recirculation. Compared with the crossflow
induced-draft cooling tower, the vertical height from the installation level to the inlet of the
water-spraying nozzles in a counterflow tower is greater and, therefore, requires a higher pump
head.
Crossflow Induced-Draft Cooling Towers
In a crossflow induced-draft cooling tower, as shown in Fig. 10.13b, the fan is also located downstream
from the fill at the top exit. The fill is installed at the same level as the air intake. Air enters
the tower from the side louvers and moves horizontally through the fill and the drift eliminator. Air
is then turned upward and finally discharged at the top exit. Water sprays from the nozzles, falls
across the fill, and forms a crossflow arrangement with the airstream.
The crossflow induced-draft cooling tower has a greater air intake area. Because of the crossflow
arrangement, the tower can be considerably lower than the counterflow tower. However, the risk of
recirculation of tower exhaust air increases.
10.34 CHAPTER TEN
Counterflow Forced-Draft Cooling Towers
In a counterflow forced-draft cooling tower, as shown in Fig. 10.13c, the fan is positioned at the
bottom air intake, i.e., on the upstream side of the fill. Condenser water sprays over the fill from the
top and falls down to the water basin. Air is forced across the fill and comes in direct contact with
REFRIGERATION SYSTEMS: COMPONENTS 10.35
FIGURE 10.13 Cooling towers: (a) counterflow induced-draft; (b) crossflow induced-draft; (c) counterflow forceddraft.
the water. Because of the evaporation of the water, its temperature gradually decreases as it flows
down along the fill in a counterflow arrangement with air. In the air stream, large water droplets are
intercepted near the exit by the eliminator. Finally, the airstream containing drift is discharged at the
top opening.
Because the fan is located near the ground level, the vibration of the counterflow forced-draft
tower is small compared with that of the induced-draft tower. Also, if the centrifugal fan blows toward
the water surface, there is a better evaporative cooling effect over the water basin. However,
the disadvantages of this type of cooling tower include the uneven distribution of air flowing
through the fill, which is caused by the forced-draft fan. In addition, the high intake velocity may
recapture a portion of the warm and humid exhaust air. Counterflow forced-draft cooling towers are
often used in small and medium-size installations.
The preceding types of cooling towers, which use the evaporation of water to cool the condenser
water, are sometimes called wet towers. There is also a kind of cooling tower called a dry tower. It
is essentially a dry cooler, a finned coil and induced fan combination that cools the condenser water
flowing inside the tubes.
10.10 THERMAL ANALYSIS OF COOLING TOWERS
Technological Terms
The following terms are commonly used when referring to the performance of a cooling tower:
Approach. Temperature difference between the temperature of the condenser water leaving the
tower and the wet-bulb temperature of the air entering the tower.
Blowdown.Water discharged to the drain periodically to avoid buildup of dissolved solids.
Fill. The structure that forms the heat-transfer surface within the tower. Water from the condenser
or coil is distributed along the flow passages of the fill down to the water basin.
Makeup.Water added to the circulating water to compensate for the loss of water to evaporation,
drift, and blowdown.
Range. Temperature difference between the temperature of condenser water entering the tower
Tte and the temperature leaving the cooling tower Ttl, both in °F (°C).
Tower Coefficient
Baker and Shryock (1961) developed the following theory for the thermal analysis of the cooling
tower. Because the makeup water in a typical cooling tower is usually about 2 percent of the
circulating condenser water, if the heat energy difference between the makeup water and the
blowdown and the drift losses are ignored, and if the enthalpy increase of the water from the addition
of liquid water is ignored, then the energy balance between condenser water and air can be
calculated as
(10.22)
where mass flow rate of water and air, lb/ h (kg / s)
cpw  specific heat of water, Btu / lb°F (kJ /kg°C)
Tw  temperature of condenser water, °F (°C)
ha  enthalpy of air, Btu / lb (kJ / kg)
If the thermal resistance of the saturated air film that separates the condenser water and the
airstream is ignored, the combined heat and mass transfer from the air-water interface (the saturated
m? w, m? a 
m?w cpw dTw  m?a dha
10.36 CHAPTER TEN
air film that surrounds the condenser water droplets) to the bulk airstream can be evaluated as
(10.23)
where Km  mass-transfer coefficient, lb /h  ft2 (kg / s m2)
hs  enthalpy of saturated air film, Btu / lb (kJ / kg)
A  surface area at air-water interface, ft2 (m2)
In Eq. (10.23), the change of enthalpy, or total heat of air, consists of changes in sensible heat
and latent heat. Consider a cooling tower with a fill volume V and a contact surface area A  aV,
ft2 (m2). Here a is the surface area of fill per unit volume. Also let K  Km / cpw. Then combining
Eqs. (10.22) and (10.23), we get
(10.24)
The integrated value of Eq. (10.24) is generally known as the tower coefficient, or the number of
transfer units NTU, of the cooling tower.
Heat- and Mass-Transfer Process for Counterflow Cooling Tower
Figure 10.14 shows the heat- and mass-transfer process between condenser water and air in a counterflow
cooling tower with a water-air ratio . In Fig. 10.14a, water is cooled from the
entering temperature Twe  85 °F (29.4°C) to the leaving temperature Twl  75°F (23.9°C) at the
exit. The temperature of the saturated air film corresponding to the condenser water temperature
drops along the saturation curve and can be represented by the section we-wl.
Air enters the counterflow tower at a dry-bulb Tae  80°F (26.7°C) and a wet-bulb Tae  68°F
(20°C). Because the temperature of the entering air is higher than the temperature of leaving water,
that is, Tae  Twl, first air is evaporatively cooled until the air temperature approaches the water
temperature. After that, air is humidified and heated because the water temperature is higher. Air is
essentially saturated at the top exit of the tower. The humidifying and heating process of air is illustrated
by the curve ae-al shown in Fig. 10.14a. The driving potential of heat and mass transfer at
the condenser water inlet is indicated by the difference in enthalpy between the saturated air film
and the air, or hs, we 	 hal. Above the water basin after the fill, it is denoted by hs, wl 	 hae.
For humidification with evaporative cooling and heating, the sensible heat transfer from condenser
water to air is only about 5 percent of the total heat transfer.
In Fig. 10.14b, for the same counterflow cooling tower, air enters at Tae  72°F (22.2°C) and at
Tae  70°F (21.1°C); water may be cooled from 86.5 to 76.5°F (30.3 to 24.7°C). Because Twl  Tae,
air is humidified and heated along a curve nearly parallel to the saturation curve. The driving potential
is still the difference in enthalpy between a saturated air film and air. Air is also approximately
saturated at the top exit and approaches point al on the saturation curve. Under this circumstance,
the sensible heat transfer is about 23 percent of the total heat transfer. This value indicates that even
on rainy days, when air enters the tower in a near-saturated condition, there is still latent and sensible
heat transfer between water and air. The driving potential is mainly due to the higher enthalpy
of the saturated air film surrounding the condenser water droplets, which are at a higher temperature
than the contacted air.
Tower Capacity, Size, and Coefficient
Equation (10.23) can also be expressed as
(10.25) m?wcpw(Twe 	 Twl)  KmaV(hs 	 ha)
m?w / m?a  1.2
KaV
m?w
 Tw2
Tw1
dTw
hs 	 ha
m?a dha  Km(hs 	 ha) dA
REFRIGERATION SYSTEMS: COMPONENTS 10.37
In Eq. (10.25), the total heat removed from the condenser water represents the
cooling tower capacity in Btu / h (W) if is in lb /h (kg / s). Tower capacity must meet the required
total heat rejection at the water-cooled condenser.
Tower size is indicated mainly by the volume of the fill V, which includes both the crosssectional
area and the depth of the fill. Strictly speaking, heat and mass transfer also occurs in the
space between the fill and the water basin as well as between the fill and the water spraying. For
simplicity, these can be considered as being included in the volume of fill.
The tower coefficient , lb°F/ Btu (kg°C/ J), actually indicates the heat-transfer unit or
size of the fill. It is the primary factor that influences the effectiveness of the cooling tower. An increase
of tower capacity may be attributable to a larger tower size or a higher tower coefficient or
both. For a fixed water circulation rate, an increase in tower coefficient may be caused by a larger
tower size or a better fill configuration.
KaV / m?w
m?w
m?wcpw(Twe 	 Twl)
10.38 CHAPTER TEN
FIGURE 10.14 Heat- and mass-transfer process for a counterflow cooling tower: (a) entering air at 80°F drybulb
and 68°F wet-bulb temperature.
By using the numerical integration method, the tower coefficient for a counterflow cooling tower
can be calculated as
(10.26)
For crossflow cooling towers, the variation in air and water temperature is a function not only of the
depth of the fill but also of the transverse position of the fill. A theoretical thermal analysis for
crossflow towers is far more complicated than that for counterflow towers.
KaV
m?w

Tw2 	 Tw1
hs 	 ha
REFRIGERATION SYSTEMS: COMPONENTS 10.39
FIGURE 10.14 (Continued) (b) entering air at 72°F dry-bulb and 70°F wet-bulb temperature.
10.11 COOLING TOWER PERFORMANCE
Factors That Affect Cooling Tower Performance
The criteria for selecting a cooling tower for a water-cooled condenser are the effective removal
of total heat rejected at the condenser and the minimization of the sum of power consumption in
compressors, condenser fans, and condenser water pumps. Proper selection of tower range, waterair
ratio, approach, fill configuration, and water distribution system directly affects the performance
of a cooling tower.
A cooling tower used for an air conditioning system is rated on the following conditions:
 Unit of heat rejection at condenser: 1 condenser ton  15,000 Btu /h (4.395 kW)
 Water circulation rate: 3 gpm per condenser ton (0.014 L / s per kW heat rejected)
 Entering condenser water temperature: 95°F (35°C)
 Leaving condenser water temperature: 85°F (29.4°C)
 Outdoor wet-bulb temperature: To: 78°F (25.6°C)
 Range: 10°F (5.6°C)
 Approach: 7°F (3.9°C)
Range and Water Circulating Rate
Many cooling towers are custom-made; i.e., the manufacturer varies the fill configuration, water circulation
rate, and airflow rate to meet the operating characteristics required by the customer. The
range Twe 	 Twl, °F (°C), and the corresponding water circulating rate , lb/ min or gpm, are the
primary parameters that should be specified during design.
The range depends on the heat rejected from a water-cooled condenser Qrej, Btu /h (W), and the circulation
rate of condenser water , lb/ h (kg / s), or , gpm. Their relationship can be expressed as
(10.27)
The greater the water circulation rate, the smaller the range. In contrast, a smaller results in a
greater range. During the selection of a cooling tower at the design stage, Qrej, To, and Ttl are usually
known and fixed. Under such circumstances, a smaller range Tte 	 Ttl and a greater water circulating
rate mean the following:
 A lower condensing pressure and temperature
 A greater pumping energy consumption because of the greater
 A greater tower size because of smaller hs 	 ha
 A higher airflow rate for a fixed water-air ratio
A greater range and a smaller result in a higher condensing pressure, a lower pumping energy, a
smaller tower size, and a lower airflow rate.
The range is usually determined at design conditions, i.e., at summer outdoor design wet-bulb
temperature To and a design Qrej. At a lower To and part-load operation, both the range and Ttl will
be lower. In recent years, there have been projects that used a range greater than 10°F (5.6°C) for
air conditioning systems, such as 12°F (6.7°C) or even 15°F (8.3°C). A life-cycle cost analysis is
recommended to determine the optimum range at specific conditions, especially for cooling towers
whose annual operating time exceeds 3000 h.
m?w
m?w
m?w
m? w
Tte 	 Ttl 
Qrej
cpw m?w

Qrej
500V?gal
V?
gal m? w
m? w
10.40 CHAPTER TEN
Tower Coefficient and Water-Air Ratio
Cooling tower performance is often determined from field tests and presented in the form of an empirical
correlation as
(10.28)
where C  constant and z  depth of the fill, ft. In Eq. (10.28), the average value of exponent n
varies from 	0.4 to 	0.65. According to Webb (1984a, 1984b), the exponent m varies between 0.7
and 1. From the correlation in Eq. (10.28), Bernier (1994) showed that if all the other conditions
remain the same and only the water-air ratio varies from 1.17 to 0.46, then the tower coeffi-
cient increases 58 percent.
The face velocity of air that is calculated based on the cross-sectional area of the cooling tower
perpendicular to the airflow is usually 300 to 400 fpm (1.5 to 2 m / s) with a total pressure drop
across a counterflow induced-draft cooling tower of about 1 in. WC (250 Pa). This gives a mass
flow density of air between 1250 and 1700 lb /h ft2 (6100 and 8296 kg /hm2).
When the mass flow density of water drops to 500 lb /h ft2 (2440 kg /hm2), the water cannot
be spread sufficiently over the fill. On the other hand, when the mass flow density of water approaches
3000 lb /h ft2 (14,640 kg /hm2), the fill becomes flooding. For air conditioning systems,
the mass flow density of water usually varies between 1000 and 2000 lb /h ft2 (4880 and 9760
kg/hm2), and the water-air ratio may vary between 0.7 and 1.5. For a water circulation rate of 3
gpm (0.189 L / s) per condenser ton, and , the airflow rate is about 360 cfm (170 L / s) per
condenser ton.
Approach
Approach determines the temperature of condenser water leaving the tower Ttl corresponding to a
local outdoor design wet-bulb To. If the range and Qrej are fixed values, a closer approach always
means a lower Ttl and a lower condensing pressure and temperature.
Approach is closely related to the tower coefficient and the size of the tower. A closer approach
always means a larger tower, and a greater approach results in a smaller tower. For cooling towers
used for air conditioning systems, approach usually varies from 5 to 12°F (2.8 to 6.7°C). When a
cooling tower is connected with a precooling coil forming a water-side economizer (which is covered
in detail in Sec. 21.2), a closer approach, such as 5°F (2.8°C), means a greater percentage of
operating hours will be available to replace refrigeration and thereby reduce energy costs.
If other conditions remain the same, when the approach of a typical cooling tower drops from 10
to 5°F (5.6 to 2.8°C), the tower size may increase around 65 percent. An approach of 5°F (2.8°C)
requires a very large tower, which is not economical and is therefore not recommended.
Outdoor Wet-Bulb Temperature
Because heat and mass transfer at the fill is based on the difference in enthalpy between the saturated
film and the air, the wet-bulb temperature of the outdoor air is the primary parameter that affects
the performance of the cooling tower. It is best to use either a 0.4 percent or 1 percent summer
outdoor design wet-bulb temperature to guarantee tower performance during most of the operating
period in summer, depending on the requirements of the project.
If two different refrigeration systems at different locations have the same range and approach,
the one with a higher design outdoor wet-bulb temperature will have a smaller tower size than the
system with a lower To. This is because the enthalpy of the saturated air film is higher, so hs 	 ha is
greater in the system with a higher design To. However, Ttl is lower with a lower To.
m?w / m?a  1
KaV / m?w
m?w / m?a
KaV
m?w
 C  m?w
m?a n
z	m
REFRIGERATION SYSTEMS: COMPONENTS 10.41
Fill Configuration and Water Distribution
Fill configuration has a direct effect on the tower coefficient. An efficient fill has a greater surface
area per unit volume a, which means more contact surfaces, a longer contact time, and intimate
contact between air and water. Currently, two types of fill are widely used in cooling towers for
air conditioning systems: splash bars, as shown in Fig. 10.15a, and the more recently developed
10.42 CHAPTER TEN
FIGURE 10.15 Two types of fill for cooling towers: (a) splash bars;
(b) PVC cellular film.
cellular film, shown in Fig. 10.15b. Splash bars have the advantage of not being easily plugged by
dirt contained in the water; the splashing action can also redistribute the water at various levels.
However, field practice has shown that when wooden splash bars are replaced by polyvinyl chloride
(PVC) cellular film fill, the temperature of condenser water leaving the tower drops 2°F (1.1°C)
compared to the value before retrofit. Contaminated water must be strained or treated before it is
cooled in a PVC cellular film fill.
An even water distribution system is very important to tower performance, especially for counterflow
towers. A recently developed large-orifice, square plastic nozzle is essentially nonclogging
and can provide an even pattern of spraying over the fill. It also has a far greater spraying capacity
than small-orifice conical nozzles, and so its maintenance and cleaning are simple.
Construction Materials
Hot-dip galvanized steel construction provides adequate corrosion protection for most applications
at a low cost. PVC and fiberglass-reinforced polyester are often used for fills, drift eliminators, inlet
louvers, inlet screens, and the water basin. It is recommended to use stainless steel for the suction
strainer for the cooling tower.
10.12 PART-LOAD OPERATION AND OPTIMUM CONTROL
Part-Load Operation
If the water circulating rate , the water-air ratio , and the outdoor wet-bulb temperature To
remain the same, the performance of a counterflow cooling tower shows the following results when
it is operated at part-load (a load ratio less than 1):
 A smaller range due to lower Qrej, p
 A closer approach because of a fixed tower coefficient and heat-transfer surface and a lower Qrej, p
 A lower Tte and Ttl
The load ratio Rload is defined as
(10.29)
where Qrej, p, Qrej, f  heat rejection at condenser at part-load and full-load operation, respectively,
Btu/ h (W). If , , and Qrej remain the same, only To falls to a lower temperature than the
design condition. Then we note the following:
 A smaller range
 A smaller hs 	 ha because of lower enthalpy of saturated air film at lower To; also a lower Tte and
Ttl
 A greater approach because of a smaller hs 	 ha
If and remain the same, both To and Qrej, p drop at part-load operation. In addition to a
smaller range and closer approach, there is a further drop of Tte and Ttl.
Optimum Control
During part-load operation, the capacity of a cooling tower should be controlled for effective operation
of the refrigeration system at minimum energy consumption. In a typical cooling tower, the
m?w / m?a m?w
m?w / m?a m?w
Rload 
Qrej, p
Qrej, f
m?w / m?a m?w
REFRIGERATION SYSTEMS: COMPONENTS 10.43
electric power input to the tower fans is approximately 10 percent of the power input to the centrifugal
compressors; and the power input to the circulating pumps for condenser water is about 2 to
5 percent of the power input to the centrifugal compressors.
One earlier strategy for capacity control of cooling towers during part-load operation is to maintain
a fixed temperature Ttl for condenser water leaving the tower. This strategy does not minimize
the sum of power input to the compressors, tower fans, and circulating pumps when both Ttl and the
condensing temperature can be lowered. When To falls, a lower condenser water temperature leaving
the tower Ttl and entering the water-cooled condenser is always energy-saving, except when low
Ttl causes a low condensing pressure which cannot provide effective operation of the refrigeration
system refrigerant flow from the condenser and evaporator and the required evaporation (refrigeration
effect) in the evaporator.
In ASHRAE Journal (no. 3, 1997), “The HVAC&R ‘Evolution’ Continues,” it was reported that
a centrifugal chiller operated with an entering condenser water temperature of 55°F (12.8°C), a
minimum entering condenser water temperature. A better strategy of capacity control of cooling
towers for water-cooled condensers in a centrifugal chiller is to lower Ttl until Ttl falls to a limit, the
minimum entering condenser water temperature Tcon, m. If Ttl falls below this limit, the refrigeration
system cannot be effectively operated.
Most tower fans in current use are two- or three-speed fans. The key points for a nearly optimum
control of cooling towers, according to Braun and Diderrich (1990), are as follows:
 Use sequencing of tower fans to maintain a possibly lowered Ttl  Tcon, min during part-load operation
to minimize the sum of power input to the compressors and tower fans.
 Heat rejection Qrej, p determines the sequencing of tower fans, and Qrej, p, is measured by the product
of the range and the water circulation rate.
 Establish a simple relationship between the tower capacity and the sequencing of the tower fan.
 Use an open-loop control (see Chap. 5) for a stable operation.
 When an increase of tower capacity is required, the speed of tower fan operated at the lowest
speed should be raised first. Similarly, during a decrease of tower capacity, the highest tower fan
speed should be the first one to be reduced.
 The speed of the tower fan should not be reduced in sequence to maintain a Ttl lower than Tcon, min.
Example 10.2. A counterflow induced-draft cooling tower is required to cool the condenser water
from 95 to 85°F (35 to 29.4°C) at a design outdoor wet-bulb temperature of 78°F (25.6°C). If the
water-air ratio of this tower is 1.2, do the following:
1. Calculate the tower coefficient at this design wet-bulb temperature.
2. If the outdoor wet-bulb temperature, water circulation rate, and airflow rate remain the same as
in the design conditions, calculate the condenser water entering and leaving the tower if the rate
of heat rejection at the water-cooled condenser has fallen to 50 percent of its design value.
Take the specific heat for water as cpw  1 Btu / lb °F (4187 J /kg °C).
Solution. To find the enthalpy of saturated air film hs, the enthalpy of ambient air ha, and the
driving potentials hs 	 ha, do these calculations:
 Divide the temperature range of condenser water into 1°F (0.56°C) divisions such as 85, 86, . . .,
93, 94, 95°F. From App. Table B.1, the enthalpy of the saturated air film that surrounds the water
droplets hs at various temperatures can be determined.
 Find the enthalpies of the air that contacts with the condenser water at various temperatures ha by
means of Eq. (10.22), starting with incoming air at a wet-bulb To  78°F (25.6°C) in contact
with the coldest condenser water at temperature 85°F (29.4°C).
Based on Eq. (10.26), the tower coefficient or fill surface area required to cool the condenser
water from 86 to 85°F and raise the air enthalpy from that at To  78°F to ha78 	 1.2 is
10.44 CHAPTER TEN
(Tw2 	 Tw1 )/(hs 	 ha), or their average, 1/2[1/ (hs86 	 ha78) 	 1/(hs85 	 ha78)]. The tower coeffi-
cient of the cooling tower that cools the condenser water from 95 to 85°F is the summation or
integration of the tower coefficient of 95 to 94°F, . . ., 86 to 85°F.
From App. Table B.1, at a wet-bulb temperature of 78°F, the enthalpy of saturated air is 41.59
Btu/ lb. Because the enthalpy of the saturated air film at a water temperature 85°F is 49.45 Btu / lb,
the enthalpy difference between the saturated air film at 85°F and air at 78°F wet-bulb temperature
is hs 	 ha  49.45 	 41.59  7.86 Btu / lb, and 1 / (hs 	 ha)  0.1272 lb / Btu (0.0609 kg / kJ).
From Eq. (10.22), for each degree of increase in condenser water temperature, the corresponding
increase in the enthalpy of air is
Then at a condenser water temperature of 86°F, the enthalpy of saturated air film is 50.68 Btu / lb,
and the enthalpy of the corresponding air in contact with this condenser water is
ha  41.59 	 1.2  42.79 Btu / lb
The enthalpy difference between the saturated air film and air at Tw  86°F is then
hs 	 ha  50.68 	 42.79  7.89 Btu / lb
and
1
hs 	 ha

1
7.89
 0.1267 lb / Btu (0.0606 kg / kJ)
dha 
m? w
m? a
cpw dTw  (1.2 1) dTw
REFRIGERATION SYSTEMS: COMPONENTS 10.45
TABLE 10.2 Numerical Integration of a Counterflow Cooling Tower for Example 10.2
Cooling Enthalpy of
water saturated air Air enthalpy hs – ha,
temperature, °F film hs, Btu / lb ha, Btu / lb Btu / lb
Design conditions
85 49.45 41.59 7.86 0.1272
0.1270 0.1270
86 50.68 42.79 7.89 0.1267
0.1262 0.2532
87 51.95 43.99 7.96 0.1256
0.1249 0.3781
88 53.25 45.19 8.06 0.1241
0.1231 0.5012
89 54.58 46.39 8.19 0.1221
0.1209 0.6221
90 55.95 47.59 8.36 0.1196
0.1182 0.7403
91 57.36 48.79 8.57 0.1167
0.1151 0.8554
92 58.79 49.99 8.80 0.1136
0.1118 0.9672
93 60.27 51.19 9.08 0.1101
0.1082 1.0754
94 61.79 52.39 9.40 0.1064
0.1045 1.1799
95 63.34 53.59 9.75 0.1026
Part-load operation (50 percent design load)
82 45.91 41.59 4.32 0.2315
0.2328 0.2328
83 47.06 42.79 4.27 0.2341
0.2350 0.4678
84 48.23 43.99 4.24 0.2358
0.2402 0.7080
85 49.45 45.19 4.26 0.2347
0.2338 0.9418
86 50.68 46.39 4.29 0.2331
0.2312 1.1730
87 51.95 47.89 4.36 0.2294
,
lb /Btu
1
hs 	 ha
Average
,
Tw
hs 	 ha
,
lb /Btu °F
 Tw
hs 	 ha 
The average value of Tw /(hs 	 ha) between water temperature 85 and 86°F is 1(0.1272 	
0.1267) /2  0.1270 lb  °F/ Btu (0.0338 kg  °C/ kJ). Similarly, values of ha, hs 	 ha, and Tw /(hs 	 ha)
at other water temperatures can be calculated and are listed in Table 10.2.
Using the numerical integration method, we find the tower coefficient is
0.1270 	 0.1262 	 0.1249 	 0.1231 	 0.1209
	 0.1182 	 0.1151 	 0.1118 	 0.1082 	 0.1045
 1.1799 lb  °F/ Btu (0.3137 kg  °C/ kJ)
At 50 percent part-load operation, the tower coefficient , water circulation rate To, and the
water-air ratio remain the same. Heat rejection at the condenser Qrej, p  0.5 Qrej, f because the range
has been dropped approximately from 10°F to 0.5 10°F  5°F. By trial and error, we find that
when the condenser water is cooled from 87 to 82°F (30.6 to 27.8°C), the calculated tower coeffi-
cient is 1.1730 lb  °F/ Btu (0.3119 kg  °C/ kJ), which is approximately the same as the value at the
design condition, or 1.1799 lb  °F/ Btu (0.3137 kg  °C/ kJ). The results are also listed in Table 10.2.
10.13 OPERATING CONSIDERATIONS FOR COOLING TOWERS
Recirculation and Interference
Most cooling towers are installed outdoors. Location and layout of the cooling towers should be
carefully chosen so that recirculation, interference, and insufficient airflow are minimized.
The recapture of a portion of warm and humid exhaust air at the intake by the same cooling
tower is called recirculation. Recirculation causes an increase in entering wet-bulb temperature and,
therefore, degrades the tower performance. Recirculation is mainly caused by insufficient exhaust
air velocity, or insufficient vertical distance between the intake and the exit. When a wind encounters
a cooling tower, a pressure lower than atmospheric is formed on the leeward side of the tower.
Thus, the exhaust air is extracted downward toward the air intake. If the exhaust air velocity is not
high enough, recirculation occurs.
If the exhaust air of a cooling tower or warm air from other heat sources located upstream is
extracted by another cooling tower downstream, such a phenomenon is called interference. Interference
can be minimized by considering the summer prevailing wind direction and by properly
planning the layout of the cooling towers.
When the space between the intakes of two cooling towers or the space between the wall and the
intake is insufficient, or if a barrier is used to shield the view or noise, airflow may be limited, causing
an airflow rate lower than the required rate. Insufficient airflow impairs the effectiveness of heat and
mass transfer and results in a greater approach and range and, therefore, a higher condensing pressure.
Freeze Protection and Fogging
When a cooling tower is operated at locations where the outdoor temperature may drop below 32°F
(0°C), freeze protection should be provided to avoid the formation of ice on components of the
cooling towers (such as intake louvers) which may seriously restrict the airflow.
An electric immersion heater is usually installed in the water basin for freeze protection. In idle
towers, the fan should not be run and water flow should be completely shut off. Condenser water may
bypass the fill and be distributed directly over the inside surface of the intake louvers through a special
piping arrangement. This action eliminates the accumulation of ice on the surfaces of the louvers.
When warmer and essentially saturated air that is exhausted from the cooling tower is mixed
with the cooler ambient air in cold winter, the mixture becomes oversaturated, and the excess water
KaV / m?w
KaV
m?w
 
Tw
hs 	 ha

10.46 CHAPTER TEN
vapor condenses as fog, which is often called fogging or plume. Fogging limits visibility and may
form ice on nearby highways. Fogging should be minimized by locating the cooling towers so that
visible plumes and their influence on the nearby environment are not objectionable.
Blowdown
To avoid the concentration of impurities contained in water beyond a certain limit, a small percentage
of water in a condenser water system is often purposely drained off or discarded. Such a treatment
is called blowdown or bleed-off. The amount of blowdown is usually 0.8 to 1.0 percent of the
total water circulating flow. Because drift in a cooling tower varies from 0.1 to 0.2 percent, if the
evaporation loss in a typical location is 1 percent, the amount of makeup water is about 0.9 	
0.1 	 1.0  2 percent. The practice of blowdown or bleed-off is widely used in condenser water
systems.
If simple blowdown is inadequate to control scale formation, chemicals can be added to inhibit
corrosion and to limit microbiological growth. Water treatment to reduce scale formation and corrosion,
and to control microbiological growth, is discussed in Chap. 7.
Legionnaires’ Disease
In summer of 1976, American Legion conventioneers in Philadelphia suffered an epidemic in which
34 of the victims died. Later at the end of the same year, scientists at the Centers for Disease Control
in Atlanta discovered the bacterium that caused Legionnaires’ disease and called it Legionella.
According to Millar et al. (1997), there are two types of disease due to Legionella: the pneumonia
experienced by the victim with 105°F (40.6°C) fever which can lead to respiratory failure and
death; and the less severe, less frequent Pontiac fever whose patient recovers after a few days of
treatment. The estimate of annual cases of Legionnaires’ disease in the United States may be 10,000
to 100,000 cases, and the death rate may range from 500 to 2000.
Legionnaires’ disease is always aquired from the enviroment. The HVAC&R sources generally
associated with the outbreak of Legionnaires’ disease are aerosol producers and are associated with
reservoirs and water-spraying systems such as cooling towers, evaporative condensers, humidifiers,
and direct evaporative coolers. A water source is contaminated with Legionella, this water is
aerosolized into mist, and a victim inhales the contaminated mist—these are the necessary steps to
produce an outbreak of Legionnaires’ disease. Most transmission of Legionnaires’ disease occurs
indoors.
Legionella is a living organism. Its amplification, or growth, is preventable. Regular maintenance
of reservoirs and water-spraying systems, ensuring cleanliness, proper pH value and temperature
of water, and adequate level of residual biocides, is an effective measure to prevent Legionella
growth.
The location of a cooling tower or an evaporative condenser should be carefully arranged so that
the drift from the cooling tower or evaporative condenser is not extracted into the outdoor air intakes
of the air-handling units or packaged units.
In addition to the prevention of Legionella growth in the source, the current approach of
legionellosis control includes the identification of outbreaks, the investigation of potential sources,
biological tests, and decontamination of the source.
Maintenance of Cooling Towers
Maintenance of cooling towers includes the following:
 Inspection of the tower fan, condenser water pump, fill, water basin, and the drift carryover to ensure
satisfactory operation
REFRIGERATION SYSTEMS: COMPONENTS 10.47
 Periodic blowdown, cleaning of wetted surfaces, and effective water treatments to prevent formation
of scale and dirt as well as the growth of microorganisms
 Periodic lubrication of moving parts
The manufacturer’s maintenance schedule should be implemented. Otherwise, a maintenance document
must be established.
If noise from the tower fan exceeds the limits specified by local codes, sound attenuators at
the fan exit and at the intake louver should be installed. Details of sound control are discussed in
Chap. 19.
10.14 HEAT REJECTION SYSTEMS
Types of Heat Rejection System
A heat rejection system is a system that extracts condensing heat from refrigerant in the condenser
of a refrigeration system, uses some of or all the condensing heat for winter heating or other purposes,
if possible, or transfers it to a heat sink, usually atmospheric air or surface water. A heat rejection
system consists of a condenser, heat exchangers, fan(s) or pump, piping work, and ductwork
if necessary. Four types of heat rejection systems are commonly used in vapor compression refrigeration
systems, as shown in Fig. 10.16:
1. Water-cooled condensers using lake, river, ground, or seawater directly. A water-filtering device
is often required to prevent dirt and solid particles from fouling the condenser tubes. Water treatment
may be required depending on the quality of the cooling water.
2. Water-cooled condensers using recirculation water from the cooling towers. This type of rejection
system is most widely used in large refrigeration plants for commercial buildings. Condenser
water leaves the water-cooled condenser and enters the top of the cooling tower. It is
sprayed over the fills and cooled to a lower temperature by means of evaporation. The condenser
water then passes through a strainer and is chemically treated. Finally, it is extracted by a water
pump and discharged to the water-cooled condenser again.
3. Air-cooled condensers which are widely used in small and medium-size refrigeration plants or
in locations where water is scarce.
4. Evaporative condensers, which extract condensing heat from the coil directly by means of the
evaporative cooling effect.
Comparison between Various Systems
Systems should be compared at summer design conditions at a specific location with the following
outdoor air and water temperatures:
 Outdoor dry-bulb temperature: 95°F (35°C)
 Outdoor wet-bulb temperature: 78°F (25.6°C)
 Lake, river, or seawater temperature: 78°F (25.6°C)
The evaporating temperature and refrigeration capacity are the same for all the heat rejection systems.
During comparison, factors such as condensing temperature, cooling air and water flow rate,
maintenance, and initial and energy costs should be considered. The initial cost is compared on the
basis of means mechanical cost data for a self-contained rooftop unit of about 32 tons (113 kW)
of cooling capacity. The energy cost is compared for the whole refrigeration plant. In addition to
the factors mentioned above, designers should recognize that water is becoming scarcer in many
10.48 CHAPTER TEN
REFRIGERATION SYSTEMS: COMPONENTS 10.49
FIGURE 10.16 Heat rejection systems: (a) water-cooled condenser; (b) water-cooled condenser and cooling
tower; (c) air-cooled condenser; (d) evaporative condenser.
locations in the United States. Whether the released condensing heat can be easily used for winter
heating or the equipment itself can be used for energy conservation is also a factor that should be
considered. For example, a cooling tower can be used as part of a water economizer to replace part
of the refrigeration. The results of this comparison are listed in Table 10.3 for reference.
For small refrigeration systems, air-cooled condensers are still widely used. For large refrigeration
plants where condensing heat is not used for winter heating, the evaporative condenser should be
compared with a water-cooled condenser and cooling tower combination through detailed analysis.
ASHRAE/IESNA Standard 90.1-1999
ASHRAE/IESNA Standard 90.1-1999 defines the cooling tower performance as the maximum
flow rating of the tower divide by the fan nameplate rated motor power, in gpm / hp (L / s per kW).
For cooling towers, the rating conditions are: 95°F (35°C) entering water, 85°F (29.4°C) leaving
water, and 75°F (23.9°C) wet bulb of the outdoor air. Standard 90.1-1999 mandates the performance
required for the total system heat rejection capacity at rated conditions for cooling towers as
of October 29, 2001:
Propeller or axial fan cooling towers  38.2 gpm / hp (3.23 L / s per kW)
Centrifugal fan cooling towers  20.0 gpm / hp (1.69 L / s per kW)
ASHRAE/IESNA Standard 90.1-1999 defines the air-cooled condenser performance as the heat
rejected from the refrigerant divided by the fan nameplate rated motor power, in Btu /hhp (COP).
10.50 CHAPTER TEN
TABLE 10.3 Comparison of Various Heat Rejection Systems at Summer Design Conditions
Water-cooled Water-cooled condenser Air-cooled Evaporative
condenser and cooling tower condenser condenser
Condensing temperature, °F 95 102 120 100
Cooling air volume flow Smaller than airrate,
cfm/ ton* 600–1200 cooled condenser
Cooling water, gpm/ ton* 3 3 1.6–2
Makeup water, gal / (hton)* 2.4 2.4
Maintenance Periodic cleaning Monthly inspection
of coil when and cleaning of coil
outdoor air
is not clean
Initial cost (refrigeration plant) Depending on Lower than Higher Higher
initial cost on air-cooled
water intake condenser
Energy consumption Lower Lower than Higher at Lower
(refrigeration plant) air-cooled design load
condenser
Condensing heat used Easier Best applied Between water- Difficult
for winter heating or cooled and
equipment itself used evaporative
for evaporative cooling condenser
Application Large Medium Medium and small Medium and large
and large size, or where
water is scarce
Comparison is based on outdoor 95°F dry-bulb, 78°F wet-bulb; lake, river, or seawater temperature  78°F.
*Ton refrigeration capacity at the evaporator.
The rated conditions of air-cooled condensers using HCFC-22 as test fluid are: 125°F (51.7°C) condensing
temperature, 190°F (87.8°C) entering gas temperature, 15°F (8.3°C) subcooling, and 95°F
(35°C) entering dry bulb (cooling outdoor air). Standard 90.1-1999 mandates the total system heat
rejection capacity at rated conditions for air-cooled condensers as of October 29, 2001:
Total system heat rejection capacity  176,000 Btu /hhp (19.7 COP or 0.18 kW / ton)
ASHRAE/IESNA Standard 90.1-1999 also specifies the fan speed control including condenser
fans and cooling tower fans for heat rejection systems. Each fan powered by a motor of 7.5 hp (5.6
kW) or larger can operate the fan at two-thirds of full speed or less. The fan can automatically
change the fan speed to control the leaving fluid temperature or condensing temperature /pressure
of the heat rejection system except for condenser fans serving multiple refrigerant circuits, or where
installation located in climates with greater than 7200 CDD50, or up to one-third of the fans on a
condenser or tower with multiple fans where the lead fans comply with the requirement of fan
speed control.
10.15 REFRIGERANT FLOW CONTROL DEVICES
Refrigerant control devices, or refrigerant flow control devices, include expansion valves, electric
expansion valves, float valves, capillary tubes, and multiple orifices. Multiple orifices are discussed
in Chap. 13.
In a refrigeration system, the expansion valve is an adjustable throttling device through which
the refrigerant at condensing pressure is throttled to evaporating pressure or interstage pressure. At
the same time, the expansion valve regulates its opening to feed the required amount of refrigerant
to the evaporator to meet the refrigeration load at the evaporator.
Thermostatic Expansion Valves
Thermostatic expansion valves (TXVs) are the most widely used throttling or expansion devices in
direct-expansion refrigeration systems.
Operating Characteristics. A thermostatic expansion valve regulates its refrigerant flow rate to
the evaporator according to the degree of superheat of vapor refrigerant leaving the evaporator.
Figure 10.17 shows a thermostatic expansion valve connected to a DX coil, the evaporator. A thermostatic
expansion valve consists of a valve body, a valve spring, a diaphragm, and a sensing bulb
at the outlet of the evaporator. The sensing bulb is connected to the upper part of the diaphragm by
means of a capillary tube. The outlet of the thermostatic expansion valve is then connected to various
refrigerant circuits of the DX coil through the nozzles and tubes of a refrigerant distributor.
When the liquid refrigerant passes through the small opening around the valve pin, its pressure
is reduced to the evaporating pressure. Liquid refrigerant flows through the refrigerant distributor
and tubes, and vaporizes gradually as it flows inside the copper tubes. At position x, as shown in
Fig. 10.17, all the liquid has been vaporized. By the time the vapor refrigerant reaches the outlet o
of the evaporator, it is superheated to a few degrees higher than its saturated temperature.
If the coil’s load of this DX coil increases, more refrigerant vaporizes. Consequently, point x
tends to move toward the inlet of the evaporator. This causes an increase in the degree of superheat
at the outlet o as well as an increase in the temperature of the sensing bulb. Because the sensing
bulb may be partially filled with the same type of liquid refrigerant as that in the evaporator, the
higher temperature exerts a higher saturated pressure on the top of the diaphragm. This lowers the
valve pin and widens the valve opening. More liquid refrigerant is allowed to enter the evaporator
to match the increase in refrigeration load.
If the coil’s load drops, the degree of superheat of the vapor refrigerant at the outlet of the evaporator
becomes smaller, resulting in a narrower valve opening. The liquid refrigerant feed to the
REFRIGERATION SYSTEMS: COMPONENTS 10.51
evaporator drops accordingly. The degree of superheat of the vapor refrigerant at the outlet can be
adjusted by varying the tension of the spring in the thermostatic expansion valve.
Capacity-Superheat Curve. Figure 10.18 shows the capacity-superheat curve of a thermostatic
expansion valve. In Fig. 10.18, Tss represents the static superheat when the valve is closed at noload
condition. Static superheat Tss does not have sufficient pressure to open the valve from its
closed position. Static superheat usually has a value of 4 to 8°F (2.2 to 4.4°C).
The term To indicates the opening superheat, i.e., the superheat required to open the expansion
valve from a closed position to a fully open position corresponding to the full refrigeration load or
design coil load. Opening superheat usually has a value between 6 and 8°F (3.3 and 4.4°C). The
term Top indicates the operating superheat at a specific refrigeration load Qop, Btu /h (W). The
term Trl is the sum of Tss and To and represents the superheat setting of the thermostatic expansion
valve at full-load operation. The recommended Trl is between 10 and 20°F (5.6 and 11.1°C).
Superheat in excess of 20°F (11.1°C) means that too much of the coil’s surface is producing
superheat, which reduces the coil capacity. Too little superheat is also risky because it may cause
the liquid refrigerant to flood the compressor. The term Tnom indicates nominal superheat. For
safety, Tnom is generally equal to 1.1Trl to 1.4Trl.
External Equalizer. A thermostatic expansion valve with an external equalizer has a connecting
tube that links the bottom diaphragm chamber to the evaporator outlet o, as shown in Fig. 10.17, so
that the pressure under the diaphragm is the same as that at outlet o. Because currently used DX
coils have a considerable pressure drop at the refrigerant distributor and tubes before entering the
coil, thermostatic expansion valves with external equializer tubes are often used in DX coils.
Consider a DX refrigeration system using HCFC-22 as the refrigerant. The evaporating temperature
at the inlet of the evaporator is 45°F (7.2°C), and the saturated pressure is 76 psig (524 kPa g).
If the spring pressure at full load ps is 18 psig (124 kPa g) and the pressure drop of the refrigerant
flowing through the evaporator pev is 8 psi (55 kPa), in a thermostatic expansion valve with an external
equalizer, the saturated pressure of the refrigerant at point x, denoted by px, is therefore
px  pi 	 pev  76 	 8  68 psig (469 kPag) (10.30)
The corresponding saturated temperature at point x is Tx  40°F (4.4°C). In a thermostatic expansion
valve with an external equalizer, the pressure balance at the diaphragm is
pb  px 	 ps  68 	 18  86 psig (593 kPag) (10.31)
10.52 CHAPTER TEN
FIGURE 10.17 Thermostatic expansion valve with external equalizer.
The corresponding saturated temperature Tb  51.2°F (10.7°C). Because the temperature of the
vapor refrigerant at the outlet To can be considered equal to the bulb temperature Tb, the degree of
superheat is calculated as
Trl  To 	 Tx  Tb 	 Tx  51.2 	 40  11.2°F (6.2°C) (10.32)
Liquid Charge in the Sensing Bulb
Straight and Limited Liquid Charge. When the sensing bulb is charged with an amount of liquid
with the same properties as the refrigerant in the evaporator, such a charge is known as the straight
charge. If there is only a limited amounted of liquid charge, it is known as the limited liquid charge.
All the liquid vaporizes at a predetermined temperature, as shown in Fig. 10.19a. Any increase
above this temperature results in only a small increase in pressure on the diaphragm because of the
superheated vapor.
A sensing bulb filled with limited liquid charge limits the maximum operating temperature in
the evaporator and, therefore, prevents the compressor motor from overloading.
Cross-charge. The temperature versus pressure difference T 	 p of a sensing bulb with
straight charge is shown in Fig. 10.19b. At a different evaporating temperature Tev, Tbx  Tb 	 Tx
varies accordingly. If the liquid charge in the sensing bulb is different from the refrigerant used in
the evaporator, it is known as cross-charge. The primary purpose of cross-charge is to provide a
nearly constant degree of superheat at various evaporating temperatures, as shown in Fig. 10.19c.
For a constant value of pb 	 px, the degree of superheat at various evaporating temperatures is
REFRIGERATION SYSTEMS: COMPONENTS 10.53
FIGURE 10.18 Capacity-superheat curve of a thermostatic expansion
valve.
approximately equal. This prevents the refrigerant from flooding back to the compressor at higher
evaporating temperatures.
Hunting of Thermostatic Expansion Valve. Hunting is the quick cycling of maximum and minimum
output of a controlled device. For the thermostatic expansion valve, hunting results from the
over- or underfeeding of refrigerant to the evaporator. It causes temperature and pressure fluctuations
and may reduce the capacity of the refrigeration system.
The time lag of refrigerant flow through the evaporator to the sensing bulb may cause a continuous
overshooting of the valve position. The overfed refrigerant may form unevaporated slugs. These
10.54 CHAPTER TEN
FIGURE 10.19 Thermostatic characteristics of sensing bulb: (a) limited liquid charge; (b) straight charge; (c)
cross-charge.
slugs may chill the sensing bulb and suddenly reduce the refrigerant feed. The overshooting and
sudden reduction cause hunting.
Hunting of the thermostatic expansion valve is determined by the following factors:
 Valve size. An oversized valve aggravates hunting.
 Degree of superheat. The smaller the degree of superheat, the greater the chance of hunting.
 Sensing bulb charge. Cross-charge tends to prevent hunting.
 Location of sensing bulb. A properly selected sensing bulb location often minimizes hunting. The
sensing bulb should be located at the side of the horizontal section of the suction line immediately
after the evaporator outlet, so that it does not sense the oil temperature or liquid refrigerant at the
bottom of the tube.
Electric Expansion Valves
Electric, or more truly electronic or microprocessor-controlled, expansion valves can provide more
sophisticated, effective, and energy-efficient refrigerant flow controls than thermostatic expansion
valves. Currently, three types of electric expansion valves have become widely available: step motor
valves, pulse-width-modulated valves, and analog valves.
Step Motor Valve. The step motor is an electronically controlled multiphase motor capable of
moving continuously in forward or reverse direction, as shown in Fig. 10.20a. It can also be discretely
positioned in increments of a fractional revolution. Step motors accept a series of step signals
(such as 200/s). The rotational movement is transformed to linear movement required to open
and close an expansion valve by means of a gearbox and a threaded driveshaft. The gearing is used
to increase the torque. This allows the use of a small motor without overstressing it. One of the
available step motor valves needs only 5.7 W of power. Dolin (1997) reported that a step motor
valve may have 1532 steps of resolution in 0.125 in. (3.2 mm) of travel. Each step can change the
flow rate of HCFC-22 by 0.002 lb / min or 3 Btu / h (0.001 kg /min or 0.88 W).
Pulse-Width-Modulated Valves. These are solenoid valves operated at on /off mode, i.e., either
fully opened or fully closed, as shown in Fig. 10.20b. The resolution (the act, or capability of rendering
the distinguishable parts of an object) of flow rate of refrigerant feed is indicated by the
amount of time, the pulse width, or the duration of valve opening. For instance, a pulse-width-modulated
valve may be pulsed every 5 s. If only a flow rate of 40 percent is required, the valve will
hold open for 2 s and closed for 3 s.
REFRIGERATION SYSTEMS: COMPONENTS 10.55
(a) (b) (c)
Step
motor
Valve
pin
Coil
Valve
pin
Gears
Valve
pin
Coil
FIGURE 10.20 Electric expansion valves: (a) step motor valve; (b) pulse-width-modulated valve; (c) analog
valve.
Liquid refrigerant flows at a velocity of about 100 fpm (0.5 m/ s) in the liquid line. A suddenly
closing action often causes severe vibration and water hammer. A dampening by forcing out the liquid
refrigerant above or below the valve’s plunger through a small bleed passage is an effective
means of slowing the opening and closing.
Analog Valves. Instead of fully opening or closing, an analog valve can vary the strength of the
coil’s magnetic field, so that the valve pin (or plunger) can stay at various intermediate positions, as
shown in Fig. 10.20c. Because the drive circuitry is more complicated than the digital output
required for a pulse-width-modulated or a step motor valve, the performance, such as resolution of
a analog valve may suffer compared with step motor and pulse-width-modulated valves.
Advantages of Electric Expansion Valves. Compared to the thermostatic expansion valves, the
advantages of electric expansion valves are that they
 Provide a more precise temperature control
 Provide consistent superheat control under fluctuating head pressure
 Are able to be operated at low head pressure during lower ambient air temperature
 Are more energy-efficient
Of the three types of electric expansion valves (step motor, pulse-width-modulated, and analog
valves), the performance of step motor valves has proved superior to the two others.
Float Valves
Float valves control the refrigerant feed to the evaporator to maintain a specific liquid level. The
two types of float valves used in refrigerant control are high-side float valves and low-side float
valves. A high-side float valve is located at the high pressure side of the refrigeration system, between
the compressor discharge outlet and the inlet of the thermostatic expansion valve. High-side
float valves control the flow rate of the refrigeration feed to the evaporator indirectly by maintaining
a constant liquid level in the high-pressure float chamber.
Figure 10.21a shows a high-side float valve.The valve pin is linked with a float ball in such a
way that any rise in the liquid level in the float chamber opens the valve and allows the liquid
refrigerant to pass through the valve to the evaporator. A bypass tube provides a passage for noncondensable
gas to escape from the float chamber to the evaporator.
The high-side float valve may be located above the condenser. In a sealed refrigerant circuit, the
difference in refrigerant pressure before and after the high-side float valve forces the liquid refrigerant
through the valve and into the evaporator.
Because the high-pressure float chamber only holds a small and fixed amount of liquid refrigerant,
most of the liquid refrigerant in a reciprocating refrigeration system is stored in the evaporator.
Therefore, control of refrigerant charge is essential. Excessive refrigerant charge may cause floodback.
Insufficient refrigerant charge may lower system capacity. A high-side float valve can be used
in direct-expansion or flooded evaporators.
The low-side float valve is located at the low-pressure side of the refrigeration system. The lowside
float valve controls the liquid refrigerant feed to maintain a constant liquid level in the evaporator.
Figure 10.21b shows a low-side float valve. As the liquid level in the evaporator drops, the float
ball moves downward and opens the float valve wider so that more liquid refrigerant is fed to the
evaporator. In small refrigeration systems, the float chamber is often placed directly inside the evaporator
or in an accumulator, instead of a separate low-side float chamber. Operation of the low-side
float valve may be continuous or intermittent. Control of the refrigerant charge is not as important
for a low-side float valve as for a high-side float valve.
In evaporators with high evaporating rates or limited vapor passages, foaming refrigerant
often raises the refrigerant level during operation. Therefore, the float ball should be placed in an
10.56 CHAPTER TEN
appropriate position to determine liquid levels accurately. Low-side float valves are usually used in
flooded coils or evaporators.
Capillary Tube
In most room air conditioners and many heat pumps, a capillary tube is used as the throttling
device. A capillary tube, also sometimes called a restrictor tube, is a fixed length of small-diameter
tubing lying between the outlet of the condenser and the inlet of the evaporator. It reduces the
pressure of the refrigerant from the high-pressure side to the low-pressure side of the system and
controls the flow of refrigerant to the evaporator. The inside diameter of the capillary tube Din is
between 0.05 and 0.06 in. (1.27 and 1.52 mm) with a length L from 1 in. (25 mm) to several feet.
For halocarbon refrigerants, it is usually made of copper. In recent years, there has been a tendency
to use short capillary tubes having L/Din between 3 and 20 in heat pumps. These short
capillary tubes are often made of brass or copper alloy. A strainer is generally located before the
REFRIGERATION SYSTEMS: COMPONENTS 10.57
FIGURE 10.21 Float valves: (a) high-side float valve; (b) low-side float valve.
capillary tube to prevent clogging of the tube caused by impurities and foreign matter contained in
the refrigerant.
In Aaron and Domanski (1990), tests and experiments for short capillary tubes showed that
when the downstream pressure was lower than the saturated pressure of the entering refrigerant, the
fluid inside the short tube had a metastable inner core of liquid with a surrounding two-phase annular
ring. The flow showed a very weak dependence upon the downstream pressure and could be
considered a nonideal chocked flow. The diameter of the short capillary tubes was the parameter
that most strongly affected the mass flow rate of the refrigerant. A greater differential between the
upstream and downstream pressure also had a significant effect on the mass flow rate. Chamfering
the inlet of the short capillary tube showed only a minor influence on the rate of refrigerant flow. By
using a short capillary tube as a throttling device, one can operate the cooling mode even at an outdoor
temperature of 35°F (1.7°C) if necessary.
A capillary tube has a certain capability to balance the refrigerant flow when the refrigerant load
changes. At a given condensing pressure, if the evaporating pressure is lower than normal due to a
lower refrigeration load, the greater pressure difference causes a higher mass flow rate of refrigerant
through the capillary tube. Because the refrigerant flow handled by the compressor is lower due to
the low suction pressure, a decrease in the condensing pressure, and liquid level, and seal occurs in
the condenser. Some vapor then passes through the capillary tube into the evaporator. All these reduce
the mass flow rate of refrigerant flowing through the capillary tube and form a new balance.
When the evaporating pressure becomes higher because of a greater refrigeration load, more liquid
refrigerant accumulates in the condenser. Meanwhile, the liquid charge in the evaporator is insufficient.
The effects tend to reduce the heat-transfer area in the condenser and the wetted surface
area in the evaporator. Consequently, the condensing pressure is raised, and the evaporating pressure
is lowered. A greater high- and low-side pressure difference results in a higher refrigerant flow
and forms a new balance.
REFERENCES
Aaron, D. A., and Domanski, P. A., Experimentation, Analysis, and Correlation of Refrigerant-22 Flow Through
Short Tube Restrictors, ASHRAE Transactions, 1990, Part I, pp. 729–742.
ARI Guideline—Fouling Factors, Heating / Piping / Air Conditioning, no. 2, 1998, pp. 109–110.
ASHRAE,Waterside Fouling Resistance inside Condenser Tubes, Research Note 31 (RP106), ASHRAE
Journal, June 1982, p. 61.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, ASHRAE Inc., Atlanta, GA, 1997.
ASHRAE, ASHRAE Handbook 1998, Refrigeration, ASHRAE Inc., Atlanta, GA, 1998.
Baker, D. R., and Shryock, H. A., A Comprehensive Approach to the Analysis of Cooling Tower Performance,
ASME Transactions, Journal of Heat Transfer, August 1961, p. 339.
Benner, R. L., and Ramsey, J., Evaporative Condensers, Heating / Piping / Air Conditioning, no. 8, 1987,
pp. 63–65.
Bernier, M. A., Cooling Tower Performance: Theory and Experiments, ASHRAE Transactions, 1994, Part II,
pp. 114–130.
Braun, J. E., and Diderrich, G. T., Near-Optimal Control of Cooling Towers for Chilled-Water Systems,
ASHRAE Transactions, 1990, Part II, pp. 806–813.
Braun, J. E., Klein, S. A., and Mitchell, J. W., Effective Models for Cooling Towers and Cooling Coils, ASHRAE
Transactions, 1990, Part I, pp. 164–174.
Burger, R., Cooling Tower Technology, Heating / Piping / Air Conditioning, August 1987, pp. 41–45.
Chen, J. C., Correlation for Boiling Heat Transfer to Saturated Fluids in Convective Flow, Ind. Eng. Chem.
Process Design Develop., vol. 5, no. 3, 1966 .
Cheremisinoff, N. P., and Cheremisinoff, P. N., Cooling Towers, Selection, Design and Practice, Technomic,
Lancaster, PA, 1989.
10.58 CHAPTER TEN
Cole, R. A., Avoiding Refrigeration Condenser Problems: I, Heating / Piping / Air Conditioning, no. 7, 1986,
pp. 97–108.
Dolin, B. J., Introduction and Characteristics Electric Expansion Valves, Air Conditioning, Heating & Refrigeration
News, July 7 and 14, 1997.
Eckels, S. J., and Pate, M. B., Evaporation and Condensation of HFC-134a and CFC-12 in a Smooth Tube and a
Micro-Fin Tube, ASHRAE Transactions, 1991, Part II, pp. 71–81.
Facius, T. P., Certified Cooling Tower Performance, Heating / Piping / Air Conditioning, August 1990,
pp. 66–73.
Feeley, J. C., and Morris, G. K., Legionella: Impact on Water Systems in Buildings, ASHRAE Transactions,
1991, Part II, pp. 146–149.
Galey, Jr., M. D., Cooling Tower Water Treatment Alternatives, Heating / Piping / Air Conditioning, July 1990,
pp. 55–58.
Grimm, N. R., and Rosaler, R. C., Handbook of HVAC Design, McGraw-Hill, New York, 1990.
Hensley, J. C., Cooling Tower Energy, Heating / Piping / Air Conditioning, October 1981, pp. 51–59.
Jourdan, J., Cooling Tower System Installation, Heating / Piping / Air Conditioning, January 1988, pp. 131–134.
Kirsner,W., 3 gpm/ ton Condenser Water Flow Rate: Does It Waste Energy? ASHRAE Journal, no. 2, 1996,
pp. 63–69.
Knebel, D. E., Evaporative Condensing Minimizes System Power Requirements, HPAC, no. 4, 1997,
pp. 75–84.
Leary, Jr.,W. M., Optimizing Cooling Tower Selections, Heating / Piping / Air Conditioning, August 1987,
pp. 67–68.
Mazurkiewicz, G. A., Inspectors Find “Legionella Soup” in Market’s Evaporative Condenser, Air Conditioning,
Heating, and Refrigeration News, Nov. 18, 1996, pp. 1–2.
Meitz, A. K., Clean Cooling Systems Minimize Legionella Exposure, Heating / Piping / Air Conditioning, no. 8,
1986, pp. 99–101 .
Millar, J. D., Morris, G. K., and Shelton, B. G., Legionnaires’ Disease: Seeking Efffective Prevention, ASHRAE
Journal, no. 1, 1997, pp. 22–29.
O’Neal, D. L., and Peterson, K., A Comparison of Orifice and TXV Control Characteristics during the Reverse
Cycle Defrost, ASHRAE Transactions, 1990, Part I, pp. 337–343.
Rosa, F., Legionella and Cooling Towers, Heating / Piping / Air Conditioning, February 1986, pp. 75–85.
Schlager, L. M., Pate, M. B., and Bergles, A. E., A Comparison of 150 and 300 SSU Oil Effects on Refrigerants
Evaporation and Condensation in a Smooth Tube and a Micro-Fin Tube, ASHRAE Transactions, 1989, Part I,
pp. 387–397.
Schoen, A., Resolving TEV Hunting Problems, Heating / Piping / Air Conditioning, no. 7, 1990, pp. 69–72.
Stamm, R. H., Industrial Refrigeration: Condensers, Heating / Piping / Air Conditioning, no. 11, 1984,
pp. 101–105.
Wang, S. K., Principles of Refrigeration, vol. 2, Hong Kong Polytechnic, Hong Kong, 1984 .
Webb, R. L., Shell-Side Condensation in Refrigerant Condensers, ASHRAE Transactions, 1984a, Part I,
pp. 5–25.
Webb, R. L., A Unified Theoretical Treatment for Thermal Analysis of Cooling Towers, Evaporative Condensers,
and Fluid Coolers, ASHRAE Transactions, 1984b, Part II B, pp. 398–415.
Webb, R. L., Apparao, T. R., and Choi, K. D., Prediction of Heat Duty in Flooded Refrigerant Evaporators,
ASHRAE Transactions, 1990a, Part I, pp. 339–348.
Webb, R. L., Choi, K. D., and Apparao, T. R., A Theoretical Model for Prediction of the Heat Load in Flooded
Refrigerant Evaporators, ASHRAE Transactions, 1990b, Part I, pp. 326–338.
Wong, S. P. W., and Wang, S. K., System Simulation of the Performance of a Centrifugal Chiller Using a Shelland-
Tube Type Water-Cooled Condenser and R-11 as Refrigerant, ASHRAE Transactions, 1989, Part I,
pp. 445–454.
REFRIGERATION SYSTEMS: COMPONENTS 10.59
CHAPTER 11
REFRIGERATION SYSTEMS:
RECIPROCATING, ROTARY,
SCROLL, AND SCREW
11.1
11.1 RECIPROCATING REFRIGERATION
SYSTEMS 11.2
Types of Reciprocating Refrigeration
Systems 11.2
Real Cycle of a Single-Stage
Reciprocating Refrigeration System 11.4
11.2 RECIPROCATING
COMPRESSORS 11.5
11.3 RECIPROCATING REFRIGERATION
SYSTEM COMPONENTS AND
ACCESSORIES 11.5
Compressor Components and
Lubrication 11.5
Accessories 11.8
Service Valves 11.11
11.4 REFRIGERANT PIPING 11.12
Copper Tubing 11.12
Piping Design for Reciprocating
Refrigeration Systems 11.13
Size of Copper Tube, Refrigeration Load,
and Pressure Drop 11.14
Sizing Procedure 11.14
Suction Line 11.15
Discharge Line (Hot-Gas Line) 11.20
Liquid Line 11.21
11.5 CAPACITY AND SAFETY
CONTROLS 11.24
Capacity Control 11.24
Safety Controls 11.26
11.6 PERFORMANCE OF RECIPROCATING
COMPRESSOR, EVAPORATOR, AND
CONDENSER 11.29
Performance of Reciprocating
Compressor 11.29
Performance of Evaporator 11.32
Performance of Condenser 11.33
11.7 SYSTEM BALANCE 11.34
Balance of Capacities of Selected
Components 11.35
11.8 OPERATING CHARACTERISTICS OF
AIR-COOLED, DIRECT-EXPANSION
RECIPROCATING REFRIGERATION
SYSTEM 11.36
Operating Balance at Design
Condition 11.36
Pressure Characteristics 11.37
Part-Load Operation Using a Cylinder
Unloader 11.38
Pump-Down Control 11.39
Main Problems in Direct-Expansion
Reciprocating Refrigeration
Systems 11.40
Minimum Performance and Equipment
Selection 11.42
11.9 ROTARY REFRIGERATION
SYSTEMS 11.42
Rotary Compressor 11.42
Main Components 11.43
System Performance 11.43
11.10 SCROLL REFRIGERATION
SYSTEMS 11.43
Types of Scroll Refrigeration
Systems 11.43
Scroll Compressors 11.44
Radial and Axial Compliance 11.44
Compressor Performance 11.46
Concentration Shift and
Temperature Glide 11.46
Circulating Concentration Shift and Its
Effects on Performance 11.46
Heat Exchanger Flow Configuration
11.47
Capacity Control and Part-Load
Performance 11.47
System Characteristics 11.48
Scroll Chillers 11.48
11.11 SCREW REFRIGERATION
SYSTEMS 11.50
Types of Screw Refrigeration
System 11.50
Screw Compressors 11.50
Performance of Twin-Screw
Compressor 11.52
Oil Cooling 11.53
Capacity Control 11.53
Variable Volume Ratio 11.54
Economizer 11.54
Controls 11.54
Minimum Performance 11.54
Air-Cooled Screw Chillers 11.55
REFERENCES 11.56
11.1 RECIPROCATING REFRIGERATION SYSTEMS
Types of Reciprocating Refrigeration Systems
As discussed in Sec. 9.18, currently used reciprocating, or more truly, reciprocating vapor compression,
refrigeration systems can be classified primarily into the following systems:
Air-Cooled, Reciprocating DX Cooler. Figure 11.1 shows an air-cooled, reciprocating DX cooler.
In a DX system, a direct-expansion coil is used as the evaporator, and HCFC-22 and HFC-134a are
the primary refrigerants used in new and retrofit projects. The air-cooled condenser is often combined
with the reciprocating compressor(s) to form a unit called the condensing unit. The condensing
unit is generally located outdoors. Either the DX coil is installed in a rooftop packaged unit
consisting of fan(s), filters, condensing unit, and controls, mounted on the roof of the building; or it
is a system component in a split or indoor packaged unit located indoors. In supermarkets and many
industrial applications, DX coils are also installed in refrigerated display cases and food processing
and food storage facilities. Air is forced through the DX coil by a fan for better heat transfer. DX
systems are most widely used in reciprocating and scroll refrigerating systems.
Air-Cooled Reciprocating Chiller. This kind of reciprocating refrigeration system consists of a
direct-expansion shell-and-tube liquid cooler as evaporator (as described in Sec. 10.4), an aircooled
condenser, single or multiple reciprocating compressor(s), accessories, and controls. The
entire package is often called an air-cooled reciprocating chiller, and it is shown in Fig. 11.2. As
in DX systems, HCFC-22 and HFC-134a are the primary refrigerants used in new and retrofit
projects. In the shell-and-tube liquid cooler, the refrigerant evaporates inside the metal tubes, and
11.2 CHAPTER ELEVEN
FIGURE 11.1 An air-cooled reciprocating DX cooler.
the chilled water fills the shell. All components and accessories of an air-cooled reciprocating
chiller are enclosed in a casing made of steel sheets coated with corrosion-resistant paint.
In large air-cooled reciprocating chillers, air-cooled condenser coils are installed on both sides
of the unit, and the subcooling coil is at the bottom, connected to the condensing coil through a
liquid accumulator. The air-cooled reciprocating chiller is usually located on the rooftop of the
building.
According to ARI Standard 550/590-98, air-cooled reciprocating chillers are rated the same as
centrifugal chillers discussed in Sec. 13.2. In locations where water is scarce and expensive, aircooled
reciprocating chillers are often used in central hydronic air conditioning systems or for
process cooling purposes in industrial applications with a cooling capacity less than 250 tons
(879 kW).
Liquid Overfeed Reciprocating Refrigeration Systems. Liquid overfeed reciprocating refrigeration
systems use a liquid overfeed cooler as an evaporator, as described in Sec. 10.4. Figure 11.3
shows a liquid overfeed reciprocating system. Ammonia R-717 and HCFC-22 are often used as the
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.3
FIGURE 11.2 Air-cooled reciprocating chiller.
FIGURE 11.3 Liquid overfeed reciprocating refrigeration system.
refrigerants. An additional refrigerant pump is required to circulate the refrigerant to each evaporator
at a flow rate several times greater than the actual evaporating rate. After the expansion valve,
the refrigerant is throttled to the evaporating pressure. Although liquid refrigerant directly expands
into vapor inside the evaporator’s tubes and accumulates in the low-pressure receiver, it is not a dryexpansion
evaporator. In a liquid overfeed system, a water-cooled condenser is often used.
Liquid overfeed refrigeration systems have the advantages of a higher heat-transfer coefficient
of the wetted inner surface and simpler refrigerant flow control. However, they are large and heavy
and are generally used for industrial refrigeration, food storage, and sometimes in ice storage
systems.
Multistage Reciprocating Refrigeration Systems. As described in Chap. 9, a multistage reciprocating
system could be a compound system or a cascade system. Multistage reciprocating refrigeration
systems are used for industrial applications to provide refrigeration at low temperature, such as
below 10°F.
Real Cycle of a Single-Stage Reciprocating Refrigeration System
In a reciprocating refrigeration system, the real refrigeration cycle differs from the ideal cycle
because of the pressure drop of the refrigerant when it flows through the pipelines, valves, compressor
passages, evaporator, and condenser. Also, the compression and expansion processes are not
isentropic.
Figure 11.4 shows the real cycle of a single-stage reciprocating vapor compression refrigeration
system on a p-h diagram. In a single-stage reciprocating refrigeration system, the pipeline
from the outlet of the evaporator to the inlet of the reciprocating compressor is called the suction
line. The suction line is shown by 1s1v in Figs. 11.1 and 11.4. The pipeline from the outlet of
the reciprocating compressor to the inlet of the air-cooled condenser is called the discharge line,
or hot-gas line. This is shown as 2v2 in both figures. The pipeline from the outlet of the condenser
to the inlet of the thermostatic expansion valve is called the liquid line. This is shown as 33x in
both figures.
In Fig. 11.4, the thick lines represent the real refrigeration cycle, and the thin lines the ideal
cycle. Various line segments indicate the following refrigeration processes:
11.4 CHAPTER ELEVEN
FIGURE 11.4 Real reciprocating vapor compression refrigeration cycle on
p-h diagram.
1 1s superheat in evaporator
1s1v pressure drop in suction line
1v1 pressure drop from suction valve and passage to cylinder
12 compression process in cylinder
22v pressure drop from discharge valve and passage
2v2 pressure drop in discharge line
23 condensing and subcooling processes in condenser
33x pressure drop in liquid line
3x4 throttling process in expansion valve
4 1s evaporating process in evaporator
11.2 RECIPROCATING COMPRESSORS
In a reciprocating compressor, shown in Fig. 11.5, a single-acting piston in a cylinder is driven by a
crankshaft via a connecting rod. At the top of the cylinder are a suction valve and a discharge valve.
There are usually two, three, four, or six cylinders in a reciprocating compressor.
Vapor refrigerant is drawn through the suction valve into the cylinder until the piston reaches its
lowest position. As the piston is forced upward by the crankshaft, it compresses the vapor refrigerant
to a pressure slightly higher than the discharge pressure. Hot gas opens the discharge valve and
discharges from the cylinder. The gaseous refrigerant in a reciprocating compressor is compressed
by the change of internal volume of the compression chamber caused by the reciprocating motion
of the piston in the cylinder.
The cooling capacity of a reciprocating compressor ranges from a fraction of a ton to 200 tons
(1 to 3.5 kW to 700 kW). Refrigerant HCFC-22, HFC-134a, HFC-404A, HFC-407A, and HFC-
407C in comfort and process air conditioning and R-717 (ammonia) in industrial applications are
the currently used refrigerants in reciprocating compressors because of their zero ozone depletion
factors, except HCFC-22 which will be restricted from the year 2004 due to its ozone depletion.
The maximum compression ratio pdis /psuc of a single-stage reciprocating compressor is about 7.
Volumetric efficiency of a typical reciprocating compressor v decreases from 0.92 to 0.65 when
pdis /psuc increases from 1 to 6, and the isentropic or compressor efficiency isen decreases from 0.83
to 0.75 when pdis /psuc increases from 4 to 6. Methods of capacity control during part-load operation
include on/off, cylinder unloader, and hot-gas bypass controls.
Reciprocating compressor design is now in its mature stage. There is little room for significant
improvement. Reciprocating compressors are still widely used in small and medium-size
refrigeration systems. However, they will be gradually replaced by rotary, scroll, and screw
compressors.
11.3 RECIPROCATING REFRIGERATION SYSTEM
COMPONENTS AND ACCESSORIES
Compressor Components and Lubrication
The housing of a semihermetic reciprocating compressor, shown in Fig. 11.5, is the enclosure that
contains the cylinder, pistons, crankshaft, main bearing, oil sump, crankcase, and hermetic motor,
including both the rotor and the stator. The crankcase is the bottom part of the housing of the reciprocating
compressor in which the crankshaft is housed.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.5
11.6
FIGURE 11.5 Schematic reciprocating compressor: (a) sectional
view; (b) intake stroke; (c) compression and discharge stroke. (Source:
Trane Reciprocating Refrigeration and Refrigeration Compressors.
Reprinted with permission.)
Cylinder Block and Piston. In most compressors, the cylinder block is integrated with the
crankcase and forms a single casting. The crankcase and the cylinder block are usually made of
high-grade cast iron. In medium and large reciprocating compressors, premachined cylinder liners
or sleeves are often inserted in the crankcase, and they can be replaced when worn. It costs far less
to replace cylinder liners or sleeves than to replace the cylinder block.
The pistons are generally made of aluminum, aluminum alloy, or cast iron. For aluminum
pistons, piston rings are often used to prevent leakage of gas refrigerant from the cylinder and to
prevent oil from entering the cylinder. The upper piston rings are called compression rings, and the
lower rings are called oil rings. For cast-iron pistons, a running clearance of 0.0004 in./in. (cm/cm)
cylinder diameter provides an effective seal without oil rings.
Suction and Discharge Valves. The suction valve controls the vapor refrigerant entering the
cylinder, and the discharge valve controls the hot gas discharging from the cylinder. Suction and
discharge valves are usually made of high carbon-alloy steel or stainless steel. Spring-action ring
valves are most extensively used in medium and large compressors. Ring valves are usually heattreated
to the resilience of spring steel and must be precisely ground to a perfectly flat surface. A
deflection of 0.001 in. (0.025 mm) may cause leakage.
In addition, reed valves, either free-floating or clamped at one end, are used. A reed valve opens
under discharged pressure and is closed by spring action. Reed valves are usually limited to small
valve ports. A sufficient valve opening area must be provided to reduce pressure loss. A valve may
open up to 0.01 in. (0.25 mm) during suction or discharge. The maximum velocity of HCFC-22
flowing through valves may reach 150 ft/ s (45 m/ s).
Oil Lubrication. Oil lubrication is necessary to form a fluid film separating the moving surfaces
to protect them from wear and corrosion. Oil is also used as a coolant to carry heat away and cools
the refrigerant. Oil provides an oil seal between the piston and cylinder and between the valve and
valve plates. In refrigeration systems, mineral and synthetic oils are used for lubrication. HCFC-22
is partly oil-miscible so that the refrigerant-oil mixture can return to the compressor. HFCs and
HFC blends are not miscible with mineral oil. Polyolester-based synthetic oil should be used for
lubrication. Oil must be free of solids such as wax, must be chemically stable, and must have a
suitable viscosity. In reciprocating compressors using HCFC-22 and halogenated refrigerants, a
Saybolt seconds universal (SSU) viscosity between 150 and 300 is recommended by ASHRAE
Handbook 1994, Refrigeration Systems and Applications. Kinematic viscosity of 1 ft2 / s (92,900
mm2 / s) is approximately equal to 4.3  105 SSU.
In medium and large reciprocating compressors, forced-feed lubrication through a positive
displacement vane or gear oil pump is usually used. Lubricant is fed to the main bearing surfaces and
crankshaft through oil passages. An oil mist formed by the crankshaft rotation lubricates the cylinder
wall and piston pins. The oil pump is often mounted on one end of the crankshaft. An oil strainer
submerged directly in the oil sump is connected to the suction intake of the oil pump. Because the
vapor refrigerant in the crankcase is always at the suction pressure, the pressure lift of the oil pump is
equal to the difference of the discharge pressure at the oil pump and the suction pressure.
In small reciprocating compressors, splash lubrication may be used. Splash lubrication uses the
rotation of the crankshaft and the connecting rod to splash oil onto the bearing surfaces, cylinder
walls, and other moving parts. Often, oil passages, oil rings, and other devices are used to increase
oil splashing. An adequate oil level in the crankcase is necessary for satisfactory operation. Splash
lubrication is simple. Its disadvantage is that bearing clearance must be larger than normal, which
increases noise during operation.
For HCFC-22 and other halocarbon refrigerants, lubrication of suction and discharge valves is
achieved because the refrigerant is miscible with the lubrication oil. Refrigerant piping design must
accommodate the return of oil from evaporators and condensers. Oil should be separated from the
suction vapor at the compressor and returned to the crankcase.
Crankcase Heater. When the reciprocating compressor is installed outdoors, refrigerants may migrate
to the compressor from the indoor evaporator during the shutdown period of the refrigeration
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.7
system. This is mainly because the temperature and refrigerant pressure in the indoor evaporator
are higher than in the outdoor compressor. If a considerable amount of refrigerant accumulates in
the crankcase, it will dilute the lubrication oil. When the compressor starts to operate again, the
foaming oil-refrigerant mixture may form a slug, resulting in a serious loss of oil from the
crankcase. Slug entering the compression chamber may also damage the valves. For medium and
large reciprocating compressors, a built-in crankcase heater is often installed to keep the oil temperature
in the crankcase 15 to 25°F (8.3 to 13.9°C) higher than that in the rest of the system and thus
prevent refrigerant migration.
Accessories
Liquid Receiver. If not all the refrigerant in the system can be condensed and stored in the
condenser during the shutdown period, a high-pressure-side liquid receiver is needed to provide
auxiliary refrigerant storage space, as shown in Fig. 11.6a. Consequently, if a shell-and-tube watercooled
condenser is equipped in a refrigeration system, a liquid receiver is not necessary. When
an evaporative condenser is used, a liquid receiver is often installed to supplement its small storage
capacity.
There are two types of receivers: through-type and surge-type receivers. In a through-type receiver,
the liquid flows from the condenser to the receiver, and the pressure in the receiver is always
lower than that at the condenser outlet. For a surge-type receiver, its purpose is to allow liquid to
flow directly to the expansion valve and remain subcooled. Refer to ASHRAE Handbook 1998,
Refrigeration, for details.
The liquid receiver is usually a cylindrical tank made of steel sheets. The top of the tank is connected
to the liquid line from the condenser. It often has a liquid discharge line connected to the
evaporator by a shutoff valve. A vent pipe combined with a fusible plug is mounted on the receiver
to prevent the liquid receiver from bursting when exposed to high temperatures (as in a fire). In
most installations, it is common to charge the refrigerant in the receiver 80 percent full during the
shutdown period, and the remaining 20 percent is used to allow for liquid expansion.
To provide space for refrigerant storage during shutdown and for liquid expansion, saturated
pressure exists at the interface of the vapor and liquid refrigerant inside a through-type liquid receiver.
No subcooling occurs at the outlet of the through-type liquid receiver unless a subcooling
coil is provided after the receiver or a surge-type receiver is used. Because the subcooling liquid is
not exposed to gaseous refrigerant in a surge-type liquid receiver, the liquid remains subcooled after
flowing through it.
Liquid-Suction Heat Exchanger. A liquid-suction heat exchanger is usually mounted across the
suction and liquid lines when the hot-gas discharge temperature of the refrigerant (such as HFC-
134a) does not damage the lubrication oil or discharge valves. For ammonia and usually HCFC-22,
the superheat of the suction vapor in the heat exchanger may raise the discharge temperature of the
hot gas from the compressor to an unacceptable level, so liquid-suction heat exchangers are not
used for these refrigerants.
The purposes of a liquid-suction heat exchanger are as follows:
 To subcool the liquid refrigerant and increase the refrigerant effect
 To boil liquid refrigerant in the suction line and prevent liquid refrigerant from entering the reciprocating
compressor
 To reduce the flash gas in the liquid line and ensure maximum capacity for the thermostatic expansion
valve
Liquid-suction heat exchangers can be constructed as shell-and-coil, shell-and-tube, parallel
combined liquid and suction lines, or tube-in-tube heat exchangers. In a typical shell-and-tube heat
exchanger, as shown in Fig. 11.6, the hot liquid runs through the shell. The cold suction gas flows
11.8 CHAPTER ELEVEN
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.9
FIGURE 11.6 Liquid suction heat exchanger: (a) schematic diagram; (b) heat exchanger.
into the inner tube and penetrates the perforations at high velocity. Liquid refrigerant is subcooled
between 7 and 15°F (3.9 and 8.3°C) in the heat exchanger where the heat is absorbed by the cold
vapor. The pressure drop in the suction vapor should not exceed 2 psi when it flows through the
heat exchanger.
Filter Dryer and Strainer. Moisture may freeze in the expansion valve, especially in a lowtemperature
system, and impair expansion valve operation. Foreign matter in the refrigeration
system may also damage the expansion valve, control valves, and compressor. Moisture and foreign
matter must be removed from the refrigeration system by means of a filter dryer and strainer.
Figure 11.7a shows a filter dryer. It contains a molded, porous core made of material with high
moisture affinity, such as activated alumina, and acid-neutralizing agents to remove moisture and
11.10 CHAPTER ELEVEN
FIGURE 11.7 (a) Filter dryer and (b) sight glass.
foreign matter from the refrigerant. The core should be replaced after a certain period of operation.
The filter dryer is usually installed at the liquid line immediately before the expansion valve and
shutoff valve.
The strainer contains a fine-mesh screen to remove copper fillings, dirt, and other debris that
may be in the system at start-up. The construction of a strainer is similar to that of a filter dryer. Instead
of a porous core, there is a fine-mesh wire screen. The strainer is often installed in the suction
line just before the compressor.
Sight Glass. A sight glass is a small glass port used to observe the condition of refrigerant flow in
the liquid line. Sight glasses are available with one or two ports. A double-port sight glass, as
shown in Fig. 11.7b, is recommended because light enters the opposite port and makes observations
easier. The sight glass is located just before the expansion valve. Bubbles seen through the sight
glass indicate the presence of the flash gas instead of liquid refrigerant. The presence of flash gas
always indicates that the evaporator’s capacity is reduced because of a shortage of refrigerant or insufficient
subcooling.
A moisture-detecting chemical is built into the sight chamber behind the sight glass. If excessive
moisture is present in HCFC-22, the chemical turns pink. If the moisture is below the safe limit, the
color is green.
In addition to these accessories, there are pressure gauges to indicate suction pressure, discharge
pressure, and oil pressure. A liquid level indicator with gauge glasses is used to observe the liquid
level in the receiver. It is often installed on the receiver or on the liquid line. A hot gas muffler may
be used to smooth the pulsation of hot gas discharged from the compressor. The muffler should be
located as close to the compressor as possible.
Most modern reciprocating compressors minimize oil pumping. Except in systems where oil
removal is necessary, as in low-temperature systems, or for refrigerant that is not oil-miscible, oil
separators are seldom used.
Service Valves
Solenoid Valves. Solenoid valves are widely used in the suction, discharge, and liquid lines, and
in places where equipment and components are isolated for maintenance and repair. They are often
automatically controlled by the direct digital control (DDC) and other controls.
A solenid valve in refrigerant flow control is shown in Fig. 11.8a.When the electromagnetic or
solenoid coil is energized, the plunger and the attached valve pin are lifted upward to open the valve
port fully. The valve is closed by gravity or spring action. For medium-size valves, a pilot-operated
solenoid valve is often used. In this type of valve, the solenoid coil opens a pilot port, which produces
a pressure imbalance across the diaphragm, forces it upward, and opens the main port. When
the solenoid coil deenergizes, the plunger closes the pilot port. The pressure across the diaphragm is
balanced through an equalizer hole.
Manual Shutoff Valves. These valves are used to shut off or fully open the refrigerant flow manually.
Figure 11.7b shows a manual shutoff angle valve.
Most manual shutoff valves have packing around the stem. They also include back seat construction:
The valve seat closes against a second seat to prevent refrigerant leakage through the
valve stem and back seat port.
Pressure Relief Valves. In refrigeration systems, refrigerant pressure may rise to levels that could
result in an explosion in the event of fire, malfunction, or shutdown. As in water systems, pressure
relief valves in refrigeration systems are used to relieve the refrigerant pressure when it exceeds a
predetermined limit.
A typical sping-loaded pressure relief valve is shown in Fig. 11.8c. It is usually mounted on the
liquid line or on the liquid receiver. The pressure required to open the relief valve is adjusted by
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.11
the spring. Once it is set, the spring is sealed for protection. When the pressure of the refrigerant
exerted on the rubber disk is greater than the spring load, the valve port opens and the refrigerant
pressure is relieved through the upper outlet.
Refrigerant Charge Valve. A refrigerant charge valve is a manually operated shutoff valve used
only during refrigerant charge. It is often mounted on the liquid line at the outlet of the condenser
or on the outlet side of the receiver.
Check valves for refrigerant flow control are similar in construction and operation to the water
valves described in Chap. 7.
11.4 REFRIGERANT PIPING
Copper Tubing
Refrigerant pipes transport refrigerant through the compressor, condenser, expansion valve, and
evaporator to provide the required refrigeration effect. For halocarbon refrigerants, the refrigerant
pipes are usually made of copper. However, if ammonia is used as the refrigerant, the pipes are
always made of steel.
As listed in Table 11.1, three types of copper tubing are used for refrigerant piping. Type K
is heavy-duty, and type M is lightweight. Type L is the standard copper tubing most widely
used in refrigeration systems. Copper tubing installed in refrigeration systems should be entirely
free of dirt, scale, and oxides. The open ends of clean new tubes should be capped to keep contaminants
out.
11.12 CHAPTER ELEVEN
FIGURE 11.8 Service valves: (a) solenoid valve; (b) manual shutoff valve (angle type); (c) pressure relief valve.
Piping Design for Reciprocating Refrigeration Systems
Packaged units with factory-assembled and -integrated refrigeration systems have become more
popular, and air-cooled reciprocating packaged units or chillers contain a reciprocating compressor,
liquid coolers, air-cooled condensers, expansion valves, and controls all in one package. However,
many field built-up or split systems need proper refrigeration piping design to provide adequate
cooling for air conditioning and industrial applications.
A proper refrigeration piping design must satisfy the following requirements:
 Transport the required amount of refrigerant to the evaporator, compressors, condenser, and throttling
devices.
 Provide optimum pressure drop in the refrigeration lines, i.e., the most economical maximum
pressure drop in the suction, discharge, and liquid lines. Current practice limits the maximum
pressure drop in suction, discharge, and liquid lines corresponding to a change of saturated temperature
	Tsat as follows:
 For halocarbon refrigeration systems, oil should be miscible with liquid refrigerant, and oil
entrained in the refrigerant should be returned to the crankcase of the compressor to maintain an
adequate level for proper lubrication. Refrigerant piping design must accommodate the flow of both
refrigerant and oil. Oil does not mix with gaseous refrigerant. Transportation of entrained oil in hot
gas or vapor refrigerant is accomplished by ensuring a minimum refrigerant velocity not less than
500 fpm (2.5 m/s) in horizontal refrigerant lines and not less than 1000 fpm (5 m/s) in vertical
refrigerant risers. Horizontal refrigerant lines must be pitched in the direction of refrigerant flow.
 Flashing of liquid refrigerant must not occur before the expansion valve. Adequate subcooling
should be provided to offset the pressure drop of the liquid line and its accessories.
 Piping design should be configured to prevent liquid slugging, in which liquid refrigerant, oil, or a
combination of both floods back to the compressor.
 Maintain a clean and dry refrigeration piping system.
	Tsat, °F HCFC-22 pressure drop, psi (kPa)
Suction line 2 2.91 (20) at Tsuc  40°F (4.4°C)
Discharge line 1 3.05 (21) at Tcon  105°F (40.6°C)
Liquid line 1 3.05 (21) at Tcon  105°F (40.6°C)
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.13
TABLE 11.1 Correction Factors for Suction, Discharge, and Liquid Lines for HCFC-22
or Tsuc Other than 40°F and Tcon Other than 105°F
Correction factors for Tsuc other than 40°F
Tsuc, °F 50 40 30 20 10 0 10 20
Fsuc,s 1.20 1.00 0.84 0.70 0.58 0.47 0.38 0.30
Fdis,s 1.01 1.00 0.99 0.97 0.96 0.95 0.93 0.91
Correction factors for Tcon other than 105°F
Tcon, °F 80 90 100 105 110 120 130 140
Fsuc,c 1.11 1.07 1.03 1.00 0.97 0.90 0.86 0.80
Fdis,c 0.79 0.88 0.95 1.00 1.04 1.10 1.18 1.26
Fliq,c 1.12 1.07 1.02 1.00 0.98 0.93 0.88 0.83
Note: Fliq,s  Fdis,s
Size of Copper Tube, Refrigeration Load, and Pressure Drop
During the sizing of refrigerant piping, the system refrigeration load Qrl, in tons of refrigeration, is
a known value. Because the refrigerant velocity , the refrigerant mass flow rate
and the inner surface area of the copper tubing Ai  	Di
2/4. Here represents the
density of the refrigerant, and 	h is the enthalpy difference of the refrigerant leaving and entering
the evaporator. For a specific size of copper tubing, the relationship between the outside diameter D
and inside diameter Di is a fixed value. From the Darcey-Weisbach equation, the pressure drop in
the refrigerant piping can be expressed as
(11.1)
where gc  dimensional constant, 32.2 lbm
ft / lbf 
 s2
L  length of piping, ft (m)
Qrl  refrigeration load, Btu /min (W)
	h  enthalpy difference of refrigerant, Btu/ lb (kJ /kg)
	p  pressure loss, lbf/ft2 (kPa)
In Eq. (11.1), constant Cpipe  K [8/ (	2gc)]1/5. The friction factor of refrigerant f can be calculated
by the Colebrook equation and is covered in Chap. 18. If the absolute roughness of the copper tube
is taken as 0.000005 ft (0.0015 mm), f has only a minor influence on the outside diameter of the
copper tubing D. Constant K takes into account the difference between D and Di for a specific size
of copper tubing.
The size or outside diameter of the copper tubing D is
 Directly proportional to the length of the refrigeration pipe L to the power
 Directly proportional to refrigeration load Qrl to the power
 Inversely proportional to the density of the suction vapor or hot gas to the power, which is affected
by suction and condensing temperature, respectively
 Inversely proportional to the enthalpy difference 	h (which is closely related to the suction temperature
Tsuc and condensing temperature Tcon) to the power
 Inversely proportional to the maximum allowable pressure drop 	p to the power
Sizing Procedure
Refrigerant piping design and size should be determined as follows:
1. Make an optimum refrigerant piping layout, and measure the length of the piping.
2. Find the correction factor of the refrigeration load from Table 11.1, based on the actual suction
and condensing temperatures.
3. Estimate the equivalent length of refrigerant piping including the pipe fittings and accessories
Leq, ft (m). It is usually 1.5 to 5 times the measured straight length of the piping, depending on
the number of fittings and accessories.
4. Based on the corrected refrigeration load and Leq, determine the tentative diameter of the copper
tubing from refrigeration load versus equivalent length Qrl-Leq charts (see Figs. 11.10, 11.11,
and 11.12). All suction, discharge, and liquid line Qrl-Leq are plotted based on the data given in
ASHRAE Handbook 1998, Refrigeration.
15
25
15
25
15
 Cpipe fLQrl
2
(	h)2	p 1/5
D  K 8
	2gc
fLQrl
2
(	h)2	p 1/5
m?r  Qrl /	h
v  m?r /( Ai)
11.14 CHAPTER ELEVEN
5. Find the equivalent length of the pipe fittings and accessories listed in Tables 11.2 and 11.3,
based on the tentative diameter of the copper tubing. Recalculate the equivalent length of the
refrigeration piping including the pipe fittings and accessories.
6. Check the size of the copper tubing from the Qrl-Leq charts based on the calculated Leq. If it is
equal to the tentative value, the tentative diameter is the required diameter. If not, adjust the
tentative diameter.
7. Determine whether the size of the refrigeration pipe meets the oil entrainment requirement when
the refrigeration load is reduced to its minimum load at part-load operation. If it does not meet
the requirement, add a double riser.
Suction Line
The suction line is the most critical refrigerant line in the piping design. Vapor refrigerant and entrained
oil in the evaporator should return to the compressor free of liquid slugs.
Oil Trap and Piping Pitch. In the suction line, an oil trap should be installed at the bottom of a
single riser, as shown in Fig. 11.9a. The sensing bulb of the thermostatic expansion valve is often
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.15
TABLE 11.2 Pressure Drop of Valves Expressed in Equivalent Length of Straight Pipe, in Feet, for Refrigeration Systems
(Screwed, Welded, Flanged, and Flared Connections) Losses are for all valves in fully open position.
Nominal pipe or
tube size, in. Globe* 60°-Y 45°-Y Angle* Gate† Swing check‡ Lift check
17 8 6 6 0.6 5
18 9 7 7 0.7 6 Globe and
22 11 9 9 0.9 8 vertical lift
1 29 15 12 12 1.0 10 same as
1 38 20 15 15 1.5 14 globe
1 43 24 18 18 1.8 16 valves§
2 55 30 24 24 2.3 20
69 35 29 29 2.8 25
3 84 43 35 35 3.2 30
100 50 41 41 4.0 35
4 120 58 47 47 4.5 40
5 140 71 58 58 6 50
6 170 88 70 70 7 60
8 220 115 85 85 9 80
10 280 145 105 105 12 100 Angle lift
12 320 165 130 130 13 120 same as
14 360 185 155 155 15 135 angle
16 410 210 180 180 17 150 valve
18 460 240 200 200 19 165
20 520 275 235 235 22 200
24 610 320 265 265 25 240
*These losses do not apply to valves with needle point type seats.
†Regular and short pattern plug cock valves, when fully open, have same loss as gate valve. For valve losses of short pattern plug cocks above 6
in. check manufacturer.
‡ Losses also apply to the in-line, ball type check valve.
§For Y pattern globe lift check valve with seat approximately equal to the nominal pipe diameter, use values of 60°  Y valve for loss.
Source: ASHRAE Handbook 1990, Refrigeration Systems and Applications. Reprinted with permission.
312
212
12
14
34
12
38
installed between the coil outlet and the oil trap. This oil trap helps to prevent liquid accumulation
under the sensing bulb while the compressor is turned off. The accumulated liquid under the sensing
bulb causes abnormal operation of the expansion valve when the compressor is started again.
An oil trap consists of two elbows connected with a short stub of pipe.
If the evaporator is located above the compressor, the horizontal section of suction line should
be pitched toward the compressor at least in. each 10 ft (13 mm each 3 m) of pipe length.
Double Riser. When the compressor is located above the evaporator, if the compressor is
equipped with a capacity control or unloading mechanism to reduce the refrigeration load Qrl so
that the minimum Qrl in the vertical riser is less than the value listed in Tables 11.4 and 11.5, a double
riser is needed instead of a single vertical riser, as shown in Fig. 11.9b.
A double riser consists of a small riser, a large riser, and a oil trap that connects the lower ends
of the risers. During full-load operation, the vapor velocity in both risers is high enough to entrain
oil with the gaseous refrigerant, and there is no oil in the oil trap. However, when the Qrl and vapor
12
11.16 CHAPTER ELEVEN
TABLE 11.3 Fitting Losses in Equivalent Feet of Pipe
(Screwed, Welded, Flanged, Flared, and Brazed Connections)
Smooth bend elbows Smooth bend tees
Flow-
90° 90° Long- 90° 45° 45° 180° through
Straight-through flow
Std* radius† Street* Std* Street* Std* branch No Reduced Reduced
Nominal
reduction
pipe or
tube size,
in.
1.4 0.9 2.3 0.7 1.1 2.3 2.7 0.9 1.2 1.4
1.6 1.0 2.5 0.8 1.3 2.5 3.0 1.0 1.4 1.6
2.0 1.4 3.2 0.9 1.6 3.2 4.0 1.4 1.9 2.0
1 2.6 1.7 4.1 1.3 2.1 4.1 5.0 1.7 2.2 2.6
1 3.3 2.3 5.6 1.7 3.0 5.6 7.0 2.3 3.1 3.3
1 4.0 2.6 6.3 2.1 3.4 6.3 8.0 2.6 3.7 4.0
2 5.0 3.3 8.2 2.6 4.5 8.2 10.0 3.3 4.7 5.0
2 6.0 4.1 10.0 3.2 5.2 10.0 12.0 4.1 5.6 6.0
3 7.5 5.0 12.0 4.0 6.4 12.0 15.0 5.0 7.0 7.5
3 9.0 5.9 15.0 4.7 7.3 15.0 18.0 5.9 8.0 9.0
4 10.0 6.7 17.0 5.2 8.5 17.0 21.0 6.7 9.0 10.0
5 13.0 8.2 21.0 6.5 11.0 21.0 25.0 8.2 12.0 13.0
6 16.0 10.0 25.0 7.9 13.0 25.0 30.0 10.0 14.0 16.0
8 20.0 13.0 — 10.0 — 33.0 40.0 13.0 18.0 20.0
10 25.0 16.0 — 13.0 — 42.0 50.0 16.0 23.0 25.0
12 30.0 19.0 — 16.0 — 50.0 60.0 19.0 26.0 30.0
14 34.0 23.0 — 18.0 — 55.0 68.0 23.0 30.0 34.0
16 38.0 26.0 — 20.0 — 62.0 78.0 26.0 35.0 38.0
18 42.0 29.0 — 23.0 — 70.0 85.0 29.0 40.0 42.0
20 50.0 33.0 — 26.0 — 81.0 100.0 33.0 44.0 50.0
24 60.0 40.0 — 30.0 — 94.0 115.0 40.0 50.0 60.0
*R/D approximately equal to 1.
†R/D approximately equal to 1.5.
Source: ASHRAE Handbook 1998, Refrigeration Systems and Applications. Reprinted with permission.
12
12
12
14
34
12
38
12
14
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.17
velocity drop so that oil is no longer entrained in the large riser at part-load operation because
of cylinder unloading or on/off control, oil accumulates in the oil trap. This oil blocks the vapor
passage to the large riser. The vapor velocity in the small riser therefore increases to a velocity high
enough to carry the entrained oil to the compressor.
If the vertical riser has a rise greater than 25 ft (7.5 m), an additional oil trap for each 25-ft (7.5-
m) rise is required. Thus, the upper and lower sections are drained separately, and oil leaves the
traps in quantities that will not damage the compressor.
Maximum Pressure Drop. In a suction line using HCFC-22 as the refrigerant, a pressure drop
equivalent to 2°F (1.1°C) change in saturation temperature between 40 and 50°F (4.4 and 10°C) is
about 3 psi (21 kPa). When HFC-134a is used as the refrigerant, a 2°F (1.1°C) equivalent is about 2
psi (14 kPa); and if ammonia R-717 is used as the refrigerant, a 1°F (0.56°C) equivalent is about
1.5 psi (7 kPa).
Suction Line Sizing Chart. Figure 11.10 shows the Qrl-Leq suction line sizing chart. The abscissa
of this chart is the refrigeration load Qrl that the evaporated refrigerant in the evaporator can provide
at a corresponding mass flow rate, in tons of refrigeration, at 40°F (4.4°C) suction and 105°F
(40.6°C) condensing temperature. The ordinate is the equivalent pipe length Leq, ft (m). Table 11.4
indicates the minimum refrigeration load in the suction riser so that the suction vapor velocity vsuc
inside the copper tubing can still carry the oil.
Suction temperature is the saturation temperature of the refrigerant at the inlet of the compressor.
It is usually 2°F (1.1°C) lower than Tx in the evaporator or the DX coil, depending on the pressure
drop of the suction line. When the suction temperature Tsuc  40°F (4.4°C) and the condensing temperature
Tcon  105°F (40.6°C), the corrected refrigeration load Qrl,c, in tons, for sizing the suction
line can be found by dividing Qrl by correction factors Fsuc,s and Fsuc,c, as listed in Table 11.1, that is,
(11.2) Qrl,c 
Qrl
Fsuc,sFsuc,c
FIGURE 11.9 (a) Single riser, (b). double riser, and (c) additional oil trap.
where Fsuc,s  correction factor for refrigeration load in suction line when Tsuc  40°F (4.4°C)
Fsuc,c  correction factor for refrigeration load in suction line when Tcon  105°F (40.6°C)
From Eq. (9.24), refrigeration load Qrl, in tons, is given as
where  density of suction vapor, lb/ft3
v  volumetric efficiency
Here, represents the piston displacement, which is discussed later in this chapter. It is a
fixed value for a specific compressor at a certain speed. Volumetric efficiency v is affected slightly
by the compression ratio pdis /psuc. Therefore, correction factor Fsuc,s for the suction line is a
combined index that takes into account the effect of (hlv-hen) when Tsuc  40°F (4.4°C). If Tsuc 
40°F (4.4°C), then Fsuc,s  1, and Qrl,c Qrl. The correction factor Fsuc,c for the suction line
takes into account the effect of hlv - hen when Tcon  105°F. If Tcon  105°F, then Fsuc,c 1 and
Qrl,c  Qrl.
Suction Line Sizing. Consider the size of the suction line of a reciprocating refrigeration system
in a rooftop packaged unit with a refrigeration load of 40 tons (141 kW) at 45°F (7.2°C) suction
temperature and 115°F (46.1°C) condensing temperature with a subcooling of 10°F (5.6°C). There
are two refrigeration circuits and two compressors, each with a capacity of 20 tons (70 kW). Each
compressor can be unloaded to 50 percent of its full-load capacity.
The suction line has a total pipe length, including horizontal lines and vertical riser of 3 ft
(0.9 m), of 30 measured ft (9 m). There is also a manual shutoff angle valve and 10 short-radius
elbows.
V?
p
Qrl  m?r (hlv  hen)  V?p v(hlv  hen)
11.18 CHAPTER ELEVEN
FIGURE 11.10 Suction line sizing Qrl--Leq chart for HCFC-22 at 40°F suction temperature and 105°F condensing
temperature (type L copper tube, outside diameter, 2°F change in saturated temperature).
1. From Table 11.1, for a suction temperature of 45°F (7.2°C) and a condensing temperature of
115°F (46.1°C), Fsuc,s is 1.10 and Fsuc,c  0.93. The corrected refrigeration load Qrl for the suction
line is therefore 20 / (1.1  0.93)  19.6 tons (68.7 kW).
2. Assume an equivalent length of the suction line Leq of 90 ft (27 m). From Fig. 11.10, for a
refrigeration load of 19.7 tons and Leq  90 ft (27 m), the next-higher value of the tentative outside
diameter of the suction line is in. (54 mm), a nominal diameter of 2 in. (50 mm).
3. Recalculate the equivalent length, based on a copper tubing outside diameter of in. (54 mm).
The pressure drops of the refrigerant piping fittings and the accessories expressed as equivalent
length of pipe are listed in Table 11.2. The calculated total equivalent length Leq is
10 short radius elbows 10  5  50 ft (15 m)
1 angle valve 1  24  24 ft (7.2 m)
Measured length of pipe 30 ft (9 m)
Total 104 ft (31.2 m)
4. From the capacity-Leq chart, at Qrl  19.7 tons (68.7 kW) and Leq  104 ft (31.2 m), the diameter
of copper tubing is the same as the tentative diameter in. (54 mm). Consequently, in.
(54 mm) should be the size of the suction line.
218 218
218
218
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.19
TABLE 11.4 Minimum Refrigeration Load for Oil Entrainment up Suction Risers of Type L Copper Tubing, in Tons
Pipe OD, in.
0.500 0.625 0.750 0.875 1.123 1.375 1.625 2.125 2.625 3.125 3.625 4.125
Suction
Saturated gas Area, in.2
temperature, temperature,
Refrigerant °F °F 0.146 0.233 0.348 0.484 0.825 1.256 1.780 3.094 4.770 6.812 9.213 11.970
HCFC-22 40.0 30.0 0.067 0.119 0.197 0.298 0.580 0.981 1.52 3.03 5.20 8.12 11.8 16.4
10.0 0.065 0.117 0.194 0.292 0.570 0.963 1.49 2.97 5.11 7.97 11.6 16.1
10.0 0.066 0.118 0.195 0.295 0.575 0.972 1.50 3.00 5.15 8.04 11.7 16.3
20.0 10.0 0.087 0.156 0.258 0.389 0.758 1.28 1.98 3.96 6.80 10.6 15.5 21.5
10.0 0.085 0.153 0.253 0.362 0.744 1.26 1.95 3.88 6.67 10.4 15.2 21.1
30.0 0.086 0.154 0.254 0.383 0.747 1.26 1.95 3.90 6.69 10.4 15.2 21.1
0.0 10.0 0.111 0.199 0.328 0.496 0.986 1.63 2.53 5.04 8.66 13.5 19.7 27.4
30.0 0.108 0.194 0.320 0.484 0.942 1.59 2.46 4.92 8.45 13.2 19.2 26.7
50.0 0.109 0.195 0.322 0.486 0.946 1.60 2.47 4.94 8.48 13.2 19.3 26.8
20.0 30.0 0.136 0.244 0.403 0.608 1.18 2.00 3.10 6.18 10.6 16.6 24.2 33.5
50.0 0.135 0.242 0.399 0.603 1.17 1.99 3.07 6.13 10.5 16.4 24.0 33.3
70.0 0.135 0.242 0.400 0.605 1.18 1.99 3.08 6.15 10.6 16.5 24.0 33.3
40.0 50.0 0.167 0.300 0.495 0.748 1.46 2.46 3.81 7.60 13.1 20.4 29.7 41.3
70.0 0.165 0.296 0.488 0.737 1.44 2.43 3.75 7.49 12.9 20.1 29.3 40.7
90.0 0.165 0.296 0.488 0.738 1.44 2.43 3.76 7.50 12.9 20.1 29.3 40.7
Note: The capacity in tons is based on 90°F liquid temperature and superheat as indicated by the listed temperatures. For other liquid line temperatures,
use correction factors in the table below.
Liquid temperature, °F
Refrigerant 50 60 70 80 100 110 120 130 140
HCFC-22 1.17 1.14 1.10 1.06 0.98 0.94 0.89 0.85 0.80
Source: ASHRAE Handbook 1998, Refrigeration Systems and Applications.
Reprinted with permission.
5. The minimum capacity of compressor in each refrigeration circuit is 0.5  20.7  10.4 tons
(36.4 kW). From Table 11.4, when the liquid temperature is 115  10  105°F (40.6°C), the
correction factor at 105°F is 0.96. The allowable minimum refrigeration load for a suction riser
with entrained oil is 7.9  0.96  7.6 tons, which is smaller than 10.4 tons. Therefore, a double
riser is not needed.
Discharge Line (Hot-Gas Line)
The design and sizing of the discharge line (hot-gas line) are similar to those of the suction line. Oil
return and optimum pressure drop are the primary considerations.
In the discharge line, oil does not mix with gaseous refrigerant. It is entrained in the gaseous refrigerant
at the discharge outlet of the compressor as oil mist. The horizontal pipe run should be
pitched away from the compressor at a pitch not less than in. for each 10 ft (13 mm for each 3 m)
of pipe run. The minimum refrigeration loads for hot-gas risers to carry entrained oil to the condenser
are listed in Table 11.5. There would be an oil trap at the bottom of the vertical riser. When the compressor
is unloaded to part-load operation, if the minimum Qrl is less than that listed in Table 11.5, a
double riser should be used. An additional oil trap should be added for each 25 ft (7.6 m ) of pipe rise.
In the discharge line, the saturated temperature difference corresponding to the maximum pressure
drop is 1°F (0.56°C) for halocarbon refrigerants. The maximum pressure drop is about 3 psi
(21 kPa) for both HCFC-22 and HFC-134a. For ammonia R-717, a 1°F (0.56°C) saturated temperature
difference is equivalent to about 4 psi (27.6 kPa).
The procedure for sizing of the discharge line is similar to that for the suction line. Figure 11.11
shows the discharge line Qrl-Leq chart for HCFC-22 at 40°F (4.4°C) suction temperature and 105°F
(40.6°C) condensing temperature with a pressure drop corresponding to a 1°F (0.56°C) change
in saturated temperature difference. If the suction temperature Tsuc is not 40°F (4.4°C) and the
12
11.20 CHAPTER ELEVEN
TABLE 11.5 Minimum Refrigeration Load for Oil Entrainment up Hot-Gas Risers of Type L Copper Tubing, in Tons
Pipe OD, in.
0.500 0.625 0.750 0.875 1.123 1.375 1.625 2.125 2.625 3.125 3.625 4.125
Suction
Saturated gas Area, in.2
temperature, temperature,
Refrigerant °F °F 0.146 0.233 0.348 0.484 0.825 1.256 1.780 3.094 4.770 6.812 9.213 11.970
HCFC-22 80.0 110.0 0.235 0.421 0.695 1.05 2.03 3.46 5.35 10.7 18.3 28.6 41.8 57.9
140.0 0.223 0.399 0.659 0.996 1.94 3.28 5.07 10.1 17.4 27.1 39.6 54.9
170.0 0.215 0.385 0.635 0.960 1.87 3.16 4.89 9.76 16.8 26.2 38.2 52.9
90.0 120.0 0.242 0.433 0.716 1.06 2.11 3.56 5.50 11.0 18.9 29.5 43.0 59.6
150.0 0.226 0.406 0.671 1.01 1.97 3.34 5.16 10.3 17.7 27.6 40.3 55.9
180.0 0.216 0.387 0.540 0.956 1.88 3.18 4.92 9.82 16.9 26.3 38.4 53.3
100.0 130.0 0.247 0.442 0.730 1.10 2.15 3.83 5.62 11.2 19.3 30.1 43.9 60.8
160.0 0.231 0.414 0.884 1.03 2.01 3.40 5.26 10.5 18.0 28.2 41.1 57.0
190.0 0.220 0.394 0.650 0.982 1.91 3.24 3.00 9.96 17.2 26.8 39.1 54.2
110.0 140.0 0.251 0.451 0.744 1.12 2.19 3.70 5.73 11.4 19.6 30.6 44.7 62.0
170.0 0.235 0.421 0.693 1.05 2.05 3.46 3.35 10.7 18.3 28.6 41.8 57.9
200.0 0.222 0.399 0.658 0.994 1.94 3.28 5.06 10.1 17.4 27.1 39.5 54.8
120.0 150.0 0.257 0.460 0.760 1.15 2.24 3.78 5.85 11.7 20.0 31.3 45.7 63.3
180.0 0.239 0.428 0.707 1.07 2.08 3.51 5.44 10.8 18.6 29.1 42.4 58.9
210.0 0.225 0.404 0.666 1.01 1.96 3.31 5.12 10.2 17.6 27.4 40.0 55.5
Notes: The capacity in tons is based on a saturated suction temperature of 20°F with 15°F superheat at the indicated saturated condensing temperature
with 15°F subcooling. For other saturated suction temperatures with 15°F superheat, use the following correction factors:
Saturated suction temperature, °F 40 20 0 40
Correction factor 0.88 0.95 0.96 1.04
Source: ASHRAE Handbook 1998, Refrigeration Systems and Applications. Reprinted with permission.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.21
condensing temperature Tcon is not 105°F (40.6°C), the corrected refrigeration load is
(11.3)
Correction factor Fdis, s takes into account the effect of hlv-hen on in the discharge line when
Tsuc  40°F (4.4°C). If Tsuc  40°F (4.4°C), then Fdis, s  1 and qrl,c qrl.
Correction factor Fdis,c takes into account the combined effect of the density of the hot gas and
enthalpy difference hlv  hen on when Tcon  105°F (40.6°C). If Tcon  105°F (40.6°C), then
Fdis,c 1 and Qrl,c Qrl. Table 11.1 lists the correction factors Fdis,s and Fdis,c for HCFC-22 for the
discharge line.
Liquid Line
Because oil is partially miscible with liquid refrigerant such as HCFC-22, HFC-134a is miscible
with synthetic oil, and oil is separated from ammonia after the compressor, oil return is not a problem
in the liquid line. Therefore, the primary considerations in liquid line design and sizing are the
optimum pressure drop and the prevention of flashing of liquid refrigerant into vapor before the expansion
valve.
Current practice is to limit the maximum pressure drop in the liquid line corresponding to a
change of 1°F saturated temperature. For HCFC-22 it is about 3 psi (20.7 kPa), for HFC-134a it is
about 2.5 psi (17.2 kPa), and for ammonia R-717 it is about 3.5 psi (24 kPa).
To prevent the flashing of liquid to vapor, liquid refrigerant must be subcooled before it is discharged
to the liquid line. Recently designed air-cooled condensers for reciprocating refrigeration
systems can provide a subcooling of 10 to 20°F (5.6 to 11.1°C). The liquid refrigerant should be in
a subcooled state when it enters the expansion valve.
m?r
m? r
Qrl,c 
Qrl
Fdis,sFdis,c
FIGURE 11.11 Discharge line sizing Qrl--Leq chart for HCFC-22 at 40°F suction temperature and 105°F condensing
temperature (type L copper tube, outside diameter, 1°F change in saturated temperature).
If a receiver is installed after the condenser, the liquid line between the condenser and receiver
should be sized so that liquid flows from the condenser to the receiver and gas flows in the opposite
direction. This section of liquid line is usually sized based on a liquid velocity of 100 fpm (0.5 m/ s).
In the design and planning of the liquid line layout, the following factors must be considered:
 If a receiver is placed after the subcooling section of the condenser, subcooling will occur at the
outlet of the receiver only when a surge-type liquid receiver is used or if an additinal subcooling
setup is added after the receiver.
 Static pressure due to the liquid refrigerant column must be taken into account. At 105°F
(40.6°C), the density of liquid HCFC-22 is 70.3 lb/ ft3 (1125 kg/m3). For each foot of liquid column,
the increase of static pressure is therefore 	ps  70.3/144  0.488 psi (3.4 kPa), or roughly
0.5 psi (3.5 kPa) for each foot of liquid column. At the top of a 10-ft (3-m) riser, the pressure is
about 10  0.5  5 psi (35 kPa) less than at the bottom of the riser.
Figure 11.12 is a Qrl-Leq chart for a liquid line with HCFC-22 at 40°F (4.4°C) suction temperature
and 105°F (40.6°C) condensing temperature with a pressure drop corresponding to a change of 1°F
(0.56°C) saturation temperature.
For suction temperature other than 40°F (4.4°C) and condensing temperature other than 105°F
(40.6°C), the corrected refrigeration load is calculated as
(11.4)
Correction factor Fliq,s takes into account the effect of enthalpy difference hlv-hen on . Correction
factor Fliq,c also takes into account the effect of enthalpy difference on . Its values are listed in
Table 11.1. When Tcon 105°F (40.6°C), Fliq,c  1, and Qrl,c Qrl.
m?r
m?r
Qrl,c 
Qrl
Fliq, s Fliq,c
11.22 CHAPTER ELEVEN
FIGURE 11.12 Liquid line sizing Qrl--Leq chart for HCFC-22 at 40°F suction temperature and 105°F condensing
temperature (type L copper tube, outside diameter, 1°F change in saturated temperature).
Consider the reciprocating refrigeration system described in the discussion of suction line
sizing. The subcooling at the condenser outlet is 10°F (5.6°C). There are three manual shutoff angle
valves, one solenoid valve (using the pressure drop as globe valve), one sight glass of 5-ft (1.5-m)
equivalent length, and 1 filter dryer of 150-ft (45-m) equivalent length according to the manufacturer’s
data in the liquid line. Also a 5-ft (1.5-m) vertical riser is connected to the outlet of the
condenser, and the measured length of the liquid line is 40 ft (12 m). Sizing of the liquid line can be
accomplished as follows:
1. From Table 11.1, the correction factor for the refrigeration load in the liquid line at a suction
temperature of 45°F is Fliq,s  Fdis,s  1.01. Condensing temperature is 115°F and with a subcooling
of 10°F, Fliq,c  1.0. The corrected refrigeration load is then
2. Assume an equivalent length Leq  250 ft (75 m).
3. From Fig. 11.12, the Qrl-Leq chart for the liquid line, the next-higher tentative diameter of
copper tubing for the liquid line at Qrl,c  19.8 tons and Leq  250 ft is in. (29 mm).
4. Calculate the equivalent length as follows:
3 angle valves 3  12  36 ft (10.8 m)
1 solenoid valve 1  29  29 ft (8.7 m)
1 sight glass 1  5  5 ft (1.5 m)
1 filter dryer 1  150  150 ft (45 m)
Measured pipe length 40 ft (12 m)
Total 260 ft (7 m)
5. At Qrl  19.8 tons and Leq  260 ft, the next-higher diameter of the liquid line is still in.
(29 mm), the same as the tentative diameter. This is the required diameter of the liquid line.
6. For a condensing temperature of 115°F (46.1°C) and a subcooling temperature of 10°F
(5.6°C), the saturated pressure at 115  10  105°F is 210.8 psig (1453 kPag). When the pressure
of the refrigerant in the liquid line is greater than 210.8 psig (1453 kPag), flashing will not occur.
The lowest pressure occurs at the inlet of the expansion valve. This pressure can be calculated as
follows:
Condensing pressure, corresponding to 115°F 242.7 psig (1673 kPag)
Pressure drop of the liquid line 3 psi (20.6 kPa)
Pressure drop from riser 5  0.5  2.5 psi (17.2 kPa)
Liquid pressure at inlet of expansion valve 237.2 psig (1636 kPag)
The pressure at the inlet of the expansion valve is greater than 210.8 psig (1453 kPag), so liquid refrigerant
will not be flashed in the liquid line.
Parallel Connections. When two or three compressors are connected in parallel, their suction
lines should be arranged so that the suction vapor is evenly distributed to the compressors. A gas
equalizer should be connected to the crankcase of each compressor to equalize suction pressure,
and an oil equalizer should be used to equalize the oil level in all compressors.
When two or three condensers are connected in parallel, hot gas should be evenly distributed to
the condensers. If a liquid receiver is used, all condensers should be connected to a single receiver
with an equalizer between the hot-gas supply and the top of the receiver. Refer to ASHRAE Handbooks
and manufacturers’ manuals for details.
118
118
Qrl,c 
Qrl
Fliq,s Fliq,c

20
1.01  1.0
 19.8 tons (69.6 kW)
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.23
11.5 CAPACITY AND SAFETY CONTROLS
Capacity Control
Refrigeration systems for comfort air conditioning or industrial applications need capacity control
to meet refrigeration load variations during operation. In vapor compression refrigeration systems,
control of the refrigerant flowing through the compressor is widely used as the primary refrigeration
capacity control. For reciprocating refrigeration systems using DX coils as evaporators, row or
intertwined face control of DX coil capacity is often employed in conjunction with compressor capacity
control for better performance. In medium-size and large systems, two separate refrigerant
circuits are often used so that refrigerant flow in one can be cut off when the refrigeration load
drops below 50 percent of design load.
There are four basic methods of controlling the capacity of reciprocating refrigeration systems
by reducing the refrigerant flow rate during part-load operation.
On/Off Control. On/off or start /stop control is widely used to control the refrigerant flow and
the capacity of the reciprocating compressors in residential air conditioners. For reciprocating
refrigeration systems equipped with multiple compressors, as used in supermarkets, on/off control
is often the suitable choice. On/off control is simple and inexpensive, but it produces a greater
variation in the space air temperature and relative humidity maintained by the refrigeration system
than other types of control do. In addition, cyclic operation may lead to a start /stop cyclic loss for
the evaporator and condenser, which are located indoors and outdoors separately. Cyclic loss is
discussed in the next chapter. On/off control also results in increased wear and tear on compressor
components and shortens service life.
Cylinder Unloader. For multiple-cylinder reciprocating compressors, a cylinder unloader is often
used for capacity control. Cylinder unloaders reduce the capacity of compressors by bypassing the
compressed gas to the suction chamber, blocking the suction or discharge valve, or closing the
suction valve late or early. Recently developed cylinder unloaders include a solenoid valve that
unloads the cylinder(s) in response to any one of the following:
 Variation of discharge air temperature in a packaged unit
 Change in return chilled water temperature in a water chiller
 Changes in suction pressure in the evaporator
Figure 11.13 shows a typical cylinder unloader. When a temperature sensor in a packaged unit
detects a drop in the discharge air temperature, the controller energizes the solenoid, draws the
plunger away from the connecting port, allows the compressed gas to flow through a side passage,
pushes the piston and valve stem of the unloader away from the solenoid, and opens the unloader
valve. The compressed gas from the cylinder then flows to the suction side of the compressor. The
two cylinders under the same head space are unloaded.
When the discharge air temperature rises, the plunger of the solenoid covers the connecting port.
The pressure of the compressed gas exerted on the disk closes the unloader valve. Consequently, the
increased pressure in the space above the cylinder opens the check valve and permits the discharge
gas to enter the discharge line. The two cylinders are now loaded.
A cylinder unloader is a form of step control. Typically, the capacity of an eight-cylinder reciprocating
compressor can be controlled in four steps: 100 percent, 75 percent, 50 percent, and 25
percent.
Speed Modulation. Recent developments in solid-state electronics may allow adjustablefrequency
ac invertor-driven variable-speed motors to be used for capacity control in reciprocating
compressors. Commercial variable-speed air conditioners with a capacity of few tons of refrigeration
were available in Japan in the 1980s.
11.24 CHAPTER ELEVEN
FIGURE 11.13 Cylinder unloader: (a) unloaded cylinders; (b) loaded cylinders.
Invertor-driven variable-speed reciprocating compressors have the advantage of: lower start /stop
or cycling losses, stable temperature control, and better system performance. They also have several
drawbacks, such as cooling of the hermetic motor at low refrigerant flow rates, noise at high rotation
speeds, and vibration at low speeds. In addition, they require a specially designed expansion
valve for low-speed operation because a condensing pressure below a certain limit may affect the
normal operation of the expansion valve.
Hot-Gas Bypass Control. In hot-gas bypass control, according to the sensed suction pressure, a
certain amount of hot gas that balances the reduction in refrigeration load bypasses the condenser
and the expansion valve, and enters the low-pressure side directly. In a DX reciprocating refrigeration
system, bypassed hot gas is often introduced between the expansion valve and the liquid refrigerant
distributor, as shown in Fig. 11.14.
Although hot-gas bypass control permits modulation of system capacity to a very low percentage,
this method does not save energy at reduced refrigeration load. It is often used as a safety
device to protect the evaporator or DX coil from frosting or excessive cycling of the compressor
during on/off control when the refrigeration load is less than 20 to 25 percent of design capacity, or
for industrial applications where strict temperature or humidity control is essential.
ASHRAE/IESNA Standard 90.1-1999 specifies that cooling systems shall not use hot-gas bypass
or other evaporator pressure control systems except for (1) systems designed and equipped
with multiple steps of unloading or continuous capacity modulation or (2) packaged systems with
cooling capacities not grearer than 90,000 Btu/h (26 kW). When the cooling system rated capacity
is less than or equal to 240,000 Btu /h (70 kW), the maximum hot-gas capacity is 50 percent of the
total capacity. When the cooling system rated capacity is greater than 240,000 Btu/h (70 kW), the
maximum hot-gas capacity is 25 percent of total capacity.
Safety Controls
Safety controls are used to shut down the refrigeration system during a malfunction and to prevent
damage of system components. In reciprocating refrigeration systems, safety controls include lowpressure,
high-pressure, low-temperature, frost, oil pressure failure, and motor overload controls.
Low-Pressure and High-Pressure Controls. The purpose of low-pressure control is to stop the
compressor when the suction pressure drops below a preset value or when the refrigerant flow
rate is too low to cool the compressor motor. Figure 11.15a shows a typical low-pressure control
11.26 CHAPTER ELEVEN
FIGURE 11.14 Hot-gas bypass control for a DX reciprocating system.
mechanism. When the suction pressure falls below a certain limit, the spring pushes the blade
downward, opens the motor circuit, and stops the compressor. When the suction pressure increases,
the bellows expand, thus closing the contact of the motor circuit and restarting the compressor. The
two adjusting screws are used to set the cut-out and cut-in pressures. Cut-out pressure is the pressure
at which the compressor stops, and cut-in pressure is the pressure at which the compressor
starts again.
The purpose of high-pressure control is to stop the compressor when the discharge pressure of
the hot gas approaches a dangerous level. Figure 11.15b shows a typical high-pressure control
mechanism. If the discharge pressure reaches a certain limit, the bellows expand so that the blade
opens the motor circuit contact and the compressor stops. When the discharge pressure drops to a
safe level, the bellows contract and close the contact, and the compressor starts again. As in a lowpressure
control, two adjusting screws are used to set the cut-out and cut-in pressures.
In small refrigeration systems, low-pressure and high-pressure controls are often combined to
form a dual-pressure control.
Low-Temperature Control. The purpose of a low-temperature control is to prevent the temperature
of chilled water in the liquid cooler from falling below a certain limit, to protect the water from
freezing. Freezing water damages the liquid cooler. Figure 11.16 shows a typical low-temperature
control system. The sensing bulb senses the chilled water leaving temperature. When the temperature
falls below a limit, contraction of the bellows opens the motor circuit and stops the compressor.
When the chilled water temperature rises above the limit, the expansion of the bellows closes the
circuit and restarts the compressor.
Frost Control. The purpose of frost control is to prevent the formation of frost on the outer surface
of the DX coil. Usually, a temperature sensor is used to sense the outer surface temperature of
the DX coil. When the temperature drops to 32°F (0°C), the controller actuates a relay, which opens
the circuit and stops the compressor.
Oil Pressure Failure Control. The purpose of the oil pressure failure control is to stop the compressor
when the oil pressure drops below a certain limit and fails to lubricate the main bearings
and other components. Figure 11.17 shows a typical oil pressure failure control mechanism. When
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.27
FIGURE 11.15 (a) Low pressure control and (b) high pressure control.
the compressor starts, points a and b of the differential pressure switch and points c and d of the
timer switch are in contact with each other. The differential pressure switch is affected by the differential
pressure between oil pump discharge pressure and crankcase suction pressure. If the pressure
of oil discharged from the oil pump does not reach a predetermined level within a certain time interval,
the low oil pressure keeps the differential pressure switch a and b in a close contact position.
11.28 CHAPTER ELEVEN
FIGURE 11.16 Low-temperature control.
FIGURE 11.17 Oil pressure failure control.
The heater is energized, which causes the bimetal plate to deviate from the vertical position. The
bimetal plate no longer keeps the timer switch d in contact with c, and the compressor stops.
If the oil pressure reaches the preset value within the predetermined time interval, a does not
make contact with b. The breaking in contact of a and b deenergizes the heater before the bimetal
plate starts to bend. The bimetal plate still holds the timer switch d in contact with c, and the compressor
continues to operate.
If the oil pressure drops below the preset value and does not rise again within a certain time interval,
the energized heater causes the bimetal plate to bend, opens the contact of the timer switch,
and stops the compressor.
Motor Overload Control. The purpose of motor overload control is to protect the motor from
dangerous overheating caused by a long period of overloading or failure of the compressor to start.
Thermal protectors with bimetal or thermistor sensors are installed inside the motor or mounted on
the outer surface of the motor shell. These sensors sense the temperature or both the temperature
and electric current supplied to the motor windings. If dangerous overheating and overloading
occur, a controller breaks the circuit.
Recently developed microprocessor-operated DDC system controllers combine all safety
control mechanisms with other control systems of a reciprocating refrigeration system in one
package. The DDC system controller senses, analyzes, and operates the safety control systems in
coordination with other systems with far more functions and more sophisticated comprehensive
controls.
11.6 PERFORMANCE OF RECIPROCATING COMPRESSOR,
EVAPORATOR, AND CONDENSER
Performance of Reciprocating Compressor
There are two important parameters in the performance analysis of a reciprocating compressor:
refrigeration capacity Qrc, Btu /h or tons (W or kW) of refrigeration (as described in Sec. 9.10);
and power input Pin, hp (kW); or energy efficiency ratio EER, Btu/h (W) (as defined in
Sec. 9.20). Because Qrc and Pin are always affected by suction pressure psuc and discharge
pressure pcon (both in psig), it is convenient to relate the performance of the reciprocating compressor
to suction temperature Tsuc (in °F) and to the condensing temperature Tcon. The saturated
pressure at Tev is equal to pev. Suction pressure is approximately equal to the saturated pressure
of Tev  2, and discharge pressure is approximately equal to the saturated pressure of Tcon  1,
both in °F (°C).
Consider an accessible hermetic compressor with the following operating and construction characteristics:
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.29
HCFC-22
8
2.65 in. (67.3 mm)
2.25 in. (57.2 mm)
1750
40 °F (4.4°C)
115°F (46.1°C)
10 °F (5.6°C)
10 °F (5.6°C)
Refrigerant
Number of cylinders
Bore of cylinder
Stroke of piston
Revolutions per minute
Suction temperature
Condensing temperature
Subcooling
Superheating
Refrigeration Capacity. In reciprocating compressors, the piston only performs compression
work during the upward stroke in a complete rotation of the crankshaft. Piston displacement ,
cfm [m3/(60 s)], can be calculated as
(11.5)
where Dcy  diameter of the cylinder, ft (m)
Lst  stroke of cylinder, ft (m)
N  number of cylinders
n  rotating speed of compressor, rpm
From Eq. (9.72), the mass flow rate of the refrigerant , so from Eq. (9.25), the refrigeration
capacity Qrc, Btu /h (W), of a reciprocating compressor can be calculated as
(11.6)
where suc  density of suction vapor, lb/ft3 (kg/m3)
h1  enthalpy of superheated vapor leaving evaporator, Btu / lb (J /kg)
h4  enthalpy of subcooled liquid entering evaporator, Btu / lb (J /kg)
In the eight-cylinder reciprocating compressor using HCFC-22 as refrigerant described above, volumetric
efficiency v depends mainly on compression ratio pdis /psuc. The smaller the pdis /psuc and the
lower the reexpansion of clearance volume, cylinder heating, and leakages through valves and pistons,
the greater the v. Figure 11.18a shows the v versus pdis /psuc curve for this compressor. For a
suction temperature of 40°F (4.4°C), the absolute saturated pressure is 68.5  14.7  83.2 psia
(574 kPa abs.). For a condensing temperature of 115°F (46.1°C), the absolute condensing pressure
is 243  14.7  257.7 psia (1777 kPa abs.). The compression ratio is therefore pdis /psuc  (257.7
 3)/83.2  3.13. From Fig. 11.18a, v  0.81.
The upper part of Fig. 11.18a shows the refrigeration capacity versus suction temperature Qrc-
Tsuc curves. If Tcon remains constant and Tsuc increases, the refrigeration capacity rises accordingly
for the following reasons:
 For a specific compressor rotating at a specific speed, the piston displacement is constant.
 Volumetric efficiency increases as Tsuc rises.
 Density of the suction vapor refrigerant suc increases as Tsuc increases because of higher saturated
pressure.
 For a constant Tcon and a fixed amount of subcooling, enthalpy h4 is also constant. Enthalpy of
superheated refrigerant leaving evaporator at a fixed degree of superheat h1 increases nearly
linearly between 40 and 80°F (4.4 and 26.7°C) as Tsuc rises. This also causes an increase of
refrigeration effect Qrf  h1  h4. These factors result in a very slight upward bend in the
Qrc - Tsuc curve.
If the suction temperature Tsuc is at a constant value, an increase in condensing temperature Tcon results
in a decrease in v and a decrease in Qrf because of a smaller h4. Consequently, Qrc decreases
as Tcon rises, and Qrc increases as Tcon falls, as shown in Fig. 11.18.
Power Input. The refrigeration compressor consumes most of the energy input to a refrigeration
system. In a single-stage reciprocating refrigeration system, power input Pin, kW, to the compressor
can be calculated from Eqs. (9.71) and (9.72):
(11.7) Pin 
V?
p v suc(h2  h1)
56.85 com
V?
p
Qrc  60V?p v suc(h1  h4)
m?r  V?p v suc
V?
p 
	Dcy
2
LstNn
4
V?
p
11.30 CHAPTER ELEVEN
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.31
FIGURE 11.18 Cooling capacity, power input, and volumetric efficiency of a typical reciprocating compressor; (a)
v–pdis /psuc curve; (b) Qrc-Tsuc and Pin-Tsuc curves.
where h2  enthalpy of hot gas discharged from compressor if compression process is isentropic,
Btu/ lb
com  compressor efficiency, as defined in Sec. 9.20
If Tcon remains constant and suction temperature Tsuc increases, both v and suc rise and h2  h1
drops as Tsuc increases. Compressor efficiency com also increases within a rather narrow range
(0.75 to 0.80) when Tsuc changes from 5 to 50°F (2.8 to 10°C), and the com-Tsuc curve becomes
nearly horizontal as Tsuc exceeds 35°F (1.7°C). Because the increase in v and suc has a greater
effect on Pin than the drop in h2 -h1 and the change of com, the Pin-Tsuc curve is slightly convex,
indicating that Pin increases as Tsuc rises, as shown in Fig. 11.18b. The Pin-Tsuc curves are signifi-
cantly flatter than Qrc-Tsuc curves.
If Tsuc remains constant and Tcon increases, v falls and h2  h1 becomes larger. The effect of the
increase in h2  h1 is greater than that of the fall of v. The result is a higher Pin.
Performance of Evaporator
Cooling or refrigeration capacity Qrc, Btu /h (W), is the most important parameter of an evaporator.
For both DX coils and flooded liquid cooler evaporators, it is more convenient to use the following
equation for performance analysis:
(11.8)
where A  evaporating surface area, ft2 (m2)
U  overall heat-transfer coefficient, Btu/h
 ft2 
°F (W/m2 
°C)
	Tm  log-mean temperature difference, °F (°C)
For a DX coil or a flooded liquid cooler, surface area A is a constant and is always based on the total
outer surface area Ao including the tubes and fins. For a liquid cooler with direct-expansion
refrigerant feed, the surface area is the inner surface area of tubes plus inserts or enhanced surfaces.
Overall heat-transfer coefficient U depends mainly on fluid velocity, heat flux and the configuration
at the evaporating surface, a dry or wet surface on the air side, and the construction of the
DX coil or liquid cooler. In DX coils, the air velocity of a VAV system changes only when the system
load varies. In a direct-expansion liquid cooler, chilled water velocity remains constant during
operation.
The log-mean temperature difference 	Tm is the temperature difference between the air temperature
and the temperature of evaporating refrigerant inside the DX coil, or between the temperature
of chilled water in the shell and the refrigerant in the liquid cooler. If the air temperature and its
enthalpy entering the DX coil or the temperature of chilled water entering the liquid cooler both remain
constant, then an increase of suction temperature Tsuc reduces 	Tm and a fall of Tsuc increases
	Tm. The cooling capacity of a DX coil or liquid cooler is then inversely proportional to Tsuc.
Medium-size or large reciprocating refrigeration systems usually have two DX coils or two liquid
coolers. Figure 11.19 compares the Qrc-Tsuc curves of two selected dual DX coils and two selected
dual direct-expansion liquid coolers. As Qrc is linearly proportional to Tsuc, the Qrc-Tsuc curves are
straight lines. The selected larger DX coil has a total face area of 2  21 ft2 (3.9 m2), and the smaller
coil has a total face area of 2  15 ft2 (2.8 m2). The selected larger direct-expansion liquid cooler has
a total tube area of 172 ft2 (16 m2), and the smaller one has a total tube area of 90 ft2 (8.4 m2).
When the entering air is at 80°F (26.7°C) dry-bulb and 67°F (19.4°C) wet-bulb temperature,
each 1 ft2 (0.093 m2) of face area of the DX coils has a total cooling capacity of 18,000 Btu/h
(5274 W) at 40°F (4.4°C) suction temperature, 115°F (46.1°C) condensing temperature, and 600
fpm (3 m/ s) of air velocity.
Because air is directly cooled in the DX coil, whereas chilled water is cooled before the air in a
liquid cooler, the suction and evaporating temperatures in liquid coolers are lower than those in DX
coils. Because the 	Tm of a liquid cooler is far smaller than that of a DX coil, the Qrc-Tsuc curves of
Qrc  AU 	Tm
11.32 CHAPTER ELEVEN
direct-expansion liquid coolers have a much steeper negative slope than those of DX coils. For two
evaporators at the same operating condition and Qrc, the larger evaporator, which has a greater evaporating
surface area, has a higher Tsuc than the smaller one, as shown in Fig. 11.19.
Performance of Condenser
As for an evaporator, the performance of a condenser can be expressed by the following formula:
(11.9)
where Qrej  total heat rejection at condenser, Btu /h (W). As described in Sec. 10.5, Qrej  Frej Qrc.
Here Frej is the heat rejection factor.
In two condensers of the same type with different condensing surface areas, if their operating
conditions are the same, the condensing temperature Tcon, in °F (°C), of the condenser with a larger
condensing surface is lower than that of the smaller condenser. For a specific condenser, although
the heat flux at the condensing surface and fluid velocity affect the overall heat-transfer coefficient
U value, the log-mean temperature difference 	Tm is still the dominating factor that influences the
performance of the condenser.
Figure 11.20 shows the Qrej-Tcon curves for condensers. Because Qrej is linearly proportional to
Tcon, Qrej-Tcon curves are straight lines. Air-cooled condensers have a greater 	Tm than water-cooled
condensers, so the slope of the Qrej-Tcon curves for air-cooled condensers is smaller than that of
water-cooled condensers.
Qrej  AU 	Tm
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.33
FIGURE 11.19 Performance curves Qrc-Tsuc for DX coils and directexpansion
liquid coolers.
11.7 SYSTEM BALANCE
Because the refrigeration cycle of a reciprocating refrigerating system is a closed cycle, if the system
is operated in a continuous and steady state (i.e., in an equilibrium state), according to the principle
of continuity of mass and energy balance, the mass flow rates of refrigerant flowing through
the evaporator, compressor, condenser, and expansion or float valve must all be equal. Also the total
amount of energy supplied to the refrigeration system must be approximately equal to the total energy
rejected from the system.
A continuous and steady state means that the flow is continuous, and the properties of the refrigerant
at any point in the refrigeration system do not vary over time. Therefore, during the design of
a refrigeration system, the system components selected should have equal or approximately equal
mass flow rates of refrigerant at stable operation conditions.
Of the four main system components (compressor, evaporator, condenser, and expansion valve)
in a reciprocating refrigeration system, the expansion valve adjusts its mass flow rate of refrigerant
according to the degree of superheat of the vapor leaving the evaporator in the DX coil and the DX
11.34 CHAPTER ELEVEN
FIGURE 11.20 Performance curves Qrej-Tcon for condensers.
liquid cooler. Although the displacement of a selected compressor is a constant, the compressor
can adjust its mass flow rates within a certain range because of the increase or decrease of the density
of suction vapor suc and hot gas dis as Tev and Tcon vary.
There are two kinds of imbalance problems: imbalance of capacities of selected system components
and imbalance caused by system load deviations from design conditions at part-load operation.
Part-load operation is discussed later in this chapter.
Balance of Capacities of Selected Components
If the imbalance in required capacities of the selected system components is small, the refrigeration
system adjusts its evaporating and condensing temperatures to reach a system balance slightly
different from the required design operating parameters. If the capacities of the system components
are seriously imbalanced, then either the refrigeration system will operate at parameters very
different from the required conditions or the system will fail to operate or even be damaged by
unacceptable operating parameters such as heavy overloading, low evaporating temperature, or high
pressure cutout.
Consider a built-up rooftop reciprocating packaged system. The design refrigeration load Qrc is
60 tons ( 211 kW) at a suction temperature Tsuc  40°F (4.4°C) and a condensing temperature
Tcon  115°F (46.1°C). If two compressors are selected, each has a Qrc of 31.6 tons (111 kW). Each
compressor has eight cylinders, and the capacity of the unit can be controlled at 100, 75, 50, 25, and
0 percent. The Qrc-Tsuc curve of a single compressor is shown in Fig. 11.18.
Two DX coils are selected to match with the compressors, each with a face area of 21 ft2 (1.95
m2) with two expansion valves and two refrigerant distributors for intertwined face control. Their
Qrc-Tsuc curves are shown in Fig. 11.19.
When these two compressors are connected with two 21-ft2 (1.95-m2) DX coils, the operating
point of each compressor–DX coil combination must be the intersecting point Y of their Qrc-Tsuc
curves, as shown in Fig. 11.21. At the intersecting point Y representing the design condition, Qrc 
63 tons (222 kW) and Tsuc  40.2°F (4.5°C). The energy efficiency ratio EER is 11.8.
If two 15-ft2 (1.4-m2) DX coils are selected to form the compressor–DX coil combination,
the intersecting point then moves to Y, with a total cooling capacity about 59.5 tons (209 kW) at
Tsuc  38°F (3.3°C). The drop in Tsuc necessitates a greater energy input because of the higher
V?
p
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.35
FIGURE 11.21 System balance of selected components for a reciprocating refrigeration system.
compression ratio. The EER drops to 11.5. It is preferable to select two DX coils each with a face
area of 21 ft2 (1.95 m2) to reduce energy input.
Because the Tcon of an air-cooled condenser affects its total heat rejection Qrej, often Tcon is estimated
first when an air-cooled condenser is selected to determine Qrej for the compressor–DX coil
combination. If Tcon is estimated at 115°F (46.1°C), from Fig. 11.21, for a power input of 64 kW to
two compressors, the required total heat rejection of the selected air-cooled condenser is
In Fig. 11.21, draw a line from point Y to point Z whose Qrej  974,000 Btu/h (285,500 W). Draw
horizontal line ZZ intersecting the Qrej-Tcon curve of condenser size 120 at Z. At point Z, Tcon 
113.3°F (45.2°C). If a smaller condenser 095 is selected, then the intersecting point is Z and the
corresponding Tcon is 117.5°F (47.5°C), which necessitates a higher Pin to the compressor.
Using a larger condenser, the final selected capacity of this direct-expansion reciprocating
refrigeration system may be as follows:
Qrej  Qrc  3413Pin  63  12,000  64  3413  974,400 Btu / h (285,500 W)
11.36 CHAPTER ELEVEN
64 tons (225 kW)
40°F (4.4°C)
114°F (45.6°C)
64 kW
986,000 Btu/h (288,900 W)
Refrigeration load
Suction temperature
Condensing temperature
Power input to the compressor motor
Total heat rejection Qrej
For thermostatic expansion valves, the selected capacity must be greater than the cooling capacity
of the compressor at a pressure drop across the expansion valve of 75 to 120 psi (517 to 827 kPa)
for HCFC-22 at design conditions.
11.8 OPERATING CHARACTERISTICS OF AIR-COOLED,
DIRECT-EXPANSION RECIPROCATING REFRIGERATION SYSTEM
Operating Balance at Design Condition
Consider an air-cooled direct-expansion reciprocating refrigeration system in a rooftop packaged
system with the same characteristics as described in the last section and shown in Fig. 11.21. If the
design refrigeration load from the evaporation of HCFC-22 inside the two DX coils is only 60 tons
or 720,000 Btu/h (211 kW) instead of the selected capacity of 64 tons (225 kW), then the volume
flow rate of the piston displacement to extract suction vapor is greater than that of the vaporized
refrigerant in the evaporator. This imbalance lowers the evaporating pressure and temperature in the
evaporator and, therefore, increases the specific volume vsuc of the suction vapor and the log-mean
temperature difference as well as the heat transfer to the evaporator until is balanced with the
volume flow rate of refrigerant evaporized in the evaporator , cfm (m/ s). A new equilibrium between
the compressor and evaporator is reached
(11.10)
where  mass flow rate of refrigerant evaporized in evaporator, lb/min (kg/min)
Qrc  refrigeration load at evaporator, tons
qrf  refrigeration effect, Btu/lb
The refrigerant discharged from the compressor then enters the air-cooled condenser at mass
flow rate . Because the selected condenser can condense a mass flow rate of refrigerant equal to m?com
m?ev
m?com  V?p v suc  m?ev 
V?
ev
vsuc

200Qrc
qrf
V?
ev
V?
p
V?
p
a Qrej of about 986,400 Btu/h (289,000 W), a mass flow rate of refrigerant greater than is condensed
in the air-cooled condenser and causes a drop in condensing pressure pcon and condensing
temperature Tcon. As Tcon drops, 	Tm and Qrej both fall accordingly until the condensed mass flow
rate of refrigerant equals the mass flow rate of refrigerant discharged from the compressor and
a new balance is again formed in the condenser.
Pressure Characteristics
Figure 11.22 shows the pressure characteristics of a typical air-cooled, direct-expansion reciprocating
refrigeration system using DX coils and HCFC-22 as refrigerant. At design conditions, subcooled
liquid refrigerant at a temperature of 105°F (40.6°C) and a pressure of 238 psig (1641
kPa
 g) is fed to the thermostatic expansion valve. Because the saturated pressure of HCFC-22 at
105°F (40.6°C) is 210 psig (1448 kPa
 g), which is smaller than 238 psig (1641 kPa
 g), no flashing
will occur before the inlet of the expansion valve.
The pressure drop across the expansion valve is 100 psi (690 kPa), and the pressure drop in the
refrigerant distributor and tube is 59 psi (407 kPa). A fraction of liquid refrigerant is vaporized in
the expansion valve, distributor, and tubes. Refrigerant with a quality of about 0.15 enters the evaporator,
the DX coil, at a temperature of 47°F (8.3°C) and a pressure of 79 psig (545 kPa
g).
In the DX coil, all the liquid refrigerant is vaporized. The pressure drop across the refrigerant
circuits of the DX coil is 8 psi (55 kPa). At point x, the evaporating temperature is 42°F (5.6°C), and
the evaporating pressure is 71 psig (490 kPa
 g). At the outlet of the evaporator, vapor refrigerant
m?com
m?com
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.37
FIGURE 11.22 Pressure characteristics of a typical air-cooled direct-expansion reciprocating refrigeration system
using HCFC-22 as a refrigerant.
is superheated to 54°F (12.2°C), with a pressure of 71 psig (490 kPa
 g). The degree of superheat is
12°F (6.7°C).
The pressure drop at the suction line is 3 psi (20.7 kPa) corresponding to a 2°F (1.1°C) change
in saturated temperature. Suction vapor enters the reciprocating compressor at a suction temperature
of 40°F (4.4°C) and a pressure of 68.5 psig (472 kPa
g). In the compressor, the pressure is
raised to the hot-gas discharge pressure of 250 psig (1724 kPa
g) with a temperature of about
185°F (85°C).
In the discharge line, the pressure drop is 3 psi (20.7 kPa), corresponding to a 1°F (0.56°C)
change in saturated temperature. At the inlet of the air-cooled condenser, the pressure of the hot gas
is 247 psig (1703 kPa
g). Hot gas is desuperheated, condensed, and subcooled in the air-cooled
condenser at a condensing temperature of 115°F (46.1°C) and a pressure of 243 psig (1676 kPa
 g).
Liquid refrigerant is discharged from the subcooling coil of the air-cooled condenser at a temperature
of 105°F (40.6°C) and a pressure of 241 psig (1662 kPa
 g). The degree of subcooling is 10°F
(5.6°C).
In the liquid line, the pressure drop is 3 psi (20.7 kPa), corresponding to 1°F (0.56°C) change in
saturated temperature. Liquid refrigerant may enter the thermostatic expansion valve at a temperature
of 105°F (40.6°C) and a pressure of 238 psig (1641 kPa
 g).
Of the total system pressure drop of 250  68.5  181.5 psi (1248 kPa), a pressure drop of 159
psi (1096 kPa), or 88 percent of the total system drop, occurs in the expansion valve, distributor
nozzle, and tubes. During part-load operation, when outdoor temperature is lower, this pressure
drop may decrease considerably.
Part-Load Operation Using a Cylinder Unloader
The cylinders of a modern DX refrigeration system in a rooftop packaged unit using a microprocessor-
based DDC system controller can be loaded or unloaded successively in four successive steps
when discharge temperature Tdis rises and falls between 50 and 56°F (10 and 13.3°C) and Tdis is
maintained at 53°F (11.7°C) by means of PI control mode, as shown in the lower right-hand bottom
corner of Fig. 11.23. Cylinders are loaded in successive steps when Tdis is raised 1°F (0.56°C)
higher than the upper limit of the control band 55°F (2.8°C), and cylinders are unloaded when Tdis
is 1°F (0.56°C) lower than the lower limit of 51°F (10.6°C). Details of Tdis control for a rooftop
packaged unit are discussed in Sec. 29.6.
In the morning of a summer day, when the rooftop unit starts with a refrigeration load of Qrc 
60 tons (211 kW) from the cool-down load, all 16 cylinders are operating. The operating point may
be at Ao in Fig. 11.23 with a suction temperature Tsuc  42°F (5.6°C). As the refrigeration load
drops to a value of 58 tons (204 kW), the discharge temperature Tdis gradually falls until it is equal
to 50°F (10°C). At the same time, less refrigerant is vaporized in the DX coil because of the drop in
refrigeration load. As the 16 cylinders of the two compressors have a greater extracted volume than
the vapor refrigerant evaporated in the DX coil, the suction pressure psuc and temperature Tsuc drop
to point A on the 16-cylinder curve in Fig. 11.23. The DDC system controller unloads four cylinders
and closes one of the four liquid line solenoid valves of the two evaporators. The operating
point immediately shifts to point B on the 12-cylinder curve.
As the cooling capacity of the 12 operating cylinders is less than 58 tons (204 kW) of refrigeration
load, more vapor refrigerant is evaporated in the DX coils than can be extracted by the 12
cylinders, and the suction pressure and temperature of the refrigerant at operating point B tend to
move upward along the 12-cylinder curve to B with a Tsuc equal to about 44°F (6.7°C). At the same
time, because the cooling capacity of 12 cylinders Qrc is smaller than the 58-ton (204-kW) required
refrigeration load, Tdis rises to 56°F (13.3°C). The DDC system controller again loads four cylinders
and opens the shutoff solenoid valve in the evaporator. The operating point suddenly moves to point
A* on the 16-cylinder curve. When 16 cylinders are operating, the cooling capacity provided is
greater than 58 tons (204 kW), so point A* again moves downward along the 16-cylinder curve. At
the same time, Tdis drops to 50°F (10°C) and unloads four cylinders and opens the solenoid valve
again. The refrigeration system cycles between points A* and A with 16 cylinders and 100 percent
11.38 CHAPTER ELEVEN
of evaporator’s face area operating, and B and B with 12 cylinders and 75 percent of the face area
operating. Operating points A*, A, B, and B form a cycle. Because of the variation of percentages
of operating time on 16 cylinders and on 12 cylinders, the refrigeration load provided by the compressor
is balanced with the required 58-ton (204-kW) refrigeration load.
If the required refrigeration load drops to only 44 tons (155 kW), the operating point B on the
12-cylinder curve drops to B. Because of the fall of refrigeration load, Tdis again drops to 50°F
(10°C) and the DDC system controller then unloads to point C on the 8-cylinder curve and shuts-off
one of the two compressors and two of the solenoid valves in the two evaporators. The compressors
cycle between B, C, C, and B* until the refrigeration capacity provided is balanced with the required
refrigeration load of 44 tons (155 kW) by varying the time duration of 12-cylinder operation
and 8-cylinder operation.
Therefore, any refrigeration load from design load to a very low part-load can be balanced by
unloading the cylinders, shutting off the compressors, and deenergizing the face area of the evaporator.
When the required refrigeration load is greater than the refrigeration provided by the compressor
and the DX coils, the cylinders are loaded and the DX coils are energized successively in
the similar manner.
Pump-Down Control
Pump-down control is an effective means of preventing migration of refrigerant from the evaporator to
the crankcase of the reciprocating compressor during the shutdown period. When a compressor starts,
migrated refrigerant tends to mix with the oil and forms slugs which may damage the compressor.
When the discharge air temperature Tdis, or suction pressure drops below a preset limit and is
sensed by a sensor, the DDC system controller deenergizes the liquid line solenoid valve, as shown
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.39
FIGURE 11.23 Part-load capacity control of a DX reciprocating refrigeration system using a cylinder unloader.
in Fig. 11.22, and the valve closes. The compressor pumps all gaseous refrigerant to the condenser,
where it condenses to liquid. When the vapor pressure in the evaporator falls below a certain value,
the low-pressure control breaks the electric circuit and stops the compressor.
If cooling is again required, then a temperature sensor senses the increase of Tdis or a pressure
sensor senses the rise of suction pressure over a preset value, and the DDC system controller energizes
the liquid line solenoid valve. Liquid refrigerant then flows into the evaporator where it evaporates.
Vapor pressure builds up. As the pressure exceeds the cut-in pressure of the low-pressure
control, the compressor starts again.
Pump-down control acts as a capacity control by controlling Tdis or the suction pressure. It also
acts as a safety control to protect the compressor from damage from liquid slugs. Pump-down control
is widely used in DX reciprocating refrigeration systems.
When the outdoor air temperature is low, in order to maintain proper operation of the thermostatic
expansion valve, low ambient air control should be used to modulate the dampers or cycling, or
even shut off the condenser fans, as described in Sec. 10.7.
Main Problems in Direct-Expansion Reciprocating Refrigeration Systems
Liquid Slugging. Liquid slugging is a mixture of by-product of corroded metallic surface with the
refrigerant and oil. It is formed under the following conditions:
 Liquid refrigerant floods back from the evaporator to the crankcase of the compressor because of
either an insufficient degree of superheat or hunting of the thermostatic expansion valve. Liquid
refrigerant dilutes the lubrication oil.
 Liquid refrigerant migrates from the warmer indoor evaporator to the colder outdoor compressor
during a shutdown period in split packaged units.
 A considerable amount of oil returns to the crankcase from the oil traps.
 When air and water are presented in any appreciable amount as contaminants in a refrigeration
system, the chemical action involving these contaminants (the refrigerant and oil) often occurs,
forming corrosive acid and producing sludge.
A liquid slug is incompressible. When a liquid slug enters the compression chamber, it may damage
the valves, pistons, and other components.
An efficient filter dryer is essential to remove the moisture and prevents the formation of corrosive
acid and sludge. Pump-down control and crankcase heater are effective means of preventing
refrigerant migration in split packaged units. A proper degree of superheat is essential to prevent
liquid refrigerant floodback to the compressor. Proper sizing of the thermostatic expansion valve
and proper location of the sensing bulb prevent hunting of the thermostatic expansion valve.
Oil Return. For oil-miscible halocarbon refrigerants, the system design must include a means of
returning oil from the condenser and evaporator to the compressor.
Compressor Short Cycling. Short cycling is on /off control of the compressor or loading or unloading
cylinders within too short a time period (less than 3 min). It is mainly a result of close tolerance
control or improper design of on/off capacity control. If the compressor is short-cycled, the
cycling action may pump oil away from the compressor, causing insufficient oil lubrication. It may
also damage system components. In addition, the short cycle does not allow sufficient time to restabilize
the system.
Compressor short cycling can be solved by using a time delay for restart, or by overlapping unloading
and loading schedules in step capacity control.
Defrosting. In a direct-expansion reciprocating refrigeration system, if the surface temperature of
the DX coil is 32°F (0°C) and lower, frost accumulates on the coil surface. Because frost impedes
11.40 CHAPTER ELEVEN
air passage and reduces the rate of heat transfer of the coil, it must be removed periodically. The
process of removing frost from the evaporator is called defrosting.
If air enters the coil at a temperature above 36°F (2.2°C), the ambient air flowing through the
DX coil can be used for defrosting. However, in low-temperature refrigeration systems, if the ambient
air temperature is also below 32°F (0°C), an electric heating element may be used as a simple
and effective way to defrost the coil. Hot-gas defrosting is effective, but must be planned during the
design stage to ensure an adequate quantity of hot gas at a pressure high enough to flow through
the coil. After defrosting, the high-pressure condensate should be returned to the liquid line, and the
pressure of the desuperheated gas should be reduced gradually and released to the suction line.
The defrosting cycle can be controlled by sensing the pressure or temperature difference of air
entering and leaving the DX coil over a fixed time interval.
Proper Refrigerant Charge. A proper amount of refrigerant charge is neccessary to ensure a
proper DX system performance. Too much refrigerant charge will reduce the condensing area and
therefore have a higher condensing pressure. An insufficient charge will decrease the required evaporation
in the DX coil and thus have less than the required cooling capacity. A new refrigeration
system should be charged with the exact amount of refrigerant specified by the manufacturer. For
an operating DX system, a practical and convenient way to ensure a proper refrigerant charge is to
check the discharge/suction pressure by means of pressure gauges and the degree of superheat at
the outlet of the DX coil through its outer surface temperaure measurement.
If the condensing temperature of an air-cooled DX reciprocating system at an outdoor air temperature
of 95°F (35°C) is 125°F (51.7°C), the discharge pressure is about 2934  297 psia
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.41
TABLE 11.6 Electrically Operated Packaged Units and Condensing Units—Minimum Efficiency Requirements
Minimum Efficiency as of Test
Equipment type Size category Rating condition efficiency* 10/29/2001* procedure
Packaged units,  65,000 Btu/h† Split system 10.0 SEER 10.0 SEER ARI 210/240
air cooled Single package 9.7 SEER 9.7 SEER
 65,000 Btu/h and Split system and 8.9 EER§ 10.3 EER§
 135,000 Btu/h single package
 135,000 Btu/h and Split system and 8.5 EER§ 9.7 EER§ ARI 340/360
 240,000 Btu/h single package
 240,000 Btu/h and Split system and 8.5 EER§ 9.5 EER§
 760,000 Btu/h single package 7.5 IPLV§ 9.7 IPLV§
 760,000 Btu/h Split system and 8.2 EER§ 9.2 EER§
single package 7.5 IPLV§ 9.4 IPLV§
Packaged units water,  65,000 Btu/h Split system and 9.3 EER 12.1 EER ARI 210/240
and evaporatively cooled single package
 65,000 Btu/h and Split system and 10.5 EER§ 11.5 EER§
 135,000 Btu/h single package
 135,000 Btu/h and Split system and 9.6 EER§ 11.0 EER§ ARI 340/360
 240,000 Btu/h single package
 240,000 Btu/h Split system and 9.6 EER§ 11.0 EER§
single package 9.0 IPLV§ 10.3 IPLV§
Condensing units,  135,000 Btu/h 9.9 EER 10.1 EER ARI 365
air cooled 11.0 IPLV 11.2 IPLV
Condensing units, water,  135,000 Btu/h 12.9 EER 13.1 EER
or evaporatively cooled 12.9 IPLV 13.1 IPLV
*IPLVs are only applicable to equipment with capacity modulation.
†Single-phase air-cooled air conditioners  65,000 Btu/h are regulated by NAECA. SEER values are those set by NAECA.
§Deduct 0.2 from the required EERs and IPLVs for units with a heating section other than electric resistance heat.
Source: ASHRAE/IESNA Standard 90.1-1999. Reprinted by permission.
(2048 kPa abs.), and the suction pressure is 913  88 psia (607 kPa abs.) with a suction temperature
of 45°F (7.2°C). A 10 to 12°F (5.6 to 6.7°C) superheat at the outlet of the DX coil is considered
suitable. Then at an outdoor air temperature of 95°F (35°C), if the discharge/suction pressure
of the operating air-cooled DX system is around 282/73 psig (1944/503 kPag) and the superheat is
between 10 and 12°F (5.6 to 6.7°C), the refrigerant charge is considered all right.
Minimum Performance and Equipment Selection
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency requirements of aircooled,
water-cooled, and evaporatively cooled packaged units for direct-expansion reciprocating
and scroll refrigeration systems. The requirements are expressed in terms of energy efficiency ratio
(EER) and integrated part-load value (IPLV) as listed in Table 11.6. When multiple rating conditions
or performance requirements are provided, the equipment shall satisfy all stated requirements
unless otherwise exempted. The selected air-cooled packaged unit should meet the minimum performance
guidelines at both design and part-load operation.
11.9 ROTARY REFRIGERATION SYSTEMS
Rotary refrigeration systems using rotary compressors are widely adopted in small refrigeration
systems such as refrigerators, room air conditioners, and room heat pumps having a cooling capacity
from a fraction of a horsepower to about 4 hp (0.75 kW to 3.0 kW).
Rotary Compressor
The rolling-piston rotary compressor, shown in Fig. 11.24, is the most widely used rotary compressor
in small refrigeration systems today. A rolling piston is mounted on an eccentric shaft, and a
11.42 CHAPTER ELEVEN
FIGURE 11.24 A rolling-piston rotary compressor.
fixed vane sliding in a slot remains in contact with the roller. As the piston rotates, the vane is in
reciprocating motion. Vapor refrigerant enters the compression chamber through the suction inlet
and is compressed by the eccentric motion of the roller. When the rolling piston is in contact with
the top of the cylindrical housing, the hot gas is squeezed out through the discharge valve. Internal
leakage depends on clearance of mating parts, surface speed, surface finish, and oil viscocity.
HCFC-22 is the most widely used refrigerant in rotary compressors. HFC-407C and HFC-410A
with ozone depletion potential equal to zero will become the alternative refrigerants to replace
HCFC-22 after the year 2020.
A rotary compressor has a small clearance volume and therefore greater volumetric efficiency
and smaller reexpansion losses than a reciprocating compressor. The mechanical efficiency mec of
a typical rotary compressor operating at a compression ratio of 3.5 is about 0.87.
Main Components
Similar to a reciprocating refrigeration system, a rotary refrigeration system consists of the following
main components in addition to the rotary compressor:
 Evaporator. A direct-expansion or DX coil is often used as the evaporator. The DX coil is usually
made of tube and corrugated fins to improve the heat transfer. Copper tubes typically of 0.5-in.
(13-mm) diameter and aluminum fins with a thickness of 0.006 in. (0.15 mm) are used. An adequate
cooling rate per unit coil face area, Btu/h
 ft2 (W/m2) is important during the selection of a DX coil.
 Condenser. An air-cooled condenser is also often made of tube and corrugated fins. The diameter
of the copper tube is usually between and in. (6.5 and 19 mm). A subcooler is sometimes
used and is located between the condenser and the capillary tube submerging in the condensate.
The condensing coil usually has one to three rows of tubes because of the lower air-side pressure
drop provided by the propeller fan.
 Restrictor. A capillary tube is the most widely used restrictor because of its low cost and reliability.
 Permanent split-capacitor motor. This type of motor provides adequate starting torque for rotary
compressors and improves the power factor during operation.
System Performance
According to ASHRAE Handbook 1996, HVAC Systems and Equipment, for an ambient temperature
of 95°F (35°C), an evaporating temperature of 35°F (1.7°C), and a condensing temperature of
130°F (54.4°C) with a subcooling temperature of 15°F (8.3°C), the EER of a rotary refrigeration
system for cooling is 9.0 (COP  2.7). Because of its rotary motion, a rotary refrigeration system
makes less noise than a reciprocating system.
11.10 SCROLL REFRIGERATION SYSTEMS
Types of Scroll Refrigeration Systems
As in the reciprocating refrigeration systems, the currently used scroll refrigeration systems for air
conditioning are mainly for small to medium-size systems (packaged units and heat pumps) with a
cooling capacity often between slightly less than 1 ton and 25 tons (3.5 and 90 kW) used in light
commercial, industrial, and residential buildings. These include the following:
 Air-cooled, scroll DX cooler
 Air-cooled, scroll heat pumps
 Scroll chillers
34
14
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.43
Figure 11.25 shows an air-cooled, scroll DX cooler. It consists of two scroll compressors, an aircooled
condenser, two DX coils, and two thermostatic expansion valves (TXVs). A scroll refrigeration
system has almost the same refrigerant piping and safety controls as the reciprocating system
except there is no crankcase heater.
An air-source scroll heat pump has nearly the same system structure as an air-cooled, scroll DX
cooler except that a four-way reversing valve is added in the heat pump. Heat pumps are discussed
in the next chapter.
Scroll Compressors
In 1987, the first U.S. manufacturer using a computer-aided manufacturing process produced a
scroll compressor for residential air conditioning with tolerances up to 0.0001 in. (0.0025 mm). A
scroll compressor consists of two identical spiral scrolls mated face to face, as shown in Fig. 11.26.
Each scroll is bound on one side to a flat plate. The upper scroll is stationary and contains the discharge
port. The lower one moves in an orbit around the shaft center of the motor at an amplitude
equal to the orbit radius. The two scrolls are fitted together to form vapor pockets between their respective
baseplates and lines of contacts of their vane walls. During suction, vapor refrigerant enters
the space between the two scrolls through the lateral openings. The openings are then sealed off to
form trapped vapor pockets, thus completing the intake process.
During the successive revolutions of the motor shaft, the volume of the trapped vapor pockets is
reduced, and the compression process is complete when the gaseous refrigerant is compressed to its
maximum pressure and just before the compressed gaseous pocket is opened to the discharge port.
During discharge, the trapped gas pockets opens to the discharge port. All compressed hot gas is
discharged through the discharge port to the discharge line. The volume of the gas pockets is
reduced to zero. The sequence of intake, compression, and discharge occurs simultaneously in the
two trapped gas pockets.
Radial and Axial Compliance
In scroll compressors, the scrolls touch one another with force sufficient to create a seal but not
sufficient to cause wear. This is the result of radial and axial compliance. Continuous flank contact
because of the centrifugal force minimizes the gas leakage.
11.44 CHAPTER ELEVEN
DX coils
(evaporators)
TXV Sight
glass
Filter
dryer
Solenoid
valve
Charging
valve
Air-cooled
condenser
Condenser
coil
Scroll
compressors
FIGURE 11.25 An air-cooled scroll DX cooler.
In one type of radial compliance, the orbiting scroll is mounted on the sliding unloader bushing
that is fitted onto the eccentric pin of the motor shaft and thus directs the radial motion of the
orbiting scroll in and out to accommodate variations in orbit radius due to discrepancies. Radial
compliance allows the scroll members to separate in the presence of liquid refrigerant or debris and
increases system reliability.
Axial compliance is accomplished by an adjustable axial pressure that provides sealing contact
between the scroll tips and bases when the scroll is orbiting. This pressure is released when the
orbiting scroll ceases to operate. As the scroll tip bears directly on both the baseplates and the lubrication
oil between the scroll and the plates, both form an effective oil seal.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.45
FIGURE 11.26 (a) Scroll compressor and (b) scroll compression processes. (Source: ASHRAE Handbook 1996,
HVAC Systems and Equipment. Reprinted with permission.)
Compressor Performance
The cooling capacity of scroll compressors for a DX system is often between 1 and 15 tons (3.5 and
53 kW). HFC-407C and HFC-410A are the long-term alternative refrigerants, and HCFC-22 is the
interim refrigerant that can be used before 2020.
Scroll compressors have a very high volumetric efficiency v. Levins et al. (1996) showed that
v for the scroll compressor varied from 96.9 to 93.6 percent when the pressure ratio increased
from 2.77 to 3.58. The pressure ratio (or compression ratio) is defined as the ratio of discharge pressure
to the suction pressure pdis /psuc, both in psia (kPa abs.).
Scroll compressors have an isentropic efficiency isen of 70 percent when their pressure ratios
are around 3. Most of the scroll compressors produced in the late 1990s have an EER of about 11.5
(COP  3.35) which is higher than that of the rotary compressors and a standard reciprocating
compressor.
ASHRAE/IESNA Standard 90.1-1999 mandates minimum energy requirement of air-cooled,
water-cooled, and evaporatively cooled scroll DX packaged units which are the same as the reciprocating
DX packaged units as listed in Table 11.6. The minimum efficiency requirements EER of
water-cooled and evaporatively cooled DX packaged units are about 13 to 18 percent higher than
air-cooled DX packaged units at various cooling capacities listed.
Concentration Shift and Temperature Glide
Basic Mechanism. In scroll, reciprocating, and screw refrigeration systems, refrigerant blends
HFC-404A, HFC-410A, HFC-407A, and HFC-407C are used as alternative refrigerants to replace
HCFC-22 and CFC-502 for zero ozone depletion and better system performance. HFC-404A and
HFC-410A are near-azeotropic, whereas HFC-407A and HFC-407C are zeotropic or nonazeotropic
and produce significant temperature glide in condensation and evaporation.
During the phase change in the two-phase region, the evaporating liquid changes its composition
continuously and thus its bubbling point; also the condensing vapor changes its composition and
thus its dew point, as shown on the temperature-concentration diagram in Fig. 11.27. These result
in a concentration shift and temperature glide which is a temperature drop during condensation and
a temperature gain during evaporation. There are two additional causes of concentration shifts:
 Differential solubility of lubrication oil of the components of the refrigerant mixture.
 Leakage of refrigerant from the system.
Circulating Concentration Shift and Its Effects on Performance
Kruse and Wieschollek (1997) measured the concentration and mass of refrigerant in the two-phase
region in the evaporator and condenser of a DX refrigeration system during circulation of refrigerant.
The concentration shift can be calculated by means of the mass balance. For HFC-407C, a
blend of HFC-32 /HFC-125/ HFC-134a with original concentration of 23 /25/ 52 percent, HFC-32
had a circulating concentration of 24.76 percent; HFC-125, 26.44 percent; and HFC-134a, 48.81
percent. The calculated concentration shifts during circulation were 1.76, 1.44, and 3.19 percent
in weight for components HFC-32, HFC-125, and HFC-134a, respectively. The calculated performance
changes because of the circulating concentration shift of HFC-407C are as follows:
Original concentration Concentration shift
Cooling capacity, Btu /h 10,529 10,625
Mass flow rate, lb/min 2.67 2.69
Evaporating pressure, psia 62.98 64.30
Condensing pressure, psia 315.3 324.3
11.46 CHAPTER ELEVEN
As for a near-azeotropic refrigerant mixture HFC-404A, the concentration shift is small, the calculated
performance change due to circulation concentration shift is less than 1 percent and can be
neglected.
Corr et al. (1994) showed that direct-expansion evaporators typically have liquid fractions of 2
to 10 percent and result in a small shift in circulation concentration.
Heat Exchanger Flow Configuration
The temperature glide of zeotropic refrigerant blends can be utilized to increase the heat exchange
efficiency in the evaporator where the evaporating temperaure increases and in the condenser where
the condensing temperature drops. Rice (1993), using a modified Carnot cycle analysis, found that
for an air-source heat pump using zeotropic blend as refrigerant at cooling mode with a relative UA
(overall conductance-area product) level of 3, many-row crossflow has a potential COP gain of
about 7 percent, two-row countercrossflow about 10 percent, three-row countercrossflow 15 percent,
counterflow (or three-row countercrossflow) 20 percent. Hwang and Radermacher (1997)
showed that the capacity of the near-counterflow heat exchanger increases about 10 percent compared
to that of the near-parallel flow heat exchanger.
Capacity Control and Part-Load Performance
There are two types of capacity control: variable-speed and variable-displacement compressors. In
a variable-speed scroll compressor, a variable-speed drive is used to reduce the speed at part-load
operation so that the cooling capacity is reduced to meet the reduced system load. The operating
frequency often varies from 15 to 150 Hz.
In a variable-displacement scroll compressor, there are porting holes in the fixed scroll. By
connecting or disconnecting the compression chamber to the suction side through the closing and
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.47
Dewpoint
line
Bubbling
point line
Temperature
glide
(Tsv-Tsc)
Tsc
Tsv
Temperature T, F
Saturated
vapor
Saturated
liquid
0
Concentration
1
FIGURE 11.27 Temperature-concentration diagram of a zeotropic refrigerant blend.
opening of the porting holes, the cooling capacity of the scroll compressor is varied. If all porting
holes are closed, the compressor is operating at full capacity. If all the porting holes are open, the
cooling capacity drops to the minimum.
Takebayashi et al. (1994) showed the isentropic efficiency of a variable-speed scroll compressor
at a pressure ratio range between 2 and 5 during part-load operation, cooling, and heating mode operation.
Isentropic efficiency drops about 15 percent when the pressure ratio increases from 2 to 5.
System Characteristics
Compared to reciprocating and rotary compressors, the system characteristics of air-cooled, DX
scroll refrigeration systems are as follows:
 Scroll compressors tolerate liquid refrigerant and debris during compression. For most of the aircooled
DX scroll systems, the crankcase heater and suction line accumulator are often not
required.
 A scroll compressor has less than 50 percent of the parts of a reciprocating compressor of the
same size. The fewer the parts of a compressor, the higher the reliability.
 Most scroll compressors provided an EER of 11.5 (3.37 COP) in the late 1990s, which is higher
than that of a rotary compressor and is similar to that of a high-efficiency reciprocating compressor.
 A scroll heat pump shows a flatter capacity curve when outdoor air temperature To varies than a
reciprocating heat pump does. Its cooling capacity drops less when the outdoor ambient temperature
increases during summer, and the heating capacity of a scroll heat pump reduces less when
the outdoor ambient temperature decreases during winter. The advantage of a flatter capacity
curve is discussed in the next chapter.
 A scroll compressor has a lower noise than a reciprocating compressor and is about 5 dBA lower
than the reciprocating compressor of the same size and operated in the same environment.
On the other hand, scroll compressors need computer-aided precision manufacturing and have
been used in the field of air conditioning in the United States since 1987 only. There will be new
problems to be solved and potential to be improved and developed.
Scroll compressors are gradually replacing reciprocating and rotary compressors in room air
conditioners, heat pumps, and packaged units for residential and commercial buildings.
Scroll Chillers
Scroll chillers are often available with a cooling capacity of 10 to 60 tons (35 to 210 kW). They are
gradually replacing reciprocating chillers in new and retrofit installations. Scroll chillers can again
be divided into air-cooled scroll chillers and water-cooled scroll chillers.
Air-Cooled Scroll Chillers. An air-cooled scroll chiller consists of mainly an air-cooled condenser,
two or more scroll compressors, a single or dual refrigerant circuit, a thermostatic expansion
valve, a distributor, a direct-expansion liquid cooler, and controls. A direct-expansion liquid
cooler can be easily connected to dual refrigerant circuits and multiple compressors. An air-cooled
scroll chiller is usually located on the rooftop of an air conditioned building. Chilled water freeze
protection should be provided during winter operation in locations where the outdoor temperature
falls below 32°F (0°C).
An air-cooled scroll chiller always has a microprocessor-based control of the chilled water temperature
leaving the liquid cooler. There are usually two capacity control steps for small air-cooled
scroll chillers and four capacity steps for large scroll chillers. Air-cooled scroll chillers are designed
to be operated between outdoor temperatures of 40 and 115°F (4.4 and 46°C); with an additional
head pressure control damper assembly, they are permited to operate at a lower ambient temperature
down to 0°F (17.8°C).
11.48 CHAPTER ELEVEN
In addition, there are safety controls of compressor overload protection, high- and low-pressure
cutouts, low-temperature freeze protection, oil pressure control, elapse-time meter to record compressor
operating hours and number of starts, chilled water flow switch, as well as a diagnostic display.
Air-cooled scroll chillers are rated at a chilled water temperature leaving the liquid cooler at
44°F (6.7°C), 	T  10°F (5.6°C), and ambient air temperature entering the air-cooled condenser at
95°F (35°C). ASHRAE/IESNA Standard 90.1-1999 mandates the minimum performance of aircooled
chillers of cooling capacity less than 150 tons (527 kW) with a COP  2.7 (EER  9.22)
and an IPLVCOP  2.8 (IPLVEER  9.55). The EER of the air-cooled scroll chillers of one of the
leading HVAC&R manufacturers in the United States is 9.1 to 9.6, and IPLVEER is 10 to 12.5.
Water-Cooled Scroll Chillers. A water-cooled scroll chiller is similar in construction and controls
to an air-cooled scroll chiller except that a shell-and-tube water-cooled condenser is used instead of
an air-cooled condenser. A water-cooled scroll chiller is usually located in the machinery room (refrigeration
plant room) of the air conditioning building. According to ARI Standard 550 / 590-98,
air-cooled and water-cooled scroll chillers are rated at the same conditions as in a centrifugal
chiller, discussed in Sec. 13.2.
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency requirements of aircooled
and water-cooled reciprocating, scroll, and screw water chillers as listed in Table 11.7. In
Table 11.7, air-cooled reciprocating, scroll, and screw water chillers (with condensers) are specified
with the same 2.7 COPs and 2.8 IPLVs when cooling capacities are less than 150 tons (527 kW), and
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.49
TABLE 11.7 Water Chillers—Minimum Efficiency Requirements
Minimum Efficiency as of
Equipment type Size category Rating condition efficiency* 10/29/2001* Test procedure
Air cooled, with condenser,  150 tons 2.70 COP 2.80 COP ARI 550
electrically operated 2.80 IPLV 2.80 IPLV or
 150 tons 2.50 COP ARI 590
2.50 IPLV as appropriate
Air cooled, All capacities 3.10 COP 3.10 COP
without condenser, 3.20 IPLV 3.10 IPLV
electrically operated
Water cooled, electrically All capacities 3.80 COP 4.20 COP ARI 590
operated, positive displacement 3.90 IPLV 4.65 IPLV
(reciprocating)
Water cooled,  150 tons 3.80 COP 4.45 COP ARI 550
electrically operated, 3.90 IPLV 4.50 IPLV or
positive displacement ARI 590
(screw and scrolls)  150 tons and 4.20 COP 4.90 COP as appropriate
 300 tons 4.50 IPLV 4.95 IPLV
 300 tons 5.20 COP 5.50 COP
5.30 IPLV 5.60 IPLV
Water cooled, electrically  150 tons 3.80 COP 5.00 COP ARI 550
operated, centrifugal 3.90 IPLV 5.00 IPLV
 150 tons and 4.20 COP 5.55 COP
 300 tons 4.50 IPLV 5.55 IPLV
 300 tons 5.20 COP 6.10 COP
5.30 IPLV 6.10 IPLV
*The requirements do not apply for chillers used in low-temperature applications where the design leaving temperature is less than or equal to
40F (4.4C).
Source: ASHRAE/IESNA Standard 90.1-1999. Reprinted by permission.
are specified with 2.5 COP and also IPLVs when cooling capacity is greater than or equal to 150 tons
(527 kW). For water-cooled water chillers, when cooling capacities are less than 150 tons (527 kW),
minimum efficiency requirements 3.8 COP and 3.9 IPLV are the same for reciprocating, scroll, and
screw water chillers. According to the data for water-cooled scroll chillers of a capacity range between
20 to 60 tons (70 to 211kW), a leading HVAC&R manufacturer in late 1990s in the United
States listed their COPs at between 3.86 and 4.04 corresponding to a kW / ton between 0.91 and 0.87.
11.11 SCREW REFRIGERATION SYSTEMS
Types of Screw Refrigeration System
Currently used screw refrigeration systems are mostly screw chillers. Screw chillers are either
water-cooled or air-cooled. Energy-efficient water-cooled screw chillers are widely used to replace
the water-cooled reciprocating chillers employed in thermal storage systems. More energy-efficient
and more compact air-cooled screw chillers are used to replace air-cooled centrifugal chillers and
air-cooled reciprocating chillers in new and retrofit installations.
Water-cooled chillers often have a cooling capacity between 50 and 500 tons (176 and 1758
kW). A water-cooled screw chiller has five main components, as shown in Fig. 11.28. In addition to
a screw compressor, a shell-and-tube liquid cooler as evaporator, a shell-and-tube water-cooled condenser,
and orifice plates or float valves as throttling device, an oil cooler is used to cool and supply
oil to the compressor. According to ARI Standard 550 / 590-98, water-cooled and air-cooled screw
chillers can be rated the same as a centrifugal chiller, discussed in Sec. 13.2.
Screw Compressors
Screw compressors are designed to be incorporated with oil injection and have been used in industrial
applications since the 1950s. Screw compressors can be classified as twin-screw compressors
or single-screw compressors.
Twin-Screw Compressor. Twin-screw compressors are more widely used than single-screw compressors
because their isentropic efficiency is slightly higher than that of single-screw compressors
(about 3 to 4 percent). A typical twin-screw compressor, shown in Fig. 11.29a, consists of a fourlobe
male rotor and a six-lobe female rotor (4  6), or five-lobe male rotor and a seven-lobe female
rotor (5  7), a housing with suction and discharge ports, and a sliding valve for capacity control.
When the lobes are separated at the suction port, as shown in Fig. 11.29b, vapor refrigerant is
drawn in and the intake process continues until the interlobe space is out of contact with the suction
port. The volume trapped in the interlobe space within the meshing point is compressed during successive
rotations of the rotor. When the interlobe space makes contact with the discharge port, the
compressed gas discharges through the outlet.
The ratio of the volume of vapor trapped during the intake process Vin, ft3 (m3), to the volume of
trapped gaseous refrigerant discharged Vdis, ft3 (m3), is called the built-in volume ratio Vi. The application
of variable Vi instead of fixed Vi eliminates the losses associated with over- or undercompression,
especially during part-load operation.
Oil injection plays an important role in twin-screw compression. Injected oil effectively cools
the compressor and results in a lower discharge temperature. A small clearance of 0.0005 in.
(0.0127 mm) between lobes and the sealing oil minimizes leakage loss. Another important function
of the injected oil is lubrication. The cooling capacity of twin-screw compressors ranges from 20 to
1300 tons (70 to 4570 kW). Cooling capacities between 50 and 350 tons (176 and 1230 kW) are
most widely used. The pressure ratio of a screw compressor can be as high as 20 : 1 (single-stage).
HCFC-22, HFC-134a, and HFC-407C for comfort air conditioning, and R-717 for industrial applications,
are the most widely used refrigerants for twin-screw compressors.
11.50 CHAPTER ELEVEN
FIGURE 11.28 A typical screw (twin-screw compressor refrigeration system): (a) schematic diagram; (b) refrigeration
cycle; (c) undercompression; (d) overcompression.
Single-Screw Compressor. The single-screw compressor was developed in the 1960s and has
been used in refrigeration systems since the 1970s. The single-screw compressor has a single main
rotor and two gate rotors, as shown in Fig. 11.30. As the main rotor rotates, one of the flutes opens
to the suction port and is filled with suction vapor until its suction end meshes with the gate rotor.
The discharge end of this flute is covered by the rotor casing. When the main rotor turns, the meshing
of gate rotor A with the flute compresses the trapped gas and raises its pressure. When the
discharge end of this flute opens to the discharge port, compression stops and the hot gas is discharged.
In a single-screw compressor, compression occurs simultaneously in both the top and
bottom of the main rotor.
Performance of Twin-Screw Compressor
In a typical twin-screw compressor using HCFC-22 as refrigerant with Vi  2.6, v drops from 0.92
to 0.83 and isen decreases from 0.82 to 0.60 when the pressure ratio pdis /psuc increases from 3 to 8.
A twin-screw compressor is more energy-efficient than a reciprocating or a scroll compressor of the
same size, operating at the same conditions, and is less energy-efficient than a centrifugal compressor
using a water-cooled condenser. Some recently developed screw compressors can even tolerate
liquid slugging.
11.52 CHAPTER ELEVEN
FIGURE 11.29 Twin-screw compressor: (a) loading and unloading of screw compressor; (b) suction-compressiondischarge
processes. (Source: Dunham-Bush Compressors. Reprinted by permission.)
Bivens et al. (1997) tested of a screw chiller of 150-ton (527-kW) cooling capacity with both
shell-and-tube evaporator and condenser with water inside the tubes and using a zeotropic refrigerant
mixture of HFC-32/HFC-125/HFC-134a (30/10/ 60 percent). The result was that the circulating
concentration shift of HFC-32 /HFC-125/HFC-134a was 41 /17/ 42 percent; and the cooling
capacity of the screw chiller dropped 36 percent, input power increased 14 percent, and COP decreased
44 percent when it was compared with HCFC-22. Instead of a flooded evaporator, when a
screw chiller employed a direct-expansion evaporator with a refrigerant blend HFC-32 /HFC-
125/HFC-134a (30/10/ 60), the screw chiller’s cooling capacity dropped 7 percent, power input increased
8 percent, and COP decreased 14 percent when compared with HCFC-22.
Oil Cooling
As shown in Fig. 11.28, the hot-gas and oil mixture discharged from the screw compressor enters
an oil-gas separator, in which oil is separated from the gaseous refrigerant. The oil is cooled in the
oil cooler, flows through a filter, and divides into three circuits to lubricate the bearings, provides
hydraulic power to the capacity control system, and flows to the compressor. Oil in the compressor
provides a rotor-to-rotor and rotor-to-housing seal, lubricates the rotor driving mechanism, and
cools the refrigerant during compression.
In an oil cooler, oil is usually cooled by cooling water from the cooling tower or evaporative
cooler. It is sometimes cooled by river, lake, or well water, or by direct expansion of liquid refrigerant.
Capacity Control
Cooling capacity is controlled by moving a sliding valve between the discharge end of the male rotor
and the discharge port, as shown in Fig. 11.29a. The sliding valve is driven by the oil pressure.
When the sliding valve moves toward the discharge port, it opens a bypass passage on the rotor
housing. A portion of the trapped gaseous refrigerant returns to the suction port through this bypass
passage, so the mass flow rate of refrigerant and thus the cooling capacity are reduced. This kind of
capacity control is a stepless continuous modulation.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.53
FIGURE 11.30 Single-screw compressor.
Variable Volume Ratio
There are no discharge valves in a twin-screw compressor. If the pressure drop in the discharge line
is ignored, a screw compressor using built-in volume ratio Vi may cause over- or undercompression
when the discharge pressure from the compressor pdis is different from the condensing pressure
pcon, both in psig (kPa
g). Figure 11.28c shows undercompression. Undercompression causes a
volume of gas at condensing pressure to reenter the trapped volume at the beginning of the discharge
process. Figure 11.28d shows overcompression. The discharged gas reexpands to match the
condensing pressure.
In the late 1980s, twin-screw compressors with variable volume ratios were developed. The common
arrangement includes a split sliding valve, which consists of a sliding valve and a movable slide.
As the slide moves back and forth, the radial discharge port is relocated and thus varies the volume ratio
to meet the optimum requirements without over- and undercompression. For a screw refrigeration
system in which suction and condensing pressures change frequently, efficiency can be improved by up
to 20 percent by using a variable volume ratio twin-screw compressor instead of a built-in volume ratio.
Economizer
In Fig. 11.28b, if a small portion of liquid refrigerant at point 5 is flashed at an intermediate pressure
pint so that the remaining refrigerant is cooled to saturated condition at intermediate pressure,
point 8, such an arrangement significantly increases the refrigeration effect of the cycle. At the
same time, the flashed vapor refrigerant is introduced to the partially compressed gas at pint. The
compression work needed for the flashed vapor at pint is less than the compression work needed if
the liquid refrigerant is flashed at the evaporator at psuc.
Figure 11.28a is a schematic diagram of an economizer. Liquid refrigerant from the receiver at
point 5 flows through the throttling device and enters the motor shell. In the motor, a portion of
liquid refrigerant at pint is flashed into vapor to cool the motor windings and the remaining liquid
refrigerant to saturated condition point 8. The flashed vapor is then introduced to the compressed
gaseous refrigerant at pint in the compressor. The remaining liquid refrigerant flows through another
throttling device. Its pressure drops to the evaporating pressure pev and enters the evaporator at
point 9. In the evaporator, all liquid refrigerant is vaporized. Vapor refrigerant from the evaporator
is extracted by the compressor at suction pressure psuc denoted by point 1.
Controls
In a water-cooled screw chiller, a DDC system includes the following:
Status monitoring and specific controls:
Entering and leaving chilled water temperatures
Evaporator and condenser refrigerant temperatures and pressures
Compressor suction temperature
Electric current limit
Safety controls:
Motor overloading, high- and low-pressure cutouts, water freeze protection, and oil pressure
Diagnostics: operating conditions and diagnostics display
Minimum Performance
ASHRAE/IESNA Standard 90.1-1999 mandates different minimum efficiency requirements for
water-cooled screw water chillers as listed in Table 11.7:
11.54 CHAPTER ELEVEN
One of the U.S. leading HVAC&R manufacturer offered an energy use of as low as 0.59 kW/ton
(5.96 COP) for their 150- to 360-ton (527- to 1266-kW) water-cooled screw chillers at full-load and
0.52 kW/ton (6.76 COP) at part-load. These energy consumption indices were even less than the
values projected as of October 29, 2001.
Air-Cooled Screw Chillers
Air-cooled screw chillers are widely used in locations where water is scarce and the use of city water
as makeup water for cooling towers is prohibited by local codes. Because the air-cooled screw
chillers are more energy-efficient and more compact than air-cooled centrifugal chillers, and more
energy-efficient than recprocating and scroll chillers, air-cooled screw chillers become the dominant
air-cooled chillers in a cooling capacity range of 50 to 400 tons (176 to 1400 kW).
System Components. An air-cooled screw chiller consists of mainly two, three, or four screw
compressors, dry-expansion liquid cooler, air-cooled condenser, expansion valve, and a DDC unit
controller. There are two independent refrigerant circuits; each has its own compressors, expansion
valve, and refrigerant distributor and liquid cooler. This kind of setup provides greater reliability in
case there is a compressor problem.
Electric Expansion Valves. Some manufacturers offer electric expansion valves that can work
over a wider operating range, conditions such as at low ambient air temperature and at a high refrigeration
load, causing thermostatic expansion valve (TXV) hunting. Operating at a lower ambient always
saves energy. An electric expansion valve provides more precise control and gives a better
part-load performance than does TXV.
Microprocessor-Based DDC. A DDC unit controller and the correponding functional, optimizing,
and safety controls as well as diagnostics similar to that discussed for water-cooled screw
chillers are always equipped in air-cooled screw chillers.
System Performance. Air-cooled screw chillers are often manufactured at a capacity of 50 to 400
tons (176 to 1400 kW). They are rated at a chilled water leaving evaporator temperature at 44°F
(6.7°C) and a cooling air temperature entering the air-cooled condenser at 95°F (35°C). In
ASHRAE/IESNA Standard 90.1-1999, the minimum performance of air-cooled chillers with a
cooling capacity  150 tons (527 kW) is the same as that for air-cooled scroll chillers of the same
cooling capacity (see Sec. 11.10). For air-cooled chillers in a cooling capacity  150 tons, COP 
2.5 and IPLV  2.5 (EER  8.5 and IPLVEER  8.5). In the early 1990s, the EER and IPLVEER of
an air-cooled screw chiller of a U.S. leading HVAC&R manufacturer were 10.0 (2.93 COP) and
13.2 (3.87 IPLVCOP), respectively.
Location of Installation. Air-cooled screw chillers are usually located on the roof outdoors. The key
point for efficient operation is to have sufficient clearance between the condensing coils and the surroundings,
and between units, according to the manufacturers’ specifications. If the chiller is shut
down during winter, water should be drained from the evaporator in areas where outdoor temperatures
fall below 32°F (0°C). Otherwise, glycol should be added to the chilled water for freeze protection.
Minimum Efficiency Efficiency as of 10 /29/2001
Sizes COP* IPLV COP IPLV
 150 tons 3.8 3.9 4.45 4.5
 150 tons and 4.2 4.5 4.9 4.95
 300 tons
300 tons 5.2 5.3 5.5 5.6
* For a COP of 5.2, the equivalent is 0.68 kW/ton.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.55
REFERENCES
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1998, Refrigeration, Atlanta, GA, 1998.
Beseler, F., Scroll Compressor Technology Comes of Age, Heating / Piping / Air Conditioning, no. 7, 1987,
pp, 67–70.
Beseler, F., New Technology and the Helical Rotary Compressor, Heating/Piping/Air Conditioning, no.1, 1988,
pp. 127–129.
Bivens, D. B., Patron, D. M., and Yokozeki, A., Performance of R-32/R-125/R-134a Mixtures in Systems with
Accumulators or Flooded Evaporators, ASHRAE Transactions, 1997, Part I, pp. 777–780.
Coad,W. J., Refrigeration Piping for DX: Part III—Liquid Line Design, Heating / Piping / Air Conditioning,
April 1990a, pp. 88–90.
Coad,W. J., Refrigeration Piping for DX: Part IV—Suction Line Design, Heating / Piping / Air Conditioning,
May 1990b, pp. 93–95.
Cole, R. A., Reduced Head Pressure Operation of Industrial Refrigeration Systems, Heating / Piping / Air Conditioning,
May 1985, pp. 119–132.
Corr, S., Murphy, T., and Wilkinson, S., Composition Shifts of Zeotropic HFC Refrigerants in Service, ASHRAE
Transactions, 1994, Part II, pp. 538–546.
Ernst, S., Advantages of Screw Compressors, Heating / Piping / Air Conditioning, November 1987, pp. 85–104.
Grim, J. H., Basic Compressor Application, ASHRAE Transactions, 1981, Part I, pp. 837–841.
Grimm, N. R., and Rosaler, R. C., HVAC Systems and Components Handbook, McGraw-Hill, New York, 1997.
Hicks, P., and Adams, P., Jr., Saving Older Refrigeration Systems, Heating / Piping / Air Conditioning, November
1989, pp. 82–92.
Hwang,Y., Judge, J., and Radermacher, R., Experience with Refrigerant Mixtures, ASHRAE Transactions, 1997,
Part I, pp. 765–776.
Kruse, H. H., and Schroeder, M., Fundamentals of Lubrication in Refrigerating Systems and Heat Pumps,
ASHRAE Transactions, 1984, Part II B, pp. 763–783.
Kruse, H., and Wieschollek, F., Concentration Shift When Using Refrigerant Mixtures, ASHRAE Transactions,
1997, Part I, pp. 747–755.
Levins,W. P., Rice, C. K., and Baxter, V. D., Modeled and Measured Effects of Compressor Downsizing in an
Existing Air Conditioner / Heat Pump in the Cooling Mode, ASHRAE Transactions, 1996, Part II, pp. 22–33.
Patterson, N. R., Reciprocating Refrigeration Piping and Control, Heating / Piping / Air Conditioning, April
1981, pp. 78–86.
Pillis, J. W., Advancement in Refrigeration Screw Compressor Design, ASHRAE Transactions, 1986, Part I B,
pp. 219–224.
Rice, C. K., Influence of HX Size and Augmentation on Performance Potential of Mixtures in Air-to-Air Heat
Pumps, ASHRAE Transactions, 1993, Part II, pp. 665–679.
Stamm, R. H., Industrial Refrigeration Compressors, Heating / Piping / Air Conditioning, July 1984, pp. 93–111.
Strong, A. P., Hot Gas Defrost for Industrial Refrigeration, Heating / Piping / Air Conditioning, July 1988,
pp. 71–83.
Takebayashi, M., Sekigami, K., Tsubono, I., Kohsokabe, H., Suefugi, K., and Inaba, K., Performance Improvement
of a Variable-Speed Controlled Scroll Compressor for Household Air Conditioners, ASHRAE Transactions,
1994, Part I, pp. 471–475.
The Trane Company, Reciprocating Refrigeration, The Trane Company, La Crosse, WI, 1977.
The Trane Company, Refrigeration System Piping, La Crosse, WI, 1978.
The Trane Company, Refrigeration Accessories and Controls, La Crosse, WI, 1985a.
The Trane Company, Refrigeration Compressors, La Crosse, WI, 1985b.
The Trane Company, The Scroll Compressor, La Crosse, WI, 1987.
The Trane Company, Series R Centra Vac Rotary Liquid Chillers, La Crosse, WI, 1994.
Uekusa, T., Nakao, M., and Ohshim, K., Control Method of a Cooling Apparatus in Low Outdoor Air Temperatures,
ASHRAE Transactions, 1990, Part I, pp. 200–204.
11.56 CHAPTER ELEVEN
Wang, S. K., Principles of Refrigeration Engineering, vol. 3, Hong Kong Polytechnic, Hong Kong, 1984.
Yencho, J., Purging Noncondensable Gases, Heating / Piping / Air Conditioning, February 1989, pp. 75–79.
Yun, K. W., Supermarket Refrigeration Compressors, ASHRAE Transactions, 1983, Part I B, pp. 596–605.
Zimmern, B., and Sweetser, R. S., The Centrifugal Economizer: Its Match with Screw Compressors, ASHRAE
Transactions, 1986, Part I B, pp. 225–233.
Zubair, S. M., and Bahel, V., Compressor Capacity Modulation Scheme, Heating / Piping / Air Conditioning,
January 1989, pp. 135–143.
REFRIGERATION SYSTEMS: RECIPROCATING, ROTARY, SCROLL, AND SCREW 11.57
CHAPTER 12
HEAT PUMPS, HEAT
RECOVERY, GAS COOLING,
AND COGENERATION SYSTEMS
12.1
12.1 BASICS OF HEAT PUMP AND HEAT
RECOVERY 12.1
Heat Pumps 12.1
Heat Pump Cycle 12.2
Classification of Heat Pumps 12.3
HVAC&R Heat Recovery Systems 12.3
Heat Balance and Building Load
Analysis 12.4
12.2 AIR-SOURCE HEAT PUMP
SYSTEMS 12.5
System Components 12.6
Suction Line Accumulator 12.8
Operating Modes 12.9
System Performance 12.9
Cycling Loss and Degradation Factor
12.11
Minimum Performance 12.12
Defrosting 12.12
Controls 12.13
Capacity and Selection 12.13
12.3 GROUNDWATER HEAT PUMP
SYSTEMS 12.13
Groundwater Systems 12.14
Groundwater Heat Pump System for a
Hospital 12.14
Groundwater Heat Pump Systems for
Residences 12.15
12.4 GROUND-COUPLED AND SURFACE
WATER HEAT PUMP SYSTEMS 12.17
12.5 AIR-TO-AIR HEAT RECOVERY 12.19
Types of Air-to-Air Heat Recovery 12.19
Effectiveness 12.19
Fixed-Plate Heat Exchangers 12.20
Runaround Coil Loops 12.21
Rotary Heat Exchangers 12.21
Heat Pipe Heat Exchangers 12.22
Comparison between Various Air-to-Air
Heat Exchangers 12.25
12.6 GAS COOLING AND
COGENERATION 12.25
Gas Cooling 12.25
Cogeneration 12.25
Gas-Engine Chiller 12.27
Gas Engines 12.27
Exhaust Gas Heat Recovery 12.28
Engine Jacket Heat Recovery 12.28
Cogeneration Using a Gas Turbine 12.29
REFERENCES 12.29
12.1 BASICS OF HEAT PUMP AND HEAT RECOVERY
Heat Pumps
A heat pump extracts heat from a heat source and rejects heat to air or water at a higher temperature.
During summer, the heat extraction, or refrigeration effect, is the useful effect for cooling,
whereas in winter the rejected heat alone, or rejected heat plus the supplementary heating from a
heater, forms the useful effect for heating.
A heat pump is a packaged air conditioner or a packaged unit with a reversing valve or other
changeover setup. A heat pump has all the main components of an air conditioner or packaged unit:
fan, filters, compressor, evaporator, condenser, short capillary tube, and controls. The apparatus for
changing from cooling to heating or vice versa is often a reversing valve, in which the refrigerant
flow to the condenser is changed to the evaporator. Alternatively, air passage through the evaporator
may be changed over to passage through the condenser. A supplementary heater is often provided
when the heat pump capacity does not meet the required output during low outdoor temperatures.
A heat pump system consists of heat pumps and piping work; system components include heat
exchangers, heat source, heat sink, and controls to provide effective and energy-efficient heating
and cooling operations. HCFC-22, HFC-134a, and HFC-407C are the most widely used refrigerants
in new heat pumps. According to the data in the EIA’s Commercial Buildings Characteristics, for
the 57 billion ft2 (5.3 m2) of air conditioned commercial building floor area in the United States in
1992, the use of heat pumps for heating and cooling was about 15 percent (by floor space).
Heat Pump Cycle
A heat pump cycle comprises the same processes and sequencing order as a refrigeration cycle except
that the refrigeration effect q14 or qrf, and the heat pump effect q23, both in Btu / lb (J / kg), are
the useful effects, as shown in Fig. 12.1. As defined in Eqs. (9.7) and (9.9), the coefficient of performance
of a refrigeration system COPref is
(12.1)
where h4, h1  enthalpy of refrigerant entering and leaving evaporator, respectively, Btu / lb (J /kg)
Win  work input, Btu/ lb (J / kg)
The coefficient of performance of the heating effect in a heat pump system COPhp is
(12.2)
and the useful heating effect q23 can be calculated as
(12.3) q23  h2  h3  h1  h4  h2  h1
COPhp 
q23
Win
COPref 
h1  h4
W

q14
Win
12.2 CHAPTER TWELVE
FIGURE 12.1 Heat pump cycle: (a) schematic diagram; (b) cycle on p-h diagram.
where h2  enthalpy of hot gas discharged from compressor, Btu / lb (J /kg)
h3  enthalpy of subcooled liquid leaving condenser, Btu / lb (J /kg)
Here polytropic compression is a real and irreversible process. Both the subcooling of the liquid
refrigerant in the condenser and the superheating of the vapor refrigerant after the evaporator
increase the useful heating effect q23. Excessive superheating, which must be avoided, leads to a
too-high hot-gas discharge temperature and to a lower refrigeration capacity in the evaporator.
Classification of Heat Pumps
According to the types of heat sources from which heat is absorbed by the refrigerant, currently
used heat pump systems can be mainly classified into two categories: air-source and water-source
heat pump systems. Water-source heat pumps can again be subdivided into water-source, groundwater,
ground-coupled, and surface water heat pump systems. Water-source heat pump systems are
discussed in Chap. 29.
Heat pump systems are often energy-efficient cooling/heating systems. Many new technologies
currently being developed, such as engine-driven heat pumps, may significantly increase the system
performance factor of the heat pump system. Ground-coupled heat pumps with direct-expansion
ground coils provide another opportunity to increase the COP of the heat pump system.
HVAC&R Heat Recovery Systems
An HVAC&R heat recovery system converts waste heat or cooling from any HVAC&R process to
useful heat or cooling. Here heat recovery is meant in a broad sense. It includes both waste heat and
cooling recovery. An HVAC&R heat recovery system includes the following:
 The recovery of internal heat loads—such as heat energy from lights, occupants, appliances, and
equipment inside the buildings—by reclaiming the heat rejected at the condenser and absorber of
the refrigeration systems
 The recovery of heat from the flue gas of the boiler
 The recovery of heat from the exhaust gas and water jacket of the engine that drives the
HVAC&R equipment, especially engine-driven reciprocating vapor compression systems
 The recovery of heat or cooling from the exhaust air from air conditioning systems
Although heat pump systems sometimes are used to recover waste heat and convert it into a
useful effect, a heat recovery system is different from a heat pump system in two ways:
 In a heat pump system, there is only one useful effect at a time, such as the cooling effect in summer
or the heating effect during winter. In a heat recovery system, both its cooling and heating
effects may be used simultaneously.
 From Eq. (12.2), the coefficient of performance of the useful heating effect in a heat pump is
COPhp  q23 /Win, whereas for a heat recovery system, the coefficient of performance COPhr is
always higher if both cooling and heating are simultaneously used and can be calculated as
(12.4)
A heat pump is an independent unit. It can operate on its own schedule, whereas a heat recovery
system in HVAC&R is usually subordinate to a refrigeration system or to some other system that
produces the waste heat or waste cooling.
COPhr 
q14  q23
Win
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.3
Heating or cooling produced by a heat recovery system is a by-product. It depends on the operation
of the primary system. A heat recovery system can use waste heat from condenser water for
winter heating only if the centrifugal chiller is operating.
A centrifugal chiller that extracts heat from a surface water source (e.g., a lake) through its evaporator
and uses condensing heat for winter heating is a heat pump, not a heat recovery system. Heat
recovery systems that are subordinate to a centrifugal or absorption refrigeration system are discussed
in Chaps. 13 and 14, respectively. Recovery of waste heat from industrial manufacturing
processes to provide heating shows great potential for saving energy. These heat recovery systems
must be closely related to the specific requirements of corresponding manufacturing processes and
are not discussed here.
Heat Balance and Building Load Analysis
The building load consists of transmission gain or loss, solar radiation, ventilation load and infiltration
load, people, electric lights, appliances (or equipment), and heat gains from fans and
pumps. Building load is actually the load of the cooling or heating coil in an air-handling unit, or
DX coil in a packaged unit or air conditioner. On hot summer days, solar radiation and the latent
ventilation load must be included in the building load. However, both sunny and cloudy days may
occur in cold weather; therefore, building load is calculated and analyzed with and without solar
radiation in cold weather, especially for the control zones facing south in a building in northern
latitudes, or facing north in southern latitudes, where solar radiation is often a primary cooling
load on sunny days.
Figure 12.2 shows the building load analysis of a typical floor of a multistory building without
solar radiation in winter. In this figure, line ABCDE represents the building load curve at various
outdoor temperatures To, in °F (°C). Point A on this curve represents the summer design refrigeration
load Qrl, in MBtu/h (kW). During summer design load, all the space cooling loads are offset
by the cold supply air, and the condensing heat of the refrigeration system is rejected to the cooling
tower.
When the outdoor temperature To drops below 75°F (23.9°C) on the left side of point B on the
building load curve, then the following things occur:
 The perimeter zone may suffer a transmission loss.
 Solar radiation is excluded.
 The latent load of the outdoor ventilation air is no longer included in the building load because
the outdoor air is often drier than the space air.
If the outdoor temperature To  TF, hot condenser water is supplied to the heating coils in the
perimeter zone to satisfy the heating load. Here TF indicates the outdoor temperature at point F.
When the outdoor temperature drops to TD, the heat recovered from the interior zone plus the
power input to the compressor is exactly equal to the heating load of the perimeter zone. No supplementary
heating is needed. Point D is called the break-even point of the building. When To  TD,
supplementary heating is necessary to maintain a desirable space temperature.
As the outdoor temperature falls to TH, the cooling coil load in the interior zone becomes
zero. The refrigeration compressors are turned off. No recovery of condensing heat is possible.
When To 	 TH, all the heat needed for the perimeter zone will be provided by the supplementary
heating.
The area of triangle FGH indicates the heat energy recovered from the internal loads in the
interior zone, which is used to offset the heat losses in the perimeter zone by means of hot
condenser water. Similar building load curves with solar radiation for the entire building and for
the south-facing zones in the building should be calculated and analyzed in winter. For building
load with solar radiation, break-even point D will move to a lower To. Recovered heat will be
greater, and supplementary heating will be less. For south-facing zones in buildings in northern
12.4 CHAPTER TWELVE
climates, cold supply air may be required during sunny days in the perimeter zone even in
winter.
12.2 AIR-SOURCE HEAT PUMP SYSTEMS
In an air-source heat pump system, outdoor air acts as a heat source from which heat is extracted
during heating, and as a heat sink to which heat is rejected during cooling. Since air is readily available
everywhere, air-source heat pumps are the most widely used heat pumps in residential and
many commercial buildings. The cooling capacity of most air-source heat pumps is between 1 and
30 tons (3.5 and 105 kW).
Air-source heat pumps can be classified as individual room heat pumps and packaged heat
pumps. Individual room heat pumps serve only one room without ductwork. Packaged heat pumps
can be subdivided into rooftop heat pumps and split heat pumps.
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.5
FIGURE 12.2 Building load analysis of a typical floor in a multistory building without solar radiation.
System Components
Most air-source heat pumps consist of single or multiple compressors, indoor coils through which
air is conditioned, outdoor coils where heat is extracted from or rejected to the outdoor air, capillary
tubes, reversing valves that change the heating operation to a cooling operation and vice versa, an
accumulator to store liquid refrigerant, and other accessories. A typical rooftop packaged heat
pump that uses HCFC-22 as refrigerant is shown in Fig. 12.3. In this heat pump, dual circuits,
consisting of two compressors, two indoor coils, two outdoor coils, two throttling devices, and two
reversing valves, are often used for better capacity control and better defrosting control for larger
heat pumps.
Compressor. Reciprocating and scroll compressors are widely used in heat pumps. In Bucher
et al. (1990) the median service life of the compressor in a heat pump is 14.5 years, and the median
service life of a heat pump system is 19 years, depending on the conditions of operation and
12.6 CHAPTER TWELVE
FIGURE 12.3 A typical rooftop heat pump: (a) schematic diagram;
(b) reversing valve, cooling mode; (c) reversing valve, heating mode.
maintenance. Median service life is the age at which 50 percent of the units have been removed
from service and 50 percent remain in service.
Indoor Coil. In an air-source heat pump, the indoor coil is not necessarily located inside the
building. The indoor coil in a rooftop packaged heat pump is mounted on the rooftop. However, an
indoor coil always heats and cools the indoor supply air. During cooling operation, the indoor
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.7
FIGURE 12.3 (Continued)
coil acts as an evaporator. It provides the refrigeration effect to cool the mixture of outdoor and
recirculating air when the heat pump is operating in the recirculating mode. During heating
operation, the indoor coil acts as a condenser. The heat rejected from the condenser raises the
temperature of the conditioned supply air. For heat pumps using halocarbon refrigerants, the indoor
coil is usually made from copper tubing and corrugated aluminum fins.
Outdoor Coil. The outdoor coil acts as a condenser during cooling and as an evaporator to extract
heat from the outdoor atmosphere during heating. When an outdoor coil is used as a condenser, a
series-connected subcooling coil often subcools the refrigerant for better system performance. An
outdoor coil always deals with outdoor air, whether it acts as a condenser or an evaporator. Like the
indoor coil, an outdoor coil is usually made of copper tubing and aluminum fins for halocarbon
refrigerants. Plate or spine fins are often used instead of corrugated fins to avoid clogging by dust
and foreign matter.
Reversing Valve. Reversing valves are used to guide the direction of refrigerant flow when cooling
operation is changed over to heating operation or vice versa. The rearrangement of the connections
between four ways of flow—compressor suction, compressor discharge, evaporator outlet,
and condenser inlet—causes the functions of the indoor and outdoor coils to reverse. It is therefore
called a four-way reversing valve.
A typical four-way reversing valve consists of a hollow cylinder with an internal slide to position
the flow paths of the refrigerant. The slide is driven by the differential between the discharge
and suction pressure. Both are introduced to opposite ends of the cylinder by a pilot valve that is
energized by a solenoid coil. The operation of this typical four-way reversing valve is shown in
Fig. 12.3b and c.
A reversing valve is a highly reliable system component. The efficiency losses altogether including
leakage, heat transfer, and the pressure drop across the reversing valve cause a decrease of 4 to
7 percent in heat pump performance. Other accessories, such as a filter dryer, sight glass, strainer,
liquid level indicator, solenoid valves, and manual shutoff valves as well as refrigerant piping are
the same as described in Chap. 11.
Suction Line Accumulator
An accumulator is usually installed on the suction line prior to the inlet of the compressor in a heat
pump system to store the liquid refrigerant prior to the defrosting process. Bivens et al. (1997)
conducted tests on a 2.5-ton (8.8-kW) split residential heat pump with a scroll compressor and an
accumulator using HFC-407C and HCFC-22 as refrigerant. Only a slight change in concentration
shift occurred when there was no liquid refrigerant in the accumulator. There were significant
changes in concentration shift during heating when the temperatures in the evaporator and accumulator
were lower. The accumulator contained about 30 percent of the refrigerant as liquid during
high-temperature heating (outdoor air temperature at 47°F, or 8.3°C) and about 50 percent of the
refrigerant as liquid during low-temperature heating (outdoor air temperature at 17°F, or 8.3 °C).
Compared with HCFC-22 as refrigerant, the capacity of the HFC-407C heat pump increased to 101
to 105 percent of the HCFC-22 capacity during high-temperature heating, and increased to 106 to
109 percent of the HCFC-22 capacity during low-temperature heating. The COP of the HFC-407C
heat pump dropped to 0.96 to 0.97 of the HCFC-22 COP during cooling and dropped to 0.94 and
0.95 of the HCFC-22 COP during heating.
Nutter et al. (1996) conducted experiments on a scroll heat pump of 3-ton (10-kW) cooling capacity.
The heat pump was operated at heating mode of ambient air of 1.7°F (16.8°C) dry-bulb
temperature and 80 percent relative humidity in a psychrometric room. Test results showed that the
COP is higher for the heat pump without an accumulator first until a power surge at 40-min operation.
This caused the integrated cyclic COP for the heat pump without an accumulator of 2.08
which fell below that of the COP of the heat pump system with an accumulator of 2.12. Such a
12.8 CHAPTER TWELVE
phenomenon may be the result of entrained liquid refrigerant that began to adversely affect the
compressor performance by reducing discharge temperature and pressures.
Operating Modes
The operation of an air-source heat pump can be divided into cooling mode and heating mode.
Cooling Mode. When the discharge air temperature sensor detects an increase in the air
temperature above a predetermined limit at the exit of the indoor coil, cooling is required in the
air-source heat pump. The direct digital controller (DDC) unit deenergizes the pilot solenoid
valve, as shown in Fig. 12.3b. The low-pressure suction vapor is now connected to the left-hand
end of the slide, and the high-pressure hot gas between the pistons of the slide then pushes
the piston toward the left end of the cylinder. The indoor coil now acts as an evaporator and
extracts heat from the conditioned air flowing through the indoor coil. After evaporation, vapor
refrigerant from the indoor coil passes through the sliding connector of the slide and flows to the
suction line. Hot gas discharged from the compressor is led to the outdoor coil, which now acts
as a condenser.
An economizer cycle can be used when an outdoor air sensor detects the outdoor temperature
dropping below a specific limit during cooling mode. The DDC unit controller opens and modulates
the outdoor damper to admit cold outdoor air to maintain a preset discharge air temperature.
Heating Mode. When the discharge air sensor detects a drop in air temperature below a predetermined
limit at the exit of the indoor coil, heating is required. The DDC unit controller energizes the
pilot solenoid valve as shown in Fig. 12.3c. The plunger of the solenoid valve moves upward and
connects the right-hand end of the slide to the low-pressure suction line. High-pressure hot gas then
pushes the pistons toward the right-hand end of the cylinder. Consequently, the hot gas from the
compressor is discharged to the indoor coil, which now acts as a condenser. Heat is rejected to the
recirculating air, and then the hot gas is condensed into liquid form. Liquid refrigerant flows
through the capillary tube and then vaporizes in the outdoor coil, which extracts heat from outdoor
air. The outdoor coil now acts as an evaporator.
Electric heating can be used for supplementary heating. When the discharge air temperature
sensor detects a drop in air temperature further below preset limits, the electric heater can be
energized in steps to maintain the required discharge air temperature. Supplementary heating
is energized only when the space heating load cannot be offset by the heating effect of the
heat pump.
ASHRAE/IESNA Standard 90.1-1999 mandates that heat pumps equipped with internal electrical
resistance heaters shall have controls to prevent supplemental heater operation when the heating
load can be met by the heat pump alone during heating or setback recovery.
System Performance
System performance of an air-source heat pump can be illustrated by plotting the heating
coil load Qch, the cooling coil load Qcc, the heating capacity of the heat pump Qhp, the cooling
capacity of the heat pump Qrc, the coefficient of performance of the heat pump during heating
COPhp, and the energy efficiency ratio of the heat pump during cooling EERhp against the outdoor
temperature, as shown in Fig. 12.4. Coil loads and heat pump capacities all are expressed in
Btu/h (W).
The heating capacity of an air-source heat pump using a reciprocating or a scroll compressor can
be calculated as
(12.5) Qhp  60V?p suc 	v(q14  q23)  60V?p suc 	v[(h1  h4)  (h2  h3)]
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.9
where piston displacement of reciprocating compressor or suction volume flow rate, cfm
[m/ (60 s)]
h1, h4  enthalpy of refrigerant leaving and entering evaporator, Btu / lb (J /kg)
h2, h3  enthalpy of hot gas discharged from compressor and enthalpy of refrigerant leaving
condenser, Btu / lb (J / kg)
suc  density of suction vapor, lb/ ft3 (kg/m3)
	v  volumetric efficiency
The cooling capacity of an air-source heat pump Qrc can be calculated from Eq. (9.25) as described
in Chap. 9. The cooling coil load Qcc should be offset by the cooling capacity Qrc of the heat pump,
and the heating coil load Qch should be offset by the heating capacity Qhp provided by the heat
pump. All Qcc, Qrc, Qch, and Qhp are expressed in Btu/h (W).
When an air-source heat pump is operated in cooling mode during summer, the condensing temperature
Tcon and condensing pressure pcon drop as the outdoor temperature To falls. These decreases
result in higher 	v and therefore an increase in the cooling capacity of the heat pump Qcc as well as
the energy efficiency ratio of the heat pump during cooling EERhp. The fall of To also causes a decrease
in the space cooling load and the cooling coil load Qcc. The intersection of the Qrc and Qcc
curves, point D, indicates the maximum cooling coil load that can be offset by the cooling capacity
of the selected heat pump.
When the heat pump is operated in heating mode, a fall of To causes a decrease in the evaporating
temperature Tev. A lower Tev results in a lower 	v, a smaller refrigeration effect q14, and a lower
density of suction vapor suc. All these effects result in a smaller heating capacity of the heat pump
Qhp. Although the work input Win increases as To and Tev decrease, the effect of the increase in Win
on the increase of Qhp is small.
The fall of To also causes a rise in space heating load Qrh and the heating coil load Qch. The intersection
of the heating capacity Qhp curve and the heating coil load Qch curve, at point B, is the
balance point at which the heating capacity of the heat pump is equal to the required heating coil
V?
p 
12.10 CHAPTER TWELVE
FIGURE 12.4 System performance of a rooftop heat pump.
load. When the outdoor temperature To drops below this balance point (such as To  10°F or
12.2°C), as shown in Fig. 12.4, supplementary heating from the electric heater or other heat
source is required to maintain a preset discharge air temperature. The coefficient of performance of
the heat pump during heating COPhp also drops as To falls. The heat pump and supplementary heating
operate simultaneously until the COPhp drops below a certain value, such as below 1 when the
use of an electric heater is more cost-effective than to operate the heat pump. The heat pump should
be turned off. As discussed in Sec. 11.10, a scroll heat pump has a flatter capacity curve when outdoor
air temperature To varies than a reciprocating heat pump. During heating mode operation, a
flatter heating capacity Qhp curve reduces the amount of supplementary heating. During cooling
mode operation in hot summer, when To exceeds the design value, the increase of the space temperature
will be less for a flatter capacity curve Qrc than for a steeper one.
Cycling Loss and Degradation Factor
For split packaged air-source heat pumps, indoor coils are located inside the building and outdoor
coils are mounted outdoors. When an on/off control is used for the compressor, during the off period,
refrigerant tends to migrate from the warmer outdoor coil to the cooler indoor coil in summer
and from the warmer indoor coil to the cooler outdoor coil during winter. When the compressor
starts again, the transient state performance shows that a 2- to 5-min operating period of reduced
capacity is required before the heat pump can operate at full capacity. Such a loss due to cycling of
the compressor is called cycling loss.
According to the Department of Energy (DOE) test procedure (1986) and O’Neal et al. (1991),
cycling losses are illustrated by the following parameters: part-load factor (PLF), cooling load
factor (CLF), and degradation coefficient Cd. PLF can be calculated as
(12.6)
where EERcyc  energy-efficiency ratio (EER) of air-source heat pump during a whole cycle,
Btu/hW(COP)
EERss  EER of air-source heat pump if it is operated at steady-state, i.e., when compressors
and fans are operated continuously, Btu/hW(COP)
The cooling load factor can be calculated as
(12.7)
where Hcyc  total energy (cooling or heating) delivered during a cycle, Btu
Qss  steady state cooling or heating rate, Btu /h (W)
t  time of a cycle, h
The degradation coefficient Cd can be calculated as
(12.8)
Cycling losses depend on (1) the cycling rate (whether it is 2, 3, 4, or 5 cycles per hour), (2)
the indoor-outdoor temperature difference, and (3) the fraction of on-time per cycle. At design
conditions, theoretically, PLF  1. In Baxter and Moyers (1985), field tests of a typical heat
pump in an unoccupied single-family house in Knoxville, Tennessee, between 1981 and 1983
showed that the degradation coefficient Cd in the heating season is 0.26, and Cd in the cooling
season is 0.11.
Cd 
1  PLF
1  CLF
CLF 
Hcyc
Qss t
PLF 
EERcyc
EERss
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.11
Minimum Performance
To encourage the use of energy-efficient air conditioners and heat pumps, ASHRAE/IESNA Standard
90.1-1999 mandates the minimum efficiency requirements for air-source heat pumps at various
cooling capacities during cooling and heating mode as follows:
Cooling Minimum Efficency as of Heating Minimum
Size description efficiency 10/29/2001 rating condition efficiency 10/29/2001
 65,000 Btu/h Split system 10.0 SEER 10.0 SEER Split system 6.8 HSPF 6.8 HSPF
Single package 9.7 SEER 9.7 SEER Single package 6.6 HSPF 6.6 HSPF

 65,000 Btu/h and Split system and 8.9 EER 10.1 EER 47°Fdb/43°Fwb 3.0 COP 3.2 COP
 135,000 Btu/h single package outdoor air

 135,000 Btu/h and Split system and 8.5 EER 9.3 EER 47°Fdb/43°Fwb 2.9 COP 3.1 COP
 240,000 Btu/h single package outdoor air
(
 135,000 Btu/h)

 240,000 Btu/h Split system and 8.5 EER 9.0 EER
single package 7.52 IPLV 9.22 IPLV
Single-phase packaged units  65,000 Btu/h (19 kW) are regulated by U.S. National Appliance
Energy Conservation Act (NAECA) of 1987, whose HSPF rating includes all usage of internal electric
resistance heating and meets the requirements. For details, refer to ASHRAE/IESNA Standard
90.1-1999. ILPVs and part-load rating conditions are only applicable to equipment with capacity
modulation. Deduct 0.2 from the required EERs and IPLVs for units with heating devices other
than electric resistance heating.
Defrosting
Most air-source heat pumps use the reverse cycle to melt the frost that formed on the outdoor coil
during heating mode operation in cold weather. The reverse cycle defrost switches the heating mode
operation, in which the outdoor coil acts as an evaporator, to cooling mode operation, where the
outdoor coil acts as a condenser. Hot gas is forced into the outdoor coil to melt the frost that
accumulated there. After the frost is melted, the heat pump is switched back to normal heating
mode operation.
O’Neal and Peterson (1990) described the defrosting process of an air-source heat pump using
HCFC-22 as refrigerant with a short capillary tube of 0.059-in. (1.5-mm) diameter as the throttling
device. When the reversing valve was energized, the outdoor fan stopped and the defrosting cycle
began. The suction and discharge pressures equalized. The sudden decrease in pressure in the indoor
coil caused the liquid refrigerant to vaporize. The compressor became temporarily starved and
pulled the pressure in the indoor coil down to 23 psia (159 kPa abs.) about 1 min after defrosting
started. Once the suction pressure had fallen low enough, refrigerant flow began to increase. In the
interval from 1.0 to 3.5 min after defrosting, the refrigerant changed from vapor to saturated liquid
upstream of the short capillary tube. This change caused a substantial increase in refrigerant flow,
and frost was melted at the outdoor coil. After 6 min, the refrigerant flow fell 30 percent because of
the subcooling in the outdoor coil. Defrosting usually lasts a few minutes up to about 10 min, depending
on the frosting accumulation.
During defrost, a cold supply air from the indoor coil may cause a low space temperature and a
draft. Supplementary electric heating should be considered to maintain an acceptable indoor temperature.
ASHRAE Standard 90.1-1999 permits supplementary heating to be used during outdoor
coil defrost cycles.
Defrosting only takes place on the outdoor coil. The initiation and control of defrosting can be
performed by a clock or an intelligent or adaptive timer. It can also be controlled by measuring the
12.12 CHAPTER TWELVE
capacity of the unit or by measuring the temperature differential between the refrigerant inside the
outdoor coil and the ambient air. Defrosting terminates when the temperature of liquid refrigerant
leaving the outdoor coil (or the coil temperature) rises above 60°F (15.6°C).
Having or not having a suction line accumulator also affects the performance of the heat pump.
Nutter et al. (1996) showed that the refrigerant flow rate averaged 18 percent higher and had a
7 percent shorter defrost cycle for heat pumps without an accumulator than those with an accumulator.
Controls
For reciprocating heat pumps, on/off, speed modulation, and cylinder unloading (as described in
Sec. 11.5) capacity controls are generally used. For scroll heat pumps, on/off, variable-speed, and
variable-displacement modulation capacity control are usually used.
Either the discharge air temperature or the return temperature can be used as the criterion to
change automatically from cooling mode to heating mode and vice versa. A dead band of 2 to 3°F
(1.1 to 1.7°C) and a time delay are always required between cooling and heating mode operations
to prevent short cycling.
Most of the packaged heat pumps provide specific safety controls of high pressure, low pressure,
head pressure, or low ambient control; freezing protection of indoor coil; protection from
overloading; and supplementary heating. The principle and operation of these controls are the same
as described in Sec. 11.5. A microprocessor-based DDC system controller may be used to integrate
all the controls in one package and to add time delay, compressor lockout, loss of refrigerant
charge, and short-cycling protection controls to the sequence control of heat pump and gas furnace
in heating mode operation, and air economizer and refrigeration capacity control in cooling mode
operation.
Capacity and Selection
Air-source heat pump capacity is selected according to its cooling capacity because supplementary
heating may be required under winter design conditions. Also, the rated cooling capacity
at summer design conditions is often greater than the rated heating capacity at winter design
conditions.
When an air-source heat pump is installed directly inside or above the conditioned space, the
noise generated by the heat pump must be taken into consideration. Attenuation remedies should be
provided if necessary to maintain an NC curve at an acceptable level in the conditioned space.
Sound control is discussed in Chap. 19. In 1992, air-source heat pump products were available
ranging in cooling capacity from a fraction of a ton to about 40 tons (few kilowatts to 140 kW) with
an indoor airflow of 16,000 cfm (7550 L/ s).
12.3 GROUNDWATER HEAT PUMP SYSTEMS
Groundwater heat pump (GWHP) systems use well water as a heat source during heating and as a
heat sink during cooling. When the groundwater is more than 30 ft (9 m) deep, its year-round temperature
is fairy constant. Groundwater heat pump systems are usually open-loop systems. They are
mainly used in low-rise residences in northern climates such as New York or North Dakota. Sometimes
they are used for low-rise commercial buildings where groundwater is readily available and
local codes permit such usage.
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.13
Groundwater Systems
For commercial buildings, the design engineer must perform a survey and study the site and surroundings
to define the available groundwater sources. The design engineer should be fully aware
of the legalities of water rights.
A test well should be drilled to ensure the availability of groundwater. If water is corrosive, a
plate-and-frame heat exchanger may be installed to separate the groundwater and the water entering
the water coil in the water-source heat pump (WSHP).
Usually, two wells are drilled. One is the supply well, from which groundwater is extracted by
submersible pump impellers and supplied to the WSHPs. The other well is a recharge or injection
well. Groundwater discharged from the WSHP is recharged to this well. The recharged well should
be at least 100 ft (30 m) away from the supply well. Using a recharge well provides for resupply to
the groundwater and prevents the collapse of the building foundation near the supply well due to
subsidence. If the quality of groundwater meets the requirement and if local codes permit, groundwater
discharged from the WSHP can be used as the service water or can be drained to the nearby
river, lake, or canal. The groundwater intake water screen of the supply well may be located in several
levels where water can be extracted. For a small supply well for residences, the pump motor is
directly connected to the submersible pump underneath the impellers, whereas for large wells, the
motor is usually located at the top of the supply well. Information regarding groundwater use regulations
and guidelines for well separation and for the construction of supply and recharge wells are
included in Donald’s (1985) Water Source Heat Pump Handbook, published by the National Water
Well Association (NWWA).
If the temperature of the groundwater is below 50°F (10°C), direct cooling of the air in the
WSHP should be considered. If the groundwater temperature exceeds 55°F (12.8°C) and is lower
than 70°F (21.1°C), precooling of recirculating air or makeup air may be economical.
Because the groundwater system is an open-loop system, it is important to minimize the vertical
head to save pump power. Air-conditioning and Refrigeration Institute (ARI) Standard 325-85 recommends
that the groundwater pump power not exceed 60 W/gpm (950 W s /L). If the pump effi-
ciency is 0.7, the allowable head for the well pump, including static head, head loss across the
water coil or heat exchanger, valves, and piping work losses, should preferably not exceed 220 ft
WC (66 m WC). If a recharge well is used and the discharge pipe is submerged under the water
table level in the recharge well, as shown in Fig. 12.6, the groundwater system is most probably a
closed-loop system, depending on whether the underground water passage between the supply and
recharge well is connected or broken.
Groundwater Heat Pump System for a Hospital
A groundwater heat pump system described in Knipe (1983) was a 1980 retrofit project for a hospital
in Albany, New York (see Fig. 12.5). Altogether, 540 gpm (34 L/s) of groundwater was supplied
from a 12-in.- (300-mm-) diameter, 500-ft- (150-m-) deep well. During summer, the groundwater
temperature was 58°F (14.4°C), and in winter it dropped to 46°F (7.8°C). About one-third of the
groundwater was supplied to a makeup air-handling unit. The other two-thirds was sent to a heat
pump which had the following operating characteristics:
Summer:
 Well : condenser : discharged to nearby river
 Evaporator : chilled water to AHU and terminals : evaporator
Winter:
 Well : evaporator : chilled water to AHU and terminals : discharged to nearby river
 Condenser : domestic hot water preheat : perimeter heating : condenser
12.14 CHAPTER TWELVE
In summer, groundwater entered the condenser at 58°F (14.4°C) and left at 68°F (20°C). In winter,
groundwater entered the evaporator at 46°F (7.8°C), and left at 43°F (6.1°C).
The original installation had two 200-ton (700-kW) absorption chillers. Winter heating was supplied
from gas-fired boilers. After retrofit, this groundwater heat pump system, including precooling
and preheating, saved 30 percent of the energy used compared to the previous year’s expense.
Groundwater Heat Pump Systems for Residences
A typical groundwater heat pump system for a residence is shown in Fig. 12.6. Such a heat pump
system usually has a rated heating capacity from 24,000 to 60,000 Btu/h (7030 to 17,580 W).
Groundwater is extracted from a supply well by means of a submersible well pump and is forced
through a precooler or direct cooler. Then groundwater enters the water coil in the water-source
heat pump. After that, groundwater is discharged to a recharge well. If the recharge pipe is submerged
underneath the water table, as described in the previous section, such a groundwater system
is most probably a closed-loop system. In the recharge well, the water level is raised in order to
overcome the head loss required to force the groundwater discharging from the perforated pipe wall
through the water passage underground. The vertical head required is the difference between the
water levels in the supply and recharge wells, as shown in Fig. 12.6.
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.15
FIGURE 12.5 A typical groundwater heat pump system for a hospital.
The water-source heat pump for residential GWHP systems has a structure similar to the WSHP
in the water-source heat pump system (which is discussed in Chap. 29) except that a precooler or a
direct cooler may be added.
Operating parameters and characteristics for a groundwater heat pump system are as follows:
 The groundwater flow rate for a water-source heat pump should vary from 2 to 3 gpm per 12,000
Btu/h (0.036 to 0.054 L/ s per kW) heating capacity. The well pump must be properly sized. A
greater flow rate and an oversized well pump are not economical.
 The pressure drop of the groundwater system should be minimized. The pressure drop per 100 ft
(30 m) of pipe should be less than 5 ft /100 ft (5 m/100 m) of length. Unnecessary valves should
not be installed. Gate valves or ball valves should be used instead of globe valves to reduce
pressure loss. A water tank is not necessary.
 Direct cooling or precooling of recirculating air by means of groundwater increases significantly
the EER of the GWHP system.
 Water containing excessive concentrations of minerals causes deposits on heat pump water coils
that reduce the heat pump performance.
 In locations where the number of annual heating degree-days HDD65 exceeds 7000, more than
80 percent of the operating hours of the GWHP systems are for space heating.
12.16 CHAPTER TWELVE
FIGURE 12.6 A typical residential groundwater heat pump system.
 An electric heater may be used for supplementary heating in cold climates or in other locations
where it is necessary. When the heating capacity of GWHP is equal to or even greater than the
heating load operation of the electric heater must be avoided.
 Water-source heat pumps should be properly sized. They should not be operated for short-cycle durations
(cycles less than 5 min) in order to prevent cycling losses and excessive wear and tear on the
refrigeration system components. Cycle durations between 10 and 30 min are considered appropriate.
 Extraction and discharge of groundwater must comply with local codes and regulations.
 The temperature of groundwater tends to increase with its use.
A parameter called the seasonal performance factor (SPF) is often used to assess the performance
of a groundwater heat pump system. SPF is defined as
(12.9)
where Qsup  all heat supplied by GWHP during heating season, Btu (kJ)
Qcons  energy consumed during heating season, Btu (kJ)
For groundwater heat systems with a cooling capacity Qrc  135,000 Btu/h (40 kW),
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency requirements of 11.0
EER for water-source heat pumps using groundwater during cooling mode when the entering water
is 70°F (21.1°C). As of October 29, 2001, the minimum efficiency requirement is 16.2 EER when
the entering water is 59°C (15°C). During heating mode, the minimum efficiency requirement is 3.4
COP when the entering water temperature is 70°F (21.1°C). As of October 29, 2001, the minimum
efficiency requirement is 3.6 COP when the entering water temperature is 50°F (10°C).
A groundwater heat pump system has a fairly constant COP even if the outdoor air temperature
varies. According to Rackliffe and Schabel (1986), the SPF and EER for 15 single-family houses in
New York State from 1982 to 1984 were as follows:
SPF 
Qsup
Qcons
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.17
1.9 to 3 average 2.3
5.6 to 14 average 9.2
System SPF (heating)
System average EER (cooling)
In many locations, groundwater heat pump systems usually have a higher SPF and seasonal EER
than air-source heat pumps. GWHP system capacity remains fairly constant at very low and very
high outdoor temperatures.
The main disadvantage of a groundwater heat pump system is its higher initial cost. More maintenance
is required for systems using water with high mineral content. If the water table is 200 ft
(60 m) or more below ground level, the residential groundwater heat pump system is no longer
energy-efficient compared to high-efficiency air-source heat pumps.
12.4 GROUND-COUPLED AND SURFACE WATER
HEAT PUMP SYSTEMS
Ground-coupled heat pump systems can be categorized as ground coil heat pump systems and
direct-expansion ground-coupled heat pump systems. Of the ground coil heat pump systems, both
horizontal and vertical coils are used. Many types of direct-expansion ground-coupled heat pump
systems are still being developed. In fact, horizontal ground coil heat pump systems are the most
widely used ground-coupled heat pump systems. A horizontal ground coil heat pump system is
shown in Fig. 12.7. It is actually a closed-loop water-source heat pump system. The water-source
heat pump has a heating capacity between 24,000 and 60,000 Btu/h (7030 and 17,580 W).
A horizontal ground coil is often made of polyethylene or polybutylene tubes of 1- to 2-in. (25-
to 50-mm) external diameter in serpentine arrangement. In Ball et al. (1983) the horizontal coil is
buried in two layers in a trench, typically at 4 and 6 ft (1.2 and 1.8 m) deep. Spacing between the
tubes varies from 2 to 8 ft (0.6 to 2.4 m). A general guideline is to use 215 to 430 ft (65 to 130 m)
of copper/ steel tubing for each ton (3.5 kW) of capacity for cooling. When a horizontal ground coil
heat pump used in cooling mode was subjected to a heat rejection temperature that exceeded 100°F
(37.8°C), some cracking of polybutylene pipe occurred.
12.18 CHAPTER TWELVE
FIGURE 12.7 Ground-coupled heat pump system: (a) Schematic diagram; (b) grid-type ground coil.
A vertical ground coil is buried from 30 to 200 ft (6 to 60 m) deep in drilled holes. About 200 to
250 ft (60 to 75 m) of pipe is needed for each 12,000 Btu/h (3.5 kW) of heating capacity. A vertical
ground coil requires less land area than a horizontal ground coil.
If the temperature of the fluid circulated in the water coil and the horizontal ground coil could
drop below 32°F (0°C), then aqueous solutions of ethylene, propylene glycol, or calcium chloride
should be used to protect the system from freezing. A water flow rate between 2 and 3.5 gpm per
12,000 Btu/h (0.036 and 0.063 L/ s per kW) heating capacity is usually used.
In northern climates, horizontal ground coil heat pump systems have a heating SPF of 2.5 to 3
and a cooling seasonal EER of 10.5 to 13.5. There is no significant difference in SPF between
horizontal and vertical ground coils. According to Hughes et al. (1985), for newly constructed
projects in upstate New York, a ground coil heat pump system has a simple payback period of 5 to
10 years.
Based on ASHRAE RP-863 data, Cane et al. (1996) summarized and analyzed 12 groundcoupled
heat pump systems. The average installed heat pump capacity per 1000 ft2 (93 m2) of gross
floor area is 2.4 tons /1000 ft2 (0.091 kW/m2), the average length of the ground coil pipe per ton of
cooling capacity is 287 ft (88 m), the average electric energy use was 14.7 kWh/ ft2 year, and the
capital cost for the HVAC&R system was $6.6/ ft2.
For water-source heat pumps that use surface water such as lake water as a heat source and sink,
a plastic or copper coil, or sometimes a plate-and-frame heat exchanger, is often used to form a
closed-loop system to prevent fouling of the water coil.
12.5 AIR-TO-AIR HEAT RECOVERY
Types of Air-to-Air Heat Recovery
In HVAC&R systems, it is always beneficial if the cooling effect of the exhaust air can be used to
cool and dehumidify the incoming outdoor air during summer, and if the heating effect can be used
to heat the cold outdoor air during winter. Air-to-air heat recovery is used to recover the cooling and
heating energy from the exhaust air for the sake of saving energy. In an air-to-air heat recovery
system, the exhaust air is the airstream used to provide cooling capacity or heating energy. The
other airstream used to extract heat energy from or release heat energy to the exhaust airstream is
the outdoor airstream, which has a greater temperature and enthalpy difference between these
airstreams than the supply airstream (which is often a mixture of outdoor and recirculating
airstreams).
ASHRAE/IESNA Standard 90.1-1999 specifies individual air systems that have both a design
air supply capacity of 5000 cfm (2360 L/ s) or greater and have a minimum outside air supply of 70
percent or greater of the design supply volume flow rate shall have an energy (heat) recovery system
with at least 50 percent recovery effectiveness. This means that the change of the enthalpy of
the outdoor air supply is equal to 50 percent of the difference between the outdoor and exhaust air
during design conditions.
Exceptions to having an energy recovery system include the contamination of the intake outdoor
airstream or the energy recovery system is not cost effective.
Currently used air-to-air heat recovery devices are mainly divided into the following categories:
fixed-plate heat exchangers, runaround coil loops, rotary wheel, and heat pipes.
Effectiveness
According to ASHRAE Standard 84-1991, Method of Testing Air-to-Air Heat Exchangers, the
performance of the sensible energy transfer (dry-bulb temperature), latent energy transfer (humidity
ratio), and total energy transfer of an air-to-air heat recovery device is assessed by its effectiveness
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.19
, which is defined as
(12.10)
The effectiveness of sensible heat recovered T can be calculated as
(12.11)
and (12.12)
where The, Thl  air temperature of warm airstream entering and leaving heat exchanger,
°F (°C)
Tce, Tcl  air temperature of cold airstream entering and leaving heat exchanger, °F (°C)
Ch, Cc  heat capacity rate of warm and cold airstream, Btu/h°F (W/°C)
Cmin  heat capacity rate of airstream of smaller mass flow rate among these two
airstreams, Btu/h°F (W/ °C)
mass flow rate of warm and cold airstreams and of airstream of smaller mass
flow rate, lb/h (kg/ s)
cpa, h, cpa, c, cpa, min  specific heat of warm and cold airstreams and of airstream of smaller mass
flow rate, Btu / lb°F (J /kg°C)
During summer cooling, the outdoor air intake in an AHU or packaged unit (PU) is the warm
airstream in an air-to-air heat recovery heat exchanger and the cold airstream in the exhaust
airstream. The exhaust airstreams is usually the smaller mass flow rate of these two airstreams
because of the space positive pressure and additional space exhaust systems. During winter heating,
the exhaust airstream is the warm airstream and is also the airstream of smaller mass flow rate. The
outdoor airstream is the cold airstream.
The effectiveness of total heat recovered in an air-to-air heat recovery heat exchanger h can also
be calculated by Eq. (12.11), except that the air temperature T, in °F (°C), should be replaced by air
enthalpy h, in Btu/ lb (J / kg). If effectivenesses T and h are known (from manufacturer’s data), the
latent heat recovered can then be calculated and is the difference between the total and sensible heat
recovered in the air-to-air heat recovery heat exchanger.
Fixed-Plate Heat Exchangers
The schematic diagram of a fixed-plate heat exchanger is shown in Fig. 12.8a. Alternate layers of
fixed plates are separated and sealed and form two isolated airstream passages, outdoor and exhaust
air. Only sensible heat is transferred from the warm airstream to the cold airstream. A counterflow
or crossflow airflow arrangement is often used.
Aluminum and plastic are the most commonly used materials for plates because of their corrosive
resistance and ease of fabrication. Steel alloys are used for higher-temperature applications.
They are spaced from 0.1 to 0.5 in. (2.5 to 13 mm) apart. Heat resistance through the plates is small
compared to their surface resistance. Condensate drains are often equipped in fixed-plate heat
exchangers. These drains are also used to drain the wastewater after a water wash. Face air velocity
through the fixed plates are usually between 200 and 1000 fpm (1 and 5 m/s) with a pressure drop
from 0.1 to 1 in. (25 to 250 Pa) WC. Most fixed-plate heat exchangers have a capacity of 50 to
10,000 cfm (24 to 4720 L/ s). Fixed-plate heat exchangers have a sensible heat recovered effectiveness
between 0.5 and 0.8.
m? h m?c, m? min 
Ch  m?hcpa,h, Cc  m?ccpa,c, Cmin  m?mincpa,min

Cc(Tcl  Tce)
Cmin(The  Tce)
T 
Ch(The  Thl)
Cmin(The  Tce)
 
actual transfer (moisture or energy)
maximum possible transfer between airstreams
12.20 CHAPTER TWELVE
There is no moving part in a fixed-plate heat exchanger. Because two airstreams are in indirect
contact with each other, neither airstream will contaminate the other. No cross-leakage occurs
between these two airstreams.
Runaround Coil Loops
A runaround coil loop is also called a coil energy recovery loop and is shown in Fig. 12.8b. There
are two extended surface finned-tube water coils. One is located in the outdoor air passage and the
other in the exhaust air passage. These two coils are connected by water pipes and form a closedloop
system with a circulating water pump and an expansion water tank.
For locations where the outdoor air temperature is lower than 32°F (0°C), antifreeze fluid should
be used to prevent freezing. A three-way control valve is often installed to prevent frosting of the
exhaust air coil. This control valve will supply water to the exhaust air coil at a temperature of 30°F
(1.1°C) or above, or other preset temperatures by mixing a portion of bypassing warm water from
the exhaust air coil with the cold water from the outdoor air coil during winter heating.
Runaround coil loops recover sensible heat only. They are more flexible vis-?-vis the layout of
the outdoor air and exhaust air passages. The face velocity of outdoor air and exhaust air coils is
usually between 300 and 600 fpm (1.5 and 3 m/ s), and their pressure drops often range from 0.4 to
1.0 in. WC (100 to 250 Pa). Runaround coil loops usually have a sensible heat effectiveness between
0.45 and 0.65.
Rotary Heat Exchangers
Rotary heat exchangers, also called rotary wheels, can be classified into rotary sensible heat exchangers,
in which only sensible heat is recovered, and rotary enthalpy exchangers, in which both
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.21
Exhaust
air
Exhaust
air coil Outdoor
air coil
Expansion
tank
Outdoor
air
Outdoor
airstream
Exhaust
airstream
Exhaust
airstream
Outdoor
airstream
Circulating
pump
(b)
(a)
1
1
T
FIGURE 12.8 (a) A fixed-plate heat exchanger; (b) a runaround coil loop heat exchanger.
sensible heat and latent heat are recovered. A rotary heat exchanger is a rotary wheel with a depth
of 4 to 8 in. (100 to 200 mm); Fig. 12.9a and b shows a rotary heat exchanger. The wheel has support
spokes, typically 16, and its front surface is cut into segments. A rotary heat exchanger is often
equally divided into two separately sealed sections: an outdoor airstream section and an exhaust
airstream section. In a rotary sensible heat exchanger, it is filled with an air-penetrable medium with
a large internal surface area such as aluminum, monel metal, or stainless-steel corrugated wire
mesh, at a density of about 4 lb/ ft3 (64 kg/m3). In an enthalpy exchanger, a desiccant such as
lithium chloride is impregnated in the porous fiberglass matrix, or solid desiccants such as silica
gels in granular form are lined with substrate sheets at a center-to-center spacing of about 30 mil.
To prevent the dripping of LiCl from the fiberglass matrix, the ratio of absorbed water to the mass
of desiccant should be less than 10 lb H2O/lb LiCl (10 kg H2O/ kg LiCl). Each 1 lb (1 kg) of fiberglass
matrix holds about 0.10 lb (0.10 kg) of LiCl. The rotating speed of a rotary enthalpy heat
exchanger varies from 0.5 to 10 rph (revolutions per hour).
There are two airstreams in a rotary heat exchanger: an outdoor airstream and an exhaust
airstream. In summer cooling, the outdoor airstream is the warm airstream and the exhaust airstream
is the cold airstream, whereas in winter heating, the outdoor airstream is the cold airstream and the
exhaust airstream is the warm airstream. Outdoor and exhaust airstreams flow in parallel but opposite
directions for a counterflow heat transfer. The sensible heat-transfer process of a rotary sensible heat
exchanger is shown in Fig. 12.9c, and the enthalpy transfer process of a rotary enthalpy heat
exchanger is shown in Fig. 12.9d.
The blockage of the face area by desiccant, substrates, of support spokes for an enthalpy heat
exchanger is about 28 percent. The face velocity of rotary heat exchangers varies from 500 to 700
fpm (2.5 to 3.5 m/ s). At a face velocity of 500 fpm (2.5 m/ s), the pressure drop of air flowing
through the rotary heat exchanger varies from 0.45 to 0.65 in. WC (112 to 162 Pa), depending
mainly on the structure of the heat-transfer medium.
At a face velocity of 500 to 1000 fpm (2.5 to 5 m/s), the sensible heat effectiveness of a rotary
heat exchanger varies from 0.5 to 0.8, and the total heat effectiveness varies from 0.55 to 0.85. The
rotating speed of a rotary sensible heat exchanger is between 10 and 25 rpm, whereas for a rotary
total heat exchanger, its rotating speed varies between 0.5 and 10 rph. A higher face velocity means
that a higher rate of transfer, higher pressure drop, and a larger volume flow rate result in a lower
effectiveness. A lower face velocity results in a lower pressure drop, smaller volume flow rate, and a
higher effectiveness.
The maximum diameter of a rotary heat exchanger is 14 ft (4.2 m). The upper limit of the
volume flow rate for a single unit is about 65,000 cfm (30,670 L / s). A wheel larger than that would
be too difficult to ship and install.
Heat Pipe Heat Exchangers
A heat pipe is often used as a sensible indirect heat exchanger. It consists of many heat pipes
arranged in rows along the direction of airflow, as shown in Fig. 12.10a. Each sealed heat pipe
contains a volatile fluid, as shown in Fig. 12.10b. When one end of the pipe, the hot end or evaporative
section, absorbs heat from the airstream flowing over the pipe, then the volatile fluid inside the
pipe vaporizes. The vapor then moves to the other end (the cold end or condensation section)
because of high saturated pressure in the evaporative section. After condensing heat is released
to another airstream that flows over the other end of the pipe, the vapor inside the pipe condenses
to liquid form and is drawn back to the evaporative section by gravity. It has then completed an
evaporation/condensation cycle. The heat pipes are often slightly tilted to enable the condensed
liquid to flow back to the evaporation section.
Heat pipes have an inner capillary wick structure. The outer tube is often made of aluminum,
with fins of the same material for a large heat-transfer surface. The volatile fluid inside the heat pipe
is usually a halocarbon compound refrigerant.
When two separate airstreams flow over the heat pipe heat exchanger, the warmer airstream
flows over the evaporation section and the colder airstream flows over the condensation section.
12.22 CHAPTER TWELVE
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.23
FIGURE 12.9 Rotary heat exchangers: (a) isometric view; (b) outdoor and exhaust airstreams; (c) sensible heattransfer
process; (d) enthalpy-transfer process.
These two airstreams flow in a counterflow arrangement for greater effectiveness. The airstreams
are separated by a sealed partition to prevent cross-contamination.
The performance of a heat pipe heat exchanger is indicated by its sensible heat effectiveness T.
Both the effectiveness and the airstream pressure drop p of a heat pipe heat exchanger depend
mainly on its face velocity vface, fin spacing, and the number of rows of heat pipes in the direction of
airflow. If two heat pipe heat exchangers are connected in series, the number of rows of such a
series is the total number of rows of the two heat exchangers.
The greater the total number of rows of heat pipe and fins per inch, the higher the T and p.
The lower the face velocity vface of the heat pipe exchanger, the higher the T and the lower the p.
However, the volume flow rate and the capacity of the heat pipe heat exchanger will also be smaller.
The design face velocity of a heat pipe heat exchanger is between 400 and 800 fpm (2 and 4
m/ s). The total number of heat pipe rows varies from 6 to 10 rows. Sensible heat effectiveness
varies from 0.45 to 0.75. For a heat pipe with 14 fins/ in. (1.8-mm fins) and a total of 8 rows with
a face velocity vface  500 fpm (2.5m/ s), its sensible heat effectiveness T  0.65 and its pressure
drop p is about 0.6 in. WC (150 Pa). Refer to the manufacturer’s data for detailed
information.
The capacity and rate of heat transfer of a heat pipe can be controlled by varying the slope or tilt
of the heat pipe. This adjustment increases or decreases the liquid flow inside the heat pipe and,
therefore, its capacity.
12.24 CHAPTER TWELVE
FIGURE 12.10 A heat pipe heat exchanger: (a) heat exchanger and airstreams; (b) heat pipe.
Comparison between Various Air-to-Air Heat Exchangers
According to ASHRAE Handbook 1996, HVAC Systems and Equipment, among the air-to-air heat
recovery heat exchangers:
 Fixed-plate, runaround coil loop, and rotary sensible heat exchangers are all limited to sensible
heat transfer and recovery; only the rotary enthalpy exchangers can recover latent heat from the
exhaust airstream.
 There are no moving parts in fixed-plate and heat pipe heat exchangers.
 Both heat pipe and runaround coil loop heat exchangers show no cross-contamination or crossleakage.
Fixed-plate heat exchangers have a cross-contamination and cross-leakage from 0 to 5
percent, and rotary heat exchangers from 0 to 10 percent. If cross-contamination is strictly prohibited,
a purge section can be installed to reduce it.
 The effectivenesses of fixed-plate and rotary heat exchangers are higher than those of heat pipe
and runaround coil loop heat exchangers.
 Rotary heat exchangers show the lowest air pressure drop among the various types of heat exchanger.
12.6 GAS COOLING AND COGENERATION
Gas Cooling
High electricity demand charges, high peak electricity rates, and the development of high-efficiency
direct-fired equipment, highly reliable gas engines, and more sophisticated and cost-effective
desiccant-based air conditioning systems enable gas cooling systems to compete with electric
compressors after the decline of gas cooling in the 1970s. Current gas cooling systems include the
following:
 Double-effect, direct-fired, lithium bromide (LiBr) absorption chillers and chiller / heaters
 Desiccant-based air conditioning using evaporative cooling, refrigeration, and direct-fired gas
heaters for regeneration
 Gas engine chiller systems and cogeneration
Usually, a gas cooling system has a higher initial cost and a lower operating cost than an electricitydriven
refrigeration system. Accurate calculation of the operating costs of a gas cooling system is
important in this comparison.
The LiBr absorption chillers are discussed in Chap. 14, and desiccant cooling systems are
discussed in Chap. 29. Gas-engine chiller systems and cogeneration are discussed in this section.
Cogeneration
Cogeneration is the sequential use of energy from a primary source, including natural gas, oil, and
coal, to produce power and heat. Power can be electric or mechanical power, or both. In a cogeneration
system, the sequential use of the heat released from the flue gas and engine jacket significantly
increases system efficiency and makes the cogeneration system economically attractive.
In 1978 in the United States, the Public Utility Regulatory Policies Act (PURPA) permitted the
interconnection of electric power lines of cogeneration systems with electric utility systems. This
provides flexibility for cogeneration plants. They can either use or sell their electric power to the
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.25
utility and optimize the size of the cogeneration plant by reducing or eliminating its standby generation
capacity. During the 1990s, many states, such as California and New York, were leading the
way in utility deregulation. Utility deregulation will stimulate competition in the electricity market
for lower consumer rates.
Since 1990, hundreds of cogeneration systems were developed for internal use. Internal use
means the production of both power and heat for use in settings such as hospitals, medical
centers, university campuses, public buildings, and industrial facilities, and in their installed air
conditioning systems. A successfully developed cogeneration system often relies on site technical
and economic analysis, especially for local electricity demand and electricity rates. Two kinds
of prime movers are widely used in these internal use cogeneration systems: gas engines and gas
turbines.
12.26 CHAPTER TWELVE
FIGURE 12.11 Energy flow in (a) gas-engine chiller and (b) electricity-driven chiller.
Gas-Engine Chiller
A gas-engine chiller is often a combination of a gas cooling system and a cogeneration system.
Heat released from the exhaust gas and the engine jacket cooling water are all recovered to increase
system efficiency. Figure 12.11a shows the energy flow of a gas-engine chiller system.
If the efficiency of the gas engine is 35 percent, the mechanical efficiency including the transmission
gear train is 95 percent, and the chiller’s COP  4.5, then for every 10,000 Btu (10,550 kJ)
of fuel energy input to the gas-engine chiller, there is a cooling output Qrf of
In addition, there is a heating output of 4000 Btu (4220 kJ) from the exhaust gas and engine
jacket to supply hot water or low-pressure steam for an absoption, space heating, or domestic hot
water unit.
A gas-engine chiller has a prominent advantage over a motor-driven chiller because the former
can vary its speeds at various operating conditions: high speed at overloads and low speeds at partload
operation. Engine reliability is the key to user acceptance for gas-engine chillers. Several
hundred engine-driven chillers were installed in the 1960s and early 1970s. According to reliability
records of these systems reported in Davidson and Brattin (1986), the reliability of gas-engine
chillers matches the requirements of HVAC&R systems.
Gas engines can be used to drive screw, reciprocating, or centrifugal compressors. Gas-enginedriven
screw chillers are becoming more and more popular. Many manufacturers offer packaged
units for easier field installation. The capacity of gas-engine chiller packaged units varies from 30
to 500 tons (105 to 1760 kW). Gas-engine cooling systems can also be coupled to direct-expansion
(DX) refrigeration systems and rooftop packaged units. Gas-fired internal combustion engines for
cooling in buildings follow the developments of gasoline and diesel engine technology. There are
two kinds of gas engines: heavy-duty industrial applications and light-duty automotive engines. Industrial
heavy-duty gas engines run a minimum of 30,000 h of full-load service between major
overhauls, and cost about 5 times as much as automotive engines. The service life of an automobile
engine is only 2000 to 5000 h.
Recently, manufacturers have produced packaged automotive gas-engine chillers of 150-ton
(525-kW) capacity. One manufacturer also offers one 150-ton (525-kW) gas-engine chiller and
integrated hot water absorption chiller to give a total maximum output of 180 tons (630 kW).
Automotive gas engines are suitable for compressors that require speeds far above 1800 rpm. This
packaged gas-engine chiller is also equipped with microprocessor-based controls to coordinate and
monitor the operation of the engine and chiller. An operating cost as low as one-half that of similarsize
electricity-driven units is claimed, depending on the local utility rate structure.
Engine-driven chillers are maintenance-intensive. The maintenance cost of a gas-engine chiller
may be between 10 and 20 percent of the energy cost, and it should be added to the operating cost
during economic analysis. D’zurko and Epstein (1996) made a cost analysis between gas-engine
chiller and electric chillers for offices, schools, restaurants, retail stores, and hospitals in New York
State. Most have a simple payback of 2.1 to 11.9 years.
For a gas-engine chiller with a cooling output of 15,000 Btu (15,825 kJ), if its by-product, 4000-
Btu (4220-kJ) heat output from the exhaust gas and engine jacket is also counted, then the simple
payback period is usually between 2 and 5 years.
Gas Engines
Two types of automotive engines are used to drive the chillers: gas engines operated on diesel cycle
and operated on an Otto cycle. Because a diesel engine has a higher thermal efficiency than an Otto
engine, more diesel engines are used in gas-engine chillers. Most diesel engines are operated on a
four-stroke cycle, i.e., an intake stroke, a compression stroke, a power stroke, and an exhaust stroke,
to produce power. A diesel engine can also operate on a two-stroke cycle, an intake stroke and an
Qrf  10,000  0.35  0.95  4.5  15,000 Btu (15,825 kJ)
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.27
exaust stroke, which is slightly cheaper than a four-stroke engine; however, a two-stroke engine is
less efficient and experiences greater wear than a four-stroke engine.
Reciprocating gas engines can be categorized according to the manner in which the engine is aspirated:
natural aspiration and turbocharging. A natural aspiration engine supplies air or the fuel / air
mixture to the engine cylinder at atmopheric pressure, whereas a turbocharging engine supplies
higher-pressure air or fuel / air mixture to the cylinder. Turbocharging uses the engine exhaust gas to
drive a small turbine which is connected with a centrifugal compressor. This compressor raises the
intake air pressure and quantity and thus the engine’s capacity.
In a gas engine, the primary function of a heat recovery system is to reject excess heat produced
and to exhaust flue gas during the power generation process. Effective heat rejection from the engine
jacket and the exhaust of adequate flue gas must be emphasized. The secondary function is to
recover heat from the exhaust gas and engine jacket and to use the recovered heat efficiently and
economically.
Exhaust Gas Heat Recovery
According to Orlando’s Cogeneration Design Guide (1996), the exhaust gases of the reciprocating
gas engine contain about one-third of the engine’s heat output, and approximately 50 to 75 percent
of the sensible heat can be recovered. Heat may be recovered in the form of steam or hot water.
The sensible heat recovered Qex,r, in Btu/h (W), can be calculated as
(12.13)
where volume flow rate of exhaust gas at outlet of gas engine, cfm [m3 / (60 s)]
ex  density of exhaust gas at outlet of gas engine, lb / ft3 (kg/m3)
cpa  specific heat of exhaust gas, Btu/ lb °F (J /kg°C)
Tex  temperature of exhaust gas at outlet of gas engine, °F (°C)
Trec  temperature of exhaust gas leaving heat recovery heat exchanger, °F (°C)
It is more economical to have a minimum temperature difference between the exhaust gas and the
generated steam inside the recovery heat exchanger of 100°F (55°C).
According to ASHRAE Handbook 1996, HVAC Systems and Equipment, the temperature of the
recovered steam or hot water for absorbing chillers is between 190 and 245°F (88 and 118°C),
for space heating between 120 and 250°F (49 and 120°C), and for domestic hot water between
120 and 200°F (49 and 93°C). For steam, a maximum pressure up to 8 psig (55 kPag) may be
produced.
Exhaust gas heat recovery heat exchangers are often designed to reduce engine noise transmitted
along with the exhaust gas and are called heat recovery mufflers. Heat recovery mufflers should
minimize engine backpressure to 6 in. WG (1500 Pag) for natural aspiration reciprocating engines
and to 25 to 30 in. WG (6.25 to 7.5 kPag) for turbocharged engines.
Many heat recovery mufflers are designed based on a minimum exhaust temperature of 300°F
(149°C) to avoid condensation. The temperature of the exhaust gas at part-load operation is important.
The construction of the exhaust heat recovery mufflers should provide access for inspection,
cleaning, and soot removal for diesel engines.
Engine Jacket Heat Recovery
Approximately one-third of the heat input to the reciprocating gas engine is rejected from the engine
block, heads, and exhaust manifolds to the engine jacket water coolant. Heat energy can be recovered
by either the hot water with a temperature at 250°F (120°C) or low-pressure steam at a
maximum pressure of 15 psig (103 kPag).
V?
ex 
Qex,r  60V?ex excpa (Tex  Trec)
12.28 CHAPTER TWELVE
When the engine jacket water is routed through the heat recovery muffler where additional heat
is extracted from the exhaust gas, an engine jacket heat recovery system is then combined with the
exhaust gas heat recovery system. At least a water circulating pump is required to circulate the hot
water flowing through the engine jacket and the heat recovery muffler as soon as the engine is operating.
To avoid excessive thermal stress, the temperature difference between the cooling water entering
and leaving the engine jacket should not exceed 15°F (8.3°C).
Steam can be produced at the top part of the heat recovery muffler with a steam separator and is
distributed to various process loads. Hot water from the bottom of the heat recovery muffler is then
mixed with the condensate returned from the remote steam loads. The mixture is forced through the
engine jacket and the heat recovery muffler by the circulating pump.
Sometimes, the engine coolant circuit in a heat recovery loop is separated from the process
loads by heat exchangers and forms a primary and secondary loop. A primary-secondary loop
isolates and protects the engine coolant circuit from process loads, leaks, and failures in the distribution
systems. A primary-secondary loop is especially useful for a multiple-engine installation.
A cooling tower can be connected to a heat recovery loop through a heat exchanger for the sake
of maintaining a required entering temperature of the engine jacket coolant at system part-load
operation. A coolant heater can be installed to preheat the coolant during start-up.
For space heating, hot water from the heat recovery muffler can be supplied directly to the
heaters in the conditioned space. For space cooling, an absorption chiller using recovered steam or
hot water should be installed to provide cooling for the conditioned space.
Cogeneration Using a Gas Turbine
Many cogeneration plants use a combustion gas turbine instead of a gas engine as the prime mover.
A gas turbine usually consists of a compressor section to raise the air pressure, a fuel /air mixing
and combustion chamber, and an expanding turbine section. The compressor and turbine are joined
by the same shaft. Capacity may vary from several hundred brakehorsepower (bhp) to more than
100,000 bhp (75,000 kW). Gas turbines are often connected to induction generators to produce
electric power through gear trains.
Steam boilers are often used as heat recovery units to produce steam at a pressure typically 15
psig (103 kPag) from the gas-turbine exhaust gas. Recovered heat can often be used as process heat
to operate an absorption chiller.
REFERENCES
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment , ASHRAE Inc., Atlanta, GA, 1996.
Ayres, J. M., and Lau, H., Comparison of Residential Air-to-Air Heat Pump and Air-Conditioner/Gas Furnace
Systems in 16 California Climatic Zones, ASHRAE Transactions, 1987, Part II, pp. 525–561.
Ball, D. A., Fischer, R. D., and Hodgett, D. L., Design Methods for Ground-Source Heat Pumps, ASHRAE
Transactions, 1983, Part II B, pp. 416–440.
Baxter, V. D., and Moyers, J. C., Field-Measured Cycling Frosting and Defrosting Losses for a High-Efficiency
Air Source Heat Pump, ASHRAE Transactions, 1985, Part IIB, pp. 537–554.
Bivens, D. B., Patron, D. M., and Yokozeki, A., Performance of R-32/R-125/R-134a Mixtures in Systems with
Accumulators or Flooded Evaporators, ASHRAE Transactions, 1997, Part I, pp. 777–780.
Black, G. D., An Overview of the Four-Way Refrigerant Reversing Valve, ASHRAE Transactions, 1987, Part I,
pp. 1147–1151.
Brown, M. J., Hesse, B. J., and O’Neil, R. A., Performance Monitoring Results for an Office Building
Groundwater Heat Pump System, ASHRAE Transactions, 1988, Part I, pp. 1691–1707.
Cane, R. L. D., Clemes, S. B., and Morrison, A., Operating Experience with Commercial Ground-Source Heat
Pumps—Part I, ASHRAE Transactions, 1996, Part I, pp. 911–916.
HEAT PUMPS, HEAT RECOVERY, GAS COOLING, AND COGENERATION SYSTEMS 12.29
Carrier Corporation, Packaged Rooftop Heat Pumps, Carrier Corporation, Syracuse, NY, 1980.
Davidson, K., and Brattin, H. D., Gas Cooling for Large Commercial Buildings, ASHRAE Transactions, 1986,
Part I B, pp. 910–920.
Department of Energy, 1986 Proposed Standard Test Procedure, Docket No. CAS-RM-79-102, 1986.
D’zurko, D. C., and Epstein, G. J., Comparative Analysis for Natural Gas Cooling and Space Conditioning
Technologies in New York State, ASHRAE Transactions, 1996, Part I, pp. 275–283.
Eckman, R. L., Heat Pump Defrost Controls: A Review of Past, Present, and Future Technology, ASHRAE
Transactions, 1987, Part I, pp. 1152–1156.
Goldschmidt, V. W., Effect of Cyclic Response of Residential Air Conditioners on Seasonal Performance,
ASHRAE Transactions, 1981, Part II, pp. 757–770.
Hughes, P. J., Loomis, L., O’Neil, R. A., and Rizzuto, J., Result of the Residential Earth-Coupled Heat Pump
Demonstration in Upstate New York, ASHRAE Transactions, 1985, Part II B, pp. 1307–1325.
Johnson,W. S., McGraw, B. A., Conlin, F.,Wix, S. D., and Baugh, R. N., Annual Performance of a Horizontal-
Coil Ground-Coupled Heat Pump, ASHRAE Transactions, 1986, Part I A, pp. 173–185.
Kavanaugh, S. P., Groundwater Heat Pump Performance Enhancement with Precoolers and Water Pump
Optimization, ASHRAE Transactions, 1987, Part II, pp. 1205–1218.
Knipe, E. C., Applications of Heat Pumps Using Groundwater Resources, ASHRAE Transactions, 1983,
Part II B, pp. 441–451.
Mathen, D. V., Performance Monitoring of Select Groundwater Heat Pump Installations in North Dakota,
ASHRAE Transactions, 1984, Part I B, pp. 290–303.
Mohammad-zadeh,Y., Johnson, R. R., Edwards, J. A., and Safemazandarani, P., Model Validation for Three
Ground-Coupled Heat Pumps, ASHRAE Transactions, 1989, Part II, pp. 215–221.
Mulroy,W. J., The Effect of Short Cycling and Fan Delay on the Efficiency of a Modified Residential Heat
Pump, ASHRAE Transactions, 1986, Part I B, pp. 813–826.
Niess, R. C., Applied Heat Pump Opportunities in Commercial Buildings, ASHRAE Transactions, 1989, Part II,
pp. 493–498.
Nutter, D. W., O’Neal, D., and Payne,W. V., Impact of the Suction Line Accumulator on the Frost/Defrost
Performance of an Air-Source Heat Pump with a Scroll Compressor, ASHRAE Transactions, 1996, Part I,
pp. 284–290.
O’Neal, D. L., and Katipamula, S., Performance Degradation during On-Off Cycling of Single-SP Conditioners
and Heat Pump: Model Development and Analysis, ASHRAE Transactions, 1991, Part II, pp. 316–323.
O’Neal, D. L., and Peterson, K., A Comparison of Orifice and TXV Control Characteristics during the Reverse-
Cycle Defrost, ASHRAE Transactions, 1990, Part I, pp. 337–343.
Orlando, J. A., Cogeneration Design Guide, ASHRAE Inc., Atlanta, GA, 1996.
Rackliffe, G. B., and Schabel, K. B., Groundwater Heat Pump Demonstration Results for Residential
Applications in New York State, ASHRAE Transactions, 1986, Part II A, pp. 3–19.
Rasmussen, R. W., MacArthur, J. W., Grald, E. W., and Nowakowski, G. A., Performance of Engine-Driven
Heat Pumps under Cycling Conditions, ASHRAE Transactions, 1987, Part II, pp. 1078–1090.
Robertson,W. K., Electricity and Competition, Engineered Systems, no. 11, 1996, pp. 34–41.
Virgin, D. G., and Blanchard,W. B., Cary School—25 Years of Successful Heat Pump/Heat Reclaim System
Operation, ASHRAE Transactions, 1985, Part I A, pp. 40–45.
Weinstein, A., Eisenhower, L. D., and Jones, N. S.,Water-Source Heat Pump System for Mount Vernon
Unitarian Church, ASHRAE Transactions, 1984, Part I B, pp. 304–312.
Wurm, J., and Kinast, J. A., History and Status of Engine-Driven Heat Pump Developments in the U.S.,
ASHRAE Transactions, 1987, Part II, pp. 997–1005.
12.30 CHAPTER TWELVE
CHAPTER 13
REFRIGERATION SYSTEMS:
CENTRIFUGAL
13.1
13.1 CENTRIFUGAL VAPOR
COMPRESSION REFRIGERATION
SYSTEMS 13.1
Refrigerants 13.2
Compressor 13.3
System Components 13.4
Purge Unit 13.5
Types of Centrifugal Chiller 13.7
13.2 WATER-COOLED
CENTRIFUGAL CHILLER 13.7
Refrigerant Flow 13.7
Performance Rating Conditions 13.8
13.3 CENTRIFUGAL CHILLERS
INCORPORATING HEAT RECOVERY 13.9
System Description 13.9
Operating Modes 13.9
System Characteristics 13.12
13.4 AUTOMATIC BRUSH CLEANING FOR
CONDENSERS 13.13
Principles and Operation 13.13
Effect of Brush Cleaning System 13.14
13.5 CENTRIFUGAL COMPRESSOR
PERFORMANCE MAP 13.15
Surge of Centrifugal Compressor 13.15
Centrifugal Compressor Map
at Constant Speed 13.16
Centrifugal Compressor Map
at Variable Speed 13.18
13.6 CAPACITY CONTROL OF
CENTRIFUGAL CHILLERS 13.19
Difference between Centrifugal
Compressors and Centrifugal
Fans 13.19
Required System Head at Part-Load
Operation 13.19
Capacity Control Using Inlet Vanes 13.20
Capacity Control by Variable
Speed 13.20
Comparison between Inlet Vanes
and Variable Speed 13.21
13.7 CENTRIFUGAL CHILLER
CONTROLS 13.22
Functional Controls and Optimizing
Controls 13.22
Chilled Water Leaving Temperature
Control 13.22
Condenser Water Temperature
Control 13.23
Safety Controls 13.23
Fault Detection and Diagnostics 13.24
Operator Interface 13.24
13.8 OPERATING CHARACTERISTICS OF
CENTRIFUGAL CHILLERS 13.24
Sequence of Operation of a Typical
Centrifugal Chiller 13.24
System Balance at Full Load 13.25
13.9 PART-LOAD OPERATION OF
CENTRIFUGAL CHILLER 13.25
Part-Load Operating
Characteristics 13.25
Evaporating and Condensing
Temperatures at Part-Load
Operation 13.26
Temperature Lift at Part-Load Operation
for Centrifugal Chillers 13.29
13.10 FREE REFRIGERATION 13.31
Principle of Operation 13.31
Free Refrigeration Capacity 13.32
13.11 OPERATION OF MULTIPLE-CHILLER
PLANT 13.33
Multiple-Chiller Plant 13.33
Parallel and Series Piping 13.33
Chiller Staging 13.34
Chiller Minimum Performance 13.35
Design Considerations 13.35
REFERENCES 13.36
13.1 CENTRIFUGAL VAPOR COMPRESSION
REFRIGERATION SYSTEMS
A centrifugal vapor compression refrigeration system uses a centrifugal compressor to compress
the refrigerant and provides chilled water to cool the air in air-handling units and terminals. A
centrifugal refrigeration machine usually comes as a factory-assembled packaged unit and is often
called a centrifugal chiller, as shown in Fig. 13.1.
Most centrifugal chillers are driven by open and hermetic motors. Occasionally, they may be
driven by internal combustion engines. Sometimes, large centrifugal chillers are driven by steam
and gas turbines.
Refrigerants
According to the ARI survey, there were about 80,000 large-tonnage chillers in use in the United
States in 1992. Of these, about 80 percent used CFC-11; the remaining used CFC-12, CFC-114,
HCFC-22, and others. As described in Sec. 9.4, based on ARI’s estimate (Dooley, 1997), 53 percent
of the 80,000 CFC chillers will remain in service on January 1, 2000. The actual phaseout schedule
is slower than expected.
As described in Secs. 9.4 and 9.5, HCFC-123, an interim refrigerant, will replace CFC-11 in
centrifugal chillers. In 1997, DuPont Company (manufacturer of HCFC-123) specified an allowable
exposure limit of 50 ppm for HCFC-123. HFC-134a, a long-term alternative refrigerant, will
replace CFC-12. In centrifugal chillers, the direct conversion from CFC-12 to HFC-134a requires
13.2 CHAPTER THIRTEEN
FIGURE 13.1 A typical three-stage, water-cooled centrifugal chiller.
an increase in compressor speed, a drop in cooling capacity and efficiency, and the use of synthetic
polyolester lubrication oil instead of mineral oil. Ammonia R-717 is usually used in centrifugal
compressors for industrial applications due to its lower flammability and higher toxicity.
Because a large amount of new non-CFC centrifugal chillers were installed to replace CFC
centrifugal chillers, the energy use of the centrifugal chillers, expressed in kW/ ton, is significantly
reduced. The manufacturers produced new chillers with 0.49 kW/ton (7.17 COP) in 1997.
The 1990 Clean Air Act makes the knowing vent of CFCs and HCFCs illegal as of July 1, 1992.
Recovery/ recycle equipment must be used to prevent venting during installation, operation, and
maintenance.
Compressor
A centrifugal chiller consists of a centrifugal compressor, an evaporator or liquid cooler, a condenser,
a throttling device, piping connection, controls, and possibly a purge unit and a flashing
cooler. A centrifugal compressor is a turbomachine. The total head lift, in psi (kPa), results from
the conversion of velocity pressure to static pressure. A single-stage, two-stage, or three-stage
compressor—according to the number of internally connected impellers—may be used. Although
the compression ratio Rcom  pdis /psuc of a single-stage centrifugal compressor seldom exceeds 4,
the impellers connected in series increase the compression ratio Rcom and satisfy most compression
requirements. For a refrigeration system with an evaporating temperature of 40°F (4.4°C) and a
condensing temperature of 100°F (37.8°C), HCFC-123 has a compression ratio of 3.59. Use of a
two-stage or even three-stage instead of a single-stage centrifugal compressor acts to increase the
total head lift as well as the COP of the refrigeration system (refer to Sec. 9.12).
Figure 13.2 shows a two-stage centrifugal compressor. Vapor refrigerant is drawn through the
inlet vanes and enters the first-stage impeller. The impeller compresses the vapor and discharges it
through a diffuser, where it combines with flashed vapor refrigerant from the flash cooler. The mixture
then enters the second-stage impeller, where the refrigerant is further compressed. The hot gas
is discharged into the collecting volute through a diffuser. As the gaseous refrigerant flows through
the compressor, the area of the flow passage increases. This causes a decrease in gas velocity and
therefore a conversion of velocity pressure to static pressure.
Because a high head lift is required to raise the evaporating pressure to the condensing pressure,
the hot gas discharge velocity at the exit of the second-stage impeller approaches the acoustic velocity
vac of the saturated vapor. At atmospheric pressure and a temperature of 80°F (26.7°C), for
HCFC-123, vac is 420 ft / s (128 m/s); and for HCFC-22, vac is about 600 ft / s (183 m/ s). Centrifugal
compressors need a high peripheral velocity and high rotating speed (from 1800 to 50,000 rpm).
The refrigeration capacity of a centrifugal compressor ranges from 100 to 10,000 tons (350 to
35,000 kW). It is not economical to manufacture small centrifugal compressors. Because they are
turbomachines, centrifugal compressors have higher volume flow than positive displacement compressors.
Isentropic efficiency isen reaches a maximum of about 0.83 for a typical centrifugal compressor
operated at design conditions. Its value may decrease to as low as 0.6 during part-load operation.
The volumetric efficiency v of centrifugal compressors can be estimated at almost 1.
Inlet vanes are widely used as stepless capacity control for centrifugal compressors during partload
operation. Variable-speed drives, direct steam turbines, and engine drives also provide capacity
control by varying the speed of centrifugal compressors. Sometimes, a combination of adjustablefrequency
variable-speed drive and inlet vanes is used.
Many centrifugal chillers with a refrigeration capacity less than 1200 tons (4220 kW) are
installed with hermetic compressors. Very large centrifugal chillers often employ open compressors
in which an additional motor cooling system is used.
Centrifugal compressors need high peripheral velocity to produce head lift. They can be driven
either directly by motor or through a gear train. A direct-drive compressor often has a largerdiameter
impeller because of the limitation on the rotating speed of a two-pole motor. On the other
hand, direct drive reduces energy loss because of the gear train driving mechanism. Such a loss is
usually 2 to 3 percent of the total power input to the compressor.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.3
Because centrifugal compressors are reliable and energy-efficient for water-cooled chillers, they
are the most widely used refrigeration compressors in large central air conditioning systems.
System Components
Evaporator. A flooded shell-and-tube evaporator, as described in Sec. 10.3, is usually used for a
centrifugal chiller because of its compact size and high heat-transfer characteristics.
Condenser. Water-cooled, horizontal shell-and-tube type of condensers, as described in Sec. 10.6,
are widely used because of their lower condensing pressure and easier capacity control. Only in
locations where local codes prohibit the use of city water as the makeup water in cooling towers are
air-cooled centrifugal chillers adopted.
13.4 CHAPTER THIRTEEN
FIGURE 13.2 A two-stage centrifugal compressor. (Source: Carrier Corporation. Reprinted with permission.)
Flash Cooler. A flash cooler is sometimes called an economizer. It flashes a small portion of
refrigerant at the intermediate pressure to cool the remaining liquid refrigerant to the saturated
condition, in order to provide a greater refrigeration effect, as described in Secs. 9.13 and 9.14. For
a two-stage compressor, a single-stage flash cooler is used. For a three-stage compressor, a twostage
flash cooler is used.
Orifice Plates or Float Valves. Orifice plates (arranged in series) or float valves are used as throttling
devices in centrifugal chillers that employ flooded refrigerant feed. The use of multiple ori-
fices as a throttling device is shown in Fig. 13.3. This device controls the amount of refrigerant feed
to the liquid cooler according to the pressure of the liquid refrigerant in the condenser.
During full-load operation, a certain liquid pressure is maintained before the first orifice plate.
When liquid refrigerant flows through the first orifice plate, it encounters a pressure drop. However,
the fluid pressure between the two orifice plates is still higher than the saturated pressure of the
liquid. No flash of vapor occurs in the region between the orifice plates. The second orifice plate
meters the maximum refrigerant flow.
As the refrigeration load in the centrifugal chiller drops, less refrigerant is delivered to the condenser
and evaporator, resulting in a lower condensing pressure and a higher evaporating pressure.
This reduces the pressure exerted on the liquid refrigerant at the condenser due to lower condensing
pressure and hydrostatic head. Because of the pressure drop across the first orifice plate, the fluid
pressure between the orifice plates is less than the saturated pressure of the liquid refrigerant. Flashing
of vapor occurs. Both liquid and vapor flow through the second orifice plate. Consequently, the
mass flow rate of the refrigerant flow is reduced.
There are no moving parts in an orifice plate throttling device. It is simple and reliable in
operation.
Purge Unit
At an evaporating temperature Tev  40°F (4.4°C), HCFC-123 has a saturated pressure of 5.8 psia
(40 kPa abs.) which is far lower than atmospheric pressure. Noncondensable gas such as air and
carbon dioxide may diffuse and leak into the evaporator through cracks and gaps. Because
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.5
FIGURE 13.3 Orifice plates.
13.6 CHAPTER THIRTEEN
condenser is always located at a higher level in a centrifugal chiller, any leakage of noncondensable
gases into the system usually accumulates in the upper part of the condenser.
The presence of noncondensable gases reduces the refrigerant flow as well as the rate of heat
transfer in the evaporator and the condenser. It also increases the pressure in the condenser. All these
result in a loss of refrigeration capacity and an increase of power consumption. Wanniarachchi and
Webb (1982) showed that at 60 percent load, the presence of 3 percent noncondensable gas causes a
2.6 percent increase in the power input to the compressor. The purpose of the purge unit is to purge
periodically any noncondensable gases that have accumulated in the condenser.
According to the manufacturer’s catalog in 1996, a purge unit developed in the mid-1990s is
shown in Fig. 13.4. This purge unit includes mainly a purge tank, which is a low-temperature and
low-pressure chamber with a DX coil to condense the vapor refrigerant; a condensing unit to serve
the DX coil; a vacuum pump to purge the noncondensable gases; and a filter dryer to absorb moisture
and to remove dirt. Noncondensable gases and gaseous refrigerant from the upper part of the
condenser migrate to the low-temperature and low-pressure purge tank through a pipe connected to
the bottom of the tank. When the latent heat of condensation is absorbed by the DX coil, the
gaseous refrigerant condenses to liquid and falls to the bottom of the tank. Moisture contained in
the refrigerant is separated by the water separation tube and forms the middle layer. All the noncondensable
gases accumulate at the top of the tank because of their lower densities.
The purge vacuum pump operates periodically to extract the noncondensable gases and rejects
them to the atmosphere. It may be operated automatically once a day according to the signal from
the DDC unit controller; or it may be operated in adaptive mode, i.e., adjusting its operation to
Condensing
unit
Water separation
tube
Water
bleed valve
Sight
glass
Inlet gaseous
refrigerant from
condenser
Liquid refrigerant
returned to
condenser
Compressor
Filter dryer
Purge tank
DX coil
Air
Expansion
valve
Vacuum
pump
Water
layer
Drain
FIGURE 13.4 A purge unit in a centrifugal chiller.
match the chiller leak rate. The purge unit can be operated at any time independently of the chiller’s
operation.
The efficiency of this air purge unit is 0.0049 lb of refrigerant loss per lb of air purged (0.49
percent). Dietrich (1993) and Parsnow (1993) reported that many old type purge units need a refrigerant
loss from 0.1 to 0.75 lb of refrigerant per lb of air purged. Sometimes their efficiencies are
even worse than that. To select a high-efficiency air purge unit with less refrigerant loss causes
less ozone depletion, which is always important for chillers using HCFC as refrigerant and is also
economical.
Types of Centrifugal Chiller
Centrifugal chillers may be classified according to the configuration of their main components as
follows:
 Single-stage or multistage
 Air-cooled, water-cooled, or double-bundle condenser (A double-bundle condenser has two bundles
of condensing tubes for the purpose of heat recovery.)
 Open or hermetic
 Direct-drive or gear-drive
 Capacity control provided by inlet vanes or variable-speed drive
As mentioned before, because a centrifugal compressor needs a high peripheral velocity, it is not
economical to fabricate small centrifugal machines. Their refrigeration capacities range from 100 to
10,000 tons (315 to 31,560 kW). Because of their large capacity, DX coils are not used as evaporators
to condition air at such a large size and capacity, with their unacceptable air transporting and
distribution losses.
13.2 WATER-COOLED CENTRIFUGAL CHILLER
Refrigerant Flow
A typical water-cooled centrifugal chiller that uses a three-stage, hermetic, and direct-drive compressor
is illustrated in Fig. 13.1. As described in Sec. 9.14 and shown in Fig. 9.9a and b, the flow
process of HCFC-123 is as follows:
Vapor refrigerant is extracted from the liquid cooler as a result of the suction of the compressor:
 It enters the first-stage impeller through inlet vanes.
 Hot gas from the first-stage impeller mixes with flashed vapor from the low-pressure (secondstage)
flash cooler.
 The mixture enters the second-stage impeller.
 Hot gas from the second-stage impeller mixes with flashed vapor from the high-pressure (firststage)
flash cooler.
 The mixture enters the third-stage impeller.
 Hot gas discharged from the third-stage impeller enters the water-cooled condenser.
 Hot gas is desuperheated, condensed, and subcooled to liquid form in the condenser.
 A small portion of the liquid refrigerant is used to cool the hermetic motor.
 Most of the liquid refrigerant enters the high-pressure flash cooler through a multiple-orifice
throttling device; and a small portion flashes into vapor and mixes with the hot gas discharged
from the second-stage impeller.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.7
 Most of the liquid refrigerant from the high-pressure flash cooler enters the low-pressure flash
cooler through a multiple-orifice throttling device; and a small portion flashes into vapor and
mixes with the hot gas discharged from the first-stage impeller.
 Most of the liquid refrigerant from the low-pressure flash cooler enters the evaporator through a
multiple-orifice device.
 Liquid refrigerant vaporizes to vapor in the evaporator and produces its refrigeration effect there.
Performance Rating Conditions
According to ARI Standard 550/590-98, the cooling capacity of a centrifugal chiller is rated under
the following conditions:
13.8 CHAPTER THIRTEEN
44°F (6.7°C)
44°F (6.7°C)
2.4 gpm/ton (0.43 L/ s kW)
Water temperature leaving evaporator:
100 percent load
0 percent load
Chilled water flow rate:
Percent load, entering condenser water temperature (ECWT), entering dry-bulb (EDB) temperature,
and entering wet-bulb (EWB) temperature are as follows:
Weighting of part- Water-cooled Air-cooled Condenser
Percent load load, percent ECWT, °F (°C) EDB, °F (°C) EWB, °F (°C)
100 1 85 (29.4) 95 (35.0) 75 (23.9)
75 42 75 (23.9) 80 (26.7) 68.75 (20.4)
50 45 65 (18.3) 65 (18.3) 62.5 (17.0)
25 12 65 (18.3) 55 (12.8) 56.25 (13.5)
3.0 gpm/ton (0.054 L/s kW)
water side 0.0001 h ft2 °F/Btu (0.000018
m2 C/W), air side 0
water side 0.00025 h ft  °F/Btu (0.000044
m2 C/W), air side 0
Condenser water flow rate
Fouling factor in evaporator:
Fouling factor in condenser:
The integrated part-load value (IPLV) of the centrifugal chiller at standard rating conditions can be
calculated by the following formula, as specified by ARI 550/590-98:
(13.1)
where A, B, C, D  kW/ton or COP at 100, 75, 50, and 25 percent load, respectively. If the operating
conditions are different from the standard rating conditions for the same centrifugal chiller, a
general rule for purposes of estimation is as follows:
 When the temperature of chilled water leaving the chiller Tel is between 40 and 50°F (4.4 and
10°C), for each 1°F (0.56°C) increase or decrease of Tel, there is roughly a 1.5 percent difference
in cooling capacity and energy use.
 When the entering temperature of condenser water Tce is between 80 and 90°F (26.7 and 32.2°C),
for each 1°F (0.56°C) increase or decrease of Tce, there is roughly a 1 percent difference in cooling
capacity; there is roughly a 0.6 percent difference in energy use.
IPLV 
1
0.01/A  0.42 / B  0.45/C  0.12 / D
For water-cooled centrifugal chillers with hermetic compressors, manufacturers’ products that
are currently available have a cooling capacity of 150 to 2000 tons (527 to 7032 kW). For watercooled
centrifugal chillers with open compressors, the cooling capacity varies from 150 to 10,000
tons (527 to 35,160 kW).
For hermetic compressors using vaporized liquid refrigerant to cool the motor, no additional
motor cooling system is required. Leakage of refrigerant from the bearing seals is also less than that
in the open compressor. Its main drawback is that 2 to 4 percent of the liquid refrigerant must be
used to cool the motor. Its cooling capacity is reduced, and its energy use is increased.
Unfortunately, the IPLV in ARI Standard 550/590-98 is developed to represent the average partload
efficiency of of a single chiller. Most of the chillers are used in multichiller operation. Actual
multichiller operation is different from the single-chiller operation.
13.3 CENTRIFUGAL CHILLERS INCORPORATING
HEAT RECOVERY
System Description
Many large commercial applications often require heating in the perimeter zone but cooling in the
interior zone during the winter. If the building is considered as a whole, a heat recovery system can
be used to transfer the internal heat from the interior zone to the perimeter zone to offset the winter
heating load there.
A typical centrifugal chiller incorporating a heat recovery system using a double-bundle condenser
is presented in Fig. 13.5. In a double-bundle condenser, water tubes are classified as tower
bundles and heating bundles. Heat rejected in the condenser may be either discharged into the
atmosphere through the tower bundle and cooling tower or used for heating in the coils of the perimeter
zones through the heating bundle. The tower bundle may be enclosed in the same shell with the
heating bundle, as shown in Fig. 13.5b, but separate plates are needed to divide the water circuits.
A storage tank stores the excessive hot water from the heating bundle when the building is occupied.
Hot water is then used for heating in the perimeter zone when the building is either occupied
or unoccupied. A supplementary heater using, e.g., electric power may be used to raise the temperature
of the hot water from the heating bundle, if necessary. A tower bypass line is always used to
maintain the condenser water entering the condenser at a required value.
Instead of a double-bundle condenser, an auxiliary condenser plus a standard condenser can be
used, as shown in Fig. 13.6. Hot gaseous refrigerant is forced to either the standard or the auxiliary
condenser after it is discharged from the compressor. When heating is required in the perimeter
zone, the auxiliary condenser pump is energized, and the tower bypass three-way valve bypasses
part of or all the condenser water returning to the standard condenser. Because the water returning
from the perimeter zone and entering the auxiliary condenser is colder than that entering the standard
condenser due to the heating load in the perimeter zone, the auxiliary condenser tube bundle
has a lower temperature and attracts the hot gas from the compressor. The auxiliary condenser recovers
as much heat as it can allow. The remaining rejected heat goes to the standard condenser and
is rejected to the atmosphere through the cooling tower. A balance of heat rejection between these
two condensers automatically holds.
A hot water bypass should be included if an auxiliary condenser is used, as shown in Fig. 13.5.
Such a bypass ensures that water will not circulate through the auxiliary condenser when the chiller
is deenergized. In Fig. 13.5a, the hot water pump P2 is energized only when the compressor of the
centrifugal chiller is in operation.
Operating Modes
Depending on whether heating is required in the perimeter zone, the operating modes of a centrifugal
chiller that is incorporated with a heat recovery system in a building with perimeter and interior
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.9
zones can be divided into cooling mode and heating mode operation. Both the cooling and heating
mode operations can be further divided into occupied and unoccupied periods, based on whether
the conditioned space is occupied. The typical operation of a centrifugal chiller that is incorporated
with a heat recovery system is described below.
13.10 CHAPTER THIRTEEN
FIGURE 13.5 Centrifugal chiller incorporating a heat recovery system: (a) schematic diagram; (b) double-bundle
condenser.
Occupied Cooling Mode. Cooling mode operation takes place when the space temperature in the
perimeter zone Tr, p  75°F (23.9°C). The occupied cooling mode starts when the chilled water
pump P1 is turned on (see Fig. 13.5a). Meanwhile, hot water pump P2 remains inactive; tower
pump P3 forces all the condenser water to flow through the tower bundle, modulation valve V3, and
back to the cooling tower. Chilled water is then supplied to the perimeter zone through zone pump
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.11
FIGURE 13.5 (Continued )
FIGURE 13.6 Heat recovery system using an auxiliary condenser.
P6 or standby pump P7 and to the interior zone through zone pump P4 or standby pump P5. The
chilled water supply temperature Tws is usually between 42 and 45°F (5.6 and 7.2°C). Chilled water
pump P1 and hot water pump P2 are both constant-speed, constant-flow pumps. Zone pumps P5,
P6, P7, and P8 are all variable-speed, variable-flow pumps.
Unoccupied Cooling Mode. When Tr, p  75°F (23.9°C) and the space is unoccupied, as
determined by a timer, the DDC unit controller shuts down the entire centrifugal chiller and heat
recovery system.
Occupied Heating Mode. When Tr, p 	 70°F (21.1°C) and the conditioned space is occupied, the
system is under occupied heating mode operation. The centrifugal chiller is energized, and pump
P1 supplies chilled water to the interior zone through zone pumps P4 and P5. At the same time, the
DDC unit controller actuates the two-position valve V2 to block passage of chilled water to
the perimeter zone. Hot water pump P2 is energized so that hot water from the heating bundle in
the condenser is forced through the deenergized supplementary heater and supplied to heating coils
in the perimeter zone through zone pump P6 or P7 at a temperature Tsup between 105 and 110°F
(40.6 and 43.3°C). If the rejected heat in the heating bundle is more than that required in the heating
coils in the perimeter zone, the return water temperature Tret increases before the inlet of the
heating bundle and pump P2. When sensor T2 detects such a rise in temperature, the DDC unit
controller modulates valve V1, which permits some hot water to flow to the storage tank. When
the storage tank is full of hot water and sensor T2 detects a further increase in water temperature,
the modulation valve is positioned so as to allow all the hot water to bypass the storage tank. Tower
pump P3 starts, and valve V3 is modulated to control the flow rate of condenser water to the
cooling tower.
If the outdoor temperature To is so low that heat rejection at the heating bundle is not sufficient
to meet the heating requirement at the perimeter zone, then the hot water temperature detected by
sensor T2 is less than a preset value. The tower pump is then stopped. Hot water bypasses the storage
tank, and the supplementary heater is gradually energized so as to maintain the water temperature
sensed by T2 within a predetermined range.
Unoccupied Heating Mode. When To 	 70°F (21.1°C) and the conditioned space is unoccupied,
the DDC unit controller commands the following actions:
1. The hot water temperature detected by sensor T2 is reset to a value according to outdoor temperature
To to maintain a night setback indoor temperature Tr  55°F (12.8°C) and lower.
2. Only hot water pump P2 and zone pump P6 or P7 are in operation; all other equipment ceases to
operate.
3. Valve V1 is modulated so that hot water from the storage tank alone is supplied to the heating
coils in the perimeter zone.
If T2 senses a water temperature lower than the preset value, the supplementary heater is then gradually
energized. The modulating valve V1 allows the hot water to bypass the storage tank and then
enter the heaters.
System Characteristics
First, compared with a cooling-only centrifugal chiller, the refrigeration cycle of a centrifugal
chiller incorporating a heat recovery system has a greater pressure differential between condensing
and evaporating pressure pe,c, in psi (kPa). This is because a cooling-only chiller usually has
water leaving the condenser at a temperature at 90 to 95°F (32.2 to 35°C) during summer design
conditions, whereas a chiller incorporating heat recovery needs a higher condenser leaving temperature
Tcl of 105 to 110°F (40.6 to 43.3°C) for the hot water supply in winter heating. Although the
evaporating temperature can be slightly higher in winter than in summer, an increase of 10 to 12°F
13.12 CHAPTER THIRTEEN
(5.6 to 6.7°C) in temperature difference between condensing and evaporating temperature is
roughly 10 to 15 percent more power input to the compressor.
Second, during occupied heating mode at part-load operation, to maintain a preset Tret and let
Tsup float results in a lower mean condenser water temperature Tm,con; instead maintain a preset Tsup
and let Tret float. A lower mean Tm,con reduces the power input to the compressor.
Third, to be more energy-efficient, Tcl should be reduced in part-load operation during winter
heating. Condenser water entering temperature Tce should be adjusted to provide an optimum hot
water supply temperature Tcl at part-load operation.
Fourth, an air economizer reduces the heat recovery effect. This result is due to the reduction of
the refrigeration load, which also reduces the heat recovery from condensing heat.
Finally, for the same building with equal areas of conditioned space, the heating load is smaller
than the cooling load in many locations in the United States. For a refrigeration plant with multiple
centrifugal chillers, usually only some of them include heat recovery. The remainder are coolingonly
centrifugal chillers, which are more efficient.
13.4 AUTOMATIC BRUSH CLEANING FOR CONDENSERS
Principles and Operation
The condenser water from the cooling tower is usually in an open-circuit system. As described in
Sec. 10.3, although many well-maintained cooling towers may have a long-term fouling factor for
condenser tube surfaces of 0.00025 h ft2 °F/Btu (0.000044 m2 °C/W), the fouling factor for condenser
tubes may exceed 0.0005 h ft2 °F/Btu (0.000088 m2 °C/W) when a significant amount of
suspended solids is contained in the condenser water. High fouling factors result in a high
condensing pressure and higher kW/ ton. ARI Standard 550-88 specifies a field fouling allowance
of 0.00025 h ft2 °F/Btu (0.000044 m2 °C/W) for condensers.
The purpose of an automatic brush cleaning system is to remove deposits and dirt from the
water-side tube surface of a condenser for which the fouling factor of the condenser tubes is higher
than 0.00025 h ft2 °F/Btu (0.000044 m2 °C/W). Cleaning brushes in a condenser tube of a horizontal,
two-pass, and shell-and-tube condensers are illustrated in Fig. 13.7. Special nylon brushes
are inserted inside the tube, and catch baskets are attached to each end of the tube.
When condenser water flows through the condenser tubes, the difference in the pressure across
the brush, approximately 1 psi, causes the brush to move in the direction of water flow. As the
brushes in the condenser move, they remove deposit from the inner tube surfaces. The deposits are
carried away by the condenser water. The brushes are then caught by the baskets at one end. The
time taken for the brushes to travel the whole length of the tube is about 1 to 2 min.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.13
FIGURE 13.7 Cleaning brushes in a condenser tube. (Source: Water Services of America. Reprinted with
permission.)
A special four-way valve is mounted at the condenser water entrance, as shown in Fig. 13.8a.
When the four-way valve is in normal operation, the upper half of the brushes are caught by the
baskets at the right-hand end of the condenser tube, and the lower half by the left-hand end baskets.
During normal operation, the condenser water enters the condenser through the lower entrance so
as to provide better subcooling for the liquid refrigerant accumulated at the bottom of the condenser.
After a predetermined time, a signal is given to the control system to unload the compressor by
closing the inlet guide vanes or reducing the speed of the compressor. The four-way valve reverses
the water direction, which causes the brushes to move in the opposite direction, from right to left
for the upper half brushes and from left to right for the bottom half, until they are caught by the baskets
at the ends. As soon as the brushes are caught, a signal is given to the control system to reverse
the water flow again, so that the four-way valve is in normal operation. The brushes stay in the end
baskets during normal operation for about 1 to 4 h until a signal reverses the water flow to repeat
the process. The frequency of the cleaning cycle depends on the contaminants contained in the condenser
water. The nylon brushes usually have a service life of 4 to 5 years.
Effect of Brush Cleaning System
Automatic brush cleaning systems have been found to accomplish the following results:
 The fouling factor drops to 0.00025 h ft2 °F/Btu (0.000044 m2 °C/W) or below. Consequently,
energy costs are lower.
13.14 CHAPTER THIRTEEN
FIGURE 13.8 (a) Normal and (b) reverse operation of a four-way valve in an automatic brush cleaning
system.
 The maintenance cost can be reduced because the interval for cleaning the condenser tubes is
extended from 6 to 12 months or even longer.
 Chemicals for water treatment are still required, but the quantity can be reduced appropriately
according to the actual operating conditions.
Walker (1994) reported that the Walt Disney World Dolphin Hotel used an automatic brush
cleaning system that kept the fouling factor of the condenser tubes around 0.0001 h ft2 °F/Btu
(0.000018 m2 °C/W), so that the investment of the automatic tube-cleaning system was paid back
in just slightly less than 3 years.
13.5 CENTRIFUGAL COMPRESSOR PERFORMANCE MAP
The performance of a centrifugal compressor can be illustrated by a compressor performance map
or, simply, a compressor map. Like a centrifugal fan, a compressor map includes the volume flow,
specific work or power input, efficiency, surge region, and opening of the inlet vanes. Volume flow
and system head are expressed by the compressor performance curve or, simply, compressor curve.
The centrifugal compressor can be operated either at constant speed or at variable speed.
Surge of Centrifugal Compressor
Surge is defined as the unstable operation of a centrifugal compressor or fan with vibration and
noise. Surge occurs when the compressor is unable to develop a discharge pressure that is sufficient
to satisfy the requirement at the condenser. During compressor surge, the pressure flow characteristics
of the compressor fluctuate up and down along the compressor curve, resulting in pressure and
flow pulsation and noise.
Surge first develops at the point where the slope of the pressure lift versus volume flow
curve on the compressor performance curve is zero, as shown in Fig. 13.9. Pressure lift is the total
p-V?
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.15
FIGURE 13.9 Compressor surge.
pressure rise of a centrifugal compressor, in psi (kPa). For a variable-speed centrifugal compressor,
surge first develops at a point of zero slope along the curve that characterizes compressor speed n
versus refrigeration load Qref, in tons (kW) (the n-Qref curve). This point is determined by assuming
a rotating speed, in rpm, that is lower than the minimum possible speed at a required pressure lift
and volume flow of the compressor.
A three-stage centrifugal compressor has a significantly smaller surging region than a two-stage
or a single-stage centrifugal compressor. A single-stage centrifugal compressor using inlet vanes to
control its capacity has a comparatively smaller surging region than a variable-speed compressor.
Surging may damage the components of the compressor. A centrifugal compressor should never
be operated within the surge region on a compressor map.
Centrifugal Compressor Map at Constant Speed
A compressor map of a single-stage centrifugal compressor operated at constant speed is presented
in Fig. 13.10a. The abscissa is the percentage of design volume flow of refrigerant at compressor
suction or load ratio
(13.2) LR 
Qref
Qref, d
100 V?suc /V?suc, d,
13.16 CHAPTER THIRTEEN
FIGURE 13.10 Compressor map of a single-stage centrifugal compressor: (a) constant speed, inlet vane modulation;
(b) variable speed. (Adapted with permission from ASHRAE Handbook 1988, Equipment.)
where Qref, d  refrigeration load at design condition, tons (kW). The ordinate is the ratio of
isentropic work Wisen to the isentropic work at design condition Wis, d, or Wisen /Wis, d. Isentropic
work is defined as the work required for isentropic compression per unit mass of refrigerant, as
described in Sec. 9.20. For the same compressor operated at the same conditions, the ratio
Wisen /Wis,d is approximately equal to the ratio of temperature lifts RT, l and the ratio of system heads
Rhead, that is,
(13.3) RT,1 
Tcon 	 Tev
Tcon, d 	 Tev, d
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.17
FIGURE 13.10 (Continued )
and (13.4)
where Tcon, Tcon, d  condensing temperature and condensing temperature at design condition,
respectively °F (°C)
Tev, Tev, d  evaporating temperature and evaporating temperature at design condition,
respectively °F (°C)
pcon, pcon, d  condensing pressure and condensing pressure at design condition, respectively,
psig (kPag)
pev, pev, d  evaporating pressure and evaporating pressure at design condition, respectively,
psig (kPag)
System head, or head, is the total pressure rise of the refrigerant by the compressor, in psi (kPa). In
a centrifugal refrigeration system, system head is mainly used to raise pev to pcon.
The relationship between the power input to the compressor Pin, in hp, and the isentropic work
Wisen is
(13.5)
where  volume flow of suction vapor, cfm
suc  density of suction vapor, lb/ ft3
com  compressor efficiency
In Fig. 13.10a, compression efficiency cp, as defined in Sec. 9.20, can be calculated as
(13.6)
where mec  mechanical efficiency [see Eq. (9.67)]. Compression efficiency usually drops when
the load ratio falls below 1, as shown by the constant cp contours.
In Fig. 13.10a, “open, 0.75 open, . . ., closed” indicates the fraction of opening of the inlet
vanes, to be discussed in Sec. 13.6. For each fraction of inlet vanes, there is a corresponding compressor
performance curve that relates the load ratio and Rhead.
In Fig. 13.10a, there are three required system head (pcon-pev) curves, illustrated by straight lines
OS, OT, and OW. The system head drops when the load ratio falls below 1. In the top left corner of
Fig. 13.10a, a surge line separates the surging region of this single-stage centrifugal compressor
from the normal operating region. For two- or three-stage centrifugal compressors, the surge region
is smaller.
Centrifugal Compressor Map at Variable Speed
A compressor map of a single-stage centrifugal compressor operating at variable speed is illustrated
in Fig. 13.10b. The abscissa, ordinate, and required system head curves are the same as those in the
compressor map at constant speed (see Fig. 13.10a). The inlet vanes are fully open at both design
load and part-load operations. Their main differences are these:
 Compressor curves are curves at various rotating speeds, such as 105 percent, 100 percent, . . .,
80 percent of speed at design condition, instead of curves at various inlet vane openings on a constant-
speed map.
 At lower heads, the surge region is enlarged.
cp 
com
mec
V?
suc
Pin 
V?
suc
sucWisen
42.41 com
Rhead 
pcon 	 pev
pcon, d 	 pev, d
13.18 CHAPTER THIRTEEN
13.6 CAPACITY CONTROL OF CENTRIFUGAL CHILLERS
Capacity control of a centrifugal chiller is achieved mainly by modulating the volume flow of the
refrigerant at the centrifugal compressor.
Difference between Centrifugal Compressors and Centrifugal Fans
During part-load operation, the following means are used to modulate the capacity of a centrifugal
compressor:
 Open the inlet vanes.
 Use an adjustable-frequency ac inverter to vary the compressor rotating speed. Adjustablefrequency
variable-speed drive is discussed in Sec. 15.4.
 Combine the opening of inlet vanes and variable-speed control by means of an inverter.
 Vary the speed of the steam- or gas-turbine-driven large centrifugal chillers.
A hot-gas bypass provides no energy saving at reduced system loads as specified in Sec. 11.5.
Although centrifugal compressors and centrifugal fans are both centrifugal turbomachinery, and
both centrifugal chillers and centrifugal fan duct systems use inlet vanes and variable-speed drives
to provide capacity control during part-load operation, there is an important difference between the
centrifugal chiller and centrifugal fan duct system. In a fan duct system, the total pressure of the
centrifugal fan is used mainly to overcome the friction and dynamic losses in airflow. This required
fan total pressure drops considerably when the air volume flow is reduced. In the centrifugal chiller,
on the other hand, the system head is used mainly to lift the evaporating pressure to condensing
pressure to produce the refrigerant effect. The pressure drop due to friction and dynamic losses of
refrigerant flow is only a small part of the required head. Reduced volume flow of refrigerant at
part-load operation has a minor influence on the required system head.
The primary factor that affects the system head at part-load operation is the difference between
condensing and evaporating pressures or temperatures, pcon	 pev or Tcon	 Tev. For a water-cooled
condenser using condenser water cooled from a cooling tower, the outdoor wet-bulb temperature
directly influences the entering temperature of condenser water and, therefore, the condensing
temperature and pressure.
Required System Head at Part-Load Operation
During part-load operation, because the condensing and evaporating surface areas at reduced
volume flow of refrigerant are greater than the required area, there is a drop in condensing temperature
Tcon and an increase in evaporating temperature Tev. Most probably, the outdoor wet-bulb
temperature also drops at reduced load. Therefore, the profiles for required system head at part-load
operation for a water-cooled centrifugal chiller usually fall into the following three schemes:
 Scheme A. Constant 85°F (29.4°C) condenser water entering temperature Tce and constant 44°F
(6.7°C) chilled water leaving temperature Tel: The temperature differential Tcon 	 Tev at a load
ratio equal to 1 is 66°F (36.7°C). At a load ratio of 0.2, Tcon 	 Tev is about 45°F (25°C).
 Scheme B. Constant Tel and a drop of 2.5°F (1.4°C) in Tce for each 0.1 increment of reduction of
load ratio. The quantity Tcon 	 Tev at design load is still 66°F (36.7°C), whereas at load ratio 
0.2, Tcon 	 Tev is about 26°F (14.4°C).
 Scheme C. A reset of Tel of 1°F (0.56°C) increase and a drop of 2.5°F (1.4°C) of Tce for each 0.1
increment of reduction of load ratio. Under this scheme, Tcon 	 Tev at design load is still 66°F
(36.7°C), and at 0.2 load ratio, Tcon 	 Tev is about 19°F (10.6°C).
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.19
Required system head curves for schemes A, B, and C at various load ratios are shown in
Fig. 13.10.
On a compressor map (e.g., Fig. 13.10a), the intersections of the compressor performance curve
and the required system head curve at that load ratio—points P, Q, R, . . . —are the operating
points of the centrifugal chiller. The required system head curve joins the operating points together,
to form an operating curve of the centrifugal chiller at various load ratios.
Capacity Control Using Inlet Vanes
Inlet vanes are pivoted movable vanes installed at the inlet of the impeller, as shown in Fig. 13.1. In
Fig. 13.1 only one set of inlet vanes is shown. For a three-stage centrigugal chiller with three impellers,
actually there are three sets of inlet vanes, one for each impeller, to improve the chiller’s
performance. These vanes are linked mechanically so that they can turn simultaneously around the
axis. When the opening of the inlet vanes at the inlet of the centrifugal compressor is reduced, it
throttles the volume flow of vapor refrigerant and imparts a spin or prerotation on the refrigerant
flow before it enters the impeller. If the direction of the spin is the same as the direction of rotation
of the impeller, the tangential component of the absolute velocity of the fluid elements has a positive
value. The head lift thus developed is smaller than if vapor refrigerant enters the impeller with
a radial entry. This effect produces a new performance curve at a lower head and volume flow at
each position of reduced vane opening.
The compression efficiency com decreases as the inlet vane opening angle decreases, as the vapor
refrigerant at inlet is further deviated from the radial entry and the inlet area is reduced. At high
load ratios with larger openings, compression efficiency is only slightly affected. At low load ratios,
throttling at the inlet vanes causes considerable reduction of compression efficiency. However, the
reduction of shaft power from the decreases in head lift p and volume flow of vapor refrigerant
compensates for the effect of the drop in com and shows energy saving from inlet vane capacity
modulation.
For a single-stage centrifugal compressor using inlet vanes to modulate its capacity (see
Fig. 13.10a), the following can be noted:
 If the condenser water entering temperature Tce remains at the design condition, the operating
curve hits the surge line when the load ratio drops to about 0.12.
 When the condenser water temperature drops 2.5°F (1.4°C) for every 0.1 increment of reduction
of load ratio, at a load ratio of 0.5 the compression efficiency of the compressor then drops below
50 percent.
Capacity Control by Variable Speed
When an adjustable-frequency variable-speed drive is added to a centrifugal chiller, the rotating speed
of the centrifugal compressor varies and a family of compressor curves is produced (see Fig. 13.10b).
A lower rotating speed results in a performance curve of lower head and volume flow of refrigerant.
One can see the following in Fig. 13.10b:
 When the load ratio  0.5, the compression efficiency of a centrifugal compressor using variablespeed
control is not less than 75 percent.
 If the condenser water entering temperature Tce remains at the design value, the operating curve
hits the surge line at a load ratio of about 0.27.
 If Tce drops and Tel is reset according to scheme C, the operating curve intersects the surge line at
a load ratio of about 0.18.
When capacity control is to be achieved by varying the compressor rotating speed, a reset of chilled
water leaving temperature Tel should be used.
V?
13.20 CHAPTER THIRTEEN
Comparison between Inlet Vanes and Variable Speed
Inlet vanes modulation is simple and has low initial cost. At lower load ratios, it functions as a
throttling device and produces considerable dynamic and friction losses and, consequently, a lower
compression efficiency than a variable-speed compressor. In Fig. 13.10, inlet vanes have a comparatively
smaller surge region than in a variable-speed compressor in a single-stage centrifugal
compressor when the load ratio falls below 0.3. However, a two- or three-stage centrifugal
compressor significantly reduces the surging region.
Hubbard (1997) and Bellenger (1996) showed that the annual energy use of a centrifugal
chiller with an adjustable-frequency variable-speed drive compressor is about 10 to 15 percent
lower than that of a constant-speed compressor depending on the system characteristics of the
chiller. The simple payback period for an adjustable-frequency variable-speed drive ranges from 5
to 10 years. Erth (1987) compared the power consumption of centrifugal chillers, in kW/ ton, of
constant-speed centrifugal compressors using inlet vanes and variable-speed compressors with
or without a chilled water leaving temperature Tel reset, as shown in Fig. 13.11. At a load ratio of
0.25, variable-speed control with Tel reset has a power consumption of only about one-half its
design value.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.21
FIGURE 13.11 Comparison between constant-speed and variable-speed
centrifugal chiller. (Source: R. A. Erth, Heating/Piping/ Air Conditioning,
May 1987. Reprinted with permission.)
13.7 CENTRIFUGAL CHILLER CONTROLS
Functional Controls and Optimizing Controls
A large modern centrifugal chiller plant usually includes multiple chillers to prevent the entire shutdown
of the refrigeration system when one chiller fails to operate. Control of a centrifugal chiller
plant, therefore, includes the functional controls and optimizing controls, as well as the unit control
of a single chiller and the system control of a multiple-chiller plant. Unit control consists of specific,
safety, diagnostic, and optimizing controls. Only a microprocessor-based DDC system can
possibly perform all these duties. For functional controls, a DDC unit controller monitors and
controls the following:
 Chilled water leaving temperature control
 Air purge
 Safety features, including oil pressure, low-temperature freezing protection at the evaporator, high
condensing temperature control, motor overheating, and time delaying
 Status features, including condenser water flow, chilled water flow, electric current, and refrigerant
and chilled water temperatures and pressures at key points
Chiller plant optimizing control with multiple centrifugal chillers includes the following:
 Optimize the on/off staging of the multiple chillers.
 Optimize start and stop.
 Control condenser water entering temperature.
Controls of the multiple chillers in a centrifugal chiller plant are discussed in later sections.
Chilled Water Leaving Temperature Control
Continuous modulation control is extensively used for chilled water leaving temperature Tel control
in centrifugal chillers. A temperature sensor located in the chilled water pipes at the exit of the
evaporator is used to sense Tel. The DDC unit controller either varies the position of the opening of
the inlet vanes or varies the speed of the compressor motor through an adjustable-frequency ac invertor.
The volumetric flow of refrigerant then decreases or increases, as does the capacity of the
centrifugal chiller. Chilled water leaving temperature control is actually the capacity control of a
centrifugal chiller.
The temperature of the chilled water leaving the evaporator Tel is often reset to a higher value at
reduced load according to either the outdoor temperature To or the load ratio. The load ratio can be
indicated by the chilled water return temperature. An increase in Tel may reduce the dehumidifying
capability at the cooling coil, especially at very low load ratios.
A better scheme for Tel reset is this: When the outdoor temperature is high, that is, To 
 80°F
(26.7°C), there is no reset. Only when To is less than 80°F (26.7°C) is To reset for an increase of
0.2 to 0.3°F (0.11 to 0.17°C) for each 1°F (0.56°C) drop in To. When To falls below 50°F (10°C), To
remains constant.
For energy-efficient design and operation, as discussed in Sec. 7.11, ASHRAE/IESNA Standard
90.1-1999 specifies that chilled and hot water systems with a design capacity exceeding 300,000
Btu/h (87,900 W) supplying chilled water to comfort air conditioning systems shall include
automatically reset chilled water leaving temperature according to building loads or outdoor temperature.
A proportional-integral-derivative (PID) control mode is often used for modulation control
of the chilled water leaving temperature for better performance.
13.22 CHAPTER THIRTEEN
Condenser Water Temperature Control
Interaction of the centrifugal chiller and the cooling tower must be considered when one
optimizes the condenser water temperature entering the condenser Tce so as to minimize power
consumption. The control strategy is to minimize the total power consumption of the chiller and
tower fans.
According to Cascia (1988), the condenser water temperature should be floated; i.e., it should
rise and fall with the outdoor wet-bulb temperature in agreement with the following requirements:
 Temperature Tce must be maintained above the lower limit specified by the centrifugal chiller
manufacturer, such as 60°F (15.6°C), in order to maintain a necessary head pressure for proper
operation. If the lower limit is reached, the tower fan must be turned off first; then the tower bypass
valve should be modulated to maintain Tce equal to or above the specified limit.
 If Tce reaches the high limit, possibly more tower fans should be added so that Tce does not exceed
the high limit.
A temperature sensor should be used to sense Tce. A DDC unit or system controller actuates the
cooling tower fans and the cooling tower bypass valve to maintain the predetermined Tce.
Safety Controls
Electric Demand Limit Control. Electric demand or electric current limit control is used to prevent
the electric current from exceeding a predetermined value. A motor current is measured at the
current, transformers, and a proportional signal is sent to the DDC unit controller. If the current
exceeds a predetermined limit, the controller directs the inlet vanes to close or reduces the rotating
speed of the compressor through the adjustable-frequency ac inverter. The limit could be from 40 to
100 percent of the electric demand.
Motor Protection. The compressor motor assembly is the most expensive part of the centrifugal
chiller. For a typical control that protects a motor against high temperatures, if the resistance temperature
detector in the motor stator winding senses a temperature under 165°F (73.9°C), the DDC
unit controller introduces a time delay, normally of 4 min, after the chiller switch is turned on. If
the sensed temperature exceeds 165°F (73.9°C), a start delay of 15 min is required. If the stator
winding temperature exceeds a dangerous upper limit, such as 265°F (129.4°C), the DDC unit controller
then prohibits the operation of the motor because of its high temperature.
There are also protection controls against current overload, temporary power loss, low voltage,
and phase unbalance.
Low-Temperature and High-Pressure Cutouts. Refrigerant temperature at the evaporator is detected
by a sensor. This information is sent to the DDC unit controller so that if the predetermined
low-temperature cutout is approached, inlet vanes hold or close to provide freeze protection for the
evaporator.
A pressure sensor detects the condensing pressure of the refrigerant. The DDC unit controller
outputs override the inlet vane position instruction to close the inlet vanes if the condensing pressure
reaches the predetermined limit.
High Bearing Temperature. If the bearing temperature detected by a temperature sensor exceeds
a preset value, the chiller is stopped immediately.
Short-Cycling Protection. Usually a time delay, typically of 30 min, is needed to restart a centrifugal
chiller after it has been turned off to prevent short cycling.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.23
Surge Protection. A surge detection device is used to send a signal to the DDC unit controller. A
head relief scheme is implemented by lowering the condenser water temperature. If the surge continues
beyond a predetermined time period, the compressor is shut down.
Loss of Flow. If a loss of flow in evaporator or condenser is detected by the flow switch, the
chiller will be shut down accordingly.
Air Purge. An automatic purge operation includes the operation of an air purge unit within a specific
time interval, for instance, 5 min of air purge every 2 h. Both condensable gas and condensable
vapor, mainly water vapor, leaking into the centrifugal chiller raises the purge drum pressure. If an
increase in drum pressure is detected by a pressure sensor, the period of operation of the air purge
unit is lenthened.
Fault Detection and Diagnostics
As discussed in Sec. 5.15, fault detection and diagnostics aid in detecting a problem, solving a
problem, and getting the chiller on-line quickly in case there is a problem. For a typical diagnostic
display module, more than 100 diagnostic checks and displays are available. Diagnostic modules
are also able to display the run test to verify the thermistor, potentiometer, oil pump, air purge unit,
and compressor before the chiller is started.
Operator Interface
All the inputs (such as switches and set points), output displays (such as operating mode indicators,
key parameter monitors, and status lights), and diagnostic displays should be visible at a glance on
the control display panel. They should be easy to understand and to use.
13.8 OPERATING CHARACTERISTICS OF
CENTRIFUGAL CHILLERS
Sequence of Operation of a Typical Centrifugal Chiller
Many centrifugal chillers in commercial buildings are shut down at night in summer. Usually, one
of the chillers is operated to serve areas needing 24-h service, such as computer areas. The temperature
of chilled water in the shut-down system rises during the shutdown period. Chillers are often
started earlier than air-handling units (AHUs) or terminals to pull down the chilled water temperature
so that its Tel is maintained within acceptable limits.
When the switch of a centrifugal chiller is turned on, the sequence of operations to start and run
a centrifugal chiller is as follows:
1. The sensed chilled water leaving temperature Tel is higher than the set point, indicating a demand
for refrigeration.
2. The interlocked chilled water pump switch starts the chilled water pump, and the chilled water
flow is sensed by the chilled water flow sensor.
3. A signal is issued to start the condenser water pump and the tower fans; the condenser water
flow is sensed by the flow switch.
4. A signal is sent to start the oil pump, and the oil pressure is sensed by the pressure sensor.
5. The short-cycling protection timer is checked to see whether the time period after the previous
shutdown is greater than the predetermined limit.
6. A command is sent to close the inlet vanes and is confirmed.
13.24 CHAPTER THIRTEEN
7. After the oil pressure has been established for 10 to 20 s, a start command is sent to the compressor
motor starter.
8. A successful start is followed by a display of “run normal.” Failure to perform a successful start
shuts down the compressor, and the diagnostic code is displayed for correcting the fault. After
correction and reset, the start sequence may be repeated to attempt a successful start again.
9. During “run normal” operation, the current limit, the condenser high-pressure limit, and the
evaporator low-temperature cutout are monitored to see whether these limits have been
approached. If so, the relevant controls hold or close the inlet vanes so that these limits are not
exceeded. When a surge is detected, the head relief relay is energized. If surge continues for
15 min, the compressor will shut down.
10. When the DDC unit controller originates a stop signal based on the timer’s schedule, a “preparing
to shut down” message appears as the inlet guide vanes close or the variable-speed compressor
unloads. The compressor and condenser pump motor starter are deenergized. The oil
pump will continue to run for about 10 min. Finally, the chilled water pump for this chiller is
deenergized. The chiller can be manually stopped at any time by pushing the Stop key.
System Balance at Full Load
When a centrifugal chiller is operating at full design load under the standard rating conditions, that
is, Tel  44°F (6.7°C), and chilled water volume flow gpm/ton (0.043 L/ s kW), Tce 
85°F (29.4°C) and condenser water volume flow gpm/ton (0.054 L/ s kW).
If the capacity of the selected evaporator is greater than the compressor at design conditions,
more liquid refrigerant is vaporized in the evaporator than the compressor can extract. The evaporating
pressure gradually rises until the increase in the mass flow rate due to a higher density of suction
vapor extracted by the compressor is just equal to the mass flow rate of refrigerant vaporized in
the evaporator. A new system balance forms between the evaporator and compressor at a higher
evaporating temperature Tev.
On the other hand, if the capacity of the selected compressor is greater than the evaporator, more
evaporized refrigerant is extracted than the liquid refrigerant vaporized in the evaporator. The evaporating
pressure and temperature gradually drop until the mass flow rate of suction vapor of the centrifugal
compressor is just equal to the mass flow rate of the vaporized refrigerant in the evaporator.
If the capacity of the selected condenser is greater than the compressor at design conditions, more
hot gas can be condensed to liquid form in the condenser than the compressor can supply. The result is
a drop in condensing pressure pcon and temperature Tcon. Because the condenser water entering temperature
remains unchanged, a drop in Tcon results in a lower log-mean temperature and a reduction in
the condensing capacity. A new system balance is then formed between the mass flow rate of hot gas
condensed in the condenser and that discharged from the centrifugal compressor at a lower Tcon.
On the other hand, if the capacity of the selected centrifugal compressor is greater than the condenser,
more hot gas is discharged to the condenser by the compressor than can be condensed in the
condenser. The result is a rise in condensing pressure pcon and temperature Tcon. A rise in Tcon means
a higher log-mean temperature and a greater condenser capacity, as well as a greater head lift,
which reduces the discharged volume flow rate of the hot gas from the centrifugal compressor. Both
effects are helpful to make a new system balance between the compressor and the condenser.
13.9 PART-LOAD OPERATION OF CENTRIFUGAL CHILLER
Part-Load Operating Characteristics
If we assume that a centrifugal chiller is operated initially at full design load and is in a condition of
system balance, then chilled water leaving the evaporator is controlled at a temperature of 44°F
(6.7°C), and the return temperature Tee is 60°F (15.6°C).
V?
con  3
V?
ev  2.4
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.25
If there is a reduction of coil load in the AHUs or in the terminals, the chilled water temperature
entering the evaporator Tee falls. Because the rate of heat transfer and the amount of refrigerant
evaporated are only slightly affected, Tel is lower than 44°F (6.7°C), the set point.
When the temperature sensor at the exit of the evaporator senses a drop in Tel and signals the
DDC unit controller, the positioner of the inlet vanes is actuated to close the opening. The centrifugal
compressor is now operating at less than the design head, and the mass flow rate of the refrigerant
extracted by the compressor is reduced from to . At part load, both the evaporating pressure pev
and temperature Tev tend to increase. Such an effect further reduces Tm, the boiling heat-transfer coefficient
hb, and the overall heat-transfer coefficient Uo. The increase in mass flow rate due to a smaller
head lift at part-load operation is offset by the closing of the inlet vanes of the compressor. The rate of
heat transfer and the rate of evaporation are, therefore, decreased until the mass flow rate of evaporated
refrigerant is equal to the mass flow rate of the refrigerant extracted by the centrifugal compressor.
Because the rate of condensation of refrigerant in the condenser is greater than the reduced mass
flow rate of hot gas discharged from the compressor , the condensing pressure pcon and temperature
Tcon tend to drop until the rate of condensation is equal to . Consequently, there is a reduction
of total heat rejection in the condenser.
A higher evaporating pressure and a lower condensing pressure result in a small pressure difference
pcon-pev, head lift, as well as a lower hydrostatic liquid column in the condenser. As a result,
the liquid refrigerant flow through the float valve or orifice plates is reduced.
When the compressor operates with the inlet vanes closed at a small angle, i.e., at a load ratio
between 0.8 and 1, the compression efficiency decreases only slightly. However, because a lower
head and a lower volume flow have a great effect on the work input and power consumption of the
compressor, the result is a reduction in the power input. The energy performance index, in kW/ ton,
of the compressor is reduced less than power input because of the reduction of Qref.
The reduction in the mass flow rate of the refrigerant that flows through the centrifugal compressor
and the orifice plates and the reduction of the rates of evaporation and condensation create a
new system balance during part-load operation.
For a constant-speed centrifugal chiller, power consumption, in kW/ton, rises when the load
ratio drops below 0.55 (see Fig. 13.11). This increase occurs because the fall in compression
efficiency exceeds the savings in power consumption at a lower head and flow during low load
ratios when inlet vanes are closed to a smaller opening.
Evaporating and Condensing Temperatures at Part-Load Operation
If other conditions remain the same, the greater the heat-transfer surface area Ao in the evaporator, the
higher the evaporating temperature Tev. Also, the greater Ao in the condenser, the lower the condensing
temperature Tcon. Consequently, greater Ao results in a lower temperature lift between Tcon and Tev.
From Eq. (10.9), if the thermal resistance of metal is ignored, the rate of heat transfer at the
shell-and-tube evaporator and condenser can be calculated as
(13.7)
(13.8)
where f  fin efficiency; for integrated fins on copper tubes, its value is between 0.9 and 0.95.
From Eq. (10.12), the boiling coefficient is given as
(13.9)
Meanwhile, from Eq. (10.18), the condensing coefficient can be calculated as
(13.10) hcon  Ccon 1
Qrej /Ao 1 /3
hb  C  Qev
Ao n
Uo 
1
1/( f ho)  Rf (Ao /Ai)  (Ao /Ai)(1/hi)
Q  Uo Ao Tm
m?r
m?r
m? r m? r
13.26 CHAPTER THIRTEEN
The log-mean temperature difference across the evaporating and condensing surfaces can be expressed
as
(13.11)
Then the evaporating temperature Tev can be evaluated as
(13.12)
where (13.13)
Similarly, the condensing temperature:
(13.14)
where (13.15)
where Aev, Acon  evaporating and condensing surface area, ft2 (m2)
Uoe, Uoc  overall heat-transfer coefficient based on evaporating and condensing surface
area, Btu/h ft2 °F (W/m2 °C)
Example 13.1. A centrifugal chiller using HCFC-123 as the refrigerant has the following operating
parameters for its shell-and-tube liquid cooler at design conditions:
C 
Tcl 	 Tce
Qrej /(AconUoc)
Tcon 
eCTcl 	 Tce
eC 	 1
B 
Tee 	 Tel
Qev /(AevUoe)
Tev 
eBTel 	 Tee
eB 	 1

Tcon 	 Tcl 	 (Tcon 	 Tce)
ln[(Tcon 	 Tcl) /(Tcon 	 Tce)]
Tm 
Tee 	 Tev 	 (Tel 	 Tev)
ln[(Tee 	 Tev) /(Tel 	 Tev)]
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.27
54°F (12.2°C)
44°F (6.7°C)
36°F (2.2°C)
3.5
0.00025 h ft2 °F/ Btu (0.000044 m2 °C/W)
1700 Btu /h ft2 °F (9653 W/m2 °C)
Chilled water entering temperature Tee
Chilled water leaving temperature Tel
Evaporating temperature Tev
Ratio of outer to inner surfaces Ao /Ai
Fouling factor on water side
Water-side heat-transfer coefficient hi
1. If the boiling coefficient of HCFC-123 can be calculated by hb  2.5 (Qev /Ao)0.7 and f  1,
calculate the outer evaporating surface area required in the shell-and-tube evaporator for 1 ton
(3.5 kW) of refrigeration capacity.
2. Calculate the evaporating temperature for the same evaporator at a load ratio of 0.5.
Solution
1. From Eqs. (13.8), and (13.9) and the given data, the overall heat-transfer coefficient at design
condition is

1
1/[2.5(Qev /Ao)0.7]  0.00025 3.5  3.5/1700
Uo 
1
1/hb  RfAo/Ai  (Ao /Ai)(1/hi)
from Eq. (13.11), the log-mean temperature difference at the evaporator is
Then, from Eq. (13.7),
Rearranging, we find
Solving by iteration, we have
For Qev  1 ton  12,000 Btu /h, then,
That is, for each ton of refrigeration, there should be 4.36 ft2 (0.41 m2) of outer surface area Ao for
the tubes at the evaporator.
2. At 50 percent part-load operation, Qev  0.5 12,000  6000 Btu / h. Then
and
Also,
From Eq. (13.7), then, at 50 percent part-load operation
Since the chilled water flow rate at the plant loop is constant, then at 50 percent part-load operation
Tee 	 Tel  0.5(54 	 44)  5°F. Also, from Eq. (13.13),
B 
Tee 	 Tel
Qev / (Uo Ao)

5
6000/ (182.6 4.36)
 0.663
Tm 
Qev
Uo Ao

6000
182.6 4.36
 7.54F (4.2C)
Uo 
1
1 / 393  0.00025 3.5  3.5 /1700
 182.6 Btu / h  ft2  F (1037 W/m2  C)
hb  2.5 Qev
Ao 0.7
 2.5(1375)0.7  393 Btu / h  ft2  F
Qev
Ao

6000
4.36
 1375 Btu/h  ft2
Ao 
12,000
2750
 4.36 ft2 (0.41m2)
Qev
Ao
 2750 Btu / h ft2
0.00734
Qev
Ao
  Qev
Ao 0.3
 30.83
Qev /Ao  Uo Tm 
12.33[2.5(Qev /Ao)0.7]
1  0.00734(Qev /Ao)0.7

54 	 36 	 (44 	 36)
ln[(54 	 36) /(44 	 36)]
 12.33°F (6.85°C)
Tm 
Tee 	 Tev 	 (Tel 	 Tev)
ln[(Tee 	 Tev) /(Tel 	 Tev)]
13.28 CHAPTER THIRTEEN
Therefore, from Eq. (13.12) the evaporating temperature at 50 percent part-load operation is
For part-load operation at a load ratio of 0.5, Tev rises 38.69 	 36  2.69°F (1.5°C).
Temperature Lift at Part-Load Operation for Centrifugal Chillers
Figure 13.12 is a diagram of temperature lift at various load ratios for a centrifugal chiller using
HCFC-123 as refrigerant. The chilled water temperature leaving the chiller is maintained at 44°F
(6.7°C) at all load ratios, and the outdoor wet-bulb temperature also remains constant at 75°F
(23.9°C) when the load ratio varies from 0 to 1. Fouling factors for both evaporator and condenser
are 0.00025 h ft2 °F/ Btu (0.000044 m2 C/W).

e0.663 44 	 49
e0.663 	 1
 38.69F (3.72C)
Tev 
eBTel 	 Tee
eB 	 1
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.29
FIGURE 13.12 Temperature lift at various load ratios of a centrifugal chiller.
Under design conditions, if the cooling tower approach is 9°F (5°C), the evaporating temperature
Tev is 8°F (4.4°C) lower than Tel, the condensing temperature Tcon is 8°F (4.4°C) higher than
Tcl, and Tcl is 10°F (5.6°C) higher than Tce. Therefore, the condensing temperature is
and the evaporating temperature is
The temperature lift at design condition is, therefore,
If the load ratio  0, theoretically Tcon drops to 75°F (23.9°C), and Tev rises to 44°F (6.7°C). Therefore,
the temperature lift Tlift drops to
This drop is about 31 /66  0.47, or 47 percent, of the design temperature lift Tlift, d.
The drop in temperature lift is caused by the following:
 A decrease in condenser water entering temperature Tce because of the smaller value of approach
in the cooling tower at reduced load ratios
 A drop in condenser water temperature rise (entering and leaving temperature difference) because
of the reduction in total heat rejection
 A decrease in the difference between the condensing temperature and the condenser water leaving
temperature Tcon 	 Tcl
 A decrease in the difference between the chilled water leaving temperature and the evaporating
temperature Tel 	 Tev.
Tlift  75 	 44  31F (17.2C)
Tlift, d  Tcon 	 Tev  102 	 36  66F (36.7C)
Tev  44 	 8  36F (2.2C)
Tcon  75  9  10  8  102F (38.9C)
13.30 CHAPTER THIRTEEN
Power input ratio
KW/ton 1.00 2.0
1.6
1.2
0.8
0.4
0
0.80
0.60
0.40
0.20
0.20 0.40
Load ratio
Av. LCHWT  average leaving chilled water temperature
Av. ECDWT  average entering condenser water temperature
(a)
0.60 0.80 1.00 0.2 0.4
Load ratio
(b)
0.6 0.8 1.0
0
Field measured data
Av. LCHWT  46.1F
Av. ECDWT  80.1F
Entering condenser
water temperature
85F
75F
65F
FIGURE 13.13 Relationship (a) between power input ratio and load ratio and (b) between kW/ton and load ratio
of a water-cooled centrifugal chiller.
The actual percentage of design power input may be higher or slightly lower than the load ratios,
depending on
 Whether the outdoor wet-bulb temperature is constant or drops at part-load operation
 whether there is a Tel reset
 the decrease of compression efficiency at part-load operation
Figure 13.13a is a plot of the relationship between the power input ratio Pin /Pin, d and load ratio
Qref /Qref, d at various condenser water entering temperatures of 65, 75, and 85°F (18.3, 23.9, and
29.4°C). Here, Pin and Pin, d represent the power input and design power input, in hp (kW). Likewise,
Qref and Qref,d are the refrigeration load and the design refrigeration load, in Btu / h (W). The
lower the load ratio, and the lower the condenser water entering temperature Tce, the lower the
power input ratio. For a specific load ratio, power input ratio Pin /Pin, d is alway less than the load ratio
Qref /Qref, d except when Tce  85°F (29.4°C) and Qref /Qref, d  0.4.
Figure 13.13b is a plot of the chiller efficiency curve, a curve that relates the energy use in
kW/ton of the chiller and the load ratio of a water-cooled centrifugal chiller based on Austin
(1991). The optimum loading point is at 58 percent load at ARI condenser water conditions, 80 percent
load at constant 85°F (29.4°C) condenser water entering temperature. ARI condenser water
conditions are based on a 2.5°F (1.4°C) condenser water temperature drop for each 10 percent load
decrease starting with 85°F (29.4°C) at 100 percent load. Lower condenser water temperature needs
a higher optimum load point.
13.10 FREE REFRIGERATION
Principle of Operation
Free refrigeration means the production of a refrigeration effect without the operation of a compressor.
This effect can be achieved because refrigerant tends to migrate to the area of lowest temperature
in the system when the compressor ceases to operate.
Air-free refrigeration operation for a single-stage, water-cooled centrifugal chiller is illustrated
in Fig. 13.14. When the free refrigeration is turned on, the compressor ceases to operate, the inlet
vanes are fully opened, and the valve in the passage that connects the condenser and evaporator is
opened.
When the temperature of condenser water from the cooling tower is lower than the temperature
of the chilled water leaving the evaporator, the higher saturated pressure of the refrigerant in the
evaporator forces the vaporized refrigerant to migrate to the condenser, where the saturated pressure
of the refrigerant is lower. The absorption of latent heat of vaporization from the chilled water in
the evaporator’s tube bundle causes the temperature of chilled water to drop. The condensation of
vaporized refrigerant into liquid results in the rejection of the latent heat of condensation to the condenser
water through the tube bundle in the condenser, raising the temperature of condenser water a
few degrees. Liquid refrigerant drains to the evaporator as a result of gravity. Free refrigeration is
thus achieved.
Automatic free refrigeration control can be provided by the DDC system. When the condenser
water temperature is a specified number of degrees lower than the required chilled water temperature,
the DDC unit controller starts the free refrigeration cycle. The shutoff valve is opened, and an
interlocked circuit shuts the compressor off. The free refrigeration cycle lasts as long as the
condenser water from the cooling tower is a few degrees higher than the temperature of the chilled
water leaving the evaporator, and as long as the free refrigeration capacity meets the refrigeration
load required.
For a two-stage centrifugal chiller with a flash cooler, or a three-stage chiller with a two-stage
flash cooler, there are more connecting passages and shutoff valves. The principle of operation
of free refrigeration is still the same. Once the refrigeration required is greater than the free
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.31
refrigeration capacity, as determined from the temperature of the chilled water returned to the evaporator,
the DDC unit controller closes the vapor and liquid shutoff valves and starts the compressor.
The vaporized refrigerant from the evaporator is extracted by the compressor, discharged to the
condenser, and condensed to liquid refrigerant as in normal operation.
Free Refrigeration Capacity
Free refrigeration capacity depends on the following parameters:
 The temperature difference between the chilled water leaving the evaporator and the condenser
water entering the condenser.
 The size of the evaporator.
 Condenser water flow.
 Fouling factors at the evaporator and the condenser. An allowable fouling factor is usually
0.00025 h ft2 °F/ Btu (0.000044 mC/ W). A higher fouling factor means a lower free refrigeration
capacity.
Free refrigeration capacity is always expressed as a percentage of the base capacity of the centrifugal
chiller. For a standard-size evaporator in a typical centrifugal chiller that is running in a free
refrigeration cycle, if the chilled water leaves the evaporator at 48°F (8.9°C) and the condenser
13.32 CHAPTER THIRTEEN
FIGURE 13.14 Free refrigeration operation for a single-stage, water-cooled centrifugal chiller.
water enters the condenser Ten,c at 35°F (1.7°C), then the free refrigeration capacity Qref, f can be 45
percent of the base capacity. When Ten, c is 40°F (4.4°C), Qref, f is 25 percent; and when Ten, c is 44°F
(6.7°C), Qref, f may be zero.
Free refrigeration saves energy input to the compressor and can be applied to locations where
the condenser water temperature can be at least 5°F (2.8°C) lower than the temperature of chilled
water leaving the evaporator. A free refrigeration arrangement is usually optional.
13.11 OPERATION OF MULTIPLE-CHILLER PLANT
Multiple-Chiller Plant
A large centrifugal refrigeration system is often located in a multiple-chiller plant in which two or
more centrifugal chillers are installed. Multiple chillers improve the system reliability. On the contrary,
too many smaller chillers often lower the efficiency and increase the initial cost.
A multiple-chiller water-cooled plant also comprises cooling towers, chilled water pumps, condenser
water pumps, piping, accessories, and controls. Among these, centrifugal compressors,
chilled and condenser water pumps, and cooling tower fans are energy users.
According to Bellenger and Becker (1996) and Knebel (1997), the electric power consumption
of the chilled water and condenser water pumps varies between 0.1 and 0.2 kW / ton, and the cooling
tower fans vary from 0.04 to 0.08 kW / ton depending on the efficiency of pumps and fans as
well as the head of the pumps.
Parallel and Series Piping
Multiple chillers can be piped in parallel, as shown in Fig. 13.15a. Each chiller usually has its own
chilled water pump. If a butterfly valve is installed for each chiller-pump combination, both the
cooling capacity and the chilled water flow are turned on and off in sequence and are controlled by
the DDC system controller. The flexibility and reliability of operation of parallel chillers are excellent.
Chillers can be withdrawn from operation for maintenance without affecting the others. For a
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.33
FIGURE 13.15 (a) Parallel chillers and (b) series chillers.
plant with two or more parallel piped chillers, the chiller that is turned on first is called the leading
chiller, and the other chillers are called lag chillers.
When multiple chillers are piped in series, as shown in Fig. 13.15b, there are usually only two
chillers in series, and they have the following characteristics:
 Chilled water coils are sized for a relatively higher temperature rise and lower water flow rate.
 The pressure drop of chilled water through two series-connected evaporators is additive.
 Control is more complicated in part-load operation.
In practice, multiple-chiller plants are often piped in parallel.
Chiller Staging
In a multiple-chiller plant using a plant-building loop combination, the principle of staging the
chillers, on or off, is based on these considerations: (1) The chilled water flow rate in the plant loop
is always slightly higher than the building loop; (2) staging on the chiller(s) must meet the requirements
of the increased system load; (3) power input to the chillers and chilled water pumps should
be minimum.
Cascia (1988) recommended the following for staging the chillers on or off:
 Calculate the required refrigeration load.
 Determine the load ratio of each chiller and the chiller plant.
 Calculate the part-load power consumption for each chiller, and the combination of the chiller and
water pumps including chilled water and condenser water pumps. Select the combination with the
minimum power consumption.
Austin (1991) recommended in a multiple-chiller plant to start a lag chiller when the leading chiller
exceeds its optimum load point. The leading chiller is always the most energy-efficient chiller and
the last one to turn off. The last chiller to be turned on is the least energy-efficient one and the first
chiller to turn off.
For a multiple-chiller plant:
 A higher compressor efficiency always means a lower kW/ ton energy use.
 It is always energy-efficient to operate a single chiller with its chiller pump, instead of two
chillers of the same size and their chiller pumps, if a single chiller can meet the required refrigeration
load. The saving of the power input to the chilled water pump lowers the energy use.
 If all the multiple chillers are equal in size, have the same chilled water flow rate, and have the
same full-load cooling capacity, then when two or more chillers are turned on, each must be
loaded approximately equally. This is because the chilled water entering and leaving temperature
differential Tee 	 Tel is the same, and each turned-on chiller must have the same refrigeration load
ratio Rload, which can be calculated as:
(13.16)
where Tee,d, Tel,d  temperature of chilled water entering and leaving the evaporator at design conditions,
°F (°C). For a multiple-chiller plant, it is attractive to install chillers of unequal size to
meet required low loads when only one chiller is turned on. From Eq. (13.16), when two or more
chillers are loaded, the load ratio of unequal-size chillers is still approximately the same. It is
important to note that all unequal-size chilled water pumps connected in parallel must have the
same head.
Rload 
Tee 	 Tel
Tee, d 	 Tel, d
13.34 CHAPTER THIRTEEN
Chiller Minimum Performance
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency requirements for watercooled
electrically operated centrifugal chillers as follows:
Minimum efficiency requirements do not apply for chillers in low-temperature applications where
design leaving fluid temperature is 40°F (4.4°C).
Design Considerations
Power Consumption. The following are effective means to reduce the power consumption, in
kW/ton (COP), as well as to enhance the minimum performance of centrifugal chillers:
 Raise the temperature of the chilled water leaving the evaporator Tel by using a Tel reset at reduced
loads under the condition that space relative humidity does not exceed 65 percent.
 Lower the condenser water entering temperature Tce as low as the chiller manufacturer permits.
 If the fouling factor on the water side of the condensing surface exceeds 0.00025 h ft2 °F/Btu
(0.000044 m2 °C/W), reduce the fouling factor by using a brush cleaning system or some other
means, especially in location where the outdoor air is heavily contaminated.
 Increase the surface area of the evaporator and condenser if they are cost-effective.
Many manufacturers of centrifugal chillers now offer different sizes of evaporators and condensers
to match a specific size of compressor. The designer can thus select the centrifugal chiller that
meets the designated minimum performance.
Equipment Sizing. Although a single chiller is less costly than multiple chillers, there is a danger
of total loss of service when the chiller fails to operate or needs maintenance. Multiple chillers are
usually the best choice for a large chiller plant. A plant installed with three chillers ensures a 66
percent of capacity even if one of the chillers fails.
If continuity of service is critical, a standby chiller that has a capacity equal to the largest chiller
in the refrigeration plant should be provided. A smaller chiller used during light-load conditions, especially
in after-hours operation, in addition to several equally large chillers is often beneficial. The
refrigeration load always increases as building usage grows. The possibility of future growth should
be considered.
Plant Location and Layout. Generally, the central plant should be located near the system load as
well as near to the sources of utilities. For a high-rise building, the ideal location is often the hub of
the building, near shopping and public areas. If space at the hub is not available, the chiller plant
may be located in the basement, on an intermediate floor, or on the top floor, depending on the type
of building and the architectural design.
Equipment should be laid out in an orderly arrangement for efficient piping. Vertical and lateral
clearances must be maintained. Equipment may be 10 to 12 ft (3 to 3.6 m) in height. Provision of
access platforms and ladders to valves and piping should be considered in the design stage. A codespecified
clearance of 3 to 5 ft (0.9 to 1.5 m) from electrical panels and devices must be maintained.
Size Minimum efficiency Efficiency as of 10/29/2001
	150 tons 3.8 COP, 3.9 IPLV 5.0 COP, 5.0 IPLV
150 tons and 4.2 COP, 4.5 IPLV 5.55 COP, 5.55 IPLV
	300 tons
300 tons 5.2 COP, 5.3 IPLV 6.1 COP, 6.1 IPLV
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.35
Maintenance and service facilities should also be planned in the design stage. These include
removal and replacement of components, stored space, washrooms, and janitorial facilities.
REFERENCES
ARI, The New ARI Standard 550/590-98 for Water Chilling Packages Using the Vapor Compression Cycle,
HPAC, no. 2, 1999, pp. 63–66.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, 1999.
Austin, S. B., Optimum Chiller Loading, ASHRAE Journal, no. 7, 1991 pp. 40–43.
Bellenger, L. G., and Becker, J. D., Selecting High-Efficiency Centrifugal Chillers: A System Approach, HPAC,
no. 7, 1996 pp. 41–49.
Bjorklund, A. E., Heat Recovery and Thermal Storage at a State Office Building, ASHRAE Transactions, 1987,
Part II, pp. 832–849.
Braun, J. E., Mitchell, J. W., Klein, S. A., and Beckman,W. A., Models for Variable-Speed Centrifugal Chillers,
ASHRAE Transactions, 1987, Part I, pp. 1794–1813.
Calm, J. M., Alternative Refrigerants: Challenges, Opportunities, Heating/Piping/Air Conditioning, no. 5,
1992, pp. 38–49.
Cascia, M. A., Optimizing Chiller Plant Energy Savings Using Adaptive DDC Algorithms, ASHRAE Transactions,
1988, Part II, pp. 1937–1946.
Clark, E. M., Anderson, G. G.,Wells,W. D., and Bates, R. L., Retrofitting Existing Chillers with Alternative
Refrigerants, ASHRAE Journal, April 1991, pp. 38–41.
Corr, S., Dekleva, T. W., and Savage, A. L., Retrofitting Large Refrigeration Systems with R-134a, ASHRAE
Journal, no. 2, 1993, pp. 29–33.
Dietrich,W. E., A Positive Look for the Future, ASHRAE Journal, no. 5, 1993, pp. 64–65.
Dooley, E. W., Industry Status of CFC Chiller Retrofits and Replacements, HPAC, no. 1, 1997, pp. 131–135.
Erth, R. A., Power Invertors and Off-Design Chiller Performance, Heating/Piping/Air Conditioning, November
1980, pp. 63–67.
Erth, R. A., Chilled Water Reset: Variable Speed Solution, Heating/Piping/Air Conditioning, May 1987,
pp. 79–81.
Hubbard, Jr., R. S., Evaluating Chiller Plant Energy Performance, Engineered Systems, no. 9, 1997, pp. 80–88.
Hummel, K. E., Nelson, T. P., and Thompson, P. A., Survey of the Use and Emissions of Chlorofluorocarbons
from Large Chillers, ASHRAE Transactions, 1991, Part II, pp. 416–421.
Knebel, D. E., Evaporative Condenser Minimizes System Power Requirements, HPAC, no. 4, 1997, pp. 75–84.
Landman,W. J., The Search for Chiller Efficiency, Heating/Piping/Air Conditioning, July 1983, pp. 77–81.
Lau, A. S., Beckman,W. A., and Mitchell, J. W., Development of Computerized Control Strategies for a Large
Chilled Water Plant, ASHRAE Transactions, 1985, Part I B, pp. 766–780.
Leitner, G. F., Automatic Brush Cleaning for Condenser Tubes, Heating/Piping/Air Conditioning, October
1981, pp. 68–70.
Lewis, M. A., Microprocessor Control of Centrifugal Chillers—New Choices, ASHRAE Transactions, 1990,
Part II, pp. 800–805.
Lord, R, Davis, B., Allcott, J., Cottrell, G., and Hubbard, R. S., Forum on ARI Standard 550/590-98, HPAC,
no. 5, 1999, pp. 53–60.
Newton, E. W., and Beekman, D. M., Compressor Capacity Control for System Part-Load Operation, ASHRAE
Transactions, 1980, Part I, pp. 493–503.
Parsnow, J. R., The Long Term Alternative: R-134a in Positive Pressure Chiller, ASHRAE Journal, no. 5, 1993,
pp. 54–56.
Sauer, H. J., and Howell, R. H., Design Guidelines for Use of an Economizer with Heat Recovery, ASHRAE
Transactions, 1988, Part II, pp. 1877–1894.
13.36 CHAPTER THIRTEEN
Spethmann, D. H., Optimizing Control of Multiple Chillers, ASHRAE Transactions, 1985, Part II B,
pp. 848–856.
Statt, T. G., An Overview of Ozone-Safe Refrigerants for Centrifugal Chillers, ASHRAE Transactions, 1990,
Part I, pp. 1424–1428.
Tao,W., and Janis, R. R., Modern Cooling Plant Design, Heating/Piping/Air Conditioning, May 1985,
pp. 57–81.
The Trane Company, Air Conditioning Clinic: Centrifugal Water Chillers, The Trane Company, La Crosse, WI,
1983.
The Trane Company, Applications Engineering Manual: Model CVHE CenTraVacs with Microprocessor Based
Control, La Crosse, WI, 1987.
The Trane Company, CTV-DS-15 CenTraVac Liquid Chillers, La Crosse, WI, 1996.
The Trane Company, Engineering Bulletin: Automatic Condensing Cleaning System, La Crosse, WI, 1989.
Utesch, A. L., Direct Digital Control of a Large Centrifugal Chiller, ASHRAE Transactions, 1990, Part II,
pp. 797–799.
Walker, P., On-Line Automatic Tube Cleaning, Engineered Systems, no. 10, 1994, pp. 58–62.
Wang, S. K., Principles of Refrigeration Engineering, vol. 3, Hong Kong Polytechnic, Hong Kong, 1984.
Wanniarachchi, A. S., and Webb, R. L., Noncondensable Gases in Shell-Side Refrigerant Condensers, ASHRAE
Transactions, 1982, Part II, pp. 170–184.
Water Services of America, On-Load Automatic Brush Cleaning System,Water Services of America, Inc.,
Milwaukee, WI, 1979.
Wong, S. P. W., and Wang, S. K., System Simulation of the Performance of a Centrifugal Chiller Using a Shelland-
Tube Type Water-Cooled Condenser and R-11 as Refrigerant, ASHRAE Transactions, 1989, Part I,
pp. 445–454.
REFRIGERATION SYSTEMS: CENTRIFUGAL 13.37
CHAPTER 14
REFRIGERATION SYSTEMS:
ABSORPTION
14.1
14.1 ABSORPTION SYSTEMS 14.1
Type of Absorption System 14.1
Historical Development 14.2
Cost Analysis 14.2
Applications 14.3
14.2 PROPERTIES OF AQUEOUS
LITHIUM-BROMIDE SOLUTION 14.3
Mass Balance in Solution 14.3
Vapor Pressure 14.3
Equilibrium Chart 14.4
Enthalpy-Concentration Diagram 14.5
14.3 DOUBLE-EFFECT, DIRECT-FIRED
ABSORPTION CHILLERS 14.6
System Description 14.6
Air Purge Unit 14.8
Series Flow, Parallel Flow, and Reverse
Parallel Flow 14.8
Flow of Solution and Refrigerant 14.9
14.4 PERFORMANCE OF DOUBLE-EFFECT,
DIRECT-FIRED ABSORPTION
CHILLER 14.11
Mass Flow Rate of Refrigerant
and Solution 14.11
Thermal Analysis 14.12
Coefficient of Performance 14.14
14.5 ABSORPTION CHILLER
CONTROLS 14.16
Capacity Control and Part-Load
Operation 14.16
Crystallization and Controls 14.17
Cooling Water Temperature
Control 14.17
Safety and Interlocking Controls 14.18
Monitoring and Diagnostics 14.18
14.6 OPERATING CHARACTERISTICS
AND DESIGN CONSIDERATIONS 14.18
Difference between Absorption
and Centrifugal Chillers 14.18
Evaporating Temperature 14.19
Cooling Water Entering
Temperature 14.19
Heat Removed from Absorber
and Condenser 14.19
Condensing Temperature 14.19
Corrosion Control 14.20
Rated Condition of Absorption
Chiller 14.20
Minimum Performance 14.20
14.7 ABSORPTION CHILLERHEATERS
14.20
Heating Cycle 14.21
Actual Performance 14.22
14.8 ABSORPTION HEAT PUMPS 14.22
Functions of Absorption Heat
Pump 14.22
Case Study: Series-Connected Absorption
Heat Pump 14.22
14.9 ABSORPTION HEAT
TRANSFORMER 14.24
System Description 14.24
Operating Characteristics 14.24
Coefficient of Performance 14.26
REFERENCES 14.26
14.1 ABSORPTION SYSTEMS
Type of Absorption System
Absorption systems use heat energy to produce refrigeration or heating and sometimes to elevate
the temperature of the waste heat. Aqueous lithium bromide (LiBr) is often used to absorb the refrigerant,
the water vapor, and provides a higher coefficient of performance.
Current absorption systems can be divided into the following categories:
 Absorption chillers use heat energy to provide refrigeration.
 Absorption chiller-heaters use heat energy to provide cooling or heating.
 Absorption heat pump extracts heat energy from the evaporator through the absorber, adds to the
heat input in the generator, and releases them both to the hot water in the condenser for heating.
 Absorption heat transformers elevate the temperature of the waste heat to a value higher than any
other input fluid stream supplied to the absorption heat transformer.
Historical Development
In the 1950s and 1960s, both centrifugal chillers driven by electric motors and absorption chillers
using steam as heat input to provide summer cooling were widely used in central refrigeration
plants. Steam was widely used because excess steam was available in summer in many central
plants that used steam to provide winter heating, and because energy costs were of little concern.
After the energy crisis in 1973, the price of natural gas and oil used to fuel steam boilers drastically
increased. The earliest single-stage, indirect-fired steam absorption chillers had a coefficient
of performance (COP) of only 0.6 to 0.7. They required more energy and could not compete with
electric centrifugal chillers. Many absorption chillers were replaced by centrifugal chillers in the
late 1970s and 1980s.
Because of the high investment required to build new power plants, electric utility companies
added high-demand charges and raised cost-per-unit charges during peak usage periods. In recent
years, double-effect, direct-fired absorption chillers have been developed in both Japan and the
United States with a COP approximately equal to 1.
Cost Analysis
A cost analysis is often required. Aumann (1996) recommended that in addition to the chiller itself,
the energy and the initial costs of the auxiliaries, such as condenser pumps, cooling tower, and
tower fans, be included in the cost analysis because of the higher heat rejection in the absorption
chiller. The auxiliary energy costs can be 30 percent higher for absorption chillers than for electric.
Sun (1991) compared two designs for a project with a 1200-ton (4220-kW) design refrigeration
load. One design is three 400-ton (1406-kW) electric centrifugal chillers, and the other is two
electric centrifugal chillers plus one absorption chiller.
The cost of gas for this project is $0.46 per therm in summer, from April through November, and
$0.50 per therm in winter. One therm is equal to 100,000 Btu/h (29,300 W) of heat energy input.
The cost of electricity is $0.126 per kWh during on-peak periods and $0.0433 per kW/h for the offpeak
periods. The on-peak period is the 8-h weekday daytime determined by the electric utility. The
additional electricity demand charge is $3.60 per kW.
The centrifugal chiller uses 0.7 kW/ton (5.02 COP), and the additional energy use for the auxiliaries
for the centrifugal chiller is about 0.1 kW / ton higher than that for the absorption chiller. The
double-effect, direct-fired absorption chiller uses 0.12 therm/ ton. The comparison cost ( in dollars)
of these two alternatives for the third 400-ton (1406-kW) chiller operated 8 h/day, 5 days/wk,
8 months for summer and 4 months for winter is as follows:
Absorption chiller Centrifugal chiller
Initial cost
Chiller 200,000 80,000
Cooling tower 53,000 48,000
Demand 2126 11,491
Energy
Electricity
Summer 7200 36,700
Winter 3598 18,337
14.2 CHAPTER FOURTEEN
Gas
Summer 22, 969
Winter 12,474
Total energy cost before tax $48,417 $66,528
Total energy cost after tax $54,469 $74,844
The absorption chiller has an annual energy cost saving of $20,375 and a simple payback period of
about 6 years.
Applications
According to the Air Conditioning, Heating and Refrigeration News, April 14, 1997, the shipments
of new absorption chillers from the manufacturers in the United States in 1996 numbered 579,
whereas the shipments of centrifugal and screw chillers in 1996 numbered about 9200. Absorption
chillers had a share of about 5 percent of the new large chiller market in the United States in the
mid-1900s.
Absorption chillers have the advantage of using gas and are therefore not affected by the high
electric demand charge and high unit rate at on-peak hours. Most absorption chillers use water as
refrigerant, and its ozone depletion potential is zero. Absorption chillers are advantageous to combine
with electric chillers so that electric chillers will undertake the base load and the absorption
chillers handle the load at on-peak hours. Although absorption chillers have a higher initial cost
than centrifugal compressors, in many locations in the United States where the cost ratio of electricity
to natural gas is favorable, installation of absorption chiller for use during on-peak hours or even
normal operating hours is sometimes economically beneficial. Absorption chillers may have a simple
payback of several years.
14.2 PROPERTIES OF AQUEOUS LITHIUM-BROMIDE SOLUTION
Mass Balance in Solution
The composition of a solution is generally expressed by the mass fraction of its components. In a
solution containing lithium bromide (LiBr) and water, X is used to indicate the mass fraction of
lithium bromide, i.e.,
(14.1)
where ml  mass of lithium bromide in solution, lb (kg)
mw  mass of water in solution, lb (kg)
The mass fraction of water in solution is 1  X. If the mass fraction of LiBr in a solution is expressed
as a percentage, it is known as the concentration of LiBr.
Vapor Pressure
In general, the total pressure of a solution is equal to the sum of the vapor pressures of the solute
and of the solvent. However, in the case of a lithium bromide–water solution, the vapor pressure of
pure LiBr can be ignored because its value is much smaller than that of water.
X 
ml
ml  mw
REFRIGERATION SYSTEMS: ABSORPTION 14.3
When LiBr is dissolved in water, the boiling point of the solution at a given pressure is raised.
However, if the temperature of the solution remains constant, the dissolved LiBr reduces the vapor
pressure of the solution.
When a solution is saturated, equilibrium is established. The number of molecules across the
interface from liquid to vapor per unit time is equal to the number of molecules from vapor into
liquid. If the number of liquid molecules per unit volume is reduced due to the presence of a solute
then the number of vapor molecules per unit volume is also reduced. Consequently, the vapor pressure
of the solution is decreased.
Equilibrium Chart
The properties of an aqueous lithium bromide solution, including vapor pressure, temperature, and the
mass fraction at equilibrium, may be illustrated on an equilibrium chart based on the D?hring plot, as
shown in Fig. 14.1. The ordinate of the equilibrium chart is the saturated vapor pressure of water in
log-scale millimeters of mercury absolute (mm Hg abs) and the corresponding saturated temperature
(°F). The scale is plotted on an inclined line. The abscissa of the chart is the temperature of the solution
(°F). Mass fraction or concentration lines are inclined lines and are not parallel to each other.
At the bottom of the concentration lines, there is a crystallization line or saturation line. If the
temperature of a solution of constant mass fraction of LiBr drops below this line—or if the mass
fraction of LiBr of a solution of constant temperature is higher than the saturated condition—the
part of LiBr salt exceeding the saturated condition tends to form solid crystals.
14.4 CHAPTER FOURTEEN
FIGURE 14.1 Equilibrium chart for aqueous lithium-bromide (LiBr) solution. (Source: Carrier Corporation.
Reprinted with permission.)
Enthalpy-Concentration Diagram
When water is mixed with anhydrous lithium bromide at the same temperature to form a solution
adiabatically, there is a significant increase in the temperature of the solution. If the mixing process
is to be an isothermal process, i.e., if the temperature of the process is to be kept constant, then heat
must be removed from the solution. Such a heat transfer per unit mass of solution is called the integral
heat of the solution hi, or heat of absorption, in Btu / lb (kJ / kg). Based on the common rule
of thermodynamics, hi is negative.
REFRIGERATION SYSTEMS: ABSORPTION 14.5
FIGURE 14.2 Enthalpy-concentration diagram for aqueous LiBr solution. (Source: ASHRAE Handbook 1989,
Fundamentals, Reprinted with permission.)
If an aqueous LiBr solution is formed by an isothermal process, its specific enthalpy h, in Btu / lb
(kJ / kg), may be expressed as
(14.2)
where hw, hl  specific enthalpy of water and LiBr, respectively, Btu / lb (kJ / kg). Figure 14.2 shows
the enthalpy-concentration h-X diagram of an aqueous LiBr solution. The ordinate of the diagram
is the specific enthalpy of the solution h, in Btu / lb, (kJ / kg), and the abscissa is the percent concentration
of the aqueous LiBr solution. The curved lines represent the specific enthalpy of the solution
at various concentrations of LiBr mixed with water at a constant temperature. At the right-hand side
of the diagram, there is also a crystallization line. To the right of this line, excess solid LiBr may
crystallize.
14.3 DOUBLE-EFFECT, DIRECT-FIRED ABSORPTION CHILLERS
System Description
In an absorption chiller, water is the refrigerant, and aqueous lithium bromide (LiBr) is the absorbent.
Double-effect, direct-fired absorption chillers are most commonly used today because of
their higher coefficient of performance. Double-effect means that there are two generators. Direct-
fired means that the heat input is supplied to a directly fired generator. Figure 14.3 is a schematic
diagram of a double-effect, direct-fired absorption chiller.
A double-effect, direct-fired absorption chiller consists of 12 main components: an evaporator,
absorber, high-temperature generator, low-temperature generator, condenser, vapor-liquid separator,
high-temperature heat exchanger, low-temperature heat exchanger, high-temperature solution
pump, low-temperature solution pump, refrigerant pump, and refrigerant throttling device.
Evaporator and Refrigerant Pump. An evaporator is composed of a tube bundle, an outer shell,
spraying nozzles, and a water trough. The chilled water flows inside the tubes. A refrigerant pump
sprays the liquid refrigerant through the spray nozzles over the outer surface of the tube bundle to
increase the rate of evaporation. A water trough is located at the bottom of the evaporator to maintain
a certain liquid level for circulation. A high vacuum, typically at a saturated vapor pressure of
6.8 mm Hg abs. (906 Pa abs.), is maintained to provide an evaporating temperature of about 42°F
(5.6°C).
Absorber and Solution Pumps. Absorbers are water-cooled tube bundles with cooling water
flowing inside them. In the absorber, the vaporized refrigerant, the water vapor, is extracted and absorbed
by the concentrated LiBr solution sprayed from the nozzles so that at high vacuum, 5 mm
Hg abs. (666 Pa abs.) can be maintained in the absorber. The low-temperature solution pump raises
the pressure of dilute solution from 5 mm Hg abs. (666 Pa abs.) to about 45 to 60 mm Hg abs.
(6000 to 8000 Pa abs.) in the low-temperature generator; and the high-temperature solution pump
raises the pressure of the partly concentrated solution from about 50 to 350 to 450 mm Hg abs. (6.7
to 46.7 to 60 kPa abs.). This LiBr solution pressure is sufficient to overcome the pressure drop in
the heat exchangers, pipes, and pipe fittings.
Heat Exchangers. Heat exchangers are sometimes called economizers. Shell-and-tube or plateand-
frame heat exchangers are most commonly used for greater effectiveness. There are two heat
exchangers: a low-temperature heat exchanger and a high-temperature heat exchanger. In the lowtemperature
heat exchanger, heat is transferred from the hot solution mixture discharged from the
heat exchangers to the cold diluted solution pumped from the absorber. In the high-temperature
heat exchanger, heat is transferred from the hot concentrated solution discharged from the highh
 (1  X)hw  Xhl  hi
14.6 CHAPTER FOURTEEN
temperature generator to the partly concentrated solution from the low-temperature generator. Both
heat exchangers conserve heat energy input and increase the coefficient of performance of the
absorption chiller.
Generators. In the high-temperature generator, also called a direct-fired generator, heat is
supplied from the direct-fired burner to the diluted solution, and then the water is vaporized from
the solution. Vapor refrigerant flows to the low-temperature generator and condenses to liquid
there. The latent heat released during condensation is then used to heat the diluted solution. Water is
again vaporized at the low-temperature generator. Many types of fuel can be used in the direct-
fired burner, such as natural gas, oil, or propane. Natural gas is the most widely used. In the hightemperature
generator, there is a vapor-liquid separator in which water vapor is boiled off from the
partly concentrated solution.
REFRIGERATION SYSTEMS: ABSORPTION 14.7
Direct–fired
generator
Hot
water
Cooling water
outlet
Low-temperature
generator
Purge
Chilled
water
54
Condenser
Evaporator
Evaporator spray
pump
Eductor
Absorber
spray pump
High-temp.
generator
pump
High-temperature
heat exchanger
Low-temperature
heat exchanger Low-temperature
generator
pump
Cooling
water
inlet
Hot water heater
66% 390mm Hg abs
260
95
85
91
95
44
42
305
150
180
60.2%
300 195
120 63%
6
8
5
2 4
3
6.8 mm Hg abs
1
Hg abs absorber 58.5%
5 mm
46.2
mm
Hg abs
Refrigerant
storage
50 mm
Hg abs
FIGURE 14.3 A double-effect, direct-fired, reverse parallel flow absorption chiller. (Source: ASHRAE Handbook
1998, Refrigeration. Reprinted with permission.)
Instead of heat input from a direct-fired burner, absorption chillers can use other heat sources
such as industrial waste heat or exhaust heat from engines or gas turbines.
Condenser. A condenser is a water-cooled tube bundle in which cooling water flows. In the condenser,
water is often supplied at a temperature of 80 to 90°F (26.7 to 32.2°C). Water vapor boiled
off from the low-temperature generator is condensed to liquid in the condenser, and the liquid water
condensed inside the tubes of the low-temperature generator is cooled to a lower temperature in the
condenser.
Throttling Devices. Orifices, traps, and valves are often used as throttling devices to reduce the
pressure of refrigerant and solution. An orifice plate on the liquid refrigerant line acts as an expansion
valve between the condenser and evaporator.
A single-effect absorption chiller has only one generator, while a double-effect absorption
chiller uses part of the latent heat of condensation of the vapor refrigerant in the low-temperature
generator to enhance the COP. If steam is used as the input heat source in the indirect-fired absorption
chiller, the combined efficiency of the boiler, loss in steam transport, and the heat transfer at
the high-temperature generator is significantly lower than that of the direct-fired furnace.
Air Purge Unit
Because the pressures in the evaporator, absorber, and condenser are far below atmospheric pressure,
air and other noncondensable gases leak into an LiBr absorption refrigeration system. An air
purge unit should be installed to remove these noncondensable gases and to maintain the required
pressure, temperature, and concentration in the absorption chiller.
Figure 14.4 shows a typical purge unit in an LiBr absorption refrigeration system. Such a unit
usually consists of a purge chamber, a pickup tube, a vacuum pump, a solenoid valve, and a manual
shutoff valve. When noncondensable gases leak into the system during operation, they tend to migrate
to the absorber where the pressure is lowest. Through the pickup tube, noncondensable gases
and a certain amount of water vapor flow to the purge chamber from the absorber. Water vapor is
absorbed by the LiBr solution sprayed over the tube, and the heat of absorption is removed by the
cooling water inside a bank of tubes. The solution absorbed with water vapor is returned to the
absorber through a liquid trap. Noncondensable gases are evacuated from the chamber periodically
by a vacuum pump which raises the gas pressure and discharges the gases to the outdoor atmosphere.
In addition to the air purge unit, palladium cells are required to continuously remove the small
amount of hydrogen gas produced during corrosion because thin membranes of heated palladium
are only permeable by the hydrogen gas.
Series Flow, Parallel Flow, and Reverse Parallel Flow
In a double-effect, direct-fired absorption chiller, because there are two generators, the flow of the
solution from the absorber to generators can be in series flow, parallel flow, or reverse parallel flow.
Series Flow. The solution from the absorber flows in sequence through the low-temperature heat
exchanger, high-temperature heat exchanger, high-temperature generator, high-temperature heat
exchanger, low-temperature generator, and low-temperature heat exchanger and returns to the
absorber.
Parallel Flow. The solution from the absorber is pumped both to the high-temperature generator
through the high-temperature heat exchanger and to the low-temperature generator through the lowtemperature
heat exchanger separately, and then to the low-temperature generator. The concentrated
solution from both generators flows through corresponding heat exchangers. They are recombined
and returned to the absorber.
14.8 CHAPTER FOURTEEN
REFRIGERATION SYSTEMS: ABSORPTION 14.9
Reverse Parallel Flow. As shown in Fig. 14.3, the solution from the absorber is pumped through
a low-temperature heat exchanger and flows into a low-temperature generator. After that, the solution
splits into two streams. One flows in sequence through the high-temperature heat exchanger,
high-temperature generator, and high-temperature heat exchanger again, and then combines with
the other stream discharged from the low-temperature generator. The combined stream flows
through the low-temperature heat exchanger and returns to the absorber. A reverse parallel flow system
is one of the most energy-efficient operating cycles today. It also has a more compact design.
Flow of Solution and Refrigerant
For a double-effect, direct-fired, reverse parallel flow absorption chiller, the absorption cycle shows
the properties of the solution at various concentration, temperatures, and pressures, as seen in
Fig. 14.5. Sometimes the refrigerant, water vapor, is combined with the solution; and sometimes the
refrigerant and solution are separated.
Three kinds of LiBr solution are used during operation: concentrated solution, partly concentrated
solution, and diluted solution. Concentrated solution has a higher concentration of LiBr and
is created after all the refrigerant has been boiled off from the solution. Concentrated solution is
used to absorb the water vaporized in the evaporator. The solution that leaves the low-temperature
generator is partly concentrated. Diluted solution has a lower concentration of LiBr and is the combination
of aqueous LiBr solution and absorbed water vapor.
An absorption refrigeration system can also be divided into three regions according to the pressure
of the refrigerant and solution. The high-pressure region includes the high-temperature generator;
the intermediate-pressure region comprises the low-temperature generator and condensers; and
the low-pressure region includes the absorber and evaporator. Because the evaporator and absorber
as well as the low-temperature generator and the condenser are both connected by vapor passages
FIGURE 14.4 Purge unit in a lithium bromide absorption chiller. (Source:
ASHRAE Handbook 1998, Refrigeration. Reprinted with permission.)
of adequate size, the pressure difference between the evaporator and absorber as well as between
the low-temperature generator and condenser is small.
In a typical double-effect, direct-fired, reverse parallel flow absorption chiller operating at design
load, as shown in Fig. 14.3, water usually evaporates at about 42°F (5.6°C) and at a saturated
pressure of 6.8 mm Hg abs. (906 Pa abs.) in the evaporator, point 1. Chilled water enters the evaporator
inside the tube bundle at 54°F (12.2°C) and leaves at 44°F (6.7°C). Heat is transferred from
the chilled water to the vaporized refrigerant at an amount equal to the latent heat of vaporization.
Vaporized water in the evaporator is then extracted to the absorber because of its lower vapor
pressure and is absorbed by the concentrated LiBr solution. In the absorber, concentrated solution is
supplied and sprayed at about 120°F (48.9°C) with a concentration of 63 percent, point 5. It is
cooled by the cooling water flowing inside the tube bundle at an entering temperature of 85°F
(29.4°C). As the water vapor from the evaporator is absorbed, the solution is diluted to a concentration
of about 58.5% and its temperature drops to 95°F (35°C). The vapor pressure of the solution is
then about 5 mm Hg abs. (666 Pa abs.), which is lower than the evaporating pressure of 6.8 mm Hg
abs. (906 Pa abs.). The heat of absorption is removed by the cooling water. Diluted solution is then
pumped to the low-temperature heat exchanger and the low-temperature generator by a lowtemperature
solution pump.
In the low-temperature heat exchanger, diluted solution enters at 95°F (35°C ) and leaves at
about 150°F (65.6°C). After the diluted solution enters the low-temperature generator, because of
the release of the latent heat of condensation from the condensed water vapor that flows inside the
tube bundle of the low-temperature generator, the diluted solution is first heated to a temperature of
180°F (82.2°C), and the water vapor then boils off at a pressure of about 48.5 mm Hg abs. (6.5 kPa
abs.), point 3. As the water vapor is boiled off, the LiBr solution becomes partly concentrated with
a concentration of 60.2%.
At the outlet of the low-temperature generator, partly concentrated LiBr solution is divided
into two streams: One flows through the high-temperature heat exchanger and then enters the high-
14.10 CHAPTER FOURTEEN
200
70%
40%
50%
55%
150
100
50
50 100
5
2
7 6
3 4 8
1
6	
7	
3	 4	
2	
8	
5	 1	
150 200 250 300 350
5
Refrigerant temperature, F
Solution temperature, F
Vapor pressure, mm Hg abs
10
Partload
Full load
1000
500
400
300
200
100
60
40
20
Crystallization
line
65%
60%
FIGURE 14.5 Absorption cycle of a double-effect, direct-fired, reverse parallel flow absorption chiller.
temperature generator through a high-temperature solution pump. The other stream combines with
the concentrated solution from the high-temperature heat exchanger and enters the low-temperature
heat exchanger, point 4.
In the high-temperature heat exchanger, partly concentrated solution enters at 180°F (82.2°C)
and leaves at about 260°F (126.7°C). After that, partly concentrated solution flows into the hightemperature
generator. In the high-temperature generator, the solution is heated to 305°F (152°C),
and the water vapor boils off at a pressure of about 390 Hg mm abs. (52 kPa abs.), point 7. As the
water vapor is boiled off, the LiBr solution is concentrated to 66% in the high-temperature generator.
The concentrated solution returns to the high-temperature heat exchanger and cools to about
203°F (95°C ) as it leaves the heat exchanger.
The other stream of partly concentrated LiBr solution combines with the concentrated solution
from the high-temperature heat exchanger at 203°F (95°C), and the mixture enters the lowtemperature
heat exchanger at 192°F (89°C) and a concentration of 63.0%. After it flows through
the low-temperature heat exchanger, the temperature of the solution drops to about 120°F (49°C).
The solution then enters the absorber, sprays over the cooling water tube bundle, and absorbs the
water vapor from the evaporator.
As the water vapor that boils off from the LiBr solution in the high-temperature generator at
300°F (149°C) flows through the tubes submerged in the diluted solution in the low-temperature
generator, the water vapor condenses to liquid form. Latent heat of condensation released from the
liquid refrigerant heats the diluted solution from 150 to 180°F (65 to 82°C) and boils off the water
vapor from the diluted solution at a saturated pressure of about 50 mm Hg abs. (6.7 kPa abs.). Finally,
in the low-temperature generator, the partly concentrated solution reaches a concentration of
60% and splits into two streams when it leaves at the outlet.
The condensed liquid refrigerant that flows inside the tubes in the low-temperature generator
enters the trough of the condenser at a temperature of around 195°F (90°C). Water vapor boiled
off in the low-temperature generator is extracted by the condenser’s low saturated pressure of
46.2 mm Hg abs. (6.2 kPa abs.) through the top passage and condenses to liquid as the cooling
water removes the latent heat of condensation at a temperature of about 98°F (37°C). In the condenser,
condensed liquid refrigerant from the tubes is cooled to a temperature of 98°F (37°C) and
combined with the liquid refrigerant condensed in the condenser. Liquid refrigerant is forced
through an orifice, throttled to a pressure of about 6.8 mm Hg abs. (906 Pa abs.), and returned to
the evaporator.
14.4 PERFORMANCE OF DOUBLE-EFFECT, DIRECT-FIRED
ABSORPTION CHILLER
Mass Flow Rate of Refrigerant and Solution
For a throttling process, the enthalpy of water entering the evaporator hle is equal to the enthalpy of
saturated liquid leaving the condenser hlc, both in Btu / lb (kJ / kg). The mass flow rate of refrigerant
required for each 1 ton or 12,000 Btu /h (kW) of refrigeration Qref produced in the absorption
chiller , in lb /htonref (kg / s kWref), is calculated as
(14.3)
where hve  enthalpy of saturated water vapor leaving evaporator, Btu / lb (kJ / kg). According to
the mass balance of LiBr in the diluted and concentrated solution,
(14.4)
m?s1  m?s5  m?r  m?s5  m?rl  m?rh
X1m?s1  X5ms5  X3(m?s7  m?s9)
m?r 
Qref
hve  hlc

12,000
hve  hlc
m?r
REFRIGERATION SYSTEMS: ABSORPTION 14.11
When the absorption chiller operates at steady-state condition, the heat balance analyses for the
main system components, in Btu /h ton, are shown below:
In the evaporator, for 1 ton of refrigeration load (kW) Qrl, in Btu /htonref (kW/kWref),
In the absorber, the heat of entering water vapor is equal to the heat of leaving water vapor in the
evaporator:
where h1, h5  enthalpy of solution leaving and entering absorber, Btu / lb (kJ / kg)
Qab  heat to be removed from absorber, Btu /htonref (kW/kWref)
Also,
(14.5)
where X1, X5  mass fraction of LiBr in diluted and concentrated solution leaving and entering
absorber
 mass flow rate of diluted solution per ton of refrigeration, lb /htonref (kg / s kWref)
 mass flow rate of concentrated solution entering absorber per ton of refrigeration,
lb /htonref (kg / s kWref)
 mass flow rate of partly concentrated solution per ton of refrigeration at points 7
and 9, respectively, lb/htonref (kg / s kWref)
 mass flow rate of refrigerant boiled off from low-temperature generator and hightemperature
generator per ton of refrigeration, lb /htonref (kg / s kWref)
Ccir  circulating factor
Thermal Analysis
To help determine the status of the solution in the absorption cycle of a double-effect, direct-fired,
reverse parallel flow absorption chiller as shown in Figs. 14.3 and 14.5, the state points of the cycle
are defined as follows:
m? rl, m? rh
m?s7, m?s9
m?s5
m?s1
Ccir 
X1
X5  X1
m? s5 
X1m? r
X5  X1
14.12 CHAPTER FOURTEEN
5  Concentrated solution entering the
absorber
6  Partly concentrated solution leaving
high-temperature heat exchanger
7  Concentrated solution leaving hightemperature
generator
8  Concentrated solution leaving hightemperature
heat exchanger
9  Partly concentrated solution leaving
low-temperature generator
1  Mean condition of the diluted solution
leaving the absorber
2  Dilute solution leaving the low-temperature
heat exchanger
3  Partly concentrated solution leaving
the low-temperature generator
4  The condition of the mixture of partly
concentrated and concentrated
solutions entering the low-temperature
heat exchanger
Heat of leaving heat of entering refrigeration
water vapor  liquid  load
(14.6)  m?r(hve  hlc)  Qrl m? rhve  m? rhlc
Heat of entering heat of entering

heat of heat to be removed
water vapor  solution leaving solution  from absorber
(14.7)  Qab  m? s1h1 m? r hve  m?s5 h5
In the high-temperature generator,
where  mass flow rate of water vapor boiled off in high-temperature generator per ton of refrigeration,
lb /htonref (kg / s kWref)
hvhg  enthalpy of superheated water vapor leaving high-temperature generator, Btu / lb
(kJ / kg)
 mass flow rate of concentrated solution leaving high-temperature generator per ton of
refrigeration, lb /htonref (kg / s kWref)
 mass flow rate of partly concentrated solution entering high-temperature generator per
ton of refrigeration, lb /htonref (kg / s kWref)
h6,h7  enthalpy of solution entering and leaving high-temperature generator, Btu / lb (kJ / kg)
Qhg  heat input to high-temperature generator per ton of refrigeration, Btu/htonref
(kW/kWref )
In the low-temperature generator,
where hllg  enthalpy of liquid water leaving low-temperature generator, Btu / lb (kJ / kg)
 mass flow rate of partly concentrated solution leaving low-temperature generator to
combine with concentrated solution per ton of refrigeration, lb /htonref (kg / s kWref)
 mass flow rate of partly concentrated solution leaving low-temperature generator toward
high-temperature heat exchanger per ton of refrigeration, lb /htonref
(kg / s kWref)
 mass flow rate of diluted solution entering low-temperature generator per ton of
refrigeration, lb /htonref (kg / s kWref)
h2, h3  enthalpy of solution entering and leaving low-temperature generator, Btu / lb (kJ / kg)
hfg  latent heat of vaporization in low-temperature generator, Btu / lb (kJ / kg)
 mass flow rate of water vapor boiled off in low-temperature generator per ton of
refrigeration, lb /htonref (kg / s kWref)
And
(14.10)
In the condenser,
where hvlg  enthalpy of water vapor leaving low-temperature generator, Btu / lb (kJ /kg)
Qcon  heat to be removed from condenser, Btu /htonref (kW/kWref)
m?s3  m? s6  m? s9
m? r  m? rlg  m? rhg
m?rlg
m? s2
m? s9
m? s6
m?s6
m?s7
m?rh
REFRIGERATION SYSTEMS: ABSORPTION 14.13
Heat of heat of heat of heat to be
leaving  leaving  entering  supplied to highwater
vapor solution solution temperature generator
(14.8)  Qhg  m? s6h6 m? rhhvhg  m? s7h7
Heat of heat of heat of heat of heat to water vapor
entering  leaving  leaving  entering  boiled off in lowwater
vapor water solution solution temperature generator
  (14.9)  m? rlghfg m? s2h2 (m?s6  m? s9)h3 m? rhg(hvhg  hllg)
Heat of heat of heat of heat to be
entering  entering  leaving  removed from
liquid water water vapor liquid water condenser
(14.11)  Qcon  m? rhlc m? rhghlhg  m? rlghvlg
In the high-temperature heat exchanger, if the difference of specific heat of solution is ignored, the
effectiveness 	h is given as
(14.12)
Similarly, the effectiveness of the low-temperature heat exchanger is
(14.13)
where h4, h8  enthalpy of mixture of concentrated and partly concentrated solutions entering lowtemperature
heat exchanger and enthalpy of concentrated solution leaving hightemperature
heat exchanger, respectively, Btu / lb (kJ /kg)
Because the pump power is usually small compared to the heat supplied to the generator, it is often
ignored in thermal analyses.
If the heat loss from the system components is ignored, then heat input to the absorption system
must be equal to the heat removed from the system, in Btu/htonref (kW/kWref), that is,
Qhg  Qrl  Qab  Qcon (14.14)
Coefficient of Performance
If the pump power of the solution and refrigerant pumps is excluded, the coefficient of performance
of a direct-fired absorption chiller COPc can be calculated as
(14.15)
where Qhg  heat input to high-temperature generator per ton of refrigeration, Btu/htonref
(kW/kWref)
com  combustion efficiency of direct-fired generator
There are two kinds of heat input of fuel: the higher heating value (HHV) and lower heating
value (LHV). The HHV is obtained when all the combustion products are cooled to the temperature
existing before combustion and the water vapor formed during combustion is condensed.
The LHV subtracts the latent heat of vaporization from the HHV. The LHV is about 90 percent
of the HHV. For any given COPc value, it must be checked whether the heat input is based on
HHV or LHV.
For double-effect, direct-fired, reverse parallel flow absorption chillers, the COP (HHV also including
the combustion efficiency) of current products varies between 0.92 and 1.03.
Example 14.1. A typical double-effect, direct-fired, reverse parallel flow absorption chiller, described
in Sec. 14.3 is operated under the following conditions.
The enthalpy of the solution at various points is as follows:
COPc 
refrigeration output
HHV of fuel consumption

12,000 com
Qhg

h4  h5
h4  h1
	l 
m? s5(h4  h5)
m? s5(h4  h1)

h7  h8
h7  h3
	h 
m? s7(h7  h8)
m? s7(h7  h3)
14.14 CHAPTER FOURTEEN
The mass flow rates of partly concentrated solution to the high-temperature heat exchanger and to
combine with the concentrated solution from the high-temperature heat exchanger are equal.
Finally, the amounts of water vapor boiled off from the high- and low-temperature generators
are equal.
Calculate the effectiveness of the heat exchangers and the coefficient of performance of this absorption
chiller.
Solution. From a psychrometric table, at a condensing temperature of 98°F, the enthalpy of
saturated liquid water leaving the condenser hlc  66.03 Btu/ lb, and the enthalpy of saturated water
vapor leaving the evaporator hve at an evaporating temperature of 42°F is 1079.5 Btu/ lb. From
Eq. (14.3), the mass flow rate of refrigerant per ton of refrigeration is therefore
As described in the previous section, the concentration of dilute solution from absorber is
58.5%, and the mean concentration of solution entering the absorber is 63%. From Eq. (14.5), the
mass flow rate of concentrated solution for each ton of refrigeration produced is therefore
From Eq. (14.4), the mass flow rate of diluted solution per ton of refrigeration is
From Eq. (14.13), the effectiveness of the low-temperature heat exchanger can be calculated as
From Eq. (14.12), the effectiveness of the high-temperature heat exchanger can be calculated as
The mass flow rate of water vapor boiled off from the high-temperature generator is
m? rh  0.5m? r  0.5(11.84)  5.92 lb / h  tonref
	h 
h7  h8
h7  h3

150  107.5
150  85
 0.65
	l 
h4  h5
h4  h1

97  55.5
97  42
 0.75
m? s1  m?s5  m? r  153.92  11.84  165.76 lb / h  tonref
 153.92 lb / h  tonref
m? s5 
X1
X5  X1
m? r 
0.585
0.63  0585
(11.84)
 11.84 lb / h  tonref
m? r 
12,000
hve  hlc

12,000
1079.5  66.03
Point Temperature,°F Concentration, % Enthalpy, Btu/ lb
1 95 58.5 42
2 150 58.5 67.5
3 180 60.2 85
4 192 63.0 97
5 120 63.0 55.5
6 260 60.2 122.8
7 305 66.0 150
8 203 66.0 107.5
REFRIGERATION SYSTEMS: ABSORPTION 14.15
From the psychrometric table, the enthalpy of the saturated water vapor at 300°F is hvhg 
1180.04 Btu/lb. Also the mass flow rate of partly concentrated solution at point 6 can be calculated as
And the mass flow rate of concentrated solution at point 7 can be calculated as
From Eq. (14.8), the heat input to the high-temperature generator is then calculated as
If the combustion efficiency of the direct-fired generator is 85 percent, from Eq. (14.15), the coeffi-
cient of performance of this double-effect, direct-fired, reverse parallel flow absorption chiller
(HHV) can be calculated as
If the heat losses from various system components of the absorption chiller are taken into consideration,
then the calculated COPc will be nearer to the value given by the manufacturer’s catalogs,
about 0.92 to 1.0.
14.5 ABSORPTION CHILLER CONTROLS
Capacity Control and Part-Load Operation
A double-effect, direct-fired, reverse parallel flow absorption chiller adjusts its capacity during partload
operation by modulation of the heat input to the high-temperature generator through the direct-
fired burner, or other waste heat supply. A sensor is often located at the outlet of the evaporator to
monitor the temperature of the chilled water leaving the evaporator Tel. As the system refrigeration
load falls, Tel decreases accordingly. Once a drop of Tel below a predetermined set point is sensed,
the DDC unit controller reduces the gas flow or other fuel supply to the direct-fired burner, and less
water vapor is boiled off from the solution in the generator. Between 30 and 100 percent of the
system design refrigeration load, the burner fired rate is modulated. Below 30 percent of design
load, the burner is cycled on and off, and all refrigerant and solution pumps remain on so that the
system refrigeration load is allowed to drop to 10 percent minimum of the design load. The refrigerant
pump is cycled at minimum cooling.
Consequently, the concentration of the solution entering the absorber drops, less water vapor is
extracted to the absorber, and, therefore, the rate of evaporation and refrigeration effect in the evaporator
are both reduced until they are balanced with the reduction of refrigeration load so that Tel is
maintained within acceptable limits. Meanwhile, because less water vapor has been extracted to the
absorber from the evaporator, both evaporating pressure pev and evaporating temperature Tev increase.
Although less water vapor is boiled off in the generators, the rate of heat transfer at the condensing
surface and the amount of water vapor to be condensed to liquid water in the condenser remain
the same. The condensing pressure pcon and temperature Tcon decrease until a new balance is
formed.
If Tel rises above a limit, on the other hand, more heat is provided to the generator, and the concentration
of solution and the refrigeration capacity increase and Tel again falls within preset limits.
Of course, the increase in solution concentration should not exceed the saturation limit.
COPc 
12,000 com
Qhg

12,000(0.85)
8271.4
 1.23
 6985.6  11,100  9814.2  8271.4 Btu / h  tonref
Qhg  m?rhhvhg  m? s7h7  m? s6h6  5.92(1180)  74(150)  79.92(122.8)
m?s7  m?s6  0.5m?r  79.92  0.5(11.84)  74.0 lb / h  tonref
m?s6  0.5(m?s1  0.5m?r)  0.5[165.76  0.5(11.84)]  79.92 lb / h  tonref
14.16 CHAPTER FOURTEEN
REFRIGERATION SYSTEMS: ABSORPTION 14.17
The left-hand side of Fig. 14.5 shows the part-load absorption cycle of the double-effect, direct-
fired, reverse parallel flow absorption chiller. During part-load operation, the following conditions
exist:
 The mass flow rate of refrigerant is directly proportional to the load ratio.
 The evaporating pressure pev and temperature Tev rise.
 The condensing pressure pcon and temperature Tcon will drop.
 The boiled-off temperatures in the high- and low-temperature generators will decrease.
 Heat input to the high-temperature generator is reduced.
 The drop in temperature of cooling water at a lower outdoor wet-bulb temperature lowers the condensing
temperature Tcon and, therefore, the heat input.
During part-load operation, the evaporating and condensing temperatures; the concentrations of diluted,
partly concentrated, and concentrated solutions; and the other operating parameters can be
determined from the equations based on the heat balance analyses described in Sec.14.4.
Crystallization and Controls
In an absorption chiller using aqueous LiBr as the absorbent, the LiBr in a solution of constant concentration
starts to crystallize when the solution temperature drops below the crystallization line on
the equilibrium chart. The crystals formed are pure LiBr salt. Crystallization does not harm the absorption
chiller, but it eventually decreases the concentration of the remaining solution. It is also a
symptom of malfunction, and the cause of crystallization must be determined before normal operation
can resume.
Absorption chillers are now designed to operate in the region away from the crystallization line
shown in Fig. 14.5. There are also devices available that prevent crystallization and dissolve crystals
if crystallization occurs. Crystallization is no longer the serious problem for currently designed
absorption chillers that it was before the energy crisis.
Causes of crystallization include the following:
 Air leaks into the system raise the evaporating temperature and the chilled water leaving temperature
Tel. A higher Tel increases the heat input and the solution concentration to the point of crystallization.
 When the system is operated at full load, if the temperature of the cooling water is too low, the
diluted solution temperature may fall low enough to reduce the temperature of the concentrated
solution to the point of crystallization.
 If the electric power is interrupted, the system ceases to operate. The temperature of the concentrated
solution in the heat exchanger starts to drop and may fall below the crystallization line.
Manufacturers have developed several devices to minimize the possibility of crystallization.
One such device uses a bypass valve to permit refrigerant to flow to the concentrated solution line
when conditions that can cause crystallization are detected by the sensor. A recently developed
microprocessor-based DDC unit controller uses measured temperatures and pressures at key points
to calculate the concentration of the solution, to prevent crystallization.
Cooling Water Temperature Control
Earlier absorption chiller designs required a close control of the cooling water temperature to prevent
crystallization. Current absorption chiller designs include control devices to prevent
crystallization, so many manufacturers now allow the temperature of the cooling water to fall to
lower value at part-load operation. A three-way mixing bypass valve should be installed to mix
bypass recirculating water from the condenser and to maintain the temperature of cooling water entering
the absorber at a temperature specified by the manufacturer. Cooling tower fans can be cycled
to supplement the bypass control.
Safety and Interlocking Controls
Low-Temperature Cutout. If the temperature of the refrigerant in the evaporator falls below a
preset value, the DDC unit controller shuts down the absorption chiller to protect the evaporator
from freezing. As soon as the refrigerant temperature rises above the limit, DDC unit controller
starts the chiller again.
Chilled Water Flow Switch. When the mass flow rate of chilled water falls below a limit, a
pressure-sensitive or flow-sensitive sensor alerts the DDC unit controller, which stops the absorption
chiller.
Cooling Water Flow Switch. When a loss in cooling water supply is detected by the pressure- or
flow-sensitive sensor, the DDC unit controller shuts down the absorption chiller. The chiller is
started again only when the cooling water supply is reestablished.
High-Pressure Relief Valve. A high-pressure relief valve or similar device is often installed on
the shell of the high-temperature generator to prevent the maximum pressure in the system from
exceeding a preset value.
Direct-Fired Safety Controls. Similar to that described in Secs. 8.2 and 8.3, a direct-fired absorption
chiller needs controls of high-pressure and low-pressure switches, flame ignition, and monitoring
for its burner and generator.
Interlocked Controls. Absorption chiller should be interlocked with chilled water pumps,
cooling-water pumps, and cooling tower fans so that the absorption chiller starts only when these
pieces of equipment are in normal operation.
Optimizing On /Off Control. If multiple absorption chillers are installed in a chiller plant, as in a
plant with multiple centrifugal chillers, optimizing control can be used to turn chillers on and off.
Monitoring and Diagnostics
Monitoring lights show the operating status of key parameters. Also fault detection and diagnostic
displays determine the operating problem and necessary corrections.
14.6 OPERATING CHARACTERISTICS AND DESIGN
CONSIDERATIONS
Difference between Absorption and Centrifugal Chillers
In a vapor compression refrigeration system, compressor power input is primarily affected by head
( pcon  pev) or temperature lift (Tcon  Tev) of the refrigeration cycle. In an absorption chiller,
Tcon  Tev is also a factor that influences the heat input to the high-temperature generator, but its
influence is significantly smaller than that of centrifugal chillers.
For the double-effect, direct-fired, reverse parallel flow absorption chiller described in Example
14.1, the heat input to high-temperature generator Qhg is distributed as follows:
14.18 CHAPTER FOURTEEN
Latent heat of vaporization is the primary factor that affects the heat input in an absorption chiller,
and temperature difference Tcon  Tev has only a minor effect on hfg. This is the primary difference
between an absorption chiller and a centrifugal chiller.
Evaporating Temperature
As in the centrifugal chiller, the evaporating temperature Tev and pressure pev in an absorption
chiller depend mainly on the chilled water temperature leaving the evaporator Tel. The difference
Tel  Tev in current absorption chiller design is between 2 and 5°F (1.1 and 2.8°C). A smaller
Tel  Tev means a higher COP and a large heat-transfer surface area in the evaporator. It is actually
a compromise between energy cost and investment. For Tel  44°F (6.7°C) and Tel  Tev  2°F
(1.1°C), the evaporating pressure is around 6.8 mm Hg abs. (0.268 in. Hg abs. or 906 Pa abs.).
The vapor pressure of the concentrated solution during absorption should be slightly less than
the evaporating pressure pev, so that water vapor vaporized in the evaporator can be extracted to the
absorber. Both Tev and Tel affect the cooling capacity of the absorption chiller as well as the heat
input to the high-temperature generator.
Cooling Water Entering Temperature
The temperature of cooling water entering the absorber Tca, in°F (°C), has the following effects on
the performance of an absorption chiller:
 Lower Tca means a higher cooling capacity.
 Lower Tca results in a lower Tcon.
 Lower Tca means a lower heat input to the high-temperature generator per ton of refrigeration produced
and, therefore, a high COPc.
 At high load ratios, too low of a Tca may cause crystallization.
Manufacturer recommendations should be followed. For example, one manufacturer recommends
that a minimum of Tca  72°F (22.2°C) be maintained at all load conditions.
Heat Removed from Absorber and Condenser
The total amount of heat to be removed from the absorber and condenser in a typical double-effect,
direct-fired, reverse parallel flow absorption chiller is about 1.5 times the heat rejected from the
condenser in a centrifugal chiller. The heat to be removed in the absorber is about 60 percent of the
total heat removal. Heat removed from the condenser is about 40 percent of total heat removal.
Usually, a cooling water temperature rise of 10 to 15°F (5.6 to 8.3°C) is used. For a cooling water
temperature rise of 10°F (5.6°C) when the temperature entering the absorber is 85°F (29.4°C),
the temperature of cooling water is about 91°F (32.8°C) entering the condenser and 95°F (35°C)
leaving the condenser.
Condensing Temperature
Condensing temperature Tcon depends mainly on the temperature of cooling water available and the
heat-transfer surface area. If cooling water is at 85°F (29.4°C), allowing for a temperature increase
Heat input Qhg 8271.4 Btu /htonref 100.0 percent
Latent heat of vaporization hfg 6985.6 Btu /htonref 84.5 percent
Heat of diluted solution 1285.8 Btu /htonref 15.5 percent
REFRIGERATION SYSTEMS: ABSORPTION 14.19
of 6°F (3.3°C) in the absorber and 4°F (2.2°C) in the condenser, and the difference between the
condensing temperature and cooling water temperature leaving the condenser Tcon  Tcl  3°F
(1.7°C), then condensing temperature Tcon can be calculated as
Condensing pressure pcon should be slightly lower than the boiled-off vapor pressure in the lowtemperature
generator in order to extract the vapor and be condensed in the condenser.
Corrosion Control
Many heat exchangers are typically of the shell-and-tube type. Shells are often made of mild steel
or stainless steel, and tube bundles are usually made of copper or copper-nickel alloy. Lithium bromide
is very corrosive. It attacks steel, copper, and copper alloys in the presence of air at temperatures
above 300°F (149°C). Corrosion inhibitor should be used to protect the internal components
against corrosive attacks, as specified by the manufacturer.
Rated Condition of Absorption Chiller
When one selects a double-effect, direct-fired, reverse parallel flow, water-cooled absorption chiller,
the chiller’s cooling capacity must meet the required system refrigeration load. According to ARI
Standard 560, absorption chillers are rated under the following conditions:
Tcon  85  6  4  3  98F (36.7C)
14.20 CHAPTER FOURTEEN
44°F (6.7°C)
10°F (5.6°C)
2.4 gpm / ton (0.043 L / s kW)
85°F (29.4°C)
4.4 gpm / ton (0.079 L / s kW)
0.00025 h ft2 °F/Btu (0.000044 m2 C/W)
Leaving chilled water temperature
Chilled water temperature rise
Chilled water flow rate
Entering cooling water temperature
Cooling water flow rate
Fouling factor
At design load, under operating conditions other than the standard rating conditions, both the cooling
capacity and the heat input should be modified.
Minimum Performance
ASHRAE/IESNA Standard 90.1-1999 mandates the minimum efficiency requirements for absorption
chillers at all capacities at rated conditions as follows:
14.7 ABSORPTION CHILLER-HEATERS
An absorption chiller-heater has nearly the same equipment and piping arrangement as those in an
absorption chiller except for some additional valves and switches. The heating cycle operation of an
Type Minimum efficiency Efficiency as of 10 /29/2001
Air-cooled, single effect 0.48 COP 0.60 COP
Water-cooled, single effect 0.60 COP 0.70 COP
Double-effect, indirect-fired 0.95 COP, 1.00 IPLV 1.00 COP, 1.05 IPLV
Double-effect, direct-fired 0.95 COP, 1.00 IPLV 1.00 COP, 1.00 IPLV
absorption chiller is shown in Fig. 14.6. The evaporator is now acting as a condenser to provide hot
water heating to an air conditioning space with a system efficiency comparable to that of a hot water
or warm air heating system.
Heating Cycle
During the heating cycle operation in a chiller-heater, the diluted LiBr solution at a temperature
of 195°F (90.6°C) and a concentration of 55% is extracted by the low-temperature solution pump
and forced to the low-temperature generator through the low-temperature heat exchanger. In
the low-temperature generator, the concentration of the solution remains the same, whereas the
temperature of the diluted solution increases to 205°F (96°C). The solution is extracted again by a
high-temperature solution pump and forced to the high-temperature generator through a high-temperature
heat exchanger. In the high-temperature generator, the solution is heated by a direct-fired
REFRIGERATION SYSTEMS: ABSORPTION 14.21
Direct–fired
generator
Hot
water
Low-temperature
generator
Purge
130
Condenser
Evaporator
Evaporator spray
pump off
High-temp.
generator
pump
High-temperature
heat exchanger
Low-temperature
heat exchanger
Absorber
spray pump
Low-temperature
generator
pump
Eductor
Hot water heater
230
260 60.5%
140
205
205
55%
165
255
5
6
7
3
195 Absorber 55% Refrigerant
storage
FIGURE 14.6 An absorption chiller-heater operated at heating cycle. (Source: ASHRAE Handbook 1998, Refrigeration.
Reprinted with permission.)
burner. The water vapor boils off from the solution at a temperature of 165°F (74°C), flows inside
the tube through the low-temperature generator, and enters the evaporator. The water vapor is then
condensed to water because of the removal of the latent heat of condensation by the hot water tube
bundle in which hot water is heated from 130 to 140°F (54 to 60°C). Hot water is supplied to the
conditioned space for heating at a temperature of 140°F (60°C) by a circulating pump.
The concentrated solution at 255°F (124°C) and a concentration of 60.5% flows back from the
high-temperature generator through high-temperature and low-temperature heat exchangers and enters
the absorber at 205°F (96°C). In the absorber, the solution absorbs the condensed water from
the evaporator and becomes a diluted solution of 55% concentration at 195°F (90°C). The diluted
solution is extracted by the low-temperature solution pump and repeats the heating cycle again.
Actual Performance
According to the data listed in one leading manufacturer’s catalog in the United States, the thermal
efficiency Et, in percent, of a chiller-heater during heating cycle operation is about 85 percent, including
the combustion efficiency of the direct-fired burner.
14.8 ABSORPTION HEAT PUMPS
Functions of Absorption Heat Pump
An absorption heat pump extracts heat from a low-temperature heat source, such as waste heat or
surface water, and delivers its heat output at a higher temperature for winter heating or other applications
at a coefficient of performance greater than 1. In Japan and Sweden, absorption heat pumps
have been installed in industrial and district heating plants using industrial waste heat to provide hot
water, typically at 165°F (74°C), for winter heating or other purposes at a COP between 1.2 and
1.7. Absorption heat pumps can be used either for winter heating or for cooling in summer and
heating in winter.
The coefficient of performance for cooling COPc for an absorption heat pump can be calculated
as described in Eq. (14.15). The coefficient of performance for heating for a double-effect absorption
heat pump COPh can be calculated as
(14.16)
where Qab  heat removed from absorber, Btu /h (kW)
Qcon  heat removed from condenser, Btu /h (kW)
Qhg  heat input to high-temperature generator, Btu /h (kW)
Several absorbents, or working fluids, other than aqueous LiBr solution are being developed; however,
LiBr/H2O is still the most widely used solution in absorption heat pumps.
Case Study: Series-Connected Absorption Heat Pump
Figure 14.7 is a schematic diagram of a 50-MW series-connected absorption heat pump installed in
an incineration plant in Uppsala City, Sweden, as described by Astrand (1988). This seriesconnected
absorption heat pump consists of two single-stage absorption heat pumps, each with an
evaporator, absorber, generator, condenser, heat exchanger, and solution pump. Liquid water, as the
refrigerant, evaporates in the evaporator. Water vapor is extracted by the concentrated solution
in the absorber. The heat of absorption transferred to the hot water in the absorber is then used for
COPh 
Qab  Qcon
Qhg
14.22 CHAPTER FOURTEEN
district heating. The dilute solution is pumped from the absorber to the generator through the heat
exchanger. In the generator, steam from the incineration plant boils off the water vapor from the
dilute solution. The boiled-off water vapor is extracted to the condenser and condensed to liquid
form. Latent heat of condensation is again rejected to the district heating hot water.
Concentrated solution from the generator flows to the absorber through the heat exchanger. Condensed
liquid water enters the evaporator via a throttling orifice and is sprayed over the tube bundle
in which flue gas cooling water flows from the incineration plant. After absorbing the latent heat
of vaporization from the flue gas cooling water, liquid water evaporates into water vapor in the
evaporator.
In this series-connected absorption heat pump, absorption heat pump no. 1 is operated at higher
temperatures, and heat pump no. 2 is operated at lower temperatures. During operation in 1987, the
return hot water from district heating was heated from 144 to 152°F (62.2 to 66.7°C) in the
REFRIGERATION SYSTEMS: ABSORPTION 14.23
FIGURE 14.7 Schematic diagram of series-connected absorption heat pump.
absorber and condenser of the no. 2 absorption heat pump, and from 152 to 160.5°F (66.7 to
71.4°C) in the absorber and condenser of the no. 1 absorption heat pump. In the evaporator, heat
was extracted from the low-temperature heat source, the flue gas cooling water, which entered the
absorption heat pump at a temperature of 97.7°F (36.5°C) and left at a temperature of 75.2°F
(24°C). The high-temperature heat source, steam, was supplied at 320°F (160°C) at a flow rate of
66,000 lb /h (29,940 kg / h) from the incineration plant. The average COPhp for this series-connected
absorption heat pump from July 15 through August 31, 1987, was 1.61.
14.9 ABSORPTION HEAT TRANSFORMER
Large amounts of low-temperature waste heat are released daily from many industrial plants.
Absorption heat transformers are systems that operate in a cycle opposite to that of absorption
chillers, to boost the temperature of input waste heat 40 to 100°F (22 to 56°C) higher for industrial
applications.
Most absorption heat transformers today are single-stage systems using lithium bromide as the
working fluid for industries in Japan and Europe. Two-stage absorption heat transformers for
greater temperature boosting and use with working fluids other than LiBr are being developed.
System Description
An absorption heat transformer consists of an evaporator, absorber, generator, condenser, heat
exchanger, solution pump, refrigerant pump, piping, and accessories. Figure 14.8 shows the
schematics and the absorption cycle on the equilibrium chart of a typical single-stage absorption
heat transformer using LiBr as the absorbent. In a heat transformer, because the heat is supplied
from the waste heat fluid stream entering at low temperature T1 and leaving at lower temperature T2,
water is vaporized from the evaporator and extracted by the absorbent sprayed in the absorber at
point 11. Heat of absorption is then released from the absorbent to the hot water stream entering at
a temperature T3 and raised to a temperature T4. After absorbing the water vapor, the diluted
solution at point 12 enters the heat exchanger, which is also called the recuperator. Diluted solution
leaves the heat exchanger at a lower temperature, flows through a regulating valve, and enters the
generator, or desorber, at point 13. Because of the heat input from the waste heat fluid streams, water
vapor boils off from the diluted solution and is condensed in the condenser. The concentrated
solution leaves the generator at point 14 and is pumped to a higher pressure equal to the pressure in
the absorber. Liquid water in the condenser is pumped to the evaporator for evaporation again.
Operating Characteristics
The absorption heat transformer operates at two pressure levels: high pressure, including the evaporator
and absorber; and low pressure, including the generator and condenser. There are three temperature
levels of input and output fluid streams:
 The fluid stream carrying the heat output from the absorber is at the highest temperature level.
 The heat source (the waste heat input to the evaporator and generator) is at the intermediate temperature
level.
 The cooling water in the condenser is at the lowest temperature level.
The purpose of an absorption heat transformer is to boost the temperature of the input waste heat
fluid stream, and the function of an absorption heat pump is to attain a higher COP from the lowertemperature
heat source.
14.24 CHAPTER FOURTEEN
14.25
FIGURE 14.8 Single-stage absorption heat transformer: (a) schematic diagram; (b) absorption cycle.
Coefficient of Performance
The coefficient of performance of an absorption heat transformer COPh,t can be calculated as
(14.17)
where Qev, Qg  waste heat input to evaporator and generator, Btu /h (kW). Mainly COPh, t is affected
by the amount of temperature boosting. The higher the temperature boosting, the lower the
COPh, t. Usually COPh, t varies between 0.4 and 0.5.
REFERENCES
Alefeld, G., and Ziegler, F., Advanced Heat Pump and Air-Conditioning Cycles for the Working Pair H2O/ LiBr:
Domestic and Commercial Applications, ASHRAE Transactions, 1985, Part II B, pp. 2062–2071.
ASHRAE, ASHRAE Handbook 1998, Refrigeration, ASHRAE Inc., Atlanta, GA, 1998.
ASHRAE, ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except New Low-Rise
Residential Buildings, Atlanta, 1999.
Astrand, L. E., Operating Experience with a 50 MW Absorption Heat Pump, ASHRAE Transactions, 1988,
Part I, pp. 716–722.
Aumann, D. J., Selecting Chillers in the 90s: Accounting for Hidden Costs, HPAC, no. 3, 1996, pp. 65–66.
Chuang, C. C., and Ishida, M., Comparison of Three Types of Absorption Heat Pumps Based on Energy Utilization
Diagrams, ASHRAE Transactions, 1990, Part II, pp. 275–281.
Davidson, K., and Brattin, H. D., Gas Cooling for Large Commercial Buildings, ASHRAE Transactions, 1986,
Part I B, pp. 910–920.
Davidson,W. F., and Erickson, D. C., Absorption Heat Pumping for District Heating Now Practical, ASHRAE
Transactions, 1988, Part I, pp. 707–715.
Fallek, M., Parallel Flow Chiller-Heater, ASHRAE Transactions, 1985, Part II B, pp. 2095–2102.
Gommed, K., and Grossman, G., Performance Analysis of Staged Absorption Heat Pumps:Water–Lithium
Bromide Systems, ASHRAE Transactions, 1990, Part I, pp. 1590–1598.
Grossman, G., Adiabatic Absorption and Desorption for Improvement of Temperature-Boosting Absorption
Heat Pumps, ASHRAE Transactions, 1982, Part II, pp. 359–367.
Grossman, G., Multistage Absorption Heat Transformers for Industrial Applications, ASHRAE Transactions,
1985, Part II B, pp. 2047–2061.
Holmberg, P., and Berntsson, T., Alternative Working Fluids in Heat Transformers, ASHRAE Transactions,
1990, Part I, pp. 1582–1589.
Hufford, P. E., Absorption Chillers Maximize Cogeneration Value, ASHRAE Transactions, 1991, Part I,
pp. 428–432.
Kluber, J. J., A Lesson in Savings, Engineered Systems, no. 3, 1999, pp. 78–84.
Kumar, P., and Devotta, S., Study of an Absorption Refrigeration System for Simultaneous Cooling and
Heating, ASHRAE Transactions, 1990, Part II, pp. 291–298.
Kurosawa, S., and Fujimaki, S., Development of Air-Cooled Double-Effect Gas-Fired Absorption Water Chiller-
Heater, ASHRAE Transactions, 1989, Part I, pp. 318–325.
Kurosawa, S., Nakamura, S., Development of Intelligent Gas-Fired Absorption Water Chiller-Heaters, ASHRAE
Transactions, 1987, Part II, pp. 850–865.
McLinden, M. O., and Klein, S. A., Steady-State Modeling of Absorption Heat Pumps with a Comparison to
Experiments, ASHRAE Transactions, 1985, Part II B, pp. 1793–1807.
Summerer, F., and Ziegler, F. F., An approach for Assessing the Economics of Sorption Chillers, ASHRAE
Transactions, 1998, Part I B, pp. 1413–1419.
COPh, t 
Qab
Qev  Qg
14.26 CHAPTER FOURTEEN
Sun, Tseng-Yao, Application of Gas-Fired Absorption Chillers, Heating / Piping / Air Conditioning, March 1991,
pp. 55–58.
The Trane Company, Thermachill Direct-Fired Absorption Chillers, The Trane Company, La Crosse, WI, 1992.
Wang, S. K., Principles of Refrigeration, vol. 3, Hong Kong Polytechnic, Hong Kong, 1984.
Watanabe, K., Akimoto, M., and Ohtsuka, S., An Approach to Support the Development of Absorption System
Technology in Japan, ASHRAE Transactions, 1982, Part II, pp. 492–506.
Wilkinson,W. H., What Are the Performance Limits for Double-Effect Absorption Cycles? ASHRAE
Transactions, 1987, Part II, pp. 2429–2441.
REFRIGERATION SYSTEMS: ABSORPTION 14.27
CHAPTER 15
AIR SYSTEMS: COMPONENTS—
FANS, COILS, FILTERS,
AND HUMIDIFIERS
15.1
15.1 FAN FUNDAMENTALS 15.2
Functions and Types of Fans 15.2
Fan Capacity or Volume Flow Rate 15.4
Fan Pressure 15.4
Fan Power and Fan Efficiency 15.4
Air Temperature Increase
through Fan 15.5
Fan Performance Curves 15.5
Influence of Elevation and
Temperature 15.6
15.2 CENTRIFUGAL FANS 15.7
Total Pressure Increase at Fan
Impeller 15.7
Backward-Curved Fans 15.8
Radial-Bladed Fans 15.10
Forward-Curved Fans 15.11
Tubular or In-line Fans 15.12
Unhoused Plug/Plenum Fans 15.12
Centrifugal Roof Ventilators 15.14
15.3 AXIAL FANS 15.14
Types of Axial Fans 15.14
Velocity Triangles 15.16
Performance Curves 15.18
15.4 FAN CAPACITY MODULATION 15.20
Fan Speed Modulation with Adjustable-
Frequency, Variable-Speed Drives 15.20
Inlet Vanes Modulation 15.21
Inlet Cone Modulation 15.23
Blade Pitch Modulation 15.24
15.5 FAN SURGE AND STALL 15.24
Fan Surge 15.24
Fan Stall 15.24
15.6 FAN CONSTRUCTION AND
ARRANGEMENTS 15.25
Size and Class Standards 15.25
Fan Width and Inlets 15.26
Drive Arrangements and Direction of
Discharge 15.26
High-Temperature Fans 15.27
Spark-Resistant Construction 15.28
Safety Devices 15.28
15.7 FAN SELECTION 15.29
Conditions Clarified and Factors
Considered 15.29
Estimated Fan Sound Power Level 15.30
Comparison between Various
Types of Fans 15.31
Case Study—Selection of Fans 15.32
15.8 COIL CHARACTERISTICS 15.32
Types of Coils 15.33
Fins 15.33
Contact Conductance 15.37
Water Circuits 15.39
15.9 SENSIBLE COOLING AND SENSIBLE
HEATING COILS—DRY COILS 15.39
Heat Transfer in a Sensible Cooling
Process 15.39
Surface Heat-Transfer Coefficients 15.41
Fin Efficiency 15.42
Dry Coil Effectiveness 15.42
Heating Coils 15.44
Fluid Velocity and Pressure Drop 15.44
Dry Coil at Part-Load Operation 15.44
Coil Construction Parameters 15.45
15.10 WATER COOLING COILS—DRY-WET
COILS 15.48
Dry-Wet Boundary 15.48
Dry Part 15.49
Wet Part 15.50
Dry-Wet Coil at Part-Load Operation 15.50
Selection of a Dry-Wet Coil 15.50
15.11 COIL ACCESSORIES, CONDENSATE
CONTROL, AND FREEZE-UP
PROTECTION 15.56
Coil Accessories 15.56
Coil Cleanliness 15.57
Condensate Collection and Drain
System 15.57
Coil Freeze Protection 15.58
15.12 AIR CLEANING, RATING, AND
ASSESSMENTS 15.60
Air Filtration and Industrial Air
Cleaning 15.60
Atmospheric Dust and Space Air
Contaminants 15.61
Rating and Assessments 15.61
15.13 TEST METHODS FOR AIR
FILTERS 15.62
Test Methods 15.62
Test Unit 15.64
15.14 AIR FILTERS 15.65
Filtration Mechanism 15.65
Classification of Air Filters 15.65
Coarse Air Filters 15.65
Low-Efficiency Air Filters 15.66
Medium-Efficiency Air Filters 15.67
High-Efficiency Air Filters 15.68
15.1 FAN FUNDAMENTALS
Functions and Types of Fans
A fan is the prime mover of an air system or ventilation system. It moves the air and provides continuous
airflow so that the conditioned air, space air, exhaust air, or outdoor air can be transported
from one location to another through air ducts or other air passages.
A fan is also a turbomachine in which air is usually compressed at a compression ratio Rcom not
greater than 1.07. The compression ratio, dimensionless, is defined as
(15.1)
where pdis  discharge pressure at outlet of compressor or fan, lbf / in.2 abs. or psia (kPa abs.)
psuc  suction pressure at inlet of compressor or fan, psia (kPa abs.)
A blower is usually an enclosed multiblade rotor that compresses air to a higher discharge pressure.
There is no clear distinction between a fan and a blower. Traditionally, blowers do not discharge air
at low pressure as some fans do.
A fan is driven by a motor directly (direct drive) or via belt and pulleys (belt drive). Some large
industrial fans in power plants are driven by steam or gas turbines.
Two types of fans are widely used in air conditioning and ventilation systems: centrifugal fans
and axial fans (Fig. 15.1). Fans can be mounted individually as ventilating equipment to provide
outdoor air or air movement inside a building. They can also transport air containing dust particles
or material from one place to another via air duct systems. In air conditioning systems, fans are
often installed in air-handling units, packaged units, or other air conditioning equipment.
In both centrifugal fans and axial fans, the increase of air static pressure is created by the conversion
of velocity pressure to static pressure. In centrifugal fans, air is radially discharged from the
impeller, also known as the fan wheel—air turns 90° from its inlet to its outlet. In an axial fan, the
Rcom 
pdis
psuc
15.2 CHAPTER FIFTEEN
Humidification Process 15.76
Ultrasonic Humidifiers 15.77
Case Study: White Plains Ultrasonic
Humidification Project 15.77
Centrifugal Atomizing Humidifiers 15.77
Pneumatic Atomizing Humidifiers 15.78
Wetted Element Humidifiers 15.78
15.21 AIR WASHERS 15.79
Construction of an Air Washer 15.79
Functions of an Air Washer 15.80
Performance of an Air Washer 15.80
Bypass Control 15.81
Single-Stage or Multistage 15.81
15.22 CHARACTERISTICS OF ATOMIZING
AND WETTED ELEMENT
HUMIDIFIERS 15.82
15.23 SELECTION OF HUMIDIFIERS
AND DESIGN 15.83
REFERENCES 15.85
Ultrahigh-Efficiency Filters 15.68
15.15 ELECTRONIC AIR CLEANERS 15.69
15.16 ACTIVATED CARBON
ADSORBERS 15.70
15.17 SELECTION OF AIR FILTERS 15.71
15.18 HUMIDIFICATION AND
HUMIDIFIERS 15.72
Space Relative Humidity 15.72
Humidifiers 15.72
Humidifying Load 15.72
Types of Humidifier 15.73
15.19 STEAM AND HEATING ELEMENT
HUMIDIFIERS 15.73
Steam Grid Humidifiers 15.73
Steam Humidifiers with Separators 15.74
Heating Element Humidifiers 15.75
Characteristics and Requirements of
Steam and Heating Element
Humidifiers 15.76
15.20 ATOMIZING AND WETTED
ELEMENT HUMIDIFIERS 15.76
15.3
FIGURE 15.1 Types of fans: (a) centrifugal; (b) axial; (c) cross-
flow.
direction of airflow is parallel to the axle of the fan. Additional differences in fan characteristics between
these two types of fans are discussed in the following sections.
Crossflow fans force the airflow by means of a long rotor with many vanes, as shown in
Fig. 15.1c. This type of fan has limited application in small air conditioning equipment because it is
less efficient and has a higher sound level than centrifugal fans.
Fan Capacity or Volume Flow Rate
Fan capacity or the fan volume flow rate , in cfm (m3 / s), is defined as the rate of volume flow
measured at the inlet of the fan, corresponding to a specific fan total pressure. It is usually determined
by the product of the duct velocity and the area of the duct connected to the fan inlet, according
to the test standard Air Movement and Control Association (AMCA) Standard 210-85 and
ASHRAE Standard 51-1985.
Fan volume flow rate is independent of air density . However, fan total pressure is affected by
air density. Therefore, the fan volume flow rate is normally rated at standard air conditions, i.e., dry
air at an atmospheric pressure of 14.696 psia (101,325 Pa abs.), a temperature of 70°F (21.1°C),
and a density of 0.075 lb / ft3 (1.2 kg /m3).
Fan Pressure
Fan total pressure ptf , expressed in inches of height of water column (inches WC or Pa), is the total
pressure rise of a fan, i.e., the pressure difference between the total pressure at the fan outlet pto
and the total pressure at the fan inlet pti, both in in. WG (Pag), or
(15.2)
Fan velocity pressure pvf , in in. WC (Pa), is the pressure calculated according to the mean velocity at
the fan outlet vo, in fpm (m / s). If air density  0.075 lb / ft3 (1.2 kg /m3), it can be calculated as
(15.3)
where pvo  velocity pressure at fan outlet, in. WC (Pa)
gc  dimensional constant, 32.2 lbmft / lbf  s2 (kgm/N s2)
 volume flow rate at fan outlet, cfm (m3/s)
Ao  cross-sectional area of fan outlet, ft2 (m2)
Fan static pressure psf, in in. WC (Pa), is the difference between the fan total pressure and fan velocity
pressure, or
(15.4)
where pso  static pressure at fan outlet, in. WG (Pa).
Fan Power and Fan Efficiency
Air power Pair, in hp (W), is the work done in moving the air along a conduit against a fan total
pressure ptf, in in. WC (Pa), at a fan volume flow rate of in cfm (m3/s). Because 1 hp  33,000
ft  lbf / min and 1 in. WC  5.192 lbf /ft2,
(15.5) Pair 
ptf  5.192  V?f
33,000

ptfV?f
6356
V?
f
psf  ptf 	 pvf  pto 	 pti 	 pvo  pso 	 pti
V?
o
pvf  pvo 
vo 2
2gc
  vo
40052
  V?o
4005Ao2
ptf  pto 	 pti
V?
f
15.4 CHAPTER FIFTEEN
The fan power input on the fan shaft, often called the brake horsepower Pf , can be calculated as
(15. 6)
When ptf is expressed in Pa, fan volume flow rate in m3/s, fan total efficiency t is the ratio of air
power Pair to fan power input Pf on the fan shaft, dimensionless, then constant C  1, and fan
power input Pf is expressed in W. If ptf is expressed in in. WC, fan volume flow rate is in cfm, fan
total efficiency t is a dimensionless ratio, then constant C  6356 and fan power input on the shaft
P is in hp.
From Eq. (15.6), fan total efficiency can also be calculated as
(15.7)
Fan total efficiency is a combined index of aerodynamic, volumetric, and mechanical efficiencies of
a fan. Fan static efficiency s is defined as the ratio of the product of the fan static pressure psf, in
in. WC, and the fan volume flow rate to the fan power input, i.e.,
(15.8)
Air Temperature Increase through Fan
If air density a  0.075 lb/ ft3, the specific heat of air cpa  0.243 Btu/ lb°F, and 1 hp  42.41
Btu/min, the relationship between fan power input and the air temperature increase when it flows
through the fan Tf, in °F, is given as
(15.9)
Combining Eqs. (15.6) and (15.9), then, gives
(15.10)
This air temperature rise in a fan when air flows through it is caused by the compression process
and energy losses that occur inside the fan. When air flows through the air duct, duct fittings, and
equipment, the duct friction loss and dynamic losses cause a temperature increase as mechanical
energy is converted to heat energy. However, this temperature increase in the air duct is offset by a
temperature drop caused by the expansion of air due to the reduction of static pressure along the
airflow. Therefore, it is more convenient to assume that the air temperature rise occurs because of
the friction and dynamic losses along the airflow only when air is flowing through the fan.
Fan Performance Curves
Fan characteristics can be described by certain interrelated parameters such as volume flow rate,
pressure, power, and efficiency. These characteristics are depicted graphically by fan performance
curves in Fig. 15.2.
Fan characteristic curves usually set the volume flow rate , in cfm (m3 / s), as the abscissa; and
fan total pressure ptf, fan static pressure psf, fan power input Pf, or fan total efficiency t as the
ordinate. In Fig. 15.2, there are three pressure-volume flow characteristic curves:
V?
Tf 
0.00667ptf
acpa t

0.37ptf
t
Pf 
V?
f
acpa Tf
42.41
s 
psf V?f
6356Pf
t 
ptf V?f
6356Pf
Pf 
ptfV?f
Ct

ptfV?f
6356 t
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.5
1. Total pressure-volume flow
2. Static pressure-volume flow
3. Velocity pressure-volume flow
The point where the curve intersects with ordinate is called the shutoff point. At this
point, the block tight or completely shut-off volume flow rate . Moreover, static pressure
is equal to total pressure at this point.
The volume flow rate at the point of intersection either between the and curves or
between the curve and the x axis is called free delivery, or the wide-open volume flow rate.
At the point of free delivery, ps  0, pt  pv, and fan static efficiency s  0.
In Fig. 15.2, the other curves shown are the fan total efficiency curve, the fan static effi-
ciency curve, and the fan power input curve. The efficiency curves show that at the shutoff
point, both t and s are zero, and at free delivery s is again zero. It is important that the fan be
operated near maximum efficiency as much as possible. Because of friction and dynamic losses, fan
power input at the shutoff point is not equal to zero.
Influence of Elevation and Temperature
Both elevation and temperature have an influence on air density a, in lb / ft3 (kg/m3), so they affect
the fan total pressure. The fan total pressure ptf, in in. WC (Pa), required at high elevations and
temperatures with corrections can be calculated as
(15.11) ptf  KTKel pt,s
P-V? s-V?
t-V?
ps-V?
pv-V? pt-V?
V?
block  0
pt-V? or ps-V?
pv-V?
ps-V?
pt-V?
15.6 CHAPTER FIFTEEN
FIGURE 15.2 Fan performance curves.
where pt,s  fan total pressure at standard air condition, in. WC (Pa). In Eq. (15.11), Kel is an
elevation factor and can be evaluated as
(15.12)
Here pat represents the atmospheric pressure at high elevation, in in. Hg. Between sea level and an
elevation of 5000 ft above sea level, pat can be roughly estimated as
pat  29.92 	 0.001Hel (15.13)
where Hel  elevation above sea level, ft. Temperature factor KT can be calculated as
(15.14)
where Ta  air temperature, °F (°C).
15.2 CENTRIFUGAL FANS
Total Pressure Increase at Fan Impeller
When air flows through the impeller of a centrifugal fan, its total pressure rise is closely related to
the peripheral velocities of the impeller as well as the tangential component of the peripheral velocities
entering and leaving the impeller. Figure 15.3 shows the velocity triangles at the blade inlet and
outlet of the impeller of a centrifugal fan. Let u represent the peripheral velocity of the impeller, w
the relative velocity in a direction tangential to the blade profile, and v the absolute velocity of the
fluid elements. The angle between the relative velocity w and the peripheral velocity u is called the
blade angle 	. The tangential component of the absolute velocity is indicated by vu and the radial
component by vm. Subscripts 1 and 2 denote the velocities at the inlet and outlet of the blade.
KT   530
Ta 4600.825
Kel 
pat
29.92
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.7
FIGURE 15.3 Velocity triangles at the blade inlet and outlet of a centrifugal
fan.
Air enters the impeller at the blade inlet with an absolute velocity of v1. As the fluid elements
flow through the blade passage, they accelerate. Moreover, a guiding pressure is exerted on the fluid
elements by the blades because of the rotation of the impeller. Air then leaves the blade outlet with
an absolute velocity v2.
If the energy losses due to the friction and eddies are ignored, and if the velocities are expressed
in fpm, then the total pressure rise, or fan total pressure pt, in in. WC, of the centrifugal fan when
air flows through the impeller is given as
(15.15)
where  air density, lb/ ft3
gc  dimensional constant, 32.2 lbmft / lbf  s2 (kgm/N s2)
If the impeller is deliberately designed at a specific blade inlet angle 	1 and a radial component vm1
so that the tangential component vu1  0, then the fan total pressure becomes
(15.16)
From Eqs. (15.15) and (15.16), the fan total pressure of a centrifugal fan depends mainly on the peripheral
velocity u2 and the tangential component vu2 and, therefore, the configuration of the blades.
The peripheral velocity at the blade outlet u2, often called the tip speed, in fpm (m/min), can be calculated
as
u2  
D2n (15.17)
where D2  outside diameter of fan impeller, ft (m)
n  rotational speed of impeller, rpm
For the same type of centrifugal fan, the greater the diameter, the greater the volume flow rate and
fan total pressure. For a centrifugal fan of a specific type and size, at each impeller speed, there will
be a set of different curves. The higher the speed, the greater , pt, and P.
Another important parameter that affects the characteristics of centrifugal fans is the blast area
Ablast, in ft2 (m2), as shown in Fig. 15.4. The blast area is the cross-sectional area just above the cutoff
point where discharge air is prevented from recirculating. It is always smaller than the outlet
area of the centrifugal fan Ao, in ft2 (m2). Blast area actually determines the maximum velocity discharged
from the fan outlet.
Based on the shape of the blade and the direction of air discharged from the impeller, centrifugal
fans can be categorized as backward-curved, radial-bladed, forward-curved, tubular or in-line, unhoused
or cabinet fans, and roof ventilators.
Backward-Curved Fans
In a backward-curved or backward-inclined centrifugal fan, the blade tip inclines away from the direction
of rotation of the impeller. The 	2 angle of a backward-curved centrifugal fan is less than
90°. Figure 15.3 shows the velocity triangles of a backward-curved centrifugal fan. The impeller of
a backward-curved centrifugal fan usually consists of 8 to 16 blades. For greater efficiency, the
shape of the blades is often streamlined to provide minimum flow separation and, therefore, minimum
energy losses. Backward-curved centrifugal fans with such blades are called airfoil fans, as
distinguished from fans with sheet-metal blades. The blades in a backward-curved fan are always
longer than those of a forward-curved fan. A volute or scroll casing is used. This shape converts
some of the velocity pressure to static pressure at the fan outlet.
From the velocity triangles at the blade outlet, the tangential component vu2  u2	vm2cot 	2 and
the radial component . Here b2 represents the width of the impeller at the blade tip. vm2  V?/( 
D2b2)
V?
pt-V?, P-V?, and t-V?
pt  1.66  10	6
vu2u2
 1.66  10	6
(vu2u2 	 vu1u1)
pt 
(vu2u2 	 vu1u1)
60  60  5.192gc
15.8 CHAPTER FIFTEEN
The total pressure developed is
(15.18)
where C1  constant equal to 1.66  10	6. Because 	2  90°, cot 	2 is a positive value, and the
theoretical curve shown in Eq. (15.18) is a straight line that declines at an increasing volume
flow rate, as shown in Fig. 15.4.
In actual operation, however, when the air flows through the centrifugal fan, it encounters the
following energy losses:
 Circulatory flow between the blades
 Air leakage at the inlet
 Friction between fluid particles and the blade
 Energy loss at the entrance
 Partially filled passage
These energy losses change the shape of the curve. The actual performance curve of a backward-
curved centrifugal fan is a concave curve declining toward the right-hand side, as shown in
Fig. 15.4. It can be seen that the maximum total pressure is slightly higher than that at the shutoff
condition. After reaching the maximum pt, it drops sharply as the volume flow rate increases.
pt-V?
pt-V?
pt  C1 vu2u2  C1 u2 2 	
u2V? cot 	2

D2b2

AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.9
FIGURE 15.4 Operating characteristics for a backward-curved centrifugal fan.
The fan total efficiency versus volume flow curves for backward-curved centrifugal fans
are concave. All have t  0 at block tight condition as . They also have a maximum
efficiency between 50 and 65 percent of the wide-open volume flow, as shown in Fig. 15.5. Among
these curves, the airfoil blade backward-curved centrifugal fan has the highest fan total
efficiency.
Figure 15.6 shows the fan power–volume flow curves for centrifugal fans using impellers
of the same size. Centrifugal fans need the minimum amount of power at block tight condition. This
power input is used to offset the aerodynamic and mechanical losses even when the volume flow is
zero. For backward-curved centrifugal fans, fan power P tends to increase initially as the volume
flow rate increases. Then, P reaches a maximum point, after which P declines as the volume flow
rate is further increased. The curve of a backward-curved centrifugal fan has a shape similar to
that of the curve. If the size of the fan motor is rated at or near the highest fan total efficiency,
the power required is often nearly the maximum value, and the motor needs no more fan power at
the same speed and density than at any other operating conditions. Such a curve is called a
nonoverloading curve.
Radial-Bladed Fans
The blades in a radial-bladed centrifugal fan are either straight or curved at the blade inlet. The blade
tip or blade outlet is always radial; that is, 	2  90°, as shown in Fig. 15.7a. Usually, there are 6 to
10 blades in a radial-bladed impeller. The construction of the radial blades is comparatively simple.
P-V?
pt-V?
P-V?
P-V?
t-V?
V?
 0
t-V?
15.10 CHAPTER FIFTEEN
FIGURE 15.5 Total efficiency–volume t- performance curves for centrifugal
fans.
V?
Figure 15.8 shows the curves for centrifugal fans with impellers of the same diameter.
The curve of a radial-bladed centrifugal fan is usually steeper than that of a backward-curved
fan, with a higher pt and a smaller . The fan total efficiency of a radial-bladed fan is lower than
that of a backward-curved fan. Furthermore, the fan power input always increases as the volume
flow becomes greater. Such an overloading curve indicates that the motor may be overloaded.
The radial-bladed centrifugal fan is often used in industrial applications to transport particles or
products because the spaces between the blades are not easily clogged.
Forward-Curved Fans
In a forward-curved centrifugal fan, the blade tip inclines in the direction of rotation of the impeller.
At the outlet of the blade, blade angle 	2 90°, as shown in Fig. 15.7b. Because of a greater
absolute velocity v2 at the blade outlet, the blades are shorter and the speed required to produce a
specific fan total pressure and volume flow rate is much lower than that for a backward-curved
centrifugal fan. The impeller generally has 24 to 64 blades.
If the diameters of the impellers of a forward-curved fan and a backward-curved fan are the
same, then the curve of the forward-curved centrifugal fan is flatter than that of the backward-
curved fans, as shown in Fig. 15.8. Forward-curved fans have a lower fan total efficiency than
backward-curved fans, and the maximum efficiency of the forward-curved fan occurs between 40
and 50 percent of the wide-open volume flow. The forward-curved fan power curve tends to bend
upward more steeply than that of other centrifugal fans, as shown in Fig. 15.6. Therefore, the
motor of the forward-curved fan may be overloaded when the volume flow rate is greater than the
rated value.
Forward-curved fans are the most compact centrifugal fans available in terms of volume flow
rate delivered per unit space occupied. They are also less stable, as indicated by the saddle shape of
their curve. For the same size fan wheel, forward-curved fans have a greater inlet than backward-
curved fans.
pt-V?
pt-V?
P-V?
V?
pt-V?
pt -V?
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.11
FIGURE 15.6 Power-volume performance curves for centrifugal
fans with impellers of same diameters.
P-V?
Tubular or In-line Fans
A tubular or in-line centrifugal fan generally consists of an impeller with airfoil blades, an inner
cone, a set of fixed vanes, and a cylindrical casing, as shown in Fig. 15.9. Air enters the tubular fan
at the inlet and then flows through the blade passages of the impeller. It is discharged radially
against the inner surface of the cylindrical casing and is then deflected in a direction parallel to the
axle of the fan, in order to produce a straight-through flow. The impeller of a tubular centrifugal fan
may have 6 to 12 blades. Fixed vanes are used to convert velocity pressure to static pressure to prevent
swirls and straighten the airflow.
In Fig. 15.8, it can be seen that the curve for a tubular centrifugal fan is similar to that of a
backward-curved fan. Because of the 90° turn of the airflow after the impeller, both total pressure
and total efficiency are lower than those of a backward-curved fan, and the sound power level is
greater. The fan power curve of this type of fan is also a nonoverloading curve. The main advantage
of the tubular fan is its straight-through airflow.
Unhoused Plug/Plenum Fans
Unhoused plug/plenum fans are centrifugal fans with no outer casing or scroll. The impeller is
mounted in a plenum with a flexibility of the location of the discharged outlet. Unhoused centrifugal
fans are simpler in construction and can plug into an air system where a plenum can be used as the
outer casing of the impeller, and the two ends of the plenum used as the inlet and outlet of the fan.
P-V?
pt-V?
15.12 CHAPTER FIFTEEN
FIGURE 15.7 Velocity triangles of radial-bladed and forwardcurved
centrifugal fan: (a) radial-bladed; (b) forward-curved.
Coward (1997) noted the influence of the gap between the plenum wall and the impeller on the
fan volume flow rate and fan power input as well as various types of duct entrance losses. The performance
of an unhoused plug fan should be predictable and should be carefully taken into consideration
during design.
Unhoused plug/ plenum fans with backward-curved impellers may require more power than
centrifugal fans with outer casings (fan total efficiency is usually between 58 and 63 percent). They
are lower in sound power level if the plenum is properly sized and acoustically treated. Unhoused
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.13
FIGURE 15.8 Pressure-volume performance curves for centrifugal
fans with same impeller diameter.
pt-V?
FIGURE 15.9 Tubular centrifugal fan.
plug/plenum fans are often used as return fans in many rooftop packaged units and as exhaust fans
in industrial applications.
Centrifugal Roof Ventilators
A centrifugal roof ventilator is often mounted on a roof to exhaust air from a ventilated space, as
shown in Fig. 15.10. Roof ventilators can also be used as intake devices to draw outdoor air through
the roof. In Fig. 15.10, air is extracted from ventilated space and enters the inlet cone directly or
through a connecting duct having a total pressure loss usually less than 0.2 in. WC. After flowing
through a backward-curved centrifugal impeller, air is discharged radially against the inner surface
of the hood. The sheet-metal or airfoil blades of the centrifugal roof ventilator are made of aluminum
alloy, stainless steel, or well-protected structural steel to provide good corrosive resistance
above the roof.
The centrifugal roof ventilator usually operates at a low total pressure and a large volume flow
rate, as illustrated by a downward inclined curve, as shown in Fig. 15.8. The power
curve is a nonoverloading curve.
15.3 AXIAL FANS
Types of Axial Fans
For an axial fan, a parameter called the hub ratio is closely related to its characteristics. Hub
ratio Rhub is defined as the ratio of hub diameter Dhub, in ft (m), to the tip-to-tip blade diameter or
diameter of impeller Dbt, in ft (m),
(15.19) Rhub 
Dhub
Dbt
P-V? pt-V?
15.14 CHAPTER FIFTEEN
FIGURE 15.10 Centrifugal roof ventilator.
The higher the hub ratio, the greater the conversion of velocity pressure to static pressure because
of the larger difference of the areas: .
The capacity and fan total pressure of axial fans can be increased by raising their rotating speed
or through the adjustment of the blade pitch angle to a higher value. This characteristic is important
for axial fans that are driven directly by motor, without belts.
Axial fans can be subdivided into the following three types:
Propeller Fans. In a propeller fan, an impeller having 3 to 6 blades is mounted within a circular
ring or an orifice plate, as shown in Fig. 15.11a. The blades are generally made of steel or molded
plastic and sometimes may increase in width at the blade tip. If the impeller is mounted inside an
orifice plate, the direction of airflow at the blade tip will not be parallel to the axle. Eddies may
form at the blade tips. Propeller fans are usually operated at very low static pressure with large volume
flow. They often have a hub ratio Rhub  0.15.
Tube-Axial Fans. The impeller of a tube-axial fan usually has 6 to 9 blades. It is mounted within
a cylindrical casing. The blades can be airfoil blades or curved sheet metal. Airfoil blades are

(Dbt
2 	 Dhup
2 ) /4
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.15
FIGURE 15.11 Axial fans: (a) Propeller;
(b) tube-axial; (c) vane-axial.
usually made of cast aluminum or aluminum alloy. The hub ratio Rhub is generally less than 0.3, and
the clearance between the blade tip and the casing is significantly closer than in propeller fans.
In some tube-axial fans, the blade angle can be adjusted manually when the fan is not in operation.
Thus, the volume flow and fan total pressure can be adjusted for tube-axial fans driven
directly by motor.
Vane-Axial Fans. The impeller of a vane-axial fan has 8 to 16 blades, usually airfoil blades. The hub
ratio is generally equal to or greater than 0.3 in order to increase the fan total pressure. Another important
characteristic of vane-axial fans is the installation of fixed guide vanes downstream from the
impeller, as shown in Figs. 15.11c and 15.12. These curved vanes are designed to remove swirl from
the air, straighten the airflow, and convert a portion of the velocity pressure of the rotating airflow to
static pressure. Sometimes guide vanes are also installed upstream from the impeller. Automatically
controllable pitch (blade angle) for variable-air-volume system is one of the features of vane-axial fans.
Velocity Triangles
Consider a blade section of an axial fan, as shown in Fig. 15.12. Air enters the impeller with a velocity
v1, which is exactly equal to the axial velocity at the blade inlet vm1. Because of the rotation
15.16 CHAPTER FIFTEEN
FIGURE 15.12 Velocity triangles for a vane-axial fan.
of the impeller, air leaves the blade at its outlet with a velocity v2. For a blade section whose diameter
at blade inlet equals the diameter at blade outlet, the peripheral velocities u1  u2, and the axial
velocities vm1  vm2. The total pressure rise pt, in in. WG, as air flows through the impeller is
given by the same equation as for a centrifugal fan when velocities are expressed in fpm:
To design a blade such that the total pressure developed at various radii of the blade is constant, the
blade angle near the hub of the same blade must be increased in order to provide a higher vu2 when
the radius becomes smaller. A greater hub ratio Dhub /Dbt results in a smaller variation of the blade
angles of the same blade.
After air leaves the blade and before it enters the downstream fixed guide vanes, there is a rotational
component vu2 in the direction of rotation. The function of the downstream guide vanes is to
convert this dynamic pressure to static pressure with minimum energy loss.
The static pressure developed in an axial fan is the combined effect of the following:
 The corresponding drop in relative velocity pressure
 The conversion of rotating dynamic pressure to static pressure
 The conversion of the difference in velocity pressures
where v3, v4  mean axial velocities at cross-sectional areas with and without hub, respectively,
fpm (m/ s)
C2  constant equal to 1 / (602  2  32.2  5.192)  8.3  10	7
C2 (v3 2 	 v4 2)
C2 (v2 2 	 vm2
2 )
C2 (w1 2 	 w2 2)
pt  1.66  10	6
vu2u2
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.17
FIGURE 15.13 Pressure-volume for axial fans with the same impeller
diameter.
pt-V?
Performance Curves
Figure 15.13 shows the pressure-volume flow curves for various types of axial fans with impellers
of the same diameter. Among these fans, the propeller fan has a low fan total pressure pt
and a high volume flow rate , whereas the tube-axial fan has a higher pt and the vane-axial has
the highest pt.
V?
pt-V?
15.18 CHAPTER FIFTEEN
FIGURE 15.14 Total efficiency t - curves for axial fans with the same
impeller diameter.
V ?
FIGURE 15.15 Fan power curves for axial fans with impellers of the
same diameter.
P-V?
Figure 15.14 shows the total efficiency curves for axial fans of the same diameter. Because
of the installation of airfoil blades and the downstream guide vanes, vane-axial fans have the
highest t of all axial fans. A well-designed vane-axial fan may have the same or even higher
maximum t as a backward-curved airfoil centrifugal fan. In a vane-axial fan, it is important to use
its velocity pressure, or convert it to static pressure by means of a diffuser. Tube-axial fans have
lower t. Propeller fans have the lowest t, ranging between 0.4 and 0.55, because of their simple
construction.
Figure 15.15 shows the fan power curves for axial fans. It is important to see that all axial
fans have their maximum fan power input at the shutoff or block tight condition. In this respect,
they are quite different from the centrifugal fans, which have high fan power input at large volume
flow rates. For axial fans, the greater the volume flow rate, the lower P, except in vane-axial fans,
whose curves are saddle-shaped. Because the curves for vane-axial and tube-axial fans are
nearly parallel to the curves at the operating range, they are nonoverloading curves. Motors
need not be sized according to the maximum fan power input as long as the axial fans will not operate
at shutoff condition.
Figure 15.16 shows the performance curves of , , and total efficiency t for a typical
vane-axial fan. The impeller is 43 in. (1.09 m) in diameter and rotates at a speed of 1750 rpm. From
P-V? pt-V?
pt-V?
P-V? P-V?
P-V?
t-V?
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.19
FIGURE 15.16 Controllable-pitch vane-axial fan performance curves. (Source: Buffalo Forge Company.
Reprinted with permission.)
15.20 CHAPTER FIFTEEN
the performance curves, it can be seen that the greater the blade pitch or blade angle, the higher the
fan total pressure pt and the volume flow rate.
Reverse Operation. If the rotation of an axial fan is reversed, the direction of airflow is reversed. Propeller
and tube-axial fans without guide vanes deliver about 60 to 70 percent of the volume flow rate in
the original forward direction when their rotations are reversed. Vane-axial fans with guide vanes either
downstream or upstream from the impeller do not perform efficiently when they rotate in reverse.
Tip Clearance and Number of Blades. For an axial fan, the size of the gap between the blade tip
and the casing , as shown in Fig. 15.12, or the ratio of this gap to the outside diameter of the impeller
/Dimp, has a definite effect on fan efficiency and sound power level. Here Dimp indicates the
diameter of the impeller. Clearance  and diameter Dimp must be expressed in the same units. Generally,
/Dimp should be equal to or less than 0.0025. When the gap becomes greater, the amount of
air leakage through the gap increases. Various tests have shown that fan total efficiency may drop
from 0.7 to 0.68 if /Dimp increases from 0.0025 to 0.01.
Theoretically, the more blades in an axial fan, the better the guidance of air. At the same time,
the friction between the fluid elements and the surface of the blade increases. An axial fan with a lot
of blades always produces a higher fan total pressure than a fan with fewer blades at the same volume
flow rate, if other conditions are the same.
15.4 FAN CAPACITY MODULATION
In a variable-air-volume (VAV) air system, more than 90 percent of its running hours operate at an
airflow less than the design volume flow rate. Modulating the fan capacity by providing a new
curve with lower volume flow and fan total pressure not only corresponds with load reduction
in the conditioned space, but also allows significant energy savings at part-load operation.
Four types of fan capacity modulation are widely used in air systems: fan speed, inlet vanes, fan
inlet cones, and blade pitch modulation.
Fan Speed Modulation with Adjustable-Frequency, Variable-Speed Drives
The peripheral velocity at blade tip u2 affects both the tangential component vu2 and radial component
vm2 of the airflow and, therefore, the and pt of the fan. The variation of fan speed produces
a family of similar curves, as shown in Fig. 15.17a.
Most widely used, variable-speed drives (VSDs) for fans and pumps are adjustable-frequency
VSDs. These VSDs modulate the speed of the ac motor by supplying a variable-frequency and
variable-voltage power source. The synchronous speed of the motor nm, in rpm, can be calculated as
(15.20)
where f  frequency of applied power, Hz
Np  number of poles
If the frequency varies, the motor speed changes accordingly. The voltage must be reduced with the
frequency, to maintain a specific voltage-to-frequency ratio, in order to follow the decreasing inductive
reactance of the motor.
The incoming three-phase ac power is rectified into dc supply through a rectifier. Then the dc
power supply is shaped into a pseudo-sine wave of predetermined frequency by an inverter, as shown
in Fig. 15.17b. The advantage of an ac inverter is that its speed can be reduced to one-tenth (or even
less) of its original speed. Through the change of frequency, the speed-torque curve of the motor can
nm 
120f
Np
pt-V? V?
pt-V?
be adjusted so that it is always operated near its maximum efficiency. The electronics themselves
consume little energy. They are usually about 97 to 98 percent efficient. However, because of the ac
inverter, motor efficiency degrades 5 to 7 percent, so the total energy loss is about 7 to 8 percent.
The pulse-width-modulated (PWM) inverter is the predominate design of ac inverters used in
HVAC&R systems. A PWM inverter operates from a fixed dc bus. Varying the timing on thyristors
or transistors produces pulsed output. A logic circuit and software control the switching to provide
the required variable voltage and frequency for a specific variable speed in an ac induction motor.
The advanced power electronics, called insulated gate bipolar transistors, successfully provides
high switchings with low energy lossses and a compact and quieter drive.
PWM inverters operate at a wide range of speeds and can also provide a “soft start,” which
brings the motor to the rated speed gradually to reduce abnormally high starting current. They are
now often cost-effective for large fan motors in variable-air-volume systems and pump motors in
variable-flow water systems. (They often have a simple payback of less than 3 years for motors
with a capacity equal to and greater than 25 hp.)
Variable-speed drives using adjustable-frequency ac inverters have a good operating characteristic
when they are used for centrifugal fans and pumps with impeller of backward-curved blades.
Operation of forward-curved centrifugal fans may become unstable and more difficult to control because
of their saddle-type fan characteristic curves.
Inlet Vanes Modulation
As with centrifugal compressors, inlet vanes in centrifugal fans are pivoted, movable vanes installed
at the inlet of the impeller. These vanes are linked mechanically so that they can turn
simultaneously around the axis, as shown in Fig. 15.18.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.21
FIGURE 15.17 AC inverter and curves of an airfoil fan at various
speeds: (a) curves at various speeds; (b) ac inverter.
pt-V?
Inlet vanes impart a spin on the airstream before it enters the impeller. From Eq. (15.15),
If the direction of the spin is the same as the direction of rotation of the impeller, the tangential
component vu1 has a positive value. The fan total pressure thus developed is smaller than if air
enters the impeller with a radial entry vu1  0. This produces a new set of curves at various
inlet vane opening angles with lower pt and .
The fan total efficiency t decreases as the inlet vane opening angle decreases, as the air at
inlet is further deviated from the shock-free condition and the inlet area is reduced. However, the
reduction of shaft power from the decreases of pt and compensates for the effect of the drop
in t. Energy savings are significant only when the inlet vanes are sufficiently near the eye of the
impeller.
V?
V?
pt-V?
pt  1.66  10	6
(vu2u2 	 vu1u1)
15.22 CHAPTER FIFTEEN
FIGURE 15.18 Inlet vane control of a backward-curved centrifugal fan.
Inlet vanes reduce the inlet area. For large fans, this reduced inlet area causes a pressure loss of
about 8 percent. For small fans, the additional pressure loss is far greater. If air velocity at the inlet
vi exceeds 5000 fpm (25 m / s), the fan total pressure is significantly reduced. Inlet guide vanes are
usually not preferable for small backward-curved centrifugal fans. During the commissioning of an
air-handling unit with VAV system, the available fan static pressure dropped from the specified 4 to
only 2.5 in. WC (1000 to 625 Pa) when the air velocity in the fan inlet of the backward-curved centrifugal
fan installed with inlet vanes exceeded 5000 fpm (25 m / s).
A backward-curved centrifugal fan with airfoil blades and inlet vanes has a higher ratio of reduced
volume flow power input to design volume flow power input Ppart /Pfull than a forward-curved
centrifugal fan has. However, a backward-curved airfoil fan has a higher inherent fan total
efficiency than a forward-curved fan. Therefore, for a VAV system, actual fan power input should be
compared between an airfoil fan and a forward-curved fan at reduced volume flow operation (according
to the manufacturer’s data) before the fan selection.
Use of inlet vanes for fan capacity modulation is still widespread for centrifugal fans because of
their lower installation cost. Because axial fans require maximum fan power input at shutoff condition,
inlet vanes are not suitable for axial fans.
Inlet Cone Modulation
The inlet cone of a backward-inclined or airfoil centrifugal fan can be moved so that a portion of
the impeller is inactive. Such a modulation results in a lower fan capacity and fan total pressure.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.23
FIGURE 15.19 Controllable-pitch vane-axial fan blade pitch control mechanism.
Inlet cone modulation has a lower initial cost. Although its fan energy use is higher than that of
variable-speed drive at part-load conditions, inlet cone modulation does not block the area of the inlet
eye. Inlet cone modulation is used in rooftop packaged units in which airfoil centrifugal fans are
installed.
Blade Pitch Modulation
From the velocity triangles shown in Fig. 15.12, for axial fans, the smaller the blade angle 	2, the
lower vm2 and vu2 and, therefore, the lower and pt. Figure 15.16 shows the performance diagram
of a typical controllable and adjustable-pitch vane-axial fan in which there is a corresponding
curve for each blade angle. For axial fans, blade pitch fan capacity modulation can take place both
during the rotation of the fan impeller (controllable pitch) or at rest (adjustable pitch).
Figure 15.19 shows the blade pitch control mechanism of this controllable-pitch vane-axial fan.
The blades are activated internally through level arms and a roller mechanism by pneumatic pressure
against a flexible diaphragm. The control mechanism is direct-acting; i.e., the blade angle increases
as the controlled pneumatic air pressure is increased. If the pneumatic pressure fails when
the fan is operating, blades return to their minimum blade pitch. A pilot positioner, located outside
the fan, ensures that the controlled blade pitch has the minimum deviation from the required one.
The pneumatic pressure used to actuate the blade pitch is 25 psig (172 kPag) while the pressure
used to activate the pilot positioner is 3 to 15 psig (21 to 103 kPag).
Of these four fan capacity modulating methods, inlet vane modulation is lower in installation
cost and provides significantly lower energy savings than fan speed and blade pitch modulations
at reduced airflow. Fan speed and blade pitch modulation also reduce fan noise and increase the
fan’s life.
15.5 FAN SURGE AND STALL
Fan Surge
Fan surge (or stall) occurs when air volume flow through a centrifugal fan is not sufficient to sustain
the static pressure difference between the discharge and suction sides of the fan. Air surges back
through the fan impeller, and its discharge pressure is reduced momentarily. The surge of air enables
the fan to resume its original static pressure. Fan surge will repeat if the air volume flow is not
sufficient to sustain normal operation.
The volume flow and pressure fluctuations caused by fan surge create noise and vibration.
The surge region of a backward-curved centrifugal fan is usually greater than that of a forwardcurved
fan. No fan should operate in the surge region, shown in Fig. 15.17. A large fan in surge is
intolerable. Fan room walls have been cracked from the vibration of the ducts during a large fan
surge.
Some manufacturers show fan surge regions on their fan performance diagram in catalogs. If
data and curves are not presented in the performance table or diagram, they are probably in or near
the fan surge region.
Fan Stall
Stall is an operating phenomenon that often occurs in axial fans. For airfoil blades in an axial fan,
the angle of attack is the angle between the motion of the blades and the chord line, as shown in
Fig. 15.20a. At normal operation, airflow around the streamlined airfoil blades causes a pressure
difference between the top and bottom surfaces of the blades and produces lift. Stall occurs when
pt-V?
V?
15.24 CHAPTER FIFTEEN
the smooth airflow suddenly breaks and the pressure difference across the airfoil blades decreases.
The axial fan loses its pressure capability drastically.
The stall region of an axial fan is generally located at the upper part of the fan performance diagram
beyond the region of highest fan total efficiency, as shown in Fig. 15.20b. For each blade pitch
angle, there is a corresponding stall region line. Usually, stall region is also the region in which no
performance curves are given.
15.6 FAN CONSTRUCTION AND ARRANGEMENTS
Size and Class Standards
The Air Movement Control Association (AMCA) has established standards concerning fan size and
classification. The size standard specifies the diameter of the impeller for a given fan size. There are
25 sizes, from 12.25 to 132.25 in. (311 to 3360 mm) in diameter. Each size is approximately 10
percent greater in diameter than the next-smaller size.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.25
FIGURE 15.20 Stall and stall region of an axial fan: (a) normal operation and stall; (b) stall region of an
axial fan.
Class standard specifies the construction of the fans as heavy, medium, or light structure. Construction
is closely related to the static pressure developed ps (in in. WC) by the fan and the fan
outlet velocity vo (in fpm or m / s), as shown in Fig. 15.21. Two classes of fans are widely used in
HVAC&R applications: class I and class II. As shown in Fig. 15.21, a class I fan must provide 5 in.
WC (1250 Pa) static pressure at an outlet velocity of 2300 fpm (11.5 m / s) and a static pressure of
2.5 in. WC (625 Pa) at an outlet velocity of 3200 fpm (16 m / s). A class II fan must provide 8.5 in.
WC (925 Pa) static pressure at an outlet velocity of 3000 fpm (15 m / s) and 4.25 in. WC (1063 Pa)
static pressure at 4175 fpm (20.9 m / s).
Fan Width and Inlets
For centrifugal fans, air velocity at the fan inlet vi should be optimal. A lower vi means a large inlet.
A higher vi results in a greater energy loss and therefore a lower fan total efficiency. A suitable relationship
is vi /u2  0.35 to 0.4.
A single-width single-inlet (SWSI) centrifugal fan has a single inlet cone connected to the
eye of the impeller, as shown in Fig. 15.22a. A double-width double-inlet (DWDI) centrifugal
fan has a double-width impeller and provides double inlets from both sides, as shown in
Fig. 15.22b. In such an arrangement, vi can still be maintained at an optimal value, and the
volume flow rate is approximately doubled. Because of the interaction of the two airstreams in
the impeller and the scroll, the fan efficiency of a DWDI fan may be 1 to 2 percent lower than
that of an SWSI fan.
Drive Arrangements and Direction of Discharge
Fans can be driven by motor directly or through V-belt and sheaves. In a belt-drive arrangement, fan
speed can be changed by varying the diameter of the sheaves. On the other hand, belt drive requires
3 to 5 percent more energy input than direct drive.
15.26 CHAPTER FIFTEEN
FIGURE 15.21 Class standards of fans.
Drive arrangements include the location of bearing and sometimes the position of the motor. For
centrifugal fans, there are eight standard drive arrangements, as shown in Fig. 15.23. There are only
two drive arrangements, 4 and 9, for axial fans. Arrangements 1, 2, 3, 7, and 8 can include either
belt drive or direct drive. Arrangements 4 is for direct drive only. Arrangements 9 and 10 are for
belt drive only. However, in arrangement 9 for centrifugal fans, the motor is located outside the
base, whereas in arrangement 10, the motor is located inside the base. This arrangement is often
used with a weatherproof hood for outdoor installation.
Motor location is always specified as W, X, Y, or Z position, facing the fan drive side. The
direction of rotation and discharge position of centrifugal fans are specified from the drive side of
the fan, as shown in Fig. 15.24. In single-inlet fans, the drive side is always opposite the fan inlet,
regardless of the actual position of the fan drive.
High-Temperature Fans
When fans are operated in high-temperature airstreams, the yield strength of the fan structure may
decrease. Operating temperatures are usually classified as 	20 to 200°F (	29 to 93°C) normal or
standard fan construction, 201 to 300°F (94 to 150°C), 301 to 400°F (151 to 204°C), 401 to 750°F
(205 to 400°C).
The rpm limits on classes I and II must be multiplied by a speed factor 0.9 when the operating
temperature is 401 to 500°F (201 to 260°C). A lower speed factor should be used at higher
temperatures. Cooling of the fan bearings is a significant problem. Generally, bearings should
be kept out of the airstream except when operating at room temperature. In many antifriction
bearings, grease can be used up to 200°F (93°C). Oil bearings or high-temperature grease is required
for temperatures higher than 300°F (150°C). Special coolant should be applied to the
shaft and bearings at higher temperatures. High-temperature aluminum paint can protect surface
finishes.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.27
FIGURE 15.22 Single- and double-width centrifugal fans: (a) single-width single-inlet (SWSI); (b) doublewidth
double-inlet (DWDI).
Spark-Resistant Construction
AMCA specifies three types of spark-resistant fan construction:
 Type A construction requires that all parts of the fan in contact with the air be made of nonferrous
metal.
 Type B construction requires that a fan have a nonferrous impeller and shaft rings.
 Type C construction requires that a shift of the impeller or shaft not allow the ferrous parts to rub
or strike.
The bearings of spark-resistant construction fans should not be placed in the airstream, and all fan
parts should be electrically grounded.
Safety Devices
When a fan is installed in a location where it is open to occupants or is accessible during operation,
safety devices such as belt guards and protecting screens must be installed. Even a wire mesh
installed at the outlet of a fan with a high air velocity may cause a pressure loss of 0.15 in. WC
15.28 CHAPTER FIFTEEN
FIGURE 15.23 Drive arrangements for centrifugal and axial fans and motor positions.
(37 Pa). Therefore, if the fan is covered by an enclosure or is accessible only when the fan is not in
operation (during repair or maintenance), a safety device is not required.
15.7 FAN SELECTION
Selection of a fan for a given type of air system or mechanical ventilating system actually is done in
two stages: selection of fan type and determination of fan size.
Conditions Clarified and Factors Considered
Before the selection, the following conditions must be clarified:
 Setting (in a commercial building to handle clean air at room temperature, or an industrial setting
to handle dirty air)
 Special requirements (such as high-temperature operation or spark-resistant construction)
 Function (supply fan or a return fan in an air-handling unit, or supply or exhaust fan in a ventilating
system)
 Characteristics of the air system (constant volume or variable air volume)
 Room NC curve
 Approximate annual operating hours
 Unit cost of energy at the specific location
During selection, the following factors should be considered:
 Pressure-volume flow operating characteristics. Selecting a fan to provide the required volume
flow rate and total pressure loss for an air system or a ventilating system is of prime importance.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.29
FIGURE 15.24 Direction of rotation and discharge position for centrifugal fans.
An undersized fan results in an uncontrolled indoor environment. An oversized fan wastes energy
and money.
 Fan capacity modulation. A variable-air-volume system operates at a reduced volume flow rate
during part-load operation. Effective and economical fan capacity modulation is an important factor
that affects the operation of an air system.
 Fan efficiency. Fan efficiency is closely related to the energy consumption of an air system. Fans
should be selected so that they can operate at high efficiency during as much of their operation
time as possible.
 Sound power level. Most commercial and public buildings and many industrial applications need
a quiet indoor environment. Fans are the major source of noise in an air system. Usually, the
higher the fan total efficiency, the lower the sound power level of the selected fan. A fan with a
low sound power level and sound power level at high frequencies is preferable. High-frequency
sound is more easily attenuated than low-frequency sound.
 Airflow direction. In many applications, a straight-through or in-line flow occupies less space and
simplifies layout.
 Initial cost. The initial cost of the fan modulation device, sound attenuator(s), and space occupied
by a particular type of fan, in addition to the cost of the fan itself, should be considered.
Another important factor in fan selection is the fan inlet and outlet connections. Since the effects
of these connections are closely related to the characteristics of the air system, they are covered in
Chap. 21.
In an air system, a fan is usually operated with a duct system, which may also include other
equipment and system components. A comprehensive and detailed analysis of fan operation must
consider the characteristics of the duct system.
Estimated Fan Sound Power Level
The estimated sound power level of a fan Lw, in dB, can be calculated as
(15.21)
where Kw  specific sound power level of a certain type of fan, dB
 fan volume flow rate or capacity, cfm
 1 cfm
 fan total pressure, in. WC
 1 in. WC
In Eq. (15.21), C represents the correction for off-peak operation, i.e., fan operation in the
region of the performance curve where fan efficiency is not maximum. For a fan operating at
90 to 100 percent of its maximum static efficiency, C  0; between 85 and 89 percent, C  3.
For each decrease of 10 percent of fan maximum static efficiency up to 55 percent, the C value
increases by 3 dB.
Table 15.1 lists the specific sound power level Kw and blade frequency increments BFI, in dB, of
various types of fans. Here, BFI should be added to the sound power level at blade frequency fblade,
in Hz. Blade frequency can be calculated as
(15.22)
A comparison of Kw values shows that the backward-curved centrifugal fan with airfoil blades
has the lowest Kw value at various octave bands. The differences between Kw values at various
fblade 
fan rpm  number of blades
60
pt1
pt
V?
1
V?
Lw  Kw 10 log
V?
V?
1
 20 log
pt
pt1
 C
15.30 CHAPTER FIFTEEN
octave bands for axial fans are far smaller than those of centrifugal fans. For forward-curved and radial-
bladed centrifugal fans, Kw values are higher in low frequencies and lower in high frequencies
than vane-axial and tube-axial fans.
Comparison between Various Types of Fans
Table 15.2 compares the characteristics of various types of fans. The backward-curved centrifugal
fan with airfoil blades has the highest fan total efficiency and the lowest specific sound power level,
so it is still the most widely used type of large fan in commercial, institutional, and many industrial
applications. The forward-curved centrifugal fan has a compact size, slower speed, and lighter
weight per unit volume output. It is widely used in room air conditioners, fan coils, small airhandling
units, and many packaged units. In recent years, vane-axial fans with controllable pitch,
especially used as return fans, have more applications in commercial air systems than before. For
exhaust systems that require a large volume flow rate and low fan total pressure, a propeller fan is
often the best choice.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.31
TABLE 15.1 Specific Sound Power Levels Kw for Typical Fans
Octave bands
Fan type 63 125 250 500 1K 2K 4K 8K BFI
Centrifugal
AF, BC, or BI wheel diameter
Over 36 in. 40 40 39 34 30 23 19 17 3
Under 36 in. 45 45 43 39 34 28 24 19 3
Forward-curved
All wheel diameters 53 53 43 36 36 31 26 21 2
Radial-bladed
Low pressure
(4 to 10 in. of water) 56 47 43 39 37 32 29 26 7
Medium pressure
(6 to 15 in. of water) 58 54 45 42 38 33 29 26 8
High pressure
(15 to 60 in. of water) 61 58 53 48 46 44 41 38 8
Axial fans
Vane-axial
Hub ratio 0.3 to 0.4 49 43 43 48 47 45 38 34 6
Hub ratio 0.4 to 0.6 49 43 46 43 41 36 30 28 6
Hub ratio 0.6 to 0.8 53 52 51 51 49 47 43 40 6
Tube-axial
Over 40-in. wheel diameter 51 46 47 49 47 46 39 37 7
Under 40-in. wheel diameter 48 47 49 53 52 51 43 40 7
Propeller
General ventilation 48 51 58 56 55 52 46 42 5
Note: Includes total sound power level in dB for both inlet and outlet. Values are for fans only—not packaged equipment.
AF - airfoil blades
BC - backward-curved blades
BI - backward-inclined blades
BFI - blade frequency increments
Source: ASHRAE Handbook 1991, HVAC Applications. Reprinted with permission.
Case Study—Selection of Fans
Select the appropriate type of fan for the following applications:
1. The weaving room of a textile mill
2. A convenience store
3. A high-rise office building
Solution
1. In the weaving room of a textile mill, the sound power level of the looms is far greater than the
sound power level of the fan noise transmitted through the ductwork and supply and return
grilles. Also the fan total pressure required for the air system used in the weaving room is usually
less than 3 in. WC (750 Pa). Therefore, a tube-axial or vane-axial fan would be suitable.
2. A convenience store usually requires an air system with a volume flow rate less than 5000 cfm
(2360 L/s). Usually, a small or a medium-size rooftop packaged unit is used. Forward-curved
fan is usually used in this type of packaged unit.
3. A high-rise office building may use a central hydronic VAV system or a unitary packaged VAV
system. Design criteria also vary considerably according to the location, occupancy, and design
criteria of the building. A more detailed and comprehensive analysis must be made to determine
whether backward-curved centrifugal fans with airfoil blades, forward-curved fans, vane-axial
fans, or a combination of fans should be used.
15.8 COIL CHARACTERISTICS
Coils are indirect-contact heat exchangers that transfer heat between air and another medium (such
as water, refrigerant, steam, or brine) for the purpose of heating, cooling, dehumidifying, or doing a
15.32 CHAPTER FIFTEEN
TABLE 15.2 Comparison between Various Types of Fans
Backward, airfoil Forward-curved
centrifugal fan centrifugal fan Vane-axial Propeller fan
Fan total pressure ptf Higher pt Comparatively Higher pt Low pt
lower pt
Flow rate All flow rates Larger flow rate All flow rates Larger flow rate
Fan power input Nonoverloading Overloading Nonoverloading Nonoverloading
Fan modulation Inlet vanes Inlet vanes Controllable pitch
AC inverter AC inverter AC inverter
Fan total efficiency 0.7 to 0.86 0.6 to 0.75 0.7 to 0.88 0.45 to 0.6
Sound power level Lower, higher Lw Medium, higher Lw Medium, difference Higher, higher Lw
at low frequencies at low frequencies of Lw values is at high frequencies
small at various Hz
Airflow direction 90° turn 90° turn Parallel to axle Parallel to axle
Volume and weight Greater Less Greater Medium volume and
lower weight
Initial cost Higher Medium Higher Low
Applications Large HVAC&R Lower pressure, Large HVAC&R Low-pressure, highsystems
small HVAC&R systems volume flow
systems exhaust systems
combination thereof. Chilled water and refrigerants used to cool and dehumidify moist air are
called coolants. Coils consist of tubes and external fins arranged in rows along the airflow to
provide greater surface contact. Coils are arranged in circuits to create a shorter path length. In a
coil, water and refrigerant flow inside the tubes, and air flows over the outside surface of the tubes
and fins.
Types of Coils
Coils can be classified into four categories according to the medium used.
Water Cooling Coil. A water cooling coil uses chilled water as the coolant inside the tubes. The
chilled water cools or cools and dehumidifies the moist air that flows over the external surface of
the tubes and fins, as shown in Fig. 15.25a. To maintain a higher rate of heat transfer, the air and
water normally follow a counterflow arrangement; i.e., coldest air meets the coldest water and the
warmest air meets the warmest water.
The water tubes are usually copper tubes of - to -in. (13- to 16-mm) diameter with a thickness
of 0.01 to 0.02 in. (0.25 to 0.5 mm). They are spaced at a center-to-center distance of 0.75 to
1.25 in. (19 to 31 mm) longitudinally and 1 to 1.5 in. (25 to 38 mm) transversely. The tubes may be
arranged along the airflows in 2, 3, 4, 6, or 8 rows, in either staggered or aligned form. The staggered
arrangement provides better heat transfer and a higher air pressure drop. Chilled water coils
are commonly rated at a pressure of 175 to 300 psig (1205 to 2070 kPag). Figure 15.26 shows the
structure of a water cooling coil and a DX coil.
Direct-Expansion (DX) Coil. In a direct-expansion coil, the refrigerant (usually HCFC-22, HFC-
134a, HFC-404A, HFC-410A, HFC-407A, or HFC-407C) is evaporated and expanded directly inside
the tubes to cool and dehumidify the air flowing over it, and condensation occurs on the outer
surface of the DX coil, as shown in Fig. 15.25b. This is why it is called a DX coil or wet coil. A DX
coil acts as the evaporator in a refrigerating system. In a DX coil, coolant or the refrigerant is fed into
a distributor and is evenly distributed to various tube circuits, which are made of copper tubes, typically
0.5 in. (13 mm) in diameter. Refrigerant distribution and loading to various circuits are critical
to the performance of a DX coil. After evaporation, the vapor refrigerant is discharged from the
header to the suction line. The characteristics and performance of a DX coil are covered in Sec. 10.2.
Water Heating Coil. The water heating coil shown in Fig. 15.25c is similar in construction to the
water cooling coil. There are two main differences between them. Hot water, instead of chilled water,
is used as the heating medium in a water heating coil. Also, there are fewer rows in the water
heating coil than in the water cooling coil. Generally, only two-, three-, or four-row water heating
coils are available on the market. Water heating coils are rated at pressures of 175 to 300 psig (1205
to 2070 kPag) at temperatures up to 250°F (120°C).
Steam Heating Coil. Steam heating coils use the latent heat of condensation released by steam
inside the tubes to heat outside and recirculating air, as shown in Fig. 15.25d. In a standard steam
heating coil, steam enters one end of the coil and condensate comes out the other end. For more
even distribution of steam, a baffle plate is often installed just after the inlet.
In a steam heating coil, it is important that the coil core inside the casing expand and contract
freely. The coil core is also pitched toward the return connection to facilitate drainage of the condensate.
Steam heating coils usually have a rating of 100 to 200 psig (690 to 1380 kPag) at 400°F
(205°C). Steam heating coil tubes are usually made of copper, steel, or stainless steel.
Fins
Fins are extended surfaces that are often called the coil’s secondary surface (the outer surface of
the tubes is called the primary surface of the coil). Fins are often made of aluminum, with a fin
58
12
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.33
15.34 CHAPTER FIFTEEN
FIGURE 15.25 Types of coils: (a) Water cooling; (b) direct-expansion (DX); (c) water heating; (d ) steam heating.
thickness Ft of 0.005 to 0.008 in. (0.13 to 0.2 mm). They can also be directly extruded, or shaved,
from the parent tubes. Copper, steel, or stainless-steel fins are sometimes used.
Three types of fins are widely used today:
 Continuous plate fins and corrugated plate fins, as shown in Fig. 15.27a and b. Plate fins are usually
made of aluminum.
 Smooth spiral and crimped spiral fins, used in many commercial and industrial applications, as
shown in Fig. 15.27c and d. Smooth spiral fins may be extruded from the parent tubes.
 Spined pipes, as shown in Fig. 15.27e. Spined pipe is made by shaving the spines and lifting them
from the parent tubes.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.35
FIGURE 15.26a Structure of a water cooling coil. (Source: York International Corporation. Reprinted with
permission.)
Corrugated and crimped fins create large turbulence and reduce the thickness of the boundary
layer of the airstream. Consequently, their convective heat-transfer coefficient is increased signifi-
cantly. However, corrugated and crimped fins also create a greater air-side pressure drop than plate
fins do. Various manufacturers have developed rippled corrugated fins to accommodate the higher
rate of heat transfer and the allowable air-side pressure drop.
Because spines are part of a spined pipe, they are firmly connected to the primary surface and
there is no contact resistance between the spines and tube. According to Holtzapple and Carranza
(1990), heat transfer of spine pipe per unit fan power energy is slightly better than that of plate fins
with smaller tube spacing St and small tube diameters.
Fin spacing Sf, in in. (mm), is the distance between two fins. Fin density is usually expressed in
fins per inch, and it often varies from 8 to 18 fins / in.(1.4- to 3-mm fin spacing) for coils used in air
15.36 CHAPTER FIFTEEN
FIGURE 15.26b Structure of a DX coil. (Source: York International Corporation. Reprinted with permission.)
conditioning systems. Fin spacing has a definite effect on the extended surface area and, therefore,
the capacity and air-side flow resistance of the coil.
For a finned-tube coil, finned width indicates the width of the continuous plate fin, as shown in
Fig. 15.25a, and fin length is the length of the coil with plate fins.
Contact Conductance
Thermal contact conductance between the fins and tube in a coil hcc, in Btu/hft2 °F (W/m2 °C),
is the steady heat flux across the fin-tube bonding per °F (°C) temperature difference. Contact
conductance is a function of contact pressure at the fin-tube bonding, fin thickness, fin density, tube
diameter, and hardness of the tube. It directly affects the heat-transfer capacity of a finned-tube coil.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.37
FIGURE 15.27 Types of fins: (a) Continuous plate fins; (b) corrugated
plate fins; (c) smooth spiral fins; (d ) crimped spiral fins; (e) spined pipe;
( f ) fin collar and tube bondings.
Contact between the fin and tube is a critical factor to the performance of the coil. Bonding
between the fin and tube is usually formed by mechanically expanding the tube from inside, using
a tool called a bullet to produce an interference between the fin collar and the tube, as shown in
Fig. 15.27f. Interference I, in in. (mm), is the difference between the expanded tube diameter and
the fin collar diameter before expanding. It can be also expressed as
I  Dt 2Wtk 	 Dfc (15.23)
where Dt  outside diameter of expanding tool, in. (mm)
Wtk  tube wall thickness, in. (mm)
Dfc  inner diameter of fin collar, in. (mm)
According to Wood et al. (1987), the interference of the coils from six manufacturers ranges from
0.003 to 0.0075 in. (0.08 to 0.2 mm).
15.38 CHAPTER FIFTEEN
FIGURE 15.28 Water circuits for water cooling coils. (Adapted from catalog of McQuay HVAC. Reprinted
with permission of Snyder-General Corp.)
Contact conductance is directly proportional to the magnitude of interference. Interference
determines the quality of mechanical bonding during manufacturing. A higher interference means a
better quality of mechanical bonding. Contact conductance increases as the thickness of the fin
increases. It also increases as fin density (fins per inch) increases. Contact conductance decreases
as tube diameter increases. It also decreases as the hardness of the copper tube increases in an
aluminum fin and copper tube coil.
Water Circuits
In a water cooling or water heating coil, tube feeds or water circuits determine the number of water
flow passages. The greater the finned width, the greater the number of tube feeds and thus the
greater the number of flow passages. For two finned-tube coils of the same finned width, a difference
in the number of tube feeds or water circuits means that the water flow rate, in gpm (L / s),
number of passes, and pressure drop of the chilled water, in ft WC (Pa), inside the two coils are different.
One pass means that water flows through the coil’s finned length once.
The number of serpentines ( , , , 1, , or 2) of a water cooling or a water heating coil
indicates its water flow arrangement. The greater the number of serpentines, the larger the total
cross-sectional area of the water circuits and the greater the water volume flow rate. Figure 15.28
shows five water cooling coils, each specified by the number of serpentines, water circuits, passes,
and rows, all made by the same manufacturer. In Fig. 15.28, serpentine means that at the first
row, there are eight tubes across the finned width, but only four of them are tube feeds that are connected
to the return header. For a full serpentine coil, all eight tubes in the first row are tube feeds
and connect to the return header.
15.9 SENSIBLE COOLING AND SENSIBLE HEATING
COILS—DRY COILS
Heat Transfer in a Sensible Cooling Process
In a sensible cooling process, the humidity ratio w is always constant. A sensible cooling process
only exists when the outer surface temperature of the coil Ts, in °F (°C), is equal to or higher than
the dew point of the entering air Tae, in °F (°C), that is, Ts  Tae. A sensible cooling process is
indicated by a horizontal line toward the saturation curve on the psychrometric chart, as shown in
Fig. 15.29.
If heat conduction through the coil framework is negligible, for a sensible cooling process
occurring at steady state within a certain time interval, the rate of sensible heat transfer from the
conditioned air Qcs (in Btu/ h or W) must be equal to the heat absorbed by the chilled water in the
tubes. This relationship can be expressed as
(15.24)
where  volume flow rate of conditioned air, cfm [m/ (60 s)]
a  air density, lb/ ft3 (kg/m3)
Tae, Tal  temperature of conditioning air entering and leaving sensible cooling coil, °F (°C)
cpa, cpw  specific heat of moist air and water, Btu / lb°F (J /kg°C)
Ao  total outside surface area of coil, ft2 (m2)
Uo  overall heat-transfer coefficient based on outside surface area of coil, Btu/hft2 °F
(W/m2 °C)
V?
a
 60m?w cpw(Twl 	 Twe)
 AoUo Tm  Fs Aa NrUo Tm
Qcs  60V? a acpa(Tae 	 Tal)
12
112
34
12
14
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.39
Fs  coil core surface area parameter
Aa  face area of coil, ft2 (m2)
Nr  number of rows in coil
Tm  log-mean temperature difference, °F (°C)
mass flow rate of chilled water, lb/min [kg/ (60 s)]
Twl, Twe  temperature of chilled water leaving and entering coil, °F (°C)
Because , here indicates the volume flow rate of chilled water, in cfm [m / (60 s)],
and w  water density, which can be estimated as 62.32 lb/ ft3. Also, 1 gal  0.1337 ft3 (3.785 L)
and cpw  1, so
(15.25)
where volume flow rate of chilled water, gpm (L/ s)
Coil core surface area parameter Fs is defined as
(15.26) Fs 
Ao
AaNr
V?
gal 
 500V?gal(Twl 	 Twe)
 60  0.1337V?gal
w cpw(Twl 	 Twe)
Qcs  60V?w w cpw(Twl 	 Twe)
V?
w m?w  V?w w
m?w 
15.40 CHAPTER FIFTEEN
FIGURE 15.29 Dry coils at design and part-load operation.
If the thermal resistance of copper tubes is ignored, the overall heat-transfer coefficient based on the
outside surface area of the coil Uo and the inner surface area Ui, both in Btu/h ft2 °F (W/m2 °C),
can be calculated as
(15.27)
In Eq. (15.27), the finned surface efficiency s is given as
(15.28)
where ho, hi  heat-transfer coefficient of outer surface of coil (including fins and tubes) and inner
surface of tubes, Btu/hft2 °F (W/m2 °C)
B  ratio of outside surface area Ao of coil to inner surface of tubes Ai, both in ft2 (m2)
(Ao /Ai)
Af  area of fins, ft2 (m2)
f  fin efficiency
Surface Heat-Transfer Coefficients
McQuiston (1981) developed the correlation between Chilton-Colburn j factors and parameter JP
for dry coils as follows:
(15.29)
(15.30)
where Gc  mass velocity at minimum flow area, lb /h ft2 (kg/hm2)
vmin  air velocity at minimum flow area, fpm (m/ s )
Ap  primary surface area, or tube outer surface area, ft2 (m2)
Pr  Prandtl number  cp /k; for air at temperature lower than 200°F, its value can be estimated
as 0.71
 dynamic viscosity of fluid, lb/ ft h (kg/m s)
cp  specific heat of fluid, Btu / lb°F (J /kg°C)
k  thermal conductivity of fluid, Btu /h ft  °F (W/m°C)
ReD  Reynolds number based on diameter D
The outer surface heat-transfer coefficient ho calculated from Eq. (15.29) is for smooth plate fins.
For corrugated fins, ho should be multiplied by a factor Fcor of 1.1 to 1.25 to account for the increase
in turbulence and rate of heat transfer.
Gc  60 vmin
ReD 
GcDc

JP  ReD
0.4  Ao
Ap 	0.15
js   ho
Gc cpa  Pr 2/3  0.00125 0.27JP
s  1 	
Af
Ao
(1 	
f)
Ui 
1
1/( shoB) 1/hi
Uo 
1
1/( sho) B/hi
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.41
For hot and chilled water at turbulent flow inside the tubes, the inner surface heat-transfer coeffi-
cient hi, in Btu/hft2 °F (W/m2 °C), can be calculated by the Dittus-Boelter equation:
(15.31)
where Di  inner diameter of tubes, ft (m)
kw  thermal conductivity of water, Btu /hft°F (W/m°C)
When the tube wall temperature Tt Tbulk, n  0.4; and when Tt  Tbulk, n  0.3. Here Tbulk indicates
the bulky water temperature beyond the boundary layer. All fluid properties in ReD and Pr
should be calculated at a mean temperature Tm  (Tt Tbulk)/2.
Fin Efficiency
Because the fin temperature is not uniform, fin efficiency f is used to calculate the rate of heat
transfer of a heat exchanger with fins. Fin efficiency is defined as
(15.32)
For a finned-tube coil with rectangular plate fins, fin efficiency can be calculated as recommended
in ASHRAE Handbook 1981, Fundamentals,
(15.33)
(15.34)
where f, max  maximum fin resistance number, dimensionless
Rf, Rf, max  R value and maximum R value of fin, h ft2 °F/Btu (m2 °C/W)
Ft  fin thickness, ft (m)
kf  thermal conductivity of fin material, Btu/h ft °F (W/m°C)
In Eq. (15.33), W represents a width dimension, as shown in Fig. 15.27a, in ft (m).
Figure 15.30 shows the relationship between f, max and the ratio W/ro at various SL/W values.
Here ro indicates the outer radius of the tube, and SL is the longitudinal spacing of the tubes, as
shown in Fig. 15.27b. By assuming that f / f, max  0.9 and then Rf /Rf, max  0.9, fin efficiency f
can be determined.
Dry Coil Effectiveness
As with the DX coil (wet coil) discussed in Sec.10.2, the dry coil effectiveness ?dry can be calculated as
(15.35)
The heat capacity rate of the hot fluid (moist air) Ca, in Btu/h°F (W/°C), in a sensible cooling
coil is
(15.36) Ca  60m?a cpa  60V?a a cpa

Cw(Twl 	 Twe)
Cmin(Tae 	 Twe)
dry 
Ca(Tae 	 Tal)
Cmin(Tae 	 Twe)
Rf 
1
ho
1
f 	 1
f, max 
Rf,max Ftkf
W2
f 
actual heat transfer
heat transfer when fins are at base temperature
hiDi
kw
 0.023 ReD
0.8 Pr n
15.42 CHAPTER FIFTEEN
The heat capacity of the cold fluid (chilled water) Cw, in Btu/h°F (W/°C), is
(15.37)
In Eq. (15.35), Cmin is the smaller value of Ca and Cw, in Btu/h°F (W/ °C).
In a sensible cooling coil using chilled water as coolant, is usually smaller than .
Then
(15.38)
In Eq. (15.38), Tal is usually the required off-coil (leaving coil) temperature during the design of a
sensible cooling system. Most finned coils are actually a combination of counterflow and crossflow
arrangements. Because the effectiveness of a finned coil with a counterflow arrangement is nearly
the same as that of a crossflow arrangement, the effectiveness can be calculated as
(15.39)
where C  Cmin /Cmax
Cmax  greater value of Ca and Cw, Btu/h°F (W/°C)
The number of transfer units (NTU), dimensionless, actually indicates the size of the coil and can
be calculated as
(15.40) NTU 
UA
Cmin

UoAo
Ca

UiAi
Ca
dry 
1 	 exp[-NTU(1 	 C)]
1 	 C exp[-NTU(1 	 C)]
dry 
Tae 	 Tal
Tae 	 Twe
m?wcpw m?acpa
Cw  60m?w cpw  500V?gal cpw
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.43
FIGURE 15.30 Maximum fin resistance number fmax of rectangular
fins. (Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with
permission.)
After calculation of ? of the sensible cooling coil from Eq. (15.39), the sensible cooling coil load
can be found as
(15.41)
From the energy balances between the moist air and chilled water, the temperature of conditioned
air and chilled water leaving the coil Tal and Twl, in °F (°C), can be calculated as
Tal  Tae  ?(Tae  Twe)
Twl  Twe  C(Tae  Tal) (15.42)
Heating Coils
In heating coils, only sensible heat change occurs during the heating process. The humidity ratio w
remains constant unless humidifiers are installed before or after the heating coil. For steam heating
coils, the steam temperature Tsteam  Tae. At the same time, Tsteam remains constant during the
heating process. Therefore, Cmax equals infinity, and Cmin /Cmax  0.
The system performance of both water heating and steam heating coils can also be determined from
Eqs. (15.36) through (15.42), in which the hot fluids are hot water or steam, and moist air is cold fluid.
Fluid Velocity and Pressure Drop
Air velocity calculated based on the face area of a finned-tube coil va is a primary factor that determines
the effectiveness of heat transfer, carryover of droplets of condensate in wet coils, air-side
pressure drop, and energy consumption of the system. For dry coils, there is no danger of carryover.
Their maximum face velocity is usually limited to a less than 800 fpm (4 m/ s). The maximum air
velocity calculated based on the minimum free flow area may be as high as 1400 fpm (7 m/ s). For
coils used in terminal units such as floor-mounted fan coil units, the face velocity is usually around
200 fpm (1 m/ s) so that the pressure drop across the coil is low.
In addition to the face velocity va, the air-side pressure drop pa depends also on fin-and-tube
configuration. For dry coils with a fin spacing of 12 fins/in. (2.1-mm fins), and at a face velocity
va  600 fpm (3 m/ s), pa may vary from 0.1 to 0.2 in. WC (25 to 50 Pa) per row depth.
Selection of water velocity and , in gpm (L / s), for a finned-tube coil of a given tube inner diameter
and number of water circuits is closely related to the temperature rise or drop. For a finnedtube
coil with a face area of 1 ft2 (m2), a full serpentine four-row coil may need 4.8 gpm (0.303 L / s)
of chilled water at a water temperature rise around 10°F (5.6°C), whereas a serpentine coil needs
only 2.4 gpm (0.151 L / s) and has a temperature rise of 20°F (11.1°C). Heat transfer, pressure drop,
erosion, noise, energy, maintenance space, and initial cost should be considered during the selection
of the temperature rise or drop and its corresponding water volume flow and water velocity. For
finned-tube coils, a temperature rise of 10 to 20°F (5.6 to 11.1°C) is generally used. Water-side
pressure drop is usually limited to 10 psi, about 22.5 ft WC (69 kPa). A water velocity between 2
and 6 ft / s (0.6 and 1.8 m/ s) and a pressure drop of 10 ft WC (30 kPa) should be maintained for reasonable
pump power consumption.
Dry Coil at Part-Load Operation
Because the sensible cooling and heating processes are horizontal lines on the psychrometric chart,
if the entering chilled water temperature or hot water temperature remains the same at part-load as
in full-load operation, then the reduction of water flow rates at part-load operation tends to shorten
the horizontal lines from el to elp (as shown in Fig. 15.29) to maintain the required space temperature
at part-load operation.
V?
gal
1
2
V?
gal
Qcs  60V?a 	acpa(Tae  Twe)
15.44 CHAPTER FIFTEEN
Coil Construction Parameters
Coil construction parameters of sensible cooling and heating coils (dry coils) are similar to those in
DX coils (wet coils) covered in Sec. 10.2.
Example 15.1. A counterflow sensible cooling coil using corrugated plate fins has the following
construction and operating data:
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.45
10 ft2 (0.93 m2)
12 fins / in. (2.1-mm fin spacing)
6 rows
0.528 in. (13.4 mm)
0.496 in. (12.6 mm)
90°F (32.2°C)
60°F (15.6°C)
600 fpm (3 m/ s)
28 gpm (1.77 L / s)
0.243 Btu/ lb	°F (1.02 kJ/kg	°C)
1 Btu/ lb 	°F (4.187kJ/kg	°C)
0.075 lb/ ft3 (1.2 kg/m3)
62.4 lb / ft3 (998 kg/m3)
0.0437 lb/ ft 	h (0.018 mPa	 s)
2.46 lb / f t 	h (1.02 mPa	s)
0.348 Btu/h	 ft2 	°F (0.602 W/m	°C)
Face area (finned length finned height)
Fin spacing
Number of rows
Outside diameter of copper tubes
Inside diameter of copper tubes
Entering dry-bulb temperature
Entering chilled water temperature
Face air velocity
Chilled water flow rate
Specific heat: moist air
water
Air density
Water density
At a temperature of 65°F (18.3°C):
Air dynamic viscosity 
a
Water dynamic viscosity 
w
Water thermal conductivity k
Calculate
1. Outer surface heat-transfer coefficient of coil
2. Inner surface heat-transfer coefficient of coil
3. Fin surface efficiency
4. Effectiveness of sensible cooling coil
5. Sensible cooling coil load
6. Off-coil air temperature and temperature of chilled water leaving the cooling coil
Solution
1. For corrugated fins, use a multiplying factor Fc,a  1.15. Then, from Table 10.1, parameter
Fs  14.21. For a six-row coil with a surface area of 10 ft2, the outer surface area is
Ao  14.21 1.15 6 10  981 ft2
From Table 10.1, Ao/Ai  11.4 1.15. The inner surface area of this sensible cooling coil is calculated
as
Ai 
981
11.4 1.15
 74.9 ft2
For a fin spacing of 12 fins / in., Amin /Aa  0.58. Therefore, from Eq. (15.30), the air mass velocity
at minimum flow area is
and the Reynolds number, based on the outside diameter of the tubes, is
Also parameter JP is calculated as
From Eq. (15.29), at JP  0.0231,
For smooth fins,
For corrugated fins, a multiplying factor Fcor  1.20 is used to account for the increase in turbulence.
Therefore, the outer surface heat-transfer coefficient is calculated as
2. Assume a water velocity of 4.7 ft/s inside the copper tubes. Then the Reynolds number of water,
based on the inner diameter of copper tube, is
And the Prandtl number for water is
From Eq. (15.31), the inner surface heat-transfer coefficient can be calculated as
3. For a coil with a longitudinal spacing equal to 1.083 in., ST  1.25 in., ro  0.528/2 in., kf  100
Btu/h	ft	°F, copper tubes in staggered arrangement, W  0.625 in., and SL  0.542 in. From
Fig. 15.29, at W/ro  0.625/0.264  2.37 and SL/W  0.87, 
f, max is about 0.5. Then, from Eq. (15.33)
 0.0271
Rf, max 
f, max W2
Ft kf

0.5(0.625)2
0.006 100 12
 0.023  0.348 12
0.496  (17,740)0.8(7.07)0.4  1061 Btu/h	 ft2	F (6019 W/m2	C)
hi  0.023 k
Di ReD
0.8 Pr 0.4
Pr 

cpw /k  2.46 
1
0.348
 7.07
ReD 
	vwDi

w
 62.4 4.7 3600 
0.496
2.46 12
 17,740
ho  11.92 1.2  14.3 Btu /h	 ft2	F (812 W/m2 	C)
ho, s  0.00748 4655 0.243
1
0.71
 11.92 Btu /h	 ft2	F
ho
Gc cpa
Pr 2/3  0.00125  0.27 0.0231  0.00748
JP  ReD
0.4(Ao /Ap)0.15  (4687)0.4(13.27)0.15  0.0231
ReD 
GcDc


 4655 
0.528
0.0437 12
 4687
Gc  60 	vmin  60 0.075 
600
0.58
 4655 lb / ft2 	h
15.46 CHAPTER FIFTEEN
And from Eq. (15.34),
Assume f / f, max  0.9, then Rf  0.9 0.0271  0.0244. Therefore, fin efficiency is calculated
as
From Eq. (15.28), fin surface efficiency is calculated as
4. From given data, heat capacity rates are calculated as
Minimum to maximum ratio is
Then, from Eq. (15.27),
From Eq. (15.40), the number of transfer units is
Also,
From Eq. (15.39), the effectiveness of the sensible cooling coil is
5. Then, from Eq. (15.41), the sensible cooling coil load is
 60 600 10 0.075 0.243(90  60)0.6834  134,522 Btu/h (38,415 W)
Qcs  60V?a
	acpa (Tae  Twe)

1  0.4657
1  0.4374 0.4657
 0.6834
 
1  exp [-NTU (1  C)]
1  C exp[-NTU(1  C)]
exp[-NTU(1C)  exp[1.438(1  0.4686)]  0.4657
NTU 
Uo Ao
Ca

9432
6561
 1.438

1
1/(0.763 14.3 981)  1/(1061 74.9)
 9432 Btu / h	F
UoAo 
1
1/( shoAo)  1/(hiAi)
C 
Ca
Cw

6561
15,000
 0.4374
Cw  500V?galcpw  500 30 1  15,000 Btu / h	F
Ca  60V?a 	a cpa  60 600 10 0.075 0.243  6561 Btu / h	F
s  1 
Af
Ao
(1 
f)  1  0.913(1  0.741)  0.763
f 
1
1  14.3 0.0244
 0.741
f 
1
1  hoRf, max

1
1  14.3 0.0271
 0.721
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.47
6. From Eq. (15.42), the temperature of conditioned air leaving the sensible cooling coil is
The temperature of chilled water leaving the coil is
15.10 WATER COOLING COILS—DRY-WET COILS
As in a DX coil, the airflow and water flow arrangement in a water cooling coil is usually a combination
of counterflow and crossflow. When chilled water is used as the coolant flowing inside the
copper tube of a cooling coil, the outer surface temperature of the coil at the air intake side Tse is
probably greater than the dew point of the entering air Tae, that is, Tse  Tae. The outer surface
temperature of the coil at the air leaving side Tsl  Tae. A water cooling coil is then a dry-wet
coil. It often has a partially dry surface at the air intake side and partially wet surface at the air
leaving side.
When the temperature of chilled water entering the coil Twe is low, or the dew point temperature
of the entering air Tae is high, and if Tse  Tae, the water cooling coil is then a wet coil. The differences
between a wet water cooling coil and a wet DX coil are mainly the lack of a superheated
region in a water cooling coil and the variation of chilled water temperature inside a water cooling
coil, which is often greater than the variation of evaporating temperature Tev at the two-phase region
in a DX coil.
Dry-Wet Boundary
In a dry-wet coil, there is always a dry-wet boundary that divides the coil between the dry surface
and the wet surface, as shown in Fig. 15.31. The dry-wet boundary must be determined so that the
performance of the dry surface and the wet surface can be calculated separately.
Because the outer surface temperature of the copper tubes at the dry-wet boundary Tsb is exactly
equal to the dew point of the air entering the coil Tae, that is, Tsb  Tae, the dry-wet boundary can be
determined. The dry surface extends from the dry-wet boundary to the air intake side, and the wet
surface extends to the air leaving side.
At the dry-wet boundary, if the thermal resistance of the metal tube and the heat energy released
by condensate are ignored, then
(15.43)
where Tab  temperature of airstreams at dry-wet boundary,°F (°C)
Twb  temperature of chilled water inside tubes at dry-wet boundary,°F (°C )
hi  inner surface heat-transfer coefficient, Btu/h	ft2 	°F (W/m2	°C)
In Eq. (15.43), hwet is the heat-transfer coefficient of the wetted surface, in Btu/h	 ft2 	°F
(W/m2 	°C), as discussed in Sec. 10.8. From Eq. (10.21), hwet  mho /cpa. Here ho is the heattransfer
coefficient of the outer surface, in Btu/h	 ft2 	°F (W/m2 	°C); m is the slope of air enthalpy
saturation curve, m  dhs /dTs, as shown in the upper right-hand corner of Fig. 15.31; hs denotes
the enthalpy of saturated air film, in Btu/ lb (J /kg); and Ts is the saturated temperature, in °F (°C).
At To,s  60°F (15.6°C), the ratio hwet /ho  m /cpa is about 2.8.
In Eq. (15.43), there are two unknowns, Tsb and Twb. An additional equation is needed to determine
them. At the dry-wet boundary of the dry-wet coil, there exists an energy balance between the
hwet s Ao(Tab  Tsb)  hiAo
B (Tsb  Twb)
 60  0.4686 (90  69.5)  69.6F (20.9C)
Twl  Twe  C(Tae  Tal)
Tal  Tae  (Tae  Twe)  90  0.6834(90  60)  69.5F
15.48 CHAPTER FIFTEEN
airstream and water stream
(15.44)
where  volume flow rate of conditioned air, cfm
 volume flow rate of chilled water, gpm
Because the temperature of chilled water leaving the coil Twl is a known value, from Eqs. (15.43)
and (15.44), Tab and Twb can be determined.
Dry Part
The sensible cooling capacity Qcs, in Btu/h (W), of the dry part of the dry-wet coil can be calculated
from the left-hand side of Eq. (15.44) as
(15.45)
The average temperature of the airstream in the dry part Tad, m, in °F (°C), is approximately equal to
(Tae  Tab), and the average temperature of the chilled water corresponding to the dry part of the 12
Qcs  60V?a
	a cpa(Tae  Tab)
V?
gal
V?
a
V?
a 	a cpa(Tae  Tab)  8.33V?galcpw(Twl  Twb)
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.49
FIGURE 15.31 Dry-wet boundary of a dry-wet coil.
coil Twd, m, in °F (°C), is (Twb  Twl). Unit sensible cooling capacity of each ft2 (m2) outer surface
area of the dry part of the water cooling coil, including the tube and fins, can then be calculated as
Because Uo can be calculated from Eq. (15.27), the total outer surface area of the dry part of the
coil Ao, d, in ft2 (m2) is
(15.46)
Wet Part
As for the DX coil, the cooling and dehumidifying capacity of the wet part of a dry-wet coil Qc, w,
in Btu/h (W), can be calculated as
(15.47)
where hs, we  enthalpy of saturated air film at chilled water entering temperature,°F (°C). The effectiveness
of the wet part of the dry-wet coil (wet, the number of transfer units (NTU), and the
heat capacity rates Ca and Cw can be similarly calculated by Eqs. (15.39), (15.40), (15.36), and
(15.37), respectively.
Dry-Wet Coil at Part-Load Operation
When the dry-wet coil of a constant-volume system is at part-load operation, the reduction of the
space cooling load results in a drop in space temperature. This causes the control system to modulate
the two-way valve at the inlet of the water cooling coil to reduce the mass flow rate of chilled
water flowing through the coil, as shown in Fig. 15.32. Because of the reduction of the chilled water
flow rate, the following conditions occur:
 The velocity of chilled water inside the coil tubes is reduced, and therefore the rate of heat-transfer
between the air and water decreases.
 Although the reduction of the space cooling load tends to lower the temperature of conditioned
air leaving the coil Tal, nonetheless Tal increases because of the reduction of the chilled water flow
rate and has a greater influence than the drop of the space cooling load.
 From the psychrometric analysis shown in Fig. 15.32, the space conditioning line sp-rp at partload
operation is shorter because of the reduction of space cooling load.
 The chilled water flow rate is reduced so that the off-coil temperature Tccp and supply temperature
Tsp match the reduction of space cooling load and maintain a space temperature Trp at part-load
operation.
 The higher dry-bulb and dew point temperatures of conditioned air leaving the coil cause an
increase in space relative humidity r. Sometimes, r may be raised to 70 percent during partload
operation. This is the main disadvantage in controlling space temperature by means of
chilled water flow rate modulation at part-load operation.
Selection of a Dry-Wet Coil
The construction of a water cooling coil, a dry-wet coil, is similar to that of a DX coil, covered in
Sec. 10.2. As for a DX coil, the procedure for selecting a water cooling coil according to the manufacturer’s
data is as follows:
Qc, w  wet60V?a 	a(hab  hs,we)
Ao, d 
Qcs
Uo (Tad, m  Twd, m)
Qcs
Ao, d
 Uo(Tad, m  Twd, m)
12
15.50 CHAPTER FIFTEEN
1. According to the required supply volume flow rate, choose an optimum face velocity va; it is
usually not greater than 550 fpm (2.75 m/ s) to prevent condensate carryover, wet surfaces and possible
mold growth in air system, and a 500 fpm (2.5 m/ s) or less cooling coil face velocity is preferable.
Calculate the face area Aa of the coil.
2. From the coil load, calculate the unit cooling and dehumidifying capacity per ft2 face area
Qc /Aa.
3. Determine the chilled water temperature entering the coil Twe. This is closely related to the required
condition of air leaving the coil Tal.
4. Select the optimum chilled water temperature rise Tw. A lower Tw means a greater water
flow rate, a higher water velocity in the tubes, and a greater heat-transfer coefficient. A lower Tw
also results in a greater pump power and a higher initial investment in the water system. Water temperature
rise Tw also influences air system and refrigeration system performance. It must be
matched with the expected temperature of conditioned air off the coil in the air system as well as
the chilled water temperature difference in the chiller.
The current practice is to use a Tw from 10 to 20°F (5.6 to 11.1°C) in order to reduce water
flow and save pump power. When Tw is determined, , in gpm (L/ s), can be calculated accordingly.
V?
gal
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.51
FIGURE 15.32 Psychrometric analysis of part-load operation of a dry-wet coil.
5. Check the water velocity inside the tubes vw, in ft / s (m/ s), by selecting an appropriate finned
width for the water cooling coil. Water velocity of 2 to 6 ft / s (0.6 to 1.8 m/s) typically between 3
and 5 ft / s (0.9 and 1.5 m/ s) is appropriate. As in sensible water cooling coils, the water-side pressure
drop of the coil should not exceed 10 psi, or 22.5 ft WC (69 kPa). Typically, a pressure drop of
10 ft WC (30 kPa) across the coil is considered suitable to keep the pump power and energy cost
reasonable. Many manufacturers provide the formula to calculate vw at a given finned width and the
corresponding tube feeds. Some manufacturers also provide water turbulators, which can be added
to the tubes to increase the inner surface heat-transfer coefficient if the water velocity is less than
4 ft / s (1.2 m/ s). In tubes with water turbulators, the water-side pressure drop is considerably higher
than in smooth tubes.
6. Choose the optimum number of rows and fins per inch. An even number of rows are preferable
because the supply and return header can be placed on the same side.
7. Select the coil from the manufacturer’s catalog at given air and water entering and leaving
conditions va and vw and the required Qc /Aa.
8. Determine the air-side and water-side pressure drop.
Selection of the DX coils and water cooling and heating coils is usually performed by following
the procedure given in the manufacturer’s catalog. Calculations by means of dry-wet boundary and
wet-coil effectiveness are sometimes necessary in detailed analysis and simulation in research and
development.
Example 15.2. A water cooling coil in a constant-air-volume system has a supply volume flow
rate of 11,000 cfm (5190 L/ s). Air enters the coil at a dry-bulb temperature of 80°F (26.7°C) and a
wet-bulb temperature of 67°F (27.4°C) and leaves the coil at a dry-bulb temperature of 57.5°F
(14.2°C) and a wet-bulb temperature of 56.5°F (13.6°C). Chilled water enters the coil at 45°F
(7.2°C) and is expected to leave the coil at 55°F (12.7°C). At an average temperature of 50°F
(10°C), the fluid properties are as follows:
15.52 CHAPTER FIFTEEN
0.339 Btu/h	 ft 	°F (0.587 W/m	°C)
3.09 lb / ft 	h (1.28 mPa	 s)
1.0 Btu/ lb	°F (4.187 kJ/kg	°C)
Thermal conductivity of water kw
Dynamic viscosity of water 
w
Specific heat of water cpw
Assume that the air-side heat transfer coefficient ho  13.4 Btu/h	 ft2 	 F (76.1 W/m2 	°C)
1. Select a water cooling coil from the manufacturer’s catalog.
2. Determine the dry-wet boundary of this dry-wet coil.
3. Calculate the sensible cooling capacity and the outer surface area of the dry part of this dry-wet coil.
4. Determine the cooling and dehumidifying capacity of the wet part as well as the dry-wet coil.
5. At part-load operation, if the water velocity in the copper tubes is reduced to 63 percent of the
full-load value, and air enters the coil at a wet-bulb temperature of 68°F (20°C), determine the
operating conditions of this dry-wet coil.
Solution
1. For a corrugated coil, select a face velocity va  550 fpm. From the given values, the required
face area of the coil is
From the psychrometric chart, the enthalpy of entering air at a dry-bulb temperature of 80°F and a
wet-bulb temperature of 67°F is 31.6 Btu/ lb. The enthalpy of leaving air at a dry-bulb temperature
of 57.5°F and a wet-bulb temperature of 56.5°F is 24.2 Btu/ lb. From the given conditions, the total
Aa 
V?
a
va

11,000
550
 20 ft2
cooling and dehumidifying capacity of this dry-wet coil is
Cooling and dehumidifying capacity per ft2 of face area is calculated as
Similarly, the required sensible cooling capacity per ft2 of coil face area is
From the given values, the water temperature rise is 55  45  10°F, and the water volume flow
rate is therefore
2. For a face area of 20 ft2, choose a coil with a finned width of 30 in. and finned length of 96 in.
Assume that the water velocity inside the tubes of this water cooling coil is 4 ft / s.
At an air entering condition of 80°F dry-bulb temperature and 67°F wet-bulb temperature va 
550 fpm, vw  4 ft/s, Twe  45°F, Tw  10°F, and required Qc /Aa  18.3 MBtu/h, a four-row,
12 fins / in. coil selected from manufacturer’s catalog has a cooling and dehumidifying capacity of
19.1 MBtu/h per ft2 of face area and a sensible cooling capacity per ft2 of
Here LDB represents the leaving dry-bulb temperature and LWB the leaving wet-bulb temperature,
both in °F (°C). Both the cooling capacity per ft2 of face area and the sensible cooling capacity per
ft2 of face area of this water cooling coil are greater than the required values.
From the psychrometric chart, the dew point of the entering air at 80°F (26.7°C) dry-bulb temperature
and 67°F (19.4°C) wet-bulb temperature is 60°F (15.6°C), that is, Tsb  60°F (15.6°C).
The slope of the air saturation curve at 60°F (15.6°C) is the difference in saturation enthalpy
between 59.5 and 60.5°F (15.3 and 15.8°C), or
Therefore,
m
cpa

0.69
0.243
 2.84
m  28.81  26.12  0.69 Btu / lb	F (2.89 kJ /kg 	C)
Face velocity, fpm
500 600
Water temperature rise 10°F MBtu/h LDB LWB MBtu/h LDB LWB
Water velocity 4 ft / s 18.1 56.4 55.4 20.0 57.4 56.4
 13,893 Btu/h	 ft2 (43,832 W/m2)
Qcs
Aa
 60 550 0.075 0.24380 
56.4  57.4
2 
V?
gal 
Qc
500(Twl  Twe)

366,300
500 10
 73.3 gpm (4.63 L /s)
 60 550 0.075 0.243(80  57.5)  13,532 Btu/h	 ft2
Qcs
Aa
 60va 	acpa(Tae  Tal)
Qc
Aa

366,300
20
 18,315 Btu/ h or 18.3 MBtu /h
 366,300 Btu/h
Qc  60V?a 	a(hae  hal)  60 11,000 0.075(31.6  24.2)
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.53
Using an air-side heat-transfer coefficient ho  13.4 Btu/h	ft2 	°F as in Example 15.1, from Eq.
(10.21), the heat-transfer coefficient for the wet surface is
3. On the water side,
Then
As in Example 15.1, B  11.4 1.15  13.1, from Eq. (15.43), for 1 ft2 of outer surface area at
the dry-wet boundary,
From Eq. (15.44),
According to the above simultaneous equations,
The dry-wet boundary can therefore be determined as shown in Fig. 15.31.
4. From Eq. (15.45), the sensible cooling capacity of the dry part of the dry-wet coil is
The average temperature of the airstream in the dry part is
The average temperature of the water stream corresponding to the dry part of the dry-wet coil is
Twd,m 
Twl  Twb
2

55  53.3
2
 54.15F
Tad,m 
Tae  Tab
2

80  74.8
2
 77.4F
 11,000 60 0.075 0.243 (80  74.8)  62,548 Btu/h (18,327 W)
Qcs  60V?a
	a cpa(Tae  Tab)
Tab  74.8F and Twb  53.3F (11.83C)
11,000 0.075 0.243(80  Tab)  8.33 74 1(55  Twb)
V?
a 	acpa(Tae  Tab)  8.33V?galcpw(Twl  Twb)
38.06 0.76 (Tab  60) 
838.6
13.1
(60  Twb)
hwet s(Tab  Tsb) 
hi
B
(Tsb  Twb)
 838.6 Btu /h	 ft2 	°F
hi  0.023 ReD
0.8 Pr 0.4  0.023 (12,020)0.8(9.12)0.4
Pr 

w cpw
kw

3.09 1
0.339
 9.12
ReD 
	vwDi

w

62.4 4 3600 0.496
3.09 12
 12,020
hwet 
m
cpa
ho  2.84 13.4  38.06 Btu /h	 ft2 	°F (216 W/m2	C)
15.54 CHAPTER FIFTEEN
From Eq. (15.27),
Then the outer surface area of the dry part is
5. From Eqs. (15.36) and (15.37), the heat capacity rates and their ratio are
From Eq. (15.26) and Table 10.1, for such a four-row, 12 fins / in. corrugated plate fin coil, the total
outer surface area is
The outer surface area of the wet part of the coil is
Then, from Eq. (15.27),
And from Eq. (15.40)
Also exp[-NTU(1  C)]  exp[0.7321(1  0.3251)]  0.6101
Therefore, from Eq. (15.39), the effectiveness of the wet part of the dry-wet coil is

1  0.6101
1  0.3251 0.6101
 0.4864
wet 
1  exp[-NTU(1C)]
1C exp[-NTU(1C]
NTU 
UoAo
Ca

8805
12,028
 0.7321

1
1/(0.76 13.42 1001)  13.11/(838.6 1001)
 8805 Btu/h	F
UoAo 
1
1/( sho Ao)  1/(hiAi)
Awet  Ao  Ao, d  1307  306  1001 ft2 (93 m2)
Ao  Fs AaNr  14.21 1.15 20 4  1307 ft2
C 
Cmin
Cmax

12,028
37,000
 0.3251
Cw  m?wcpw  74 500 1  37,000 Btu/h	F
Ca  m?acpa  60 11,000 0.075 0.243  12,028 Btu/h	F
 306 ft2 (28.4 m2)
Ao, d 
Qcs
Uo(Tad, m  Twd, m)

62,548
8.796(77.4  54.15)
 8.796 Btu /h	 ft2	F
Uo 
1
1/( sho)  B/hi

1
1/(0.76 13.42)  13.11/838.6
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.55
6. From the psychrometric chart, at a Tab of 74.8°F and a dew point of 60°F, hab  30.3 Btu/ lb	
Also the enthalpy of saturated air film hs,we  17.65 Btu/lb when the temperature Twe is 45°F. Then
the cooling and dehumidifying capacity of the wet part of the dry-wet coil is
The cooling capacity of the dry-wet coil is
or 367,119/20  18.4 MBtu/h	 ft2 which is close to the value 19.1 MBtu/h	 ft2 given in the manufacturer’s
catalog.
7. If the water velocity inside the copper tubes is reduced to 63 percent of the full-load value
vw  0.63 4  2.5 ft / s and the water temperature rise is increased to 16°F, from the performance
data shown in manufacturer’s catalog, at an entering dry-bulb temperature of 79°F, entering wetbulb
temperature of 68°F, and an enthalpy of 32.4 Btu/ lb, the cooling capacity is reduced from
19.1 to 15.3 MBtu/h	 ft2. Total cooling and dehumidifying capacity of the water cooling coil at part
load is
The reduction of enthalpy h during the cooling and dehumidifying process at part-load operation is
The enthalpy of the air leaving the coil is
hal  32.4  6.2  26.2 Btu/ lb
At an entering dry-bulb temperature of 79°F and wet-bulb temperature of 68°F, from the cooling
and dehumidifying curve on the psychrometric chart, the condition of air off the coil at part-load
operation is 61.3°F dry-bulb and 59.5°F wet-bulb temperature. The load ratio at part load 
15.3/19.1  0.80.
If the sensible heat ratio and system heat gain at part-load operation are the same as in full-load
operation, the air conditioning cycle at part load can then be drawn on the psychrometric chart, as
shown by the dotted line in Fig. 15.32. The space relative humidity has been increased from 55
percent at full load to 59 percent at 80 percent part-load operation. Air conditioning processes and
cycles are discussed in Chap. 20.
15.11 COIL ACCESSORIES, CONDENSATE CONTROL, AND
FREEZE-UP PROTECTION
Coil Accessories
Coil accessories include air vents, drain valves, isolation valves, flow metering valves, balancing
valves, thermometers, pressure gauge taps, condensate drain taps, and even distribution baffles.
Air Vents. These devices discharge air from the coils. Air vents are usually located at the highest
point of the coil and at the pipe risers. Manual vents are inexpensive and more reliable, whereas
automatic vents are more convenient to operate.
h 
306,000
11,000 60 0.075
 6.2 Btu / lb
Qc  15.3 1000 60 20  306,000 Btu/h (89,660 W)
Qc  Qcs  Qc,w  62,548  304,571  367,119 Btu/h (107,566 W)
 0.4864 11,000 60 0.075(30.3  17.65)  304,571 Btu/h
Qc,w  wetV?a
	a(hab  hs,we)
15.56 CHAPTER FIFTEEN
Drain and Isolating Valves. Drain valves should be provided for each coil for maintenance
and repair. Hose-end type of drain valves may be used for remote drains. Isolating valves
should be installed at both supply and return piping to cut off the water supply during repair and
maintenance.
Even Distributing Baffles. When heating coils are stacked above cooling coils, usually perforated
baffles should be installed in the path with less flow resistance to balance the difference in pressure
losses between two parallel flow paths. Heating coils usually have fewer rows and less flow resistance
than cooling coils.
Coil Cleanliness
Coil cleanliness is important for proper maintenance and operation. Coil cleanliness includes outer
surface cleanliness and internal coil care. If an air filter of a minimum of 25 percent dust-spot effi-
ciency is installed upstream from the coil without leakage air bypassing the air filter, then dirt will
not accumulate on the outer surface of the coil. If an air filter has efficiency lower than 25 percent
dust-spot rating or with many air leaks bypassing the air filter, dirt accumulation may eventually
block the air paths in a coil and significantly increase the pressure drop across the coil. In air
systems, coils usually should be inspected and cleaned every 3 months in urban areas. Blowing out
contaminants with compressed air and flushing with chemical detergent are the two cleaning methods
often used.
A chilled water system is usually a closed water system. Most of the makeup water taken from
the city water supply in the United States is well treated. If the water strainer is periodically inspected
and works properly, the inner surface of the copper tubes of a water cooling coil is usually
not contaminated.
Symonds (1997) pointed out that the internal coil care is needed after a burnout of a compressor
or a compressor motor in a packaged system. In such a condition, the DX coil is often dismantled
into segments. It is preferable to use chemical agent to flush the inner surface of the coil to clean
the existing acid, sludge, and solid particles.
Condensate Collection and Drain System
In DX coils and water cooling coils, air is often cooled and dehumidified in summer. Moisture is
removed from air and forms condensate on the outer surface of the DX coil and the wet part of the
water cooling coil. A condensate collection device must be installed, and a condensate drain line
should be designed. This is to prevent the leakage of condensate and the stagnancy of the condensate
flow which often cause the indoor air quality problems due to the growth of the bacteria and
microorganisms inside the the drain pans. Trent and Trent (1997b) discussed the problems involved
in condensate control.
Condensate Flow. The amount of condensate removed from the conditioned air depends mainly
on the type of air system (whether it is a comfort system, a processing system, a makeup system—
all outdoor air—or a recirculating system) and the location of the air system. If air enters the coil
at 80°F (26.7°C) dry-bulb temperature, 67°F (19.4°C) wet-bulb temperature and a humidity ratio
we  0.0116 lb/lb (0.0116 kg/kg) and leaving the coil at 57°F (13.9°C) dry-bulb temperature, 56°F
(13.3°C) wet-bulb temperature, and a humidity ratio wl  0.00905 lb/lb (0.00905 kg/ kg), for a
cooling and dehumidifying capacity of 1 refrigeration ton (400 cfm or 189 L/ s supply air) in a
comfort system using recirculating air, the mass flow rate of condensate can be calculated as
 0.063 lb /min or 0.0075 gpm (0.00047 L / s)
m? con  V?a 	a (we  wl)  400 0.075(0.0116  0.0095)
m?con
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.57
Drain Pan. Drain pans have been widely installed under the DX and water cooling coils to collect
condensate, as shown in Fig. 15.32. When two or more DX coils are banked vertically, there must
be an individual drain pan and drainpipe for each coil. The drainpipe of the top drain pan should be
connected to the lower main drain pan. Coils are preferably installed vertically to avoid condensate
drips from sloped coils. To prevent the carryover of the water drains down on the coil, drain pans
should be spaced at a maximum of 42 in. (1.07 m) vertically.
The size of the drain pan is nearly the same as the size of the base of the coil except that a drain
pan should extend in the direction of airflow a depth equal to 1.5 to 3 times the depth of the coil for
the installation of the drain port and the inspection of condensate flow. A larger extended drain pan
to protect the floor from condensate carryover is not a correct approach. Condensate carryover is
prevented only by choosing a coil face velocity lower than 550 fpm (2.75 m/ s) for corrugated fins.
Condensate drain pans are preferably made of 16-gauge (1.6-mm-thick) stainless steel and are
usually 2 in. (50 mm) deep. A 1-in. (25-mm) stainless-steel pipe joint should be used as the drain
port at the end of the pan wall from which the condensate drain line is connected. Drain port is
preferably located at the same side as where the chilled water enters the coil. The floor of the condensate
pan should be sloped in./ft (21 mm/m length) toward the drain port. Tests showed that
about in. (3-mm) depth of water in the condensate pan is required to overcome the surface tension
and to allow the condensate to flow freely.
Condensate drain pans should be cleaned monthly to prevent the buildup of bacteria and microorganisms
which contaminate the indoor air.
Condensate Trap. Most of the supply fans in air-handling units (AHUs) and packaged units
(PUs) are located downstream from the coil (a draw-through fan). The coil section in an AHU or
PU with a draw-through fan is always in a negative static pressure ps,coil of 1.5 to 4 in. WG
(375 to 1000 Pag). A condensate trap, as shown in Fig. 15.32, is often used to isolate the coil
section from the outside atmosphere. The height of the water column in the U tube of the condensate
trap, in in. WC, must be 1 or 2 in. (250 to 500 Pa) greater than ps,coil. When there is no
condensate in an idled DX or water cooling coil, the ambient air outside the AHU or PU can be
blocked by the water filled in the U tube. Outside ambient air is not extracted into the AHU or PU
because of the condensate trap. A condensate trap should discharge freely the liquid condensate that
has collected in the drain pan.
Condensate Drain Line. Most condensate drain lines are gravity drain systems. The difference in
water levels between the water surface in the drian pan and the water level in the discharge sump
provides the available head pava to overcome the pressure loss pdrain of the drain line, both in ft
WC (m head). The diameter of the drain line should be determined according to the calculated gpm
(L/ s) of the condensate flow rate pava and pdrain. When the coil and the drain line discharge sump
are on the same floor, the drain pan is sometimes elevated for a greater pava. The diameter of the
drain line should not be less than a minimum size (such as 1-in. or 25-mm diameter) so that it may
not be easily blocked by dirt. The drain line should be sloped away from the condensate trap at a
rate of no less than in. /ft (10 mm/m).
Condensate drain line is often made of PVC, steel, or sometimes copper. PVC or steel pipes
must be schedule 40 or heavier. For PVC pipes, fixed supports should be installed at 2- to 3-ft (0.6-
to 0.9-m) intervals, and steel pipes in 8- to 10-ft (2.4- to 3-m) intervals.
Coil Freeze Protection
One of the most common problems encountered with water and steam coils is coil freeze-up. Water
inside tubes freezes when its temperature is below the freezing point.
Air Stratification. Improper mixing of outdoor and recirculating air in the mixing box of an airhandling
unit or packaged unit often causes coil freeze-up when the outdoor air temperature is below
32°F (0°C). A mixture of 30 percent outdoor air at 0°F (17.8°C) and 70 percent recirculating
1
8
1
8
1
4
15.58 CHAPTER FIFTEEN
air at 70°F (21.1°C) theoretically gives an average temperature of 49°F (9.4°C), which is far above
the freezing point of water. However, such a temperature is only maintained by thorough mixing. If
outdoor air and recirculating air enter the mixing box of an air-handling unit, as shown in
Fig. 15.33a, the outdoor air and recirculating air will not mix thoroughly and will form air stratifi-
cation. Low-temperature outdoor air flows over part of the preheat coil, cooling coil, or even the
reheat coil, and may freeze part of the coil.
Baffles can be used to direct the airstream to promote thorough mixing, as shown in Fig. 15.33b.
The face velocity of the baffles should be between 1000 and 1500 fpm (5 and 7.5 m/ s). Temperature
sensors, installed at the coil face where the lowest air intake temperature may occur, can be
used to shut off the fan and the outdoor air damper when the temperature is below freezing. For
coils installed downstream from a supply fan, a single-inlet fan always provides better mixing than
a double-inlet fan.
Draining Water and Inner Distributor Tubes. Freeze-up of idled chilled water coils can be prevented
by draining the water inside the tube completely. Larocca (1997) recommended to blow air
from the supply fan continuously through the coils to ensure that the tubes inside become completely
dry. For most air-handling units, the fan total pressure produced by a supply fan is sufficient
to overcome the internal resistance of the coil tubes. A 3-in.- (75-mm-) diameter flexible duct is
used to connect the supply branch duct and the coil supply header via a shutoff valve. Compared
with the addition of glycol solution to water to lower the freezing point, this method saves the cost
of glycol and the labor involved in handling, mixing, and disposing of the glycol.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.59
FIGURE 15.33 Air stratification and freeze-up protection: (a) freeze-up due to air stratification; (b)
preventing freeze-up.
If a steam heating coil is installed in a location where the outdoor air temperature may drop below
32°F (0°C) in winter, as described by Delaney et al. (1984), a coil with inner distributor tubes
for freeze-up protection should be used, as shown in Fig. 15.34. Steam condensate can drain from
the outer tubes easily without blocking the steam supply.
Incorporating Freeze Plugs. Larocca (1997) suggested to incorporate freeze plugs in the coil design
(pressure relief ) to relieve the high pressure caused inside the coil tubes, bends, and header
during freezing.
Running the Water Pump. Freeze-up of water cooling coils can be prevented by running the
chilled water pump to provide a certain velocity inside the coil tubes when the outside air
temperature is below freeezing. If a water velocity of 2.5 ft / s (0.75 m / s) is maintained inside
a water cooling coil, the minimum air temperature can drop to 32°F (0°C) without freezing
the coils.
15.12 AIR CLEANING, RATING, AND ASSESSMENTS
Air Filtration and Industrial Air Cleaning
Air cleaning is the process of removing airborne particles present in the air. It can be classified into
two categories: air filtration and industrial air cleaning. Air filtration involves the removal of airborne
particles present in outdoor air as well as recirculated air from a given space. Most airborne
particles removed by air filtration are smaller than 1 
m, and the concentration of these particulates
in the conditioned airstream is often less than 2 mg /m3. The purpose of air filtration is to benefit the
health and comfort of the occupants in the conditioned space as well as maintain the cleanliness
required in manufacturing processes. Air filtration is one of the essential factors that affects the
indoor air quality of the conditioned space.
Industrial air cleaning mainly involves the removal of dust and gaseous contaminants from the
industrial manufacturing processes, and it provides pollution control to the exhaust gas and flue gas.
Air contaminants that are discharged to the outdoor environment are governed by the U.S. Environmental
Protection Agency (EPA), and those that are exhausted to the indoor working space are
regulated by the Occupational Safety and Health Administration (OSHA). The size of airborne
15.60 CHAPTER FIFTEEN
FIGURE 15.34 Steam heating coil with inner distributor tubes for
freeze-up prevention.
particles may range from  1
m to several hundred micrometers. The amount of airborne particles
present in the airstream often varies from several to 40,000 mg/m3.
In this chapter, only air filtration is discussed.
Atmospheric Dust and Space Air Contaminants
The materials other than the oxygen, nitrogen, carbon dioxide, water vapor, and rare gases present
in the air are called air contaminants. They are a mixture of dust, smoke, fumes, fogs, mists, and
living organisms. Dusts are defined as solid granular particles and fibers less than 100 
m (1 
m 
0.00004 in.) derived from natural and mechanical processes. Smoke is a solid, a liquid, or a mixture
of solid and liquid particles of sizes averaging from 0.1 to 0.3 
m. It is the product of incomplete
combustion. Fumes are solid particles smaller than 1 
m formed in the condensation of vapors.
Fogs are very small liquid particles, ranging from 2 to 60 
m, formed by the condensation of the
vapors; and mists are relatively large liquid particles, from 60 to 200 
m, produced by atomizing
and spraying processes. Living organisms include viruses, from 0.003 to 0.06 
m, and bacteria,
usually from 0.4 to 5 
m. Both of these attach themselves to a larger particle for transportation.
Fungal spores and pollens, which derive from living organisms, range from 10 to 100 
m. The mixture
of granular particles, fibers, smoke, fumes, and mists is often collectively called dust or particulates,
and a mixture of air and dust is called an aerosol.
Airborne particles that are smaller than 0.1 
m remain suspended and moving owing to brownian
movement, whereas particles that range from 0.1 to 1 
m have a negligible settling velocity.
Particles between 1 and 10 
m have a constant and significant settling velocity and tend to drop
onto a solid surface. Particles that are larger than 10 
m settle on a surface even more rapidly.
Atmospheric dust denotes dust contained in the outdoor air, or outdoor atmosphere. According to
Sec. 4.10, concentration of particle matter in atmospheric air should meet the EPA Ambient Air
Quality Standard of 260 mg/m3 within a period of 24 h. In atmospheric dust, (1) 99 percent of the
dust particles are smaller than 0.3 
m, and they constitute only 10 percent of the total weight; and
(2) 0.1 percent of the dust particles are larger than 1 
m and make up 70 percent of the total weight.
The dust concentration of an indoor conditioned space may vary considerably according to the
requirements of the space, the sources releasing dust particles, the type of equipment and devices
used in removing the dust particles, and the type of building enclosing the conditioned space. The
concentration of total airborne particles of a conditioned space usually varies, from the very low
particle count in a clean room to about 0.2 mg /m3.
Rating and Assessments
An air cleaner is a device that removes dust particles from the airstream passing through it. An air
filter is an air cleaner that removes dust particles from a conditioned airstream by means of fibrous,
metallic, or other filter media. Filter media are fabricated materials to perform air filtration.
The operating characteristics of an air filter depend on the size, shape, specific gravity, concentration,
and electric properties of the dust particles as well as the filter medium. However, the most
important factor for effective removal is the particle size, or size distributions, in a dust sample. For
air filtration in air conditioning systems, the dust particles are small and their concentrations are
comparatively low.
The rating and assessment of air filters are mainly based on the following characteristics:
 Efficiency. The efficiency of an air filter shows the effectiveness of the dust removal process and,
therefore, is the primary rating index. The efficiency of a specific filter is affected by the size of
the dust particles, dust-holding capacity mdust, or loading, in gr / ft2 (g/m2) and the test methods.
 Pressure drop. Total pressure drop of an air filter pt, in in. WC (Pa), can be calculated as
(15.48) pt  C  vf
40052
 Ca  va
40052
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.61
where C, Ca  local loss coefficients of air filter
vf  air velocity flowing through filter media, fpm
va  face velocity of air filter, fpm
The pressure drop of an air filter is affected by vf or va, as well as the loading of dust held inside the air
filter mdust. The greater vf, va, and mdust, the higher pt. At the rated volume flow rate, the pressure drop
of a clean air filter when it first starts to collect dust particles is called the initial pressure drop.
 Service life. The pressure drop of an air filter which is about to be replaced or cleaned is called its
final pressure drop. The operating period between the initial and final pressure drop is called the
service life of an air filter. When the pressure drop of an air filter is at its final pressure drop, the
dust-holding capacity of the air filter is usually the maximum dust-holding capacity mmax, d.
The efficiency of an air filter is significantly affected by the service life, or loading, of an air filter.
The rating and assessment of a specific air filter at a specific volume flow rate require complete
data on its efficiency and initial and final pressure drops.
15.13 TEST METHODS FOR AIR FILTERS
Test Methods
Since the 1940s, three test methods were widely used for air filters: weight arrestance, dust spot,
and di-octyl phthalate (DOP) methods.
Weight Arrestance Method. In ANSI/ASHRAE Standard 52.1-1992, Gravimetric and Dust Spot
Procedure for Testing Air Cleaning Devices Used in General Ventilation for Removing Particulate
Matter, standard synthetic dust particles are used as the test dust in the weight arrestance method.
By measuring the weight of dust fed and the weight gain on the membrane filter after the tested air
filter, the arrestance can be calculated in percentage as
(15.49a)
where Wmf  weight gain of membrane filter, gr (g)
Wdf  weight of dust fed, gr (g)
This method primarily determines the efficiency of a coarse or low-efficiency air filter to remove
large dust particles present in the airstream.
Dust Spot Method. The dust spot test method in Standard 52.1-1992 uses an undefined and uncontrolled
test dust, the atmospheric air. By measuring the change in light transmission of the discolored
white filter paper before and after the test air filter, the efficiency of the air filter can be calculated
as
(15.49b)
where  volume flow rate of air drawn through upstream and downstream papers, cfm (L / s)
T10, T11  initial and final light transmission of upstream filter paper
T20, T21  initial and final light transmission of downstream filter paper
This method is widely used for medium-efficiency air filters.
V?
1,V?2
E  1001 
V?
1(T20  T21)(T10)
V?
2(T10  T11)(T20)
Arrestance  1001 
Wmf
Wdf
15.62 CHAPTER FIFTEEN
DOP Method. The DOP test method was developed for the military to test the performance of
high-efficiency particulate air (HEPA) filter. According to U.S. Military Standard MIL-STD-282
(1956), a smoke cloud of uniform di-octyl phthalate (DOP) droplets 0.18 
m in diameter is
generated from the condensation of DOP vapor as the test dust. DOP vapor produced by thermal
generation is oily liquid particles. By measuring the concentration of these particles in the air sample
upstream and downstream from the tested air filter with an electronic particle counter, the penetration
P and the efficiency E of the filter, both in percent, can be calculated as
(15.50)
Unfortunately, confusion had arisen because of the uses of undefined and uncontrolled test dust
have been used to assess air filters with 1930s discoloration technology. ANSI/ASHRAE Standard
52.2-1999, Method of Testing General Ventilation Air Cleaning Devices for Removal Efficiency by
Particle Size, uses particle size and the varying-loading life-cycle method to assess air filter effi-
ciency. The primary characteristics of the new test method are as follows:
 The efficiency of an air cleaner is determined as a function of dust particle size to be removed
from the airstream by an electronic particle counter. The performance curve of an air filter is determined
according to 12 size ranges from 0.3 to 10 
m.
 The particle removal efficiency is evaluated over a multiple-loading life cycle between initial
(clean filter) and five loading stages. These efficiency curves including the minimum performance
(initial clean filter) are the basis of a composite curve, as shown in Fig. 15.35.
E  100
1  penetration
100
P  100
downstream concentration
upstream concentration
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.63
100
90
80
70
60
50
40
30
20
10
0
10.00 0.35 0.47 0.62 0.84 1.14 1.44 1.88 2.57 3.46 4.69 6.20 8.37 Efficiency, percent
Particle size, 
m
Initial
1st loading
2nd loading
3rd loading
4th loading
5th loading
Composite curve
FIGURE 15.35 Composite efficiency curves of an air filter. (Source: HPAC, August 1997. Burroughs, H.E.B.
Reprinted by permission.)
 The test method uses laboratory-generated potassium chloride dispersed into the airstream as the
test aerosol.
 Air filters are classified according to their minimum efficiency report value (MERV) into five
categories: coarse, low, medium, high, and ultrahigh efficiency.
The weight arrestance test method to assess coarse filters, the new particle size method to assess
low-, medium- and high-efficiency filters and DOP method for filters used in clean rooms and clean
spaces. Table 15.3 lists MERV for air filters.
Test Unit
The performance of an air filter is tested in a test unit. Figure 15.36 shows an air filter test unit
specified by ANSI/ASHRAE Standard 52.2-1999. The test unit consists of a blow-through fan, a
test duct, a tested filter, two samplers, a particle counter, a flow nozzle, two sets of Pitot tubes and
15.64 CHAPTER FIFTEEN
HEPA
filter
Fan
HEPA
filter
Sample
probe
Dust
feed
Sample
probe
Aerosol
injection
Test air
filter Nozzle
FIGURE 15.36 ANSI/ASHRAE Standard 52.2 specified air filter test unit.
TABLE 15.3 Minimum Efficiency Reporting Value (MERV) Parameters
Standard 52.2
Composite average particle size efficiency, % minimum
in size range, 
m
Average
efficiency arrestance, %
Minimum final resistance reporting value Range 1 Range 2 Range 3 by Standard 52.1
(MERV) 0.30–1.0 1.0–3.0 3.0–10.0 method Pa in. of water
1 n/a n/a E3  20 Aavg  65 75 0.3
2 n/a n/a E3  20 65  Aavg  70 75 0.3
3 n/a n/a E3  20 70  Aavg  75 75 0.3
4 n/a n/a E3  20 75  Aavg 75 0.3
5 n/a n/a 20  E3  35 n/a 150 0.6
6 n/a n/a 35  E3  50 n/a 150 0.6
7 n/a n/a 50  E3  70 n/a 150 0.6
8 n/a n/a 70  E3 n/a 150 0.6
9 n/a E2  50 85  E3 n/a 250 1.0
10 n/a 50  E2  65 85  E3 n/a 250 1.0
11 n/a 65  E2  80 85  E3 n/a 250 1.0
12 n/a 80  E2 90  E3 n/a 250 1.0
13 E1  75 90  E2 90  E3 n/a 350 1.4
14 75  E1  85 90  E2 90  E3 n/a 350 1.4
15 85  E1  95 90  E2 90  E3 n/a 350 1.4
16 95  E1 95  E2 95  E3 n/a 350 1.4
Note: The minimum final resistance shall be at least twice the initial resistance, or as specified above, whichever is greater.
Source: ANSI/ASHRAE Standard 52.2–1999. Reprinted with permission.
inclined manometers, two ultrahigh-efficiency air filters, and other necessary instruments and accessories.
Laboratory-generated dust is fed into the test unit from one end. The blow-through fan is
mounted at one end of the unit to maintain a positive pressure in the test unit. For details, refer to
Standard 52.2-1999.
15.14 AIR FILTERS
Filtration Mechanism
The removal or collection of dust particles in air filtration is performed by various combinations of
the following mechanisms.
 Inertial impaction. A sudden change in direction causes a collision between the dust particles, and
fibrous media.
 Straining. If the filter spaces are smaller than the size of the dust particles, the particles are
trapped.
 Diffusion. For very fine dust particles, brownian movement causes the particles to settle.
 Interception. Dust particles may follow the airstream, contact the fibrous media, and remain there.
 Electrostatic effects. Particles and the filter medium are charged to collect the dust in the airstream.
Classification of Air Filters
Except for ultrahigh efficiency filters, each category is subdivided into four MERVs, such as coarse
air filters into MERV 1, 2, 3 and 4, low-efficiency into MERV 5, 6, 7 and 8, medium-efficiency into
MERV 9, 10, 11 and 12, and high-efficiency into MERV 13, 14, 15, and 16.
Coarse Air Filters
Coarse air filters are often used to remove coarse dusts (5 to 80 
m) such as standing dust on surfaces,
pollen, and textile fibers. They are generally panel filters of MERV 1, 2, 3 and 4 with weight
arrestance efficiencies between 65 and 75 percent. The dimensions of the framework of these filters
are often 20-in. (500-mm) width by 20-in. (500-mm) height, or 24-in. (600 mm) width by 24-in.
(600-mm) height. The depth of the filter may vary from to 4 in. (13 to 100 mm).
Filter media for coarse filters are divided into three categories:
 Viscous and reusable. Some examples are corrugated wire mesh and screen strips. They are usually
coated with oil, which acts as an adhesive, to increase their dust removal ability. Detergents
may be used to wash off the dust when the filter media are to be cleaned and reused.
 Dry and reusable. Certain materials such as synthetic fibers (nylon, terylene) and polyurethane
foam can be cleaned or washed if reuse is required.
 Dry and disposable. Synthetic and cellulose fibers, and bonded glass fiber mats with most of the
glass fibers greater than 10 
m in diameter belong to this category. The air filter is discarded as
soon as the final pressure drop is reached.
The face velocity of panel filters usually lies between 300 and 600 fpm (1.5 and 3 m/ s). The
minimum final pressure drop is 0.3 in. WC (75 Pa).
Coarse air filters are widely used in window air conditioners and residential small packaged
units.
1
2
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.65
Low-Efficiency Air Filters
Low-efficiency air filters including MERV 5, 6, 7 and 8 are often used to remove dusts between 3
and 10 
m, such as spores, molds, hairspray, cement, and other solid particles. Low-efficiency air
filters are often pleated filters or sometimes bags as shown in Fig. 15.37b and c. They often have the
same size of framework as in panel filters. The pleated mat is used to extend the surface area of the
filter media.
In low-efficiency air filters, natural and synthetic fibers including glass fibers, viscous corrugated
wire mesh, bonded glass fibers, and electrostatically discharged fibers or fabrics called electret
have been used as filter media.
 For synthetic fibers and bonded glass fibers, their diameters may vary from 1 
m to several 
m.
As the efficiency increases, the diameter of the fiber is reduced and they are spaced comparatively
closer together.
15.66 CHAPTER FIFTEEN
FIGURE 15.37 Types of air filters: (a) panel filter; (b) pleated filter; (c) extended surface (bag type); (d ) rotary filter;
(e) automatic renewable rolling filter.
 Most low-efficiency filters have a face velocity of 500 fpm (2.5 m / s) to match the face velocity of
the coil in AHUs or PUs.
 Pleated filters often extend their filter area 2 to 8 times their face area, so that the air velocity penetrating
the filter media is reduced to 60 to 250 fpm (0.24 to 0.5 m/ s). A lower air velocity will
eventually increase the diffuse effect of the small particles and decrease the pressure drop across
the filter medium.
 There are three types of filters using permanently electrostatically charged fibers and fabrics
which increase the collection of small dust particles: (1) charges on the resin wool are produced
by friction during the carding process; (2) a corona charge emits positive charges on one side of a
thin polypropylene film and negative charges on the other side, and the film is shredded into
fibers; and (3) it spins a liquid polymer into fiber under the effect of a strong electric field. Because
of the combined effect of the inertial impaction, straining, diffusion, and local electrostatic
effect of electret filters, their efficiency of dust removal is enhanced. Dust loading may reduce the
efficiency of electret filters.
Low-efficiency air filters have a final minimum pressure drop of 0.06 in. WC (150 Pa). Lowefficiency
filters are widely used in packaged units and air-handling units in commercial and institutional
buildings, industrial workplaces, and more demanding residential applications.
Medium-Efficiency Air Filters
Medium-efficiency air filters including MERV 9, 10, 11 and 12 are used to remove dusts of size
1 to 3 
m such as welding fumes, Legionella bacteria and coal dusts. They are often bag and box
filters with pleated mat to extend surface area, and are made of synthetic fibers including bonded
glass fibers of diameters from <1 
m to several micrometers. MERV 9, 10, 11 and 12 have a 1 to
3 
m particle size efficiency from E2 < 50 percent (corresponding to a dust spot efficiency of 40 to
45 percent up to E2 > 80 percent (corresponding to a dust spot efficiently of 70 to 75 percent), and a
minimum final pressure drop of 1 in. WC (250 Pa). Medium-efficiency air filters are used in demanding
commercial building and industrial applications.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.67
FIGURE 15.38 Ultrahigh-efficiency particulate filter (box filter).
High-Efficiency Air Filters
High-efficiency air filters including MERV 13, 14, 15, and 16 are used to remove particles of 0.3 to
1 
m such as bacteria, viruses, cooking oil fumes, tobacco smoke, and other smoke.They have a 0.3
to 1
m particle size efficiency E1 < 75 percent (corresponding to a dust spot efficiency of 80 to 90
percent) up to E1 > 95 percent. The filter media are often made of glass fibers of submicrometer and
micrometer diameter. They are often in the form of a pleated mat in a cartridge, as shown in Fig.
15.38, or in the form of a bag filter. The air velocity flowing through the filter media is lower, and
the minimum final pressure drops across the filter media are 1.4 in. WC (350 Pa). High-efficiency
air filters are often protected by low- or medium-efficiency prefilters to extend their service life.
High-efficiency filters are widely used in air systems for hospitals, high-demand commercial buildings,
and precision manufacturing workshops.
Ultrahigh-Efficiency Air Filters
Ultrahigh-efficiency filters include high-efficiency particulate air (HEPA) filters, ultralow penetration
air (ULPA) filters, and gaseous adsorbers and chemisorbers. Activated carbon adsorbers are
discussed in Sec. 15.16, and chemisorbers are discussed in Sec. 24.5.
HEPA filters have an efficiency of 99.97 percent for dust particles  0.3 
m using the DOP test
method. ULPA filters have an efficiency of 99.999 percent for dust particles  0.12 
m using the
DOP method. A typical HEPA filter is shown in Fig. 15.38. The dimensions of this filter are 24 in.
by 24 in. by 11.5 in. (600 by 600 by 287 mm). The filter media are made of glass fibers of submicrometer
diameter that are formed into pleated paper mats. Some of the larger fibers act as the
carrier of the web. The performance of the filter medium is often assessed by an index called the
alpha value , which can be calculated as follows:
(15.51)
where pt  pressure drop of the filter medium, in. WC (Pa). Both penetration P, in percent, and
pressure drop pt are measured at an air velocity flowing through the filter medium of 10.5 fpm
(0.05 m/ s). The value is usually between 10 and 11. Because of the development of new media
with lower pressure drops, an value of 13 or even higher can now be achieved. The surface area
of the filter medium may be 50 times the face area of the ultrahigh-efficiency filter, and the rated
face velocity may vary from 190 to 390 fpm (0.95 to 1.95 m / s) for ultrahigh-efficiency filters at a
pressure drop of 0.65 to 1.35 in. WC (162 to 337 Pa) for clean filters. The face velocity of highcapacity
ultrahigh-efficiency filters can be raised to 500 fpm (2.5 m / s).
The filter media themselves have efficiencies higher than that of the mounted filter. Sealing of
the filter pack within its frame and sealing between the frame and gasket are critical factors that
affect HEPA and ULPA filter penetration and efficiency. Penetration of dusts represents aerosol
passing through the medium and through pinholes in the filter medium, as well as leaks between the
pack and the frame and between the frame and the gasket.
To extend the service life of an ultrahigh-efficiency filter, it should be protected by either a
medium-efficiency filter or two filters: a low-efficiency filter and a medium-efficiency filter located
just upstream from the ultrahigh-efficiency filter. The removal of large particles in the prefilter reduces
the dust load and prolongs the life of the ultrahigh-efficiency filter.
HEPA and ULPA filters are used to remove air contaminants such as unattached viruses, carbon
dust, combustion smoke, and radon progeny of particles in sizes  0.3 
m. They are widely used in
clean rooms and clean spaces for the microelectronics industry, pharmaceutical industry, precision
manufacturing, and operating theaters in hospitals.


(2  log P)100
25.4 pt
15.68 CHAPTER FIFTEEN
15.15 ELECTRONIC AIR CLEANERS
An electronic air cleaner uses the attraction between particles of opposite charges. Dust particles
charged within the cleaner attract and agglomerate to greater sizes at the collecting plates. They are
therefore easily removed from the airstream.
A typical electronic air cleaner is shown in Fig. 15.39. A high dc potential of 1200 V is supplied
to the ionizing field. The positive ions generated from the ionizer wire charge the dust particles.
Right after the ionizing section, the dust particles come to a collecting section, which consists of
several plates that are alternately grounded and insulated. A strong electric field is produced by supplying
a dc potential of 6000 V to these plates. The positively charged dust particles are attracted by
the grounded plates of opposite charge, and attach themselves to the plates.
Because of the numerous points of contact, the bond between particles held together by intermolecular
forces is greater than that between the particles and plates. Therefore, the dust particles
agglomerate and grow to such sizes that they are blown off and carried away by the airstream. The
agglomerates are then collected by a medium-efficiency air filter located downstream from the collecting
section.
Kemp et al. (1995) showed that electronic air cleaners are efficient for removing small particles
of 0.5 to about 8 
m. When dust particle sizes were approached and exceeded 10 
m, the filter’s
efficiency dropped from its highest peak. In a year-long test with the outdoor ambient air, no microbial
growth was observed on any of the filters.
The pressure drop across the ionizer section and collecting section is low and ranges from 0.15
to 0.25 in. WC (37 to 62 Pa) against an air velocity of 300 to 500 fpm (1.5 to 2.5 m / s). Safety
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.69
FIGURE 15.39 An electronic air cleaner.
measures must be provided for protection against the high dc potential. If the positive charges are
not removed at the collecting section, the accumulation of these positively charged dust particles
may build up a space charge. This space charge drives the charged particles to the walls and other
building envelope surfaces in the conditioned space, causing them to be smudged.
15.16 ACTIVATED CARBON ADSORBERS
Granular activated carbon adsorbers are most widely used to remove objectionable odors and irritating
vapors (including indoor volatile organic compounds, VOCs) of very small gaseous molecules
from the airstream by adsorption. Adsorption is the physical condensation of a gas or vapor onto an
activated substance. Activated substances are highly porous. One pound of extremely porous carbon
contains more than 5,000,000 ft2 (465,000 m2) of internal surface. Gas molecules diffuse to the micropores
or macropores of activated carbon, bond to these surfaces, and come in contact with the
carbon granules. One pound (0.45 kg) of activated carbon may adsorb 0.2 to 0.5 lb (0.1 to 0.25 kg)
odorous gases.
Activated carbon in the form of granules or pellets is made from coal, coconut shells, or petroleum
residues. They are heated in steam and carbon dioxide to remove foreign matter such as hydrocarbons
and to produce internal porosity. Granular activated carbon is placed in special trays,
which slide easily into position, to form activated carbon beds that are sealed into the cell housing
by face plates, as shown in Fig. 15.40. A typical carbon tray, which is 23 by 23 by in. thick (0.58
by 0.58 by 0.015 m thick), weighs 12 lb (5.4 kg). Low- or medium-efficiency air filters are used as
prefilters for protection. When air flows through a typical assembly with a face velocity of 375 to
500 fpm (1.88 to 2.5 m/ s), the corresponding pressure drops are between 0.2 and 0.3 in. WC (50
and 75 Pa). Activated carbon can also be mounted in fixed frames with perforated sheets in continuous
pleats. Adsorption capacity is defined as the amount of carbon tetrachloride adsorbed by a
given weight of activated carbon. For various odors, the adsorption capacity is also affected by the
operating temperature and humidity. In general, a higher humidity or higher operating temperature
5
8
15.70 CHAPTER FIFTEEN
FIGURE 15.40 Granular activated carbon assembly.
usually decreases the adsorption capacity of the activated carbon. The maximum operating temperature
is 100°F (37.8°C).
The removal efficiency of an activated carbon adsorber is the ratio of pollutant concentrations in
the air upstream and downstream from the absorber. Toluene is used as the surrogate for total
volatile organic compounds. The removal efficiency of the carbon bed remains relatively constant
during its service life. Service life is the operating period between when an activated carbon
absorber begins to remove pollutant and the time when the removal efficiency drops below an
acceptable minimum. When the removal efficiency drops below an acceptable minimum, it must be
replaced either by reactivation or regeneration. Reactivation is the process of removing spent
carbon and replacing it with fresh carbon. Regeneration is the process by which spent carbon is
converted to fresh carbon. Regeneration can only be performed by the activated carbon manufacturer.
The simplest way to investigate whether the carbon needs to be replaced is to detect downstream
the odor that the carbon filter is supposed to remove. A sample of carbon can also be tested
to determine its remaining capacity.
Burroughs (1997b) reported that active particle fabrics demonstrate excellent indoor air contaminant
control of VOC gases. These new fabric-based sorption filter media can be pleated into various
configurations to maximize sorption capacity and airflow.
15.17 SELECTION OF AIR FILTERS
During the selection of air filters, electronic air cleaners, and activated carbon filters to remove air
contaminants for the health and comfort of the occupants in indoor conditioned space, the following
requirements and arrangements should be carefully considered:
 Characteristics of the air contaminants, the size of the particles, and the concentration of dust particles
or irritating vapors to be removed must be identified.
 The degree of air cleanliness required in the conditioned space must be specified, especially the
design criteria for clean spaces or clean rooms.
 The minimum efficiency of the air filter at specific particle sizes and loadings must be specified.
 The initial, average, and final pressure drops during the operating period, which affect the energy
consumption and service life of the filter and the air system, must be determined.
 Service life of the air filter influences the installation cost, the pressure drop, and the efficiency of
the air filter.
The following filtration tactics are recommended:
 Select an air filter based on the dust particle size MERV to remove air contaminants.
 Burroughs (1997) recommended the use of medium-efficiency particulate filters for protection of
coils and air distribution systems, preventing the nutrition for biological growth; high-efficiency
particulate filters for control of respirable particulate and bacteria; and gaseous absorbers to control
objectionable odors and harmful volatile organic compounds.
 High-efficiency and ultrahigh-efficiency air filters, gaseous adsorbers such as granular activated
carbon, and electronic air cleaners must be protected by a prefilter of medium efficiency to extend
the life of these high-efficiency and ultrahigh-efficiency gaseous adsorbers and electronic air
cleaners. However, the addition of two low- or two medium-efficiency filters does not signifi-
cantly improve the efficiency to collect submicron dust and irritating vapors.
 Monitoring the pressure drop of the air filters and periodic maintenance of low-efficiency viscous
air filters have a direct impact on filter performance.
 Muller (1995) reported that granular activated carbon is very good to remove most hydrocarbons,
many aldehydes and organic acids, as well as nitrogen dioxide. It is not particularly effective
against sulfur oxide, nitric oxide, formaldehyde, hydrogen sulfide, or lower-molecular-weight
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.71
aldehydes and organic acids. Potassium permanganate–impregnated alumina (chemisorbers are
discussed in Sec. 24.5) is effective against sulfur and nitric oxides, hydrogen sulfide, and lowmolecular-
weight aldehydes.
15.18 HUMIDIFICATION AND HUMIDIFIERS
Space Relative Humidity
As mentioned in Sec. 4.9, for comfort air conditioning systems the space relative humidity is tolerable
between 35 and 65 percent in summer, preferably from 40 to 50 percent. During winter, for
space served by a comfort air conditioning system that is installed with a humidifier, the space relative
humidity should not exceed 30 percent except in hospitals. The actual humidifying capacity
of a humidifier should not exceed the humidifying requirement so that wet surfaces do not occur inside
the AHU, PU, or supply ducts. Wet surfaces and dirt often cause the growth of microorganisms
and thus indoor air quality problems.
For air conditioning systems without humidifiers, space relative humidity is usually not
specified.
For processing air conditioning systems, the space relative humidity should be specified as
required by the manufacturing process.
If the space temperature is maintained at 72°F (22.2°C) and only free convection is provided in
the occupied space, the space relative humidity should not exceed 27 percent when the outdoor
temperature is 30°F (1.1°C), in order to prevent condensation on the inner surface of singleglazed
windows during winter. To prevent condensation on the inner surface of double-glazed
windows, the space relative humidity should not exceed 33 percent when the outdoor temperature is
0°F (17.8°C).
Humidifiers
A humidifier adds moisture to the air. Humidifiers may (1) inject steam directly into air or add heat
and evaporate steam from water supplied to the conditioned space; (2) atomize or spray liquid water,
so that water evaporates and is added to the air; or (3) force air to flow over a wetted element so
that as water evaporates, it is added to the air as vapor. All these increase the humidity ratio of the
space air and, therefore, its relative humidity. One important index of a humidifier is its humidifying
capacity in pounds of water per hour, lb/ h (kg / h), or the rate at which water vapor is
added to the air.
Humidifying Load
Humidifying load , in lb /h or kg /h, is the amount of water vapor required to be added to the air
by a humidifier so as to maintain a predetermined space relative humidity. The humidifying load of
an air system installed with a humidifier can be calculated as (see Fig. 15.41)
(15.52)
where  supply volume flow rate of AHU or PU, cfm (m3 /min)
	s  density of supply air, lb/ ft3 (kg /m3)
wen,wlv  humidity ratio of air entering and leaving humidifier, lb/ lb (kg / kg)
ws  humidity ratio of supply air, lb/ lb (kg / kg)
wm  humidity ratio of mixture of outdoor air and recirculating air, lb/ lb (kg / kg)
V?
s
m? hu  60V?s 	s (wlv  wen)  60V?s 	s(ws  wm)
m? hu
m? cap
15.72 CHAPTER FIFTEEN
In Fig. 15.41, s is the state of supply air when the space has a heating load, and s is the condition
when the space has a cooling load, ws  ws.
The humidifying load can also be calculated from the following relationship:
(15.53)
where  supply volume flow rate of outdoor air intake, cfm (m3/min)
 volume flow rate of infiltrated air, cfm (m3/min)
 humidity ratio of space air and outdoor air, lb/ lb (kg /kg)
In Eq. (15.53), represents the space moisture gains, in lb /h (kg / h). Space moisture gains include
the latent load from the occupants, appliances, equipment, and products. The moisture gains
from the building structures are often ignored.
Types of Humidifier
According to the mechanism used for evaporation of water vapor from water, humidifiers can be classi-
fied as steam and heating element humidifiers, atomizing humidifiers, and wetted element humidifiers.
15.19 STEAM AND HEATING ELEMENT HUMIDIFIERS
At a low partial pressure when mixed with dry air, steam is the ready-made water vapor. With
proper water treatment at the boiler, steam is free of mineral dust and odor, and it does not support
the growth of bacteria that create sanitation problems. Two types of steam humidifiers are currently
used in air systems: steam grid humidifiers and steam humidifiers with separators.
Steam Grid Humidifiers
A steam grid humidifier installed inside ductwork is shown in Fig. 15.42a. A steam grid humidifier
may have a single distribution manifold or multiple manifolds. Each distribution manifold has an
m? wr
wr,wo
V?
inf
V?
o
m? hu  60(V?o  V?inf) 	o(wr  wo)  m? wr
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.73
FIGURE 15.41 Humidifying load for a typical air system.
inner steam pipe and an outer jacket. The inner steam tube is connected to a header and controlled
by a control valve. Steam is supplied at a pressure less than 10 psig (69 kPag) and throttled to a
lower pressure after the controlled valve. It then enters the inner tube and discharges through the
small holes to the outer jacket. The dry steam is again discharged through the orifices of the jacket
to the ambient airstreams to humidify them, as shown in Fig. 15.42b. The condensate inside the
jacket is discharged to a drainpipe and steam trap located at the opposite side. The inner tube and
the outer jacket are slightly pitched toward the drainpipe and steam trap.
Steam Humidifiers with Separators
A steam humidifier with a separator is illustrated in Fig. 15.43. Steam is supplied to a jacketed
distribution manifold and then enters a separating chamber with its condensate. It then flows
through a control valve, is throttled to slightly above atmospheric pressure, and enters a drying
chamber. Because of the lower pressure and temperature in the drying chamber compared with the
higher pressure and temperature in the surrounded separating chamber, the steam is superheated.
Dry steam is then discharged into the ambient airstream through the orifices of the inner steam
discharging tubes. Noise is produced mainly during the throttling of the steam at the control valve.
It is attenuated as steam flows through the silencing materials.
15.74 CHAPTER FIFTEEN
FIGURE 15.42 Steam grid humidifiers: (a) steam grid and piping connections;
(b) cross section of jacketed distribution manifold.
Heating Element Humidifiers
When a steam boiler is not available or is too expensive to install, steam humidification is still
possible by using heating element humidifiers. A heating element humidifier has a water pan with
electric or gas-fired heated elements installed at the bottom of the pan to evaporate the liquid water
to water vapor and then add the vapor to the airstream flowing over the water pan, as shown in
Fig. 15.44. The humidifying capacity of a heated element humidifier is limited by its heat-transfer
surface and the characteristics of the heating element.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.75
FIGURE 15.43 Steam humidifier with a separator. (Adapted with permission from The Armstrong Humidi-
fication Handbook, 1991.)
FIGURE 15.44 Heating element humidifier.
Characteristics and Requirements of Steam and Heating Element
Humidifiers
Steam humidifiers are widely used in commercial and industrial applications. Both steam and heating
element humidifiers need heat energy from gas, fossil fuel, or electricity to evaporate the liquid
water in the steam generator. The temperature of the ambient air or airstream in which moisture has
been added is approximately unchanged. Steam and heating element humidifiers are therefore often
called isothermal humidifiers. For a conditioned space that has a cooling load year round and needs
a cold air supply in winter, it is a waste of energy to evaporate the liquid water by using fossil fuel
instead of the excess heat gains at the conditioned space.
Control is essential in steam and heating element humidifiers. The capacity of a steam humidi-
fier is often controlled by a microprocessor-based controller by modulating the valve pin of the control
valve according to the signal of a humidity sensor and, therefore, the steam flow rate. Morton
(1996) reported an accuracy of humidity control of 5 to 7 percent for on /off control and 3 to 5
percent for modulation control. The size of the control valve should be carefully matched with the
humidifying load with an adequate turndown ratio. The control valve should be integrated with the
humidifier and steam-jacketed at supply pressure to prevent condensation. An interlocking control
should be installed to drain all condensate before the steam humidifier is started.
For an electric heating element humidifier, or simply an electric humidifier, the input wattage
can be adjusted when the space relative humidity is too high or too low. A gas-fired heating element
humidifier, or simply a gas-fired humidifier, uses infrared burners located within a heat exchanger.
A microprocessor-based controller stages and cycles the burner to vary the heat output and therefore
the humidifying capacity during part-load operation.
The makeup water supply to the steam boiler and the water pan must be well treated so that the
steam is free of mineral deposits and odor. For electric and gas-fired humidifiers, according to
DeBat (1996), recent developments, such as ionic bed technology, overcome the problem of
mineral deposits. Dissolved mineral salts begin to precipitate out of water when their concentration
becomes too high. Ionic beds consist of inert fibers that attract these precipitated minerals. The
mineral salts deposit on the ionic bed instead of on the walls of the heating chamber. When
throughly encrusted with crystalline solids, the ionic bed cartridge is easily removed and discarded.
Steam humidifiers should be located where noise is not objectionable. Access to equipment for
inspection and maintenance should be provided. Condensate drainage must be properly designed. If
the steam humidifier is located in ductwork, install a duct high-limit humidistat and an interlocked
airflow proving switch. The humidity sensor is located preferably in the return air duct or a representative
location.
15.20 ATOMIZING AND WETTED ELEMENT HUMIDIFIERS
Humidification Process
Atomizing means producing a fine spray. When liquid water is atomized, the smaller the diameter
of the water droplets, the greater the interfacial area between water and air and thus the higher the
rate of evaporation and humidification.
When air flows through an atomized water spray from an atomizing humidifier, if the difference
between the temperature of the water spray and the wet-bulb temperature of the ambient air is small,
the result is an increase in the humidity ratio of air from the addition of evaporated water vapor and a
corresponding drop in air temperature because of the absorbtion of required latent heat of vaporization
from the ambient air. Such a humidification process is an adiabatic saturation or evaporative cooling
process and follows the thermodynamic wet-bulb temperature line on the psychrometric chart.
Atomizing humidifiers can be classified, according to their configuration and mechanism of
atomizing, as ultrasonic humidifiers, centrifugal atomizing humidifiers, pneumatic atomizing humidifiers,
and air washers.
15.76 CHAPTER FIFTEEN
Ultrasonic Humidifiers
An ultrasonic humidifier uses a piezoelectric transducer to convert high-frequency mechanical electric
signals to a high-frequency ocscillation. The transducer is submerged into a water pan in which
water is well treated and demineralized. The negative oscillation produces a momentary vacuum
that causes the water to cavitate into vapor at low pressure. The positive oscillation creates a highcompression
wave which forces the water particles to leave the water surface and inject into the
airstream. Water films are then broken into a fine mist. Air containing this mist can be blown into
the space directly or through an air system by a fan.
If the water is well treated, it will be free of mineral deposits, and no filter is required downstream
from the humidifier. Ultrasonic humidifiers create less equipment noise than other atomizing
humidifiers. In the mid-1990s, ultrasonic humidification had been installed in telecommunication
projects. Ultrasonic atomizing humidifiers with small humidifying capacities have been used directly
in residences.
Case Study: White Plains Ultrasonic Humidification Project
Longo (1994) reported on a retrofit ultrasonic humidification project which is located in a 1950s
building in White Plains, New York. The humidification system in a totally new air system serves
space that houses solid-state complex multiswitching processors. Static charges created in a lowhumidity
environment can be detrimental to system components and cause failure.
The 100,000-cfm (47,200-L / s) air system consists of four supply fans (two standby), highefficiency
filters, and cooling coils with outside air economizer for free cooling. The ultrasonic
humidifiers were placed after the filters and upstream from the supply fan. There are altogether 43
ultrasonic units mounted on stainless-steel racks, each rated at 39.6 lb /h (18 kg / h) and 960 W. A
humidity sensor is installed in the return air duct. The controller turns the humidity system on one
stage at a time up to a total of six stages. A high-limit sensor set at 80 percent relative himidity
shuts down the system in case of overhumidification.
Water is treated by a series of prefilters, a reverse osmosis system, and deionization canisters.
Water quality is controlled by a resistivity probe. When water resistance reaches 0.02 M, the
deionization canisters will be changed.
After completion, the actual running results in winter are excellent. As the equipment was
installed gradually, the supply air temperature was rescheduled, and the space relative humidity was
maintained precisely at 50 percent. The initial cost of the White Plains humidity system was
$350,000. Annual operating cost was $23,000, which was less than one-tenth of the annal operating
cost of an electric heating element humidity system.
Centrifugal Atomizing Humidifiers
These humidifiers use the centrifugal force produced by a rotating device such as a rotating cone,
blades of an axial fan, rotating disk, or rotating drum to break the liquid water film into fine mist or
to fling it into fine water droplets. When air is forced through an atomizing water spray, produced
by a pulverizing fan, water vapor is evaporated and added to air as a result of the mass concentration
difference between the saturated air film at the surface of the droplets and the ambient air. The
rotating device receives liquid water either from a pressurized water supply or by dipping into the
surface of a nonpressurized supply. The humidifying capacity of a pulverizing fan depends mainly
on the volume flow rate of the fan. Because of the oversaturation characteristics of the pulverizing
fan, may vary from 50 to 150 lb /h (23 to 68 kg / h). Oversaturation is the excess amount of
water particles present in the moist air, and it is discussed in Chap. 20. Many different types of rotating
humidifiers with limiting humidifying capacities are used directly in the conditioning space
in residential and industrial applications. Most are portable.
m? cap
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.77
Maintaining a predetermined space relative humidity with a centrifugal atomizing humidifier
can be performed by modulating the flow rate of water supplied to the rotating devices.
Pneumatic Atomizing Humidifiers
A typical pneumatic atomizing nozzle consists of two concentric brass tubes, as shown in
Fig. 15.45a. The inner tube has an outer diameter of in. (2.4 mm) and an inner diameter of in.
(0.8 mm). It is connected to a water tank that can be moved to adjust the difference in water levels
between the water tank and the centerline of the nozzle. The conical outer tube has a minimum
inner diameter of in. (3.2 mm) and is connected to a compressed air line with a pressure at 15 psig
(103 kPag).
When compressed air is discharged from the annular slot at a very high velocity, it extracts
water from the inner tube and breaks the water into a very fine mist. A typical pneumatic
atomizing nozzle may have a humidifying capacity from 6 to 10 lb / h (2.7 to 7.5 kg / h).
The magnitude of is affected by the configuration of the nozzle, the pressure of the
compressed air, and the difference in water levels. In a pneumatic humidifier, is usually
controlled by adjusting the difference in water levels between the water tank and the centerline
of the nozzle.
A pneumatic humidifier produces high-frequency noise. This type of humidifier is usually applied
in industries with a certain sound level of machine noise and can be used for direct in-space
humidification.
Wetted Element Humidifiers
Wetted element humidifiers include a wetted element, such as an evaporative pad, plastic, or impregnated
cellulose, that is dipped with water from the top. Such humidifiers have been installed
in air-handling units and packaged units to humidify the air. Characteristics of wetted element
humidifiers are similar to those of wetted element evaporative coolers (see Chap. 27). Modulation
of the water supply to the water dipping device varies the humidifying capacity and maintains the
desirable space relative humidity.
m? cap
m? cap
m? cap
1
8
1
32
3
32
15.78 CHAPTER FIFTEEN
FIGURE 15.45 Pneumatic atomizing humidifier: (a) nozzle; (b) pneumatic atomizing humidifying system.
15.21 AIR WASHERS
The air washer was the first air conditioning equipment developed by Carrier in 1904, and it is still
used today to humidify, cool, and clean the air in many factories.
Construction of an Air Washer
An air washer has an outer casing, shown in Fig. 15.46a. It is usually made of plastic or galvanized
steel sheet with water-resistant paint for protection. All joints are well sealed by water-resistant
resin. A water tank either forms a part of the casing at the bottom or is installed separately on the
floor to collect or sometimes to mix the recirculating and incoming chilled water. A separately installed
water tank is usually made of steel, stainless steel, or reinforced concrete with an insulation
layer mounted on the outer surface if chilled water is used for spraying. A bank of guided baffles
installed at the entrance provides an evenly distributed air-water contact. Eliminators in the shape of
a sinusoidal curve at the exit are installed to remove entrained water droplets from the air. They are
preferably made of plastic or stainless steel for convenient cleaning and maintenance. Access doors
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.79
FIGURE 15.46 Schematic diagram of a typical air washer: (a) air washer; (b) spray nozzle.
mounted on the outer casing must be provided for inspection and maintenance. They should be well
sealed to prevent water leakage.
In most air washers, two banks of spraying nozzles face each other. Nozzles are often made of
brass, plastic, or nylon. The nozzle consists of a cover with an orifice diameter between and
in. (1.6 and 4.8 mm), a discharge chamber, and an inlet with its centerline slightly eccentric from
the centerline of the orifice, as shown in Fig. 15.46b. This offset causes the water stream to rotate
inside the discharge chamber and breaks the water into fine droplets. A larger orifice diameter does
not clog easily, and a smaller orifice can produce finer water sprays. The distance between two
spraying banks is from 3 to 4.5 ft (0.9 to 1.35 m), and the total length of the air washer varies from
4 to 7 ft (1.2 to 2.1 m).
A centrifugal water pump is used to recirculate the spraying water. For air containing much dirt
and lint, as in air conditioning systems in the textile industry, an automatic washing and collecting
water strainer, using copper or brass fine-mesh screen, often precedes the pump. The opening of
the fine-mesh screen should be smaller than the diameter of the hole of spraying nozzles to avoid
clogging.
Functions of an Air Washer
Currently, air washers are used to perform one or more of the following functions:
 Cooling and humidification
 Cooling and dehumidification
 Washing and cleaning
Whether it is a cooling and humidification or cooling and dehumidification process is determined
by the temperature of the spraying water. If recirculating water is used as the spraying water, the
temperature of water then approaches the wet-bulb temperature of the air entering the air washer,
and the air is humidified and evaporatively cooled.
If chilled water or a mixture of chilled water and recirculating water is used for spraying
water, and its temperature is lower than the dew point of the entering air, the air is cooled and
dehumidified.
Water spraying is sometimes practiced for the purpose of washing and cleaning air when an
objectionable gas is known to be soluble in water and is to be removed from the air. For example,
exhaust gas from many industrial applications may be removed by water spraying.
For humidification purposes, nozzles must have a smaller orifice and a higher water pressure,
such as a -in. (1.6-mm) diameter and a pressure of 40 psig (275 kPag). For cooling and dehumidification,
a larger orifice and lower water pressure, such as a -in. (4.8-mm) orifice and 25
psig (172 kPag), are often used.
Performance of an Air Washer
For a humidification process along the thermodynamic wet-bulb temperature line, the performance
of an air washer can be illustrated by the saturation efficiency ?sat and calculated as
(15.54)
where T1,T2  temperature of air entering and leaving air washer, °F (°C)
w1,w2  humidity ratio of air entering and leaving air washer, lb/ lb (kg / kg)
ws*, Ts*  humidity ratio and temperature of saturated air at thermodynamic wet-bulb temperature,
lb / lb (kg / kg) and °F (°C), respectively
sat 
w2  w1
w*s  w1
	
T2  T1
T*s  T1
316
116
316
116
15.80 CHAPTER FIFTEEN
For a cooling and dehumidification process, the performance of an air washer is better described by
a factor called the performance factor Fp as follows:
(15.55)
where hen,hlv  enthalpy of air entering and leaving air washer, Btu / lb (kJ / kg)
hs*  enthalpy of saturated air at chilled water spraying temperature, Btu/ lb (kJ / kg)
The performance of an air washer for a humidification process, indicated by the saturation effi-
ciency ?sat, and for a cooling and dehumidification process, expressed by Fp, depends mainly on the
contact time between air and water and the contact surface area, i.e., the water-air ratio
(which indicates the mass flow rate of the spraying water to the mass flow rate of the air flowing
through the air washer, in same units), the length of the air washer, and the direction of spray with
respect to airflow, (whether it is opposing or following the airflow).
An air washer is usually designed at an air velocity between 500 and 800 fpm (2.5 and 4 m/ s)
with respect to its cross-sectional area at the water sprayers. Its greater washer length and the
attached water tank result in a bulkier volume than that of DX coils and water cooling coils. Total
pressure loss of the airstream flowing through an air washer pt,w, in in. WC (Pa), depends mainly
on the configuration of the eliminators and the air velocity flowing through them. Usually pt,w
varies from 0.25 to 1 in. WC (62.5 to 250 Pa), and typically it is 0.5 in. WC (125 Pa). For humidifi-
cation and evaporative cooling, the water-air ratio is usually 0.3 to 0.6. For a 5-ft (1.5-m) air
washer with  0.45, that is, 4 gpm of water per 1000 cfm of air (0.534 L / s of water per
1000 L/ s of air), ?sat  0.85 to 0.9. For cooling and dehumidification, the water-air ratio is usually
0.5 to 1.2. For a 5-ft- (1.5-m-) long air washer with  0.9, that is, 8 gpm/ 1000 cfm (1 L / s
of water per 1000 L / s of air), Fp  0.5 to 0.75.
For either the humidification process or the cooling and dehumidification process in an air
washer, oversaturation always exists at the exit. Oversaturation (wo  ws*) usually varies between
0.0002 and 0.001 lb / lb (0.0002 and 0.001 kg / kg). It depends mainly on the construction of the
eliminator and the flow-through air velocity. Oversaturation is beneficial to a humidification
process with space heat gains and interferes with a cooling and dehumidification process. Because
of the existence of oversaturation, the saturation effectiveness is no longer the dominant index that
affects the performance of the air washer during humidification. Oversaturation is discussed in
Sec. 20.13.
Bypass Control
Bypass control is defined as modulating the face damper and the bypass damper of the air washer to
vary the proportion of humidifying air and bypass air in order to maintain the space relative humidity
within predetermined limits. Such a control is often used in an air washer to provide better humidity
control than modulating the water pressure or the water flow rate to the spraying nozzles.
Single-Stage or Multistage
A group of nozzles that are supplied with water at the same pressure and produce spray at the same
temperature are said to be in the same stage. A stage may have one or two spraying banks. A twostage
air washer is illustrated in Fig. 15.47.
For humidification processes or for cooling and dehumidification processes using chilled water
from the refrigeration plant, it is more economical to operate a single-stage air washer, usually with
two water-spraying banks. When low-temperature deep well or groundwater is used for cooling and
dehumidification, it is often advantageous to use two-stage, or even three-stage, water spraying to
make full use of the cooling capacity of the groundwater.
m? w /m?a
m? w /m?a
m?w /m?a
m?w /m?a
Fp 
hen  hlv
hen  hs
*
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.81
Groundwater can be used again as condenser water, or as flushing water in plumbing systems
after its discharge from the air washer following necessary water treatment. Used groundwater
should be drained to another well (recharge well) in the vicinity to prevent subsidence and damage
to the soil, which may affect the foundation of the building in the vicinity of the deep well.
The use of groundwater is regulated by local and federal environmental protection agencies.
15.22 CHARACTERISTICS OF ATOMIZING AND WETTED
ELEMENT HUMIDIFIERS
Atomizing and wetted element humidifiers, including air washers for humidification purposes have
the following operating characteristics:
 Conditioned air follows the thermodynamic wet-bulb temperature line on a psychrometric chart.
The humidity ratio of the air increases while it is evaporatively cooled. Atomizing humidifiers are
often called nonisothermal humidifiers. This characteristic indicates that atomizing and wetted
element humidifiers are especially suitable for conditioned spaces that need a cold air supply to
offset a space cooling load during or after the humidification process. The heat energy required to
evaporate the liquid water, as in steam or heating element humidifiers, has been saved. For a conditioned
space which needs a warm air supply during winter operation, the temperature of evaporatively
cooled air after an atomizing or wetted element humidification process should be raised before
supplying it to the conditioned space, so as to offset the transmission loss through the building envelope.
The heat energy required is approximately equal to that required to evaporate water in a steam
boiler or in a water pan when a steam humidifier or a heating element humidifier is used.
 For atomizing humidifiers that need a cold air supply during winter, when the humidifying
capacity is modulated, both the space relative humidity and temperature vary accordingly. Many
15.82 CHAPTER FIFTEEN
FIGURE 15.47 Two-stage air washer using deep well water.
industrial manufacturing processes need to maintain a stable space relative humidity (RH). RH is
often the controlled variable, and temperature is usually maintained within predetermined limits.
 Oversaturation exists at the exit of the atomizing humidifier even if an eliminator is used. Oversaturation
is often advantageous for humidifiers installed in conditioned spaces having excess
heat gains because it increases the humidifying capacity.
 The size of the water droplets, the local humidifying capacity, and the excessive heat gains determine
whether liquid water droplets may fall on the floor, products, or equipment. According to
field measurements, the diameter of water droplets suspended in the airstream discharged from
atomizing humidifiers is as follows:
Pneumatic atomizing humidifiers 30 to 80 
m
Centrifugal atomizing humidifiers 50 to 300 
m
When the diameter of the water droplets exceeds 80
m and the local humidifying capacity
is excessive, water may drip. Water dripping must be avoided, even if reducing the humidifying
capacity is necessary. It causes product damage and corrosion of appliances, instruments, and
equipment.
 In the water tank and water pan of atomizing and wetted element humidifiers, algae, bacteria, and
other microorganisms may grow. Water spray and the airstream discharged from the atomizing
and wetted element humidifiers may distribute the bacteria and sometimes odors. Water treatments,
regular cleaning and flushing, and periodic inspection and blowdown help to minimize
these problems. The designer must include convenient access for such maintenance.
15.23 SELECTION OF HUMIDIFIERS AND DESIGN
The following factors should be considered in selecting a humidifier:
 Energy consumption and operating cost. For a space that needs a cold air supply even in winter,
atomizing and wetted element humidifiers save energy and, therefore, have lower operating
costs. Among the atomizing humidifiers, pneumatic atomizing humidifiers require compressed air
to atomize the liquid water and, therefore, have a higher operating cost than other atomizing
humidifiers.
 Quality of humidification. Humidified air should be clean and free of odor, bacteria, particulate
matter, and water droplets to prevent microbial growth due to wetted surfaces or standing water
and improving IAQ.
 Humidifying capacity. Is a large, medium, or small humidifying capacity required? Is oversaturation
beneficial in the conditioned space?
 Capacity control. Is this an on /off control or a modulation control, i.e., a stepless control? Does
the control have a fast or slow response? What is the turndown ratio of the control valve?
 Morton (1996) recommended that the best location of a humidifier in an AHU or PU is in the first
section of supply main duct downstream from the supply fan. The next choice is between the coil
section and the supply fan if there is no oversaturation in the airstream.
 Equipment noise. For in-space humidifiers such as pneumatic humidifiers, equipment noise is a
primary concern.
 Initial cost. For a steam humidifier, the initial cost should also include the installation costs of the
steam boiler and the corresponding steam piping system.
 Maintenance. This includes the amount and cost of required maintenance work for the humidifiers.
 Space occupied. This is the volume occupied by the humidifier per unit humidifying capacity.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.83
Various types of humidifiers are compared in Table 15.4. For commercial buildings that need a
warm air supply during winter and for which a steam supply is available, a steam humidifier is
often the most suitable choice. For a small packaged unit serving a health care facility, a heating
element humidifier may be selected. For a telecommunications center with an appreciable amount
of space heat gain, an ultrasonic humidifier is often the right choice. Ultrasonic humidifiers are also
often used in residences for their comparatively low equipment noise. For a factory in which space
air contains lint and other dirt that should be cleaned, or for which a cold air supply is needed to
offset the cooling load in winter, an air washer may be the suitable choice.
When air washers or atomizing humidifiers are selected as the humidifiers, oversaturation may
result in the supply air discharged from the air washer or atomizing humidifier. As oversaturation
causes wetted surfaces and dampness along the air distribution passage, the following remedies are
recommended to prevent IAQ problems:
 All-metal (fiber-free) silencer or active silencer should be used, if required.
 Interior surfaces of ducts that contact supply air must be galvanized metal sheets.
 Duct heat gain may be beneficial to eliminate oversaturation.
 Periodic monitoring of the moisture and dirt conditions of the ducts and silencers should be
provided.
ASHRAE/IESNA Standard 90.1-1999 mandates that when a zone is served by a system or systems
with both humidification and dehumidification, means (such as limit switches, mechanical stops,
software program for DDC control) shall be provided to prevent simultaneous operation of humidi-
fication and dehumidification equipment. Exceptions are zones served by desiccant dehumidification
systems, which are used with direct evaporative cooling in series, or zones where specific
humidity levels are required such as computer rooms, museums, and hospitals. Standard 90.1-1999
also mandates that humidifiers with preheating jackets mounted in the airstream shall have an automatic
valve to shut off preheat operation when humidification is not in use.
15.84 CHAPTER FIFTEEN
TABLE 15.4 Comparison of Various Types of Humidifiers
Heating Pneumatic
Description Steam element Air washer Ultrasonic atomizing
Humidifying capacity Small to large Small Large Small to large Medium and large
Oversaturation No No Yes Possible Possible
Quality of Good Good Good Good Good
humidification
Dripping of water in No No No No No
conditioned space
Energy consumption Need more energy if Same as steam Saving energy if Saving energy if Saving energy if
and operation cost cold air supply is cold air supply is cold air supply is cold air supply is
needed, so operating needed in winter needed in winter needed in winter
cost is higher
Initial cost High High Medium Medium High
Maintenance work Low Medium High Medium Medium
and maintenance
cost
Capacity control Modulation On/off or Modulation Multiple stages Modulation
modulation
Response to control Fast Slow Medium Fast Fast
Space occupied Small Medium Large Small Medium
Equipment noise Medium Low Medium Low High
REFERENCES
ANSI/ASHRAE Standard 52.2-1999, Method of Testing General Ventilation Air-Cleaning Devices for Removal
Efficiency by Particle Size, ASHRAE Inc., Atlanta, GA, 1999.
ARI Guideline—Fouling Factors, Heating / Piping / Air Conditioning, no. 2, 1988, pp. 109–110.
Armstrong Machine Works, The Armstrong Humidification Handbook, Armstrong Machine Works, Three
Rivers, MI, 1991.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, 1997.
Avery, R. H., Selection and Uses of HEPA and ULPA Filters, Heating / Piping / Air Conditioning, no. 1, 1986,
pp. 119–123.
Branda M. R., A Primer on Adjustable Frequency Inverters, Heating / Piping / Air Conditioning, no. 8,
1984, pp. 83–87.
Braun, J. E., Klein, S. A., and Mitchell, J. W., Effective Models for Cooling Towers and Cooling Coils, ASHRAE
Transactions, 1990, Part I, pp. 164–174.
Braun, R. H., Problem and Solution to Plugging of a Finned-Tube Cooling Coil, ASHRAE Transactions, 1986,
Part I B, pp. 385–387.
Brown,W. K., Humidification by Evaporation for Control Simplicity and Energy Savings, ASHRAE Transactions,
1989, Part I, pp. 1265–1272.
Burroughs, H. E. B., IAQ: An Environmental Factor in the Indoor Habitat, HPAC, no. 2, 1997a, pp. 57–60.
Burroughs, Filtration: An Investment in IAQ, HPAC no. 8, 1997b, pp. 55–65.
Chen, J. C., Correlation for Boiling Heat Transfer to Saturated Fluids in Convective Flow, Ind. Eng. Chem.
Process Design Develop., vol. 5, no. 3, 1966.
Coward, Jr., C. W., Unhoused (Plug / Plenum) Fans: Is Their Performance Predictable? ASHRAE Journal, no. 10,
1997, pp. 71–76.
Davis,W. J.,Water Spray for Humidification and Air Flow Reduction, ASHRAE Transactions, 1989, Part II, pp.
351–356.
Deacon,W. T., Important Considerations in Computer Rooms Humidification, ASHRAE Transactions, 1989,
Part I, pp. 1273–1277.
DeBat, R. J., Humidity: The Great Equalizer, HPAC, no. 10, 1996, pp. 66–71.
Delaney, T. A., Maiocco, T. M., and Vogel, A. G., Avoiding Coil Freezeup, Heating / Piping / Air Conditioning,
no. 12, 1984, pp. 83–85.
Goldfield, J., Use Total Pressure When Selecting Fans, Heating/ Piping / Air Conditioning, no. 2, 1988, pp. 73–82.
Goswami, D., Computer Room Humidity Control, Heating / Piping / Air Conditioning, November 1984,
pp. 123–124.
Green, G. H., The Positive and Negative Effects of Building Humidification, ASHRAE Transactions, 1982,
Part I, pp. 1049–1061.
Grimm, N. R., and Rosaler R. C, Handbook of HVAC Design, McGraw-Hill, New York, 1990.
Gurock, D. R., and Aldworth, D. R., Fan-Motor Combination Saves Energy, Heating / Piping / Air Conditioning,
no. 11, 1980, pp. 53–57.
Hartman, T., Humidity Control, Heating / Piping / Air Conditioning, September 1989, pp. 111–114.
Holtzapple, M. T., and Carranza, R. G., Heat Transfer and Pressure Drop of Spined Pipe in Cross Flow, Part III:
Air Side Performance Comparison to Other Heat Exchangers, ASHRAE Transactions, 1990, Part II,
pp. 136–141.
Hunt, E., Benson, D. E., and Hopkins, L. G., Fan Efficiency vs. Unit Efficiency for Cleanroom Application,
ASHRAE Transactions, 1990, Part II, pp. 616–619.
Kemp, S. J., Kuehn, T. H., Pul, D. Y. H., Vesley, D., and Streifel, A. J., Filter Collection Efficiency and Growth
of Microorganisms on Filters Loaded with Outdoor Air, ASHRAE Transactions, 1995, Part I, pp. 229–238.
Larocca, D. V., Chilled Water Coil Freeze Protection via Internal Drying, HPAC, no. 12, 1997, pp. 67–72.
Longo, F., Ultrasonic Humidification for Telecommunications, Heating / Piping / Air Conditioning, no. 3, 1994,
pp. 65–66.
AIR SYSTEMS: COMPONENTS—FANS, COILS, FILTERS, AND HUMIDIFIERS 15.85
McQuinston, F. C., Finned Tube Heat Exchangers: State of the Art for the Air Side, ASHRAE Transactions,
1981, Part I, pp. 1077–1085.
Morton, B. W., Humidification Handbook, B. W. Morton and Dri-Steam Humidifying Co., Hopkins, MN, 1981.
Morton, C. B., Six Steps to Follow That Ensure Proper Humidification System Design and Control, ASHRAE
Transactions, 1996, Part II, pp. 618–627.
Muller, C. O., and England,W. G., Achieving Your Indoor Air Quality Goals: Which Filtration System Works
Best? ASHRAE Journal, no. 2, 1995, pp. 24–32.
Nader, J. C., and Kanis, T. W., Industrial vs. Commercial Fans, Heating / Piping / Air Conditioning, no. 2, 1983,
pp. 55–62.
National Air Filtration Association, Air Filter Test Standards, Air Conditioning, Heating & Refrigeration News,
Aug. 11, 1997, pp. 5–6.
Obler, H., Humidification Alternatives for Air Conditioning, Heating / Piping / Air Conditioning, December
1982, pp. 73–77.
Osborne,W. C., Fans, 2d ed., Pergamon Press, Oxford, England, 1977.
Remiarz, R. J., Johnson, B. R., and Agarwal, J. K., Filter Testing with Submicrometer Aerosols, ASHRAE Transactions,
1988, Part II, pp. 1850–1858.
Rivers, R. D., Interpretation and Use of Air Filter Particle-Size-Efficiency Data for General Ventilation Applications,
ASHRAE Transactions, 1988, Part II, pp. 1835–1849.
Rosenbaum, D. P., Obtaining Proper Fan Performance from Fans Installed in Air Handlers, ASHRAE Transactions,
1983, Part I B, pp. 790–794.
Sheringer, J. S., and Govan, F., How to Provide Freeze-up Protection for Heating and Cooling Coils,
Heating / Piping / Air Conditioning, no. 2, 1985, pp. 75–84.
Symonds, S., Internal Coil Care: What’s the Best Solution? Air Conditioning, Heating and Refrigeration News,
Aug. 21, 1997, pp. 9–10.
Tao,W., and Chyi, D. P., Coil Design and Selection, Heating / Piping / Air Conditioning, no. 12, 1985,
pp. 66–73.
The Trane Company, Fans and Their Application in Air Conditioning, Application Engineering Seminar, Trane
Company, La Crosse, WI, 1971.
The Trane Company, Variax Fans, La Crosse, WI, 1983.
Thornburg, D., Filter Testing and IAQ Control Move Forward, HPAC, no. 10, 1999, pp.54–56.
Trent,W., and Trent, C., Condensate Control, Engineered Systems, no. 7, 1997a, pp. 40–44.
Trent,W., and Trent, C., Seal It with More Than a Kiss, Engineered Systems, no. 9, 1997b, pp. 52–58.
Wang, S. K., Air Conditioning, vol. 3, Hong Kong Polytechnic, Hong Kong, 1987.
Wood, R. A., Sheffield, J. W., and Sauer, Jr., H. J., Thermal Contact Conductance of Finned Tubes: The Effect
of Various Parameters, ASHRAE Transactions, 1987, Part II, pp. 798–810.
15.86 CHAPTER FIFTEEN
CHAPTER 16
AIR SYSTEMS: EQUIPMENT—
AIR-HANDLING UNITS AND
PACKAGED UNITS
16.1
16.1 AIR-HANDLING UNITS 16.1
Functions of Air-Handling Units 16.1
Classifications of Air-Handling Units 16.2
16.2 MAIN COMPONENTS 16.4
Casing 16.4
Fans 16.4
Coils 16.5
Filters 16.5
Humidifiers 16.5
Outdoor Air Intake, Mixing, and Exhaust
Section 16.6
Controls 16.6
Component Layout 16.6
Coil Face Velocity 16.8
16.3 SELECTION OF AIR-HANDLING
UNITS 16.9
16.4 PACKAGED UNITS 16.12
Types of Packaged Unit 16.12
Rooftop Packaged Units 16.12
Indoor Packaged Units 16.15
Split Packaged Units 16.16
16.5 PERFORMANCE AND SELECTION OF
PACKAGED UNITS 16.17
Indoor Environmental Control 16.17
Indoor Air Quality 16.18
Scroll Compressors and Evaporative
Condensers 16.18
Controls 16.18
Minimum Performance 16.19
Selection of Packaged Units 16.19
16.6 FAN ROOM 16.24
Types of Fan Room 16.24
Layout Considerations 16.25
REFERENCES 16.28
16.1 AIR-HANDLING UNITS
Functions of Air-Handling Units
An air-handling unit (AHU) is the primary equipment in an air system of a central hydronic system;
it handles and conditions the air and distributes it to various conditioned spaces. In an AHU, the
required amounts of outdoor air and recirculating air are often mixed and conditioned. The temperature
of the discharge air is then maintained within predetermined limits by means of control
systems. After that, the conditioned supply air is provided with motive force and is distributed to
various conditioned spaces through ductwork and space diffusion devices.
Many air-handling units are modular so that they have the flexibility to add components as
required. An AHU basically consists of an outdoor air intake and mixing box section, a fan section
including a supply fan and a fan motor, a coil section with a water cooling coil, a filter section, and
a control section. A return or relief fan, a heating coil, a precooling coil, and a humidifier may also
be included depending on the application. Supply volume flow rates of AHUs vary from 2000 to
63,000 cfm (945 to 29,730 L/ s).
Whether a return fan or a relief fan should be added to an air system depends on the construction
and operating characteristics of the air system and the total pressure loss of the return
system (see Chap. 22). A heating coil is mainly used in the air-handling unit that serves the perimeter
zone, or for morning warm-up in the heating season. The use of a precooling coil, to draw
cooling water from the cooling tower as a water economizer is discussed in Chap. 21. Humidifiers
are employed for processing air conditioning and health care facilites where space humidity must
be controlled.
Classifications of Air-Handling Units
Air-handling units may be classified according to their structure, location, and conditioning characteristics.
Horizontal or Vertical Unit. In a horizontal unit, the supply fan, coils, and filters are all installed
at the same level, as shown in Fig. 16.1a. Horizontal units need more floor space for installation,
and they are mainly used as large AHUs. Most horizontal units are installed inside the fan room.
Occasionally, small horizontal units may be hung from the ceiling inside the ceiling plenum. In
such a circumstance, fan noise and vibration must be carefully controlled if the unit is adjacent to
the conditioned space.
In a vertical unit, the supply fan is not installed at the same level as the coils and filters but is often
at a higher level, as shown in Fig. 16.1b. Vertical units require less floor space. They are usually
smaller, so that the height of the coil section plus the fan section, and the height of the ductwork
that crosses over the AHU under the ceiling, is less than the head room (the height from the floor to
the ceiling or the beam of the fan room). The fan room is the room used to house AHUs and other
mechanical equipment.
Draw-Through Unit or Blow-Through Unit. In a draw-through unit, the supply fan is located
downstream from the cooling coil section, and the air is drawn through the coil section, as shown in
Fig. 16.1a and b. In a draw-through unit, conditioned air is evenly distributed over the entire surface
of the coil section. Also the discharge air from the AHU can be easily connected to a supply duct of
similar higher velocity. Draw-through units are the most widely used AHUs.
In a blow-through unit, the supply fan is located upstream from the coil section, and the air
blows through the coil section, as shown in Fig. 16.1c. Usually, a multizone air-handling unit
adopts a blow-through unit. In a multizone AHU, the coil section is divided into the hot deck and
the cold deck. The heating coil is installed in the hot deck just above the cold deck, where the cooling
coil is located. The hot deck is connected to ductwork that supplies warm air to the perimeter
zone through the warm duct. The cold deck is connected to a cold duct that supplies cold air to both
the perimeter and interior zones.
A blow-through unit also has the advantage of treating the supply fan heat gain as part of the
coil load and thus reduces the supply system heat gain.
Outdoor Air (or Makeup Air) AHU or Mixing AHU. Most mixing AHUs can be used to condition
either outdoor air only or a mixture of outdoor air and recirculating air, whereas an outdoor air AHU is
used only to condition 100 percent outdoor air, as shown in Fig. 16.1d. An outdoor air, or makeup air,
AHU is a once-through unit; there is no return air and mixing box. It may be a constant-volume system
or a variable-air-volume (VAV) system if the number of occupants varies. In an outdoor-air AHU,
the cooling coil is usually a six- to eight-row depth coil because of the greater enthalpy difference
during cooling and dehumidification in summer. Freeze protections for water coils are necessary in
locations where the outdoor temperature may be below 32°F (0°C) in winter. A heat recovery coil or a
water economizer precooling coil is often installed in makeup air AHUs for energy savings.
Single-Zone AHU or Multizone AHU. A single-zone AHU serves only a single zone. A multizone
AHU serves two or more zones, as shown in Fig. 16.1c. A zone can be a large perimeter or an
interior zone or one of the many control zones which connect to a multizone AHU through ducts
and terminals.
A multizone AHU with a hot and cold deck is now often used for a dual-duct VAV system (see
Chap. 21). Another kind of multizone unit which has many separate warm air ducts and cold air
16.2 CHAPTER SIXTEEN
16.3
FIGURE 16.1 Type of air-handling units (AHUs): (a) horizontal, draw-through unit; (b) vertical draw-through
unit; (c) blow-through unit, multizone AHU; (d ) makeup air AHU, custom-built, rooftop unit.
ducts with associated warm and cold air dampers for each of the zones became obsolete because
this kind of multizone unit wastes energy, needs complicated control, and is expensive.
Factory-Fabricated AHU or Field-Built AHU, Custom-Built or Standard Fabrication. One important
reason to use factory-fabricated AHUs or standard fabrications is their lower cost and higher
quality. Factory labor and controlled manufacturing techniques provide more efficient and betterquality
construction than field labor and assembly.
Custom-built and field-built AHUs provide more flexibility in structure, system component
arrangements, dimensions, and specialized functions than standard fabricated products. Custombuilt
and field-built AHUs also need more comprehensive, detailed specifications. Standard fabricating
products are usually less expensive and can be delivered in a shorter time.
Rooftop AHU or Indoor AHU. A rooftop AHU is an outdoor penthouse, as shown in Fig. 16.1d.
It is usually curb-mounted on the roof and should be completely weathertight. The outside casing is
usually made of heavy-gauge galvanized steel or aluminum sheets with corrosion-resistant coating
and sealant at the joints, both inside and outside. The fan motor, water valves, damper actuator linkages,
and controls are all installed inside the casing. Access doors are necessary for service and
maintenance of fans, coils, and filters. An indoor AHU is usually located in the fan room. Small
AHUs are sometimes ceiling-hung.
16.2 MAIN COMPONENTS
Because of the impact of the indoor air quality (IAQ), the design and the construction of the AHU
have been affected in many ways, as discussed in Gill (1996).
Casing
Two kinds of casings are more widely used for new AHUs today: (1) a double-wall sheet-metal casing
in which the insulation material is sandwiched between two sheet-metal panels of 1- to 2-in.
(25- to 50-mm) thickness with a U value from 0.12 to 0.25 Btu /h ft2  °F (0.68 to 1.42 W/m2 °C)
and (2) single sheet-metal panel with inner insulation layer and perforated metal liners.
Although insulating materials such as glass fibers and mineral wool are inert, when they become
wet and collect dirt, both the glass fiber and the glass fiber liner provide the site and source of
microbial growth. In addition, glass fiber liner is susceptible to deterioration and erosion over time.
With a double-wall sheet-metal casing, glass fibers are not exposed to the moisture of the ambient
air. Its inner surface can also be cleaned easily. Perforated metal liners cannot isolate the isulating
material from the ambient moisture, but they are helpful to attenuate the fan noise.
The outside surface of the casing is often coated with an ultraviolet-resistant epoxy paint. The
interior surface is better coated with a light color paint which increases the ability to spot the debris
and microbial growth. Hinged access panels to the fan, coils, and filter sections must be provided
for inspection and maintenance. Thermal break construction employs a resin bridge between the exterior
and interior panels to interrupt the through-metal path heat transfer. A well-sealed doublewall
metal panel should have an air leakage of only 1 cfm / ft2 (5.078 L / s m2) panel at a 4 in. WC
(1000 Pa) of outer and inner static pressure difference.
Fans
A double-inlet airfoil, backward-inclined centrifugal fan is often used in large AHUs with greater
cfm (L / s) and higher fan total pressure for its higher efficiency and lower noise. Vane-axial fans
with carefully designed sound-absorptive housings, sound attenuators at inlet and outlet, and other
16.4 CHAPTER SIXTEEN
attenuations can now provide a sound rating of NC 55 in the fan room. Although the forwardcurved
centrifugal fan has a lower efficiency at full load, it is more compact and its part-load operating
characteristics are better than those of a backward-inclined centrifugal fan. It is often used in
small AHUs and where cfm (L/ s) and fan total pressure are lower. For VAV systems, a dedicated
outdoor air injection fan is sometimes used to provide outdoor ventilation air according to demand
at both full and part load. An axial relief fan or an unhoused centrifugal return fan may be added as
an optional system component. A return fan is used when the total pressure loss of the return
system is considerable. This is discussed in Chap. 22. Large fans are usually belt-driven. Only small
fans are sometimes direct-driven.
As mentioned in Sec. 15.4, an adjustable-frequency variable-speed drive saves more energy than
inlet vanes for VAV systems during part-load operation. It is often cost-effective for large centrifugal
fans although a variable-speed drive is expensive. Inlet vanes are not suitable for small airfoil or
backward-inclined centrifugal fans because they block the air passage at the fan inlet.
Generally, a centrifugal fan has a higher efficiency and at the same time a lower noise. Given
two fans of the same model, both with the same cfm (L / s) and fan total pressure, a centrifugal fan
that is greater in size and slower in speed creates less noise.
Coils
In AHUs, the following types of coil are often used: water cooling coils, water heating coils, electric
heating coils, and water precooling coils. The construction and characteristics of these coils are
discussed in Chap. 15. Electric heating coils are made with nickel-chromium wire as the heating element
(see Sec. 8.4). Ceramic bushes float the heating elements, and vertical brackets prevent the
elements from sagging. In a finned tubular element sheathed construction, the electric heating coil
is usually made with a spiral fin brazed to a steel sheath. An electric heating coil is usually divided
into several stages for capacity control.
When an electric heating coil is installed inside an AHU, the manufacturer should have the
assembly tested by the Underwriters’ Laboratory (UL) to ensure that its requirements are met. Otherwise,
the heater may only be installed outside the AHU as a duct heater, with a minimum distance
of 4 ft (1.2 m) from the AHU. Various safety cutoffs and controls must be provided according to the
National Electrical Codes and other related codes.
Cooling and heating coils need periodic cleaning and freeze-up protection in locations where the
outdoor air temperature may drop below 32°F (0°C) in winter. Condensate pan and condensate
drain line must be properly designed and installed. All these are discussed in Chap. 15.
Filters
Air filtration is an important component to achieve an acceptable indoor air quality. In AHUs, earlier
low-efficiency filters of the panel type are giving way to the medium- and high-efficiency bag
type and cartridge type of filters. Carbon-activated gaseous absorption filters are also used to
remove objectionable odors or volatile organic compounds (VOCs) in buildings. Newly developed
air filters are more efficient at removing air contaminants of particle size between 0.3 and 5 m
which are lung-damaging dust.
Humidifiers
Usually, there is no humidifier installed in the AHU for comfort air conditioning systems; but the
outdoor climate is very cold in winter so that if a humidifier is not employed, the winter indoor
relative humidity may be too low. Humidifiers are necessary for health care facilities and processing
systems in pharmaceutical, semiconductor, textile, communication centers, and computer rooms.
Steam grid or electric heating element humidifiers are widely used in AHUs where a warm air
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.5
supply and humidity control are needed in winter. Ultrasonic humidifiers are often used for buildings
in which a cold air supply and humidity control are required. For industrial applications such
as textile mills where humidity control, air washing, and cold air supply are needed all year round,
an air washer is often used for these purposes.
Outdoor Air Intake, Mixing, and Exhaust Section
An outdoor air intake, mixing, and exhaust section includes an outdoor air intake, an exhaust outlet,
dampers, a mixing box, and a return fan or a relief fan. The location of outdoor air intake has a direct
impact on space IAQ.
 The outdoor air intake for each AHU should install with wind shield and louvers to prevent rain
and birds. If the AHU is located on the roof, the bottom of the outdoor intake louvers should be at
least 3 ft above the the roof.
 The outdoor air intake must be located as far away from the exhaust outlets and plumbing vent
stacks (horizontally and vertically) as possible, to prevent the intake of exhaust contaminants,
condensation, and freezing which may provide a means of growth of microorganisms. Codes and
local regulations should be followed.
 The outdoor air intake should reflect the influence of the prevailing winds.
 An outdoor intake system should be provided with air filters or even air cleaners in locations
where outdoor air contaminants exceed the National Primary Air Quality Standard, as discussed
in Sec. 4.10.
For better outdoor ventilation air control, an outdoor damper should split into two dampers: a
minimum outdoor ventilation damper and an economizer damper of 100 percent outdoor air free
cooling except in small AHUs. Both should provide a short ducted outside path for air balancing
and install with airflow measuring station, or a minimum outdoor ventilation air injection fan and
control if necessary.
Poor mixing, such as parallel outdoor and recirculating airstreams in the mixing box, causes
stratification of the mixture, as discussed in Sec. 15.11. This is one of the important causes of coil
freezing in locations where the outdoor air temperature drops below 32°F (0°C). For good mixing,
airstreams should meet at a 90° angle or opposite each other, as shown in Fig. 16.2.
Recently, there is a trend to use an unhoused plug / plenum fan as a return fan in many AHUs
and PUs. Usually, a return fan is located nearly in the center of an exhaust compartment, as shown
in Fig. 16.3. The exhaust and recirculating dampers form two sides of the exhaust compartment. A
plug / plenum fan has the advantage of discharging from both exhaust and recirculating dampers and
is quieter. After the recirculating damper, the mixture enters the coil section.
Controls
As discussed in Sec. 5.14, AHU controls include generic, specific, safety, and diagnostic controls.
The economizer control, discharge air temperature control, duct static pressure control, outdoor
ventilation air control, humidity, and warm-up or cool-down controls are discussed in Chap. 21,
22, and 23.
Component Layout
In a typical horizontal, drawn-through AHU, the layout of the components in serial order is usually
as follows:
16.6 CHAPTER SIXTEEN
1. Return or relief fan, exhaust air passage and damper (optional)
2. Mixing box with outdoor air and recirculating air dampers
3. Filters: (prefilter, optional) medium-efficiency filters
4. Preheating coil (optional)
5. Precooling coil (optional)
6. Cooling coil
7. Heating coil (optional)
8. Supply fan
9. Humidifier (optional)
10. High- or ultrahigh-efficiency filters (optional)
If there is a relief fan, it should be located on one side of the exhausting compartment or in the
relief or exhaust passage. If there is an unhoused return fan, it should be located nearly in the center
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.7
FIGURE 16.2 Mixing of outdoor and
recirculating airstreams: (a) parallel
airstreams (poor mixing); (b) airstreams at
90° (good mixing); (c) opposite airstreams
(good mixing).
of the exhaust compartment. The volume flow and fan total pressure of the return fan should be
carefully determined so that the pressure inside the exhaust compartment is positive; and the pressure
inside the mixing box must be negative in order to extract outdoor air, except an outdoor air
projected fan is installed in the outdoor air passage (see Chaps. 21 and 23).
In an AHU, to condition a mixture of outdoor and recirculating air is often simpler and less
expensive than to condition the outdoor air and recirculating air separately. Therefore, the mixing
box is usually located before the filters and the coils. A preheating coil is always located before the
water heating and cooling coils for the sake of freeze-up protection. A precooling coil is always
located before a cooling coil for a greater temperature difference between the air and water. A
steam grid humidifier is usually located after a heating coil because humidification is more effective
at a higher air temperature.
If there are ultrahigh-efficiency filters, they should be located as near to the clean room or clean
space as possible to prevent pollution from the ductwork or elsewhere. Low- or medium-efficiency
prefilters must be installed ahead of the coils.
Coil Face Velocity
The size, or more accurately, the width and height of a horizontal AHU, is mainly determined
by the face velocity of the coil. A higher coil face velocity results in a smaller coil, a higher heattransfer
coefficient, a greater pressure drop across the coil and filter, and a smaller fan room, which
directly affects the space required. On the contrary, a lower coil face velocity has a larger coil, a
lower heat-transfer coefficient, and a smaller pressure drop and fan energy use.
The maximum face velocity is usually determined according to the value required to prevent
carryover of water droplets due to condensate from a cooling coil. As mentioned in Secs. 10.2 and
16.8 CHAPTER SIXTEEN
Supply
air
Exhaust
air
Outdoor
air
Return
air
Exhaust
compartment
Unhoused
return air fan
High-efficiency
filters
Mediumefficiency
filters
Supply
fan
Coils
FIGURE 16.3 A typical AHU with unhoused plug/plenum return fan.
15.10, the cooling coil face velocity should not exceed 500 fpm (2.5 m / s) for a smooth fin coil and
550 fpm (2.75 m / s) for corrugated fins. The lower limit of the coil face velocity depends mainly on
the initial and energy cost analysis, including cost of the AHU, fan room size, the total number of
annual operating hours, and the unit rate of electric power. A lower limit of face velocities between
400 and 450 fpm (2 and 2.25 m / s) at design conditions may be considered appropriate under many
specific circumstances.
For a fan room of adequate head room, it is preferable to use a high coil to reduce the face velocity
and the pressure drop of coils and filters. Use of a high coil has little influence on the floor area of
the fan room except that when a coil is higher than 42 in. (1070 mm), a cooling coil should split into
two coils vertically and use two separate condensate pans vertically to prevent condensate carryover.
16.3 SELECTION OF AIR-HANDLING UNITS
Table 16.1 lists general data of the supply fan and coil of a typical horizontal draw-through AHU; fan B
means a class II fan. In Table 16.2, the volume flow–fan static pressure performance of this horizontal
draw-through modular AHU (unit size 30) with inlet vanes is presented. A backward-inclined centrifugal
fan of 22 -in. (565-mm) diameter is used. In Table 16.2, the following items are also listed:
 Supply volume flow rate , in cfm of standard air.
 Air velocity at fan outlet, fpm.
 Fan static pressure, in in. WC. Fan total pressure can be obtained by adding the fan velocity pressure
to the fan static pressure; velocity pressure pv, in in. WC, can be calculated by
here vout indicates the fan outlet velocity, in fpm.
 Revolutions per minute (rpm) of fan impeller.
 Brake horsepower (bhp) input to fan shaft.
The following are recommendations for selection of AHUs from a manufacturer’s catalog:
 Draw-through AHUs are widely used. For a small AHU, a vertical unit saves floor space if the
headroom of the fan room is sufficient except the AHU is ceiling-hung. For a large AHU, a horizontal
unit is often the right choice.
 The size of the AHU is selected so that the face velocity of the cooling coil vcoil is optimum. For
corrugated fins, maximum vcoil should not exceed 550 fpm (2.75m/ s).
 For large AHUs, choose a backward-curved centrifugal fan with airfoil blades or a backwardinclined
fan for higher efficiency. Select the rpm of the supply fan or supply and return fans in
order to meet the required system total pressure loss, i.e., external total pressure plus the total
pressure loss of the AHU.
 For VAV systems, an adjustable-frequency variable-speed drive should be compared with inlet
vanes via life-cycle cost analysis. Energy consumption also should be analyzed at part-load
operation. For a large AHU, an adjustable-frequency variable-speed drive may be cost-effective.
When an AHU is equipped with a small airfoil or backward-inclined centrifugal fan, inlet
vanes for capacity control are not recommended because of the extremely high air velocity at the
fan inlet.
 The required coil load is met through the variation of the number of rows of coil and the fin spacing.
An even number of rows are often used so that the inlet and the outlet of the coils are on the
same side. Four-row coils are often used for a mixing AHU, and a makeup AHU seldom uses a
coil that exceeds eight rows.
 In locations where outdoor air temperatures go below 32°F (0°C), coil freeze-up protection
attained by installing a preheating coil and the improvement of the mixing of airstreams in the
mixing box should be considered.
pv  (vout /4005)2
V?
s
14
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.9
16.10
TABLE 16.1 General Data on Supply Fans and Coils of a Typical Horizontal Draw-through AHU
Unit size number
Description 3 6 8 10 12 14 17 21 25 30
Fan data
Fan Size — — — 13 15 16 18 20 22 22
BI Shaft size (in.) — — — 1 1 1 2 2 2 2
Outlet area (ft2) — 1.41 1.90 2.31 2.79 3.39 4.14 5.05 6.30 6.30
Max. rpm/ static pressure (in. WC) — — — 3536/8 3183/8 2894/8 2616/8 2387/8 2164/8 2164/8
Motor hp range — — — 1–10 1–15 1–15 1–20 1–25 1–25 1–30
Coil data
Unit coils ( -in. tube)
Cooling Ft2 3.32 5.86 7.54 9.64 12.3 14.2 16.8 20.8 24.4 29.0
Dimensions (in.) 21  23 23  36 27  40 27  51 32  55 35  59 37  65 45  67 51  69 51  82
Heating Ft2 2.34 4.31 5.49 7.01 9.46 10.2 12.3 15.0 17.8 21.2
Dimensions (in.) 15  23 17  36 20  40 20  51 25  55 25  59 27  65 32  67 37  69 37  82
-in. or 1-in. tube coils
Ft2 2.75 4.38 6.50 8.33 11.3 12.1 14.7 16.5 19.8 23.6
Dimensions (in.) 18  22 18  35 24  39 24  50 30  54 30  58 33  64 218  66* 118  68* 118  81*
124  68 124  81
BI: Backward-inclined centrifugal fan
*Two coils are used for this unit.
Source: The Trane Company. Reprinted with permission.
5
8
1
2
11
16
11
16
7
16
3
16
15
16
15
16
11
16
1
4
1
4
1
4
1
2
1
2
16.11
TABLE 16.2 Volume Flow and Fan Static Pressure of a Typical Horizontal Draw-through Modular AHU (Size 30) with Inlet Vanes (Backward-
Inclined Centrifugal Fan of 22.25-in. Diameter)
Fan static pressure, in. WC
3.50 3.75 4.00 4.25 4.50 4.75 5.00
CFM Outlet
std. air velocity rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp
10,000 1,587 1,524 10.39 1,561 10.91 1,597 11.40 1,632 11.87 1,667 12.32 1,701 12.77 1,734 13.23
11,000 1,746 1,570 11.81 1,607 12.54 1,643 13.23 1,678 13.86 1,712 14.44 1,745 14.99 1,777 15.51
12,000 1,905 1,620 13.21 1,656 14.00 1,691 14.81 1,725 15.62 1,759 16.41 1,792 17.16 1,824 17.85
13,000 2,063 1,678 14.90 1,710 15.67 1,742 16.46 1,774 17.28 1,806 18.14 1,838 19.02 1,870 19.90
14,000 2,222 1,741 16.82 1,771 17.61 1,800 18.41 1,830 19.23 1,860 20.08 1,889 20.94 1,919 21.84
15,000 2,381 1,809 18.98 1,836 19.80 1,864 20.63 1,891 21.47 1,919 22.32 1,946 23.20 1,974 24.09
16,000 2,540 1,880 21.36 1,906 22.23 1,931 23.10 1,957 23.97 1,982 24.85 2,008 25.75 2,034 26.65
17,000 2,698 1,953 23.97 1,978 24.88 2,002 25.80 2,026 26.72 2,050 27.64 2,074 28.57 2,098 29.71
18,000 2,857 2,028 26.83 2,052 27.78 2,075 28.74 2,098 29.71
19,000 3,016 2,104 29.96
5.25 5.50 5.75 6.00 6.25 6.50
CFM Outlet
std. air velocity rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp
10,000 1,587 1,766 13.70 1,798 14.19 1,830 14.71 1,861 15.26 1,892 15.82 1,923 16.41
11,000 1,746 1,809 16.01 1,840 16.51 1,871 17.00 1,901 17.51 1,931 18.02 1,960 18.55
12,000 1,905 1,854 18.48 1,885 19.09 1,915 19.68 1,944 20.24 1,973 20.79 2,002 21.33
13,000 2,063 1,901 20.76 1,932 21.58 1,962 22.35 1,991 23.07 2,020 23.76 2,048 24.41
14,000 2,222 1,949 22.77 1,979 23.71 2,008 24.66 2,037 25.60 2,066 26.50 2,094 27.36
15,000 2,381 2,002 25.00 2,030 25.94 2,057 26.90 2,085 27.9 2,113 28.90 2,140 29.92
16,000 2,540 2,060 27.57 2,086 28.51 2,112 29.46
Outlet vel–outlet velocity, fpm
rpm–revolutions/minute
bhp–brake horsepower
Source: The Trane Company. Reprinted with permission.
 To improve indoor air quality, medium- and high-efficiency filters should be used to provide an
acceptable IAQ and to protect coils and air distribution devices. Dirty coils and condensate pans
significantly degrade the IAQ. A prefilter must be installed to extend the service life of high- and
ultrahigh-efficiency filters, or gas absorbers.
 Use an air or water economizer to save energy. Use indirect and direct evaporative coolers to
replace part of the refrigeration if they are applicable and economical. Evaporative cooling is
discussed in Chap. 27.
 Select an AHU with adequate specific and safety controls and diagnostics.
16.4 PACKAGED UNITS
Types of Packaged Unit
A packaged unit (PU) is a unitary, self-contained air conditioner. It is also the primary equipment of
a unitary packaged system. A packaged unit not only conditions the air and provides the motive
force to supply the conditioned air to the space, but also provides gas heating, or electric heating,
and refrigeration cooling from its own gas-fired furnace and refrigerating equipment or from its
own heat pump. It is actually the primary equipment in an air conditioning system.
A packaged unit is always equipped with DX coil(s) for cooling. This characteristic is the
primary difference between a packaged unit and an air-handling unit. The portion that handles
conditioned air in a packaged unit is called an airhandler, the air system of the packaged system, to
distinguish it from an air-handling unit, as shown in Fig. 16.4. Another portion is a condensing unit,
the refrigerant plant. Refrigerants HCFC-22, HFC-134a, HFC-404A, HFC-410A, HFC-407A, and
HFC-407C are now used in packaged units. Most PUs are factory-built standard fabrication units.
A packaged unit can be either enclosed in a single package or split into two units: an indoor air
handler and an outdoor condensing unit. A packaged unit can also be a packaged heat pump; most
are air-source heat pumps. In a packaged heat pump, in addition to the fan, DX coil, filters, compressors,
condensers, expansion valves and controls, there are four-way reversing valves to reverse
the refrigerant flow when cooling mode operation is changed to heating mode operation.
The construction and size of a packaged unit depend mainly on its model and cooling capacity,
in refrigeration tons or Btu / h (W). Packaged units can be classified according to their place of
installation as rooftop packaged units, indoor packaged units, and split packaged units. Among
these units, the rooftop packaged units are most widely used in commercial buildings.
Rooftop Packaged Units
A rooftop packaged unit is mounted on the roof of the conditioned space, as shown in Fig. 16.4. It
is usually enclosed in a weatherproof outer casing. The mixture of outdoor air and recirculating air
is often conditioned in the rooftop packaged unit and supplied to the conditioned space on the floors
below. Based on the types of heating and cooling sources, rooftop packaged units can be subdivided
into the following:
 Gas/ electric rooftop packaged unit. In this unit, heating is provided by a gas-fired furnace, and
cooling is provided by electric-driven reciprocating or scroll compressors.
 Electric / electric rooftop packaged unit. In this unit, heating is provided by electric heating coils
and cooling by reciprocating or scroll compressors.
 Rooftop packaged heat pump. In this unit, heating and cooling are provided by the heat pump,
with auxiliary electric heating if necessary.
16.12 CHAPTER SIXTEEN
A rooftop packaged unit is a single packaged unit composed of two main components: an air handler
and a condensing unit. Its cooling capacity may vary from 3 to 220 tons (10 to 774 kW), and
its supply volume flow rate may vary from 1200 to 80,000 cfm (565 to 37,750 L / s). An air handler
of a typical rooftop packaged unit consists of mainly a casing, an indoor fan, DX coils, filters, a
mixing box, and controls; a gas-fired heater, a relief or return fan, and a humidifier are optional. The
construction and characteristics of the casing, fans, filters, and outdoor air intake and mixing box
are similar to those discussed in AHUs.
Curb. A rooftop packaged unit is mounted on a curb which is a perimeter frame supporting the
unit. A curb is often made of galvanized or aluminized steel sheets or angle iron and sits on a deck.
It usually has additional structural support at or beneath the deck. On the top of the curb, there is
always a mating flange that matches the size of the rooftop unit with a sealing gasket to provide a
weatherproof joint. Curbs are either factory-prefabricated or field-assembled.
Curbs should be tall enough above the structural support for a sloped roof. The manufacturer
and the structural engineer should be consulted to ensure that the unit can resist the local anticipated
wind pressure. The requirements of National Roofing Contractors Association (NRCA)
should be followed.
DX Coils. For a specific model and size of rooftop packaged unit, the coil surface area is a fixed
value. DX coils are usually of two, three, and four rows (except makeup units) with a fin spacing of 12
to 17 fins/ in. (1.5- to 2.1-mm fin spacing). For large units, two separate refrigerant circuits with their
own coils, associated expansion valves, and distributors are often used for a better capacity control.
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.13
FIGURE 16.4 Cutaway view of a typical rooftop packaged unit. (Source: the Trane Company. Reprinted with
permission.)
The cooling capacity of a rooftop packaged unit should include the additional solar heat
absorbed by the unit on the rooftop. Because a DX coil is a wet coil, the same as the water cooling
coil, condensate drain pans, condensate traps, and the condensate drain line should be properly designed
and installed. Since refrigerant is used as the coolant instead of chilled water, coil freeze-up
protection is no longer necessary.
Supply, Return, and Relief or Exhaust Fans. For a rooftop packaged unit with a cooling capacity
of 10 tons (35 kW) or less, there is often only a supply (indoor) fan. For a rooftop packaged unit of
cooling capacity of 15 to 30 tons (53 to 105 kW), there is often a supply fan and a relief (exhaust)
fan. For a rooftop packaged unit of cooling capacity of 60 tons (210 kW) and greater, some manufacturers
offer a supply fan and a return fan. As in AHUs, an unhoused plug / plenum return fan is
also often located nearly in the center of an exhaust compartment in a rooftop packaged unit.
Supply and return fans in rooftop packaged units are usually belt-driven. For each refrigeration
ton (3.5 kW) of cooling capacity, a rooftop packaged unit usually has a nominal supply volume
flow rate of 350 to 450 cfm / ton (47 to 60 L / s kW). However, a rooftop packaged unit can vary its
speed of supply fan and return fan to provide various volume flow rates and fan total pressure for a
specific model and size. External pressure is the pressure loss of the duct system and terminals.
Most of the rooftop packaged units can vary their supply volume flow rate at a range between 200
and 500 cfm / ton (27 and 67 L / s kW). A maximum of 6-in. (1500-Pa) fan total static pressure or
4-in. (1000-Pa) external pressure can be provided by a supply fan in rooftop packaged units of 30
tons (105 kW) and greater.
There are also manufacturers that offer inlet vanes, inlet cone, or adjustable frequency variablespeed
drive for rooftop packaged units of 20 tons (70 kW) and greater to modulate volume flow
rates of supply and return fans at part-load operation in variable-air-volume systems. For variablespeed
drives, cooling air must be supplied to the elecronic drive mechanism to prevent encountered
high temperature because of the solar heat on the rooftop.
Gas-Fired Furnace and Electric Heating Coil. The gas burners in a rooftop packaged unit of a
heating capacity of 40,000 Btu / h (11,720 W) and greater are power burners of the induced combustion
type. A centrifugal blower is used to extract combustion air and combustion products to a
vent. There are often two tubular heat exchangers: a 16-gauge (1.6-mm-thick) stainless-steel or
aluminum-silicon alloy coated steel primary heat exchanger, and an 18-gauge (1.3-mm-thick) stainless-
steel or corrosion-resistant alloy coated steel secondary exchanger. Both are of free-floating
design to eliminate expansion and contraction stresses and noises.
Firing rate is often controlled by modulating the gas valve through a DDC controller. Safety
controls for gas-fired furnace include the proving of combustion air supply prior to ignition and
continuous electronic flame supervision.
An electric heater is often made of internally wired heavy-duty nickel-chromium elements. It is
usually divided into multiple stages or units for capacity control.
Humidifiers. A humidifier is optional. Indoor packaged units for computer room and data processing
systems are often installed with steam or heating element humidifiers in a position between
the coil section and the supply fan inlet.
As in the AHU, the outdoor intake of a rooftop unit should be shielded from the wind effect,
located as far away as possible from the contaminated air source, and the bottom of the louvers
should be kept at least 3 ft (0.9 m) from the roof.
In a rooftop packaged unit, the condensing unit mainly consists of compressors; an air-cooled,
evaporatively cooled, and water-cooled condenser; and controls.
Compressors. Semihermetic and hermetic reciprocating compressors and scroll compressors are
often used. Scroll compressors are more energy-efficient than reciprocating compressors. They
need fewer parts and are quieter. Scroll compressors are gradually replacing reciprocating compressors
in packaged units. For medium-size and large rooftop packaged units, two, three, or four
16.14 CHAPTER SIXTEEN
compressors of equal or sometimes unequal horsepower input are preferable for better capacity
control in steps.
Condensers. Multirow, -in. (10-mm) copper tube and aluminum fin air-cooled condensing coil
connected with subcooling coils are used. Condensing coil may cover the two sides of the condenser
or be in a V shape at the middle. The ratio of the face area of the condensing and subcooling
coils to the cooling capacity is often 1.5 to 2 ft2 / tonref (0.04 to 0.053 m2/kWref). Propeller fans are
used for the induced-draft condenser fan. The ratio of volume flow of cooling air to the cooling capacity
is between 600 and 900 cfm / tonref (81 and 121 L / s kWref). Air-cooled condensers are most
widely used in rooftop packaged units today for their lower initial cost and simple operation and
maintenance.
An evaporative condenser has the advantage of low condensing temperature, similar to a watercooled
condenser with cooling towers, as well as lower energy consumption but at a lower initial
cost. Today, an evaporatively cooled condenser has become one of the optional alternatives in a
rooftop packaged unit and is used more and more frequently than before.
Heat Pump. A rooftop packaged heat pump is a packaged unit installed with four-way reversing
valves to change the refrigerant flow after the compressor. In an air-source heat pump, the DX coil
is often called an indoor coil, and the air-cooled condensing coil the outdoor coil. During cooling
mode operation, hot gas from the compressor is first discharged to the outdoor coil for condensing
and subcooling. The liquid refrigerant then enters the expansion valve and indoor coil to produce
refrigeration. During the heating mode operation, the reversing valve changes its connections and
the direction of refrigerant flow. Hot gas from the compressor now enters the indoor coil to release
its condensing heat first. The liquid refrigerant is then discharged to the outdoor coil to absorb heat
from the ambient air for evaporation. Detailed operation and system performance are discussed in
Sec. 12. 2.
Indoor Packaged Units
An indoor packaged unit is also a single packaged, factory-built unit. It is usually installed indoors
inside a fan room or a machinery room, as described in Sec. 9.21. A small or medium-size indoor
packaged unit may sometimes be floor-mounted directly inside the conditioned space with or
without connected ductwork, such as the indoor packaged unit in computer rooms, as shown in
Fig. 16.5. The cooling capacity of the indoor packaged unit may vary from 3 to 100 tons (10 to 350
kW), and its supply volume flow rate from 1200 to 40,000 cfm (565 to 18,880 L / s).
Indoor packaged units can be classified as follows:
 Indoor packaged cooling units. Only cooling is provided by the DX coil.
 Indoor packaged heating / cooling units.These units not only provide cooling from the DX coil,
but also provide heating from a hot water coil, steam heating coil, or electric heater.
 Indoor packaged heat pump. When an indoor packaged unit is connected to an air-cooled condenser
and equipped with reversing valves, the change of refrigerant flow also causes the change
of cooling mode and heating mode operation and provides heating and cooling as required.
In indoor packaged units, usually only a supply fan is installed in small units. Because of the compact
size of the units, generally a forward-curved centrifugal fan is used. For large indoor packaged
units, an additional return fan section is added to extract recirculating air from the conditioned
space through the return duct.
For computer room indoor packaged units, return air entering the unit at high level and supply
air discharged at low level are common. Medium-efficiency filters or sometimes high-efficiency
filters are usually used. A steam humidifier or other type of humidifier is always an integral part of
computer room units, so as to maintain a required space relative humidity in winter and prevent
static electricity. Multiple semihermetic and hermetic reciprocating compressors or scroll compres-
38
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.15
sors—with dual refrigerant circuits—are used in medium-size and large units for capacity control.
A microprocessor-based DDC control is always used for new and retrofit indoor packaged units.
An indoor packaged unit differs from a rooftop packaged unit in condensing arrangements. Usually
there are two alternatives in indoor packaged units:
 With an air-cooled condenser, hot gas from the compressor is discharged to an air-cooled condenser
through the discharge line located on the rooftop. Liquid refrigerant is returned to the DX
coil from the air-cooled condenser through the liquid line.
 A shell-and-tube or a double-tube water-cooled condenser is installed inside the unit, and the condenser
water is supplied from the cooling tower or from other sources.
An economical analysis based on the local conditions should be made to determine whether an
air-cooled or a water-cooled condenser should be installed. If a water-cooled condenser using condenser
water from the cooling tower is used, a precooling coil is sometimes installed in the indoor
packaged unit as a component of the water-side economizer.
Split Packaged Units
A split packaged unit, sometimes called a split system, splits the packaged unit into an indoor air
handler and an outdoor condensing unit, which is most probably mounted outdoors, on the rooftop,
on a podium, or in some other adjacent place, as shown in Fig. 16.6. The indoor air handler and
outdoor condensing unit are connected by refrigerant pipes.
An air handler in a split packaged unit is similar to the air handler in a rooftop unit except that a
large air handler in a split packaged unit is usually installed inside the fan rooms, whereas the
small air handler may be hung under the ceiling. An air handler for split packaged units usually has
a cooling capacity from 3 to 80 tons (10 to 280 kW), a supply volume flow rate of 1200 to 32,000
cfm (565 to 15,100 L / s), and a maximum fan total pressure of 5.0 in. WC (1250 Pa) for mediumsize
and large units.
16.16 CHAPTER SIXTEEN
FIGURE 16.5 A typical indoor packaged unit.
Reciprocating and scroll compressors are used in split packaged units. The condenser in the outdoor
condensing unit can be either air-cooled or water-cooled. Compared with indoor packaged
units, a split packaged unit always has its compressors in its outdoor condensing unit, whereas an
indoor packaged unit has its compressor, water-cooled condenser indoors.
A split packaged heat pump is also a kind of split packaged unit. In such a unit, additional reversing
valves or changeover arrangements force the refrigerant flow from the compressor to the
outdoor air-cooled condensing coil during the cooling mode operation, and change to the indoor
DX coil during the heating mode operation. The cooling capacity of a split packaged heat pump
varies from 10 to 30 tons (35 to 105 kW), and the heating capacity ranges from 100,000 to 400,000
Btu/ h (29,300 to 117,200 W) at rated conditions.
16.5 PERFORMANCE AND SELECTION OF PACKAGED UNITS
A packaged unit is the primary equipment of a packaged air conditioning system. Its function and
performance represent the quality and capability of the packaged system. Because of the development
of microprocessor-based DDC control systems in the 1980s, and the effort toward energy conservation
since the energy crisis in 1973, the difference in indoor environmental control and indoor
air quality between a custom-built central system and a packaged system becomes smaller in the
late 1990s, according to the following analyses.
Indoor Environmental Control
Today, there are so many types, models, and sizes of packaged units, such as rooftop, indoor, and
split units with gas-fired, electric, hot water, and steam heating at a volume flow from 1200 to
48,000 cfm (565 to 22,650 L/ s) and a fan total pressure from 2.0 to 6.0 in. WC (500 to 1500 Pa)
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.17
FIGURE 16.6 A typical split packaged unit.
to meet various requirments of HVAC&R systems. Factory-fabricated and -assembled packaged
units usually have a higher quality than the field-assembled central system of nearly the same
construction.
Many manufacturers offer variable-air-volume packaged units of cooling capacity of 20 tons
(70 kWref) and greater. A VAV packaged unit maintains a required zone temperature by varying the
volume flow rate of supply air using proportional-integral control mode through each terminal.
Such a control is a modulation control, offset-free, fast response, and energy-efficient.
Some packaged units provide humidifiers to control the space relative humidity as required.
Most of the packaged units offer simultaneously operated air economizer and refrigeration compressor
at a cooling capacity of 5 tons (18 kW) and greater.
Indoor Air Quality
Adequate outdoor ventilation air, higher-efficiency air filters, and a clean and effective air system
directly affect the indoor air quality. A packaged unit is as effective at providing adequate outdoor
ventilation air for the occupants in the conditioned space as an AHU in a central system. With the
increase of the fan total pressure (external pressure), most medium-size and large packaged units
can now be installed with high-efficiency filters.
Scroll Compressors and Evaporative Condensers
Many packaged units now use scroll compressors instead of reciprocating compressors. In 1997,
one scroll compressor manufacturer in the United States announced that the energy efficiency ratio
(EER) of its product was 11.5 Btu /hW (3.37 COP). A rooftop packaged unit using an evaporative
condenser can lower the condensing temperature, approaching a value that only a water-cooled
condenser can achieve, and therefore is energy-efficient.
Controls
Microprocessor-based specific controls, safety controls, and diagnostics for 1997 manufactured
rooftop packaged units with a cooling capacity between 20 to 130 tons (70 to 455 kWref) are as
follows:
1. Discharge air temperature control
 Cooling: air economizer control
 Cooling: staging of scroll compressors
 Heating: modulation of gas valve—continuous flame supervision
 Reset based on outdoor or zone temperature
2. Duct static pressure control
 Inlet vanes /variable-speed drive control
 Maximum duct static pressure control
3. Space pressure control
4. Morning warm-up—full or cycling capacity
5. Minimum ventilation control
 Space positive pressurization, exhaust mode, and purge mode
6. DX coil frost protection
7. Occupied / Unoccupied switching—night setback
8. Low ambient compressor lockout
16.18 CHAPTER SIXTEEN
9. Compressor lead / lag control—a more balanced run time among compressors
10. Humam interface—to monitor all temperatures, pressures, humidities, inputs and outputs; to
edit set points; and to select services
11. Diagnostics—to detect faults and diagnostics (a total of 49 different diagnostics can be read
from the display)
12. Emergency stop input
A satisfactory indoor environment surely will be maintained, and at the same time the rooftop packaged
unit is energy-efficient with all these controls.
Minimum Performance
To reduce energy use in the packaged units, ASHRAE/IESNA Standard 90.1-1999 mandates minimum
efficiency requirements of various PUs as listed in Table 11.6, and also specifies the minimum
efficiency requirements for various air source heat pumps as listed in Sec. 12.2. Almost all PUs and
air source heat pumps are driven and operated electrically.
In Table 11.6, EER indicates the energy efficiency ratio, SEER the seasonal energy efficiency
ratio, IPLV the integrated part-load value, and HSPF the heating seasonal performance factor. All of
these energy index are defined in Sec. 9.20.
When using the efficiency ratings listed in Table 11.6 and Sec. 12.2 to compare different types
of HVAC&R equipment, the following conditions must be considered:
 The efficiency ratings for water-cooled equipment cannot be compared directly to those for aircooled
equipment. Water-cooled equipment does not include the energy use of required condenser
water pumps and cooling tower fans whereas air-cooled packaged unit ratings include the energy
use of condenser fan.
 The ratings for condensing units cannot be directly compared with single or split packaged units.
Condensing unit ratings do not include the energy use of fans in indoor air handlers.
 The efficiency ratings of a water chiller cannot be compared with a packaged unit using DX coil,
as the efficiency of a water chiller does not include the energy use of chilled-water pumps.
Selection of Packaged Units
The procedure of selection of PUs is different from that for an AHU because the number of rows
and fin spacing of a DX coil are fixed for a specific model and size of PU. The size of a PU is determined
primarily by the required cooling capacity of the DX coil at various operating conditions.
The cooling capacity of a typical rooftop PU is listed in Table 16.3, and Table 16.4 presents the supply
fan performance of this typical rooftop PU. The selection procedure for PUs, based on data
from manufacturers’ catalogs, can be outlined as follows:
1. Calculate the cooling coil load and sensible cooling coil load (refer to Chap. 6) of the conditioned
space that is served by the PU. For a unitary packaged system using DX coils, coil load is
equal to the refrigeration load of the refrigeration system.
Select the model and size of the PU based on the cooling capacity that is equal to or greater than
the required coil load and sensible coil load at the specified design conditions. These include the
dry- and wet-bulb temperatures of air entering the DX coil and the outdoor air entering the aircooled
condenser or evaporatively cooled condenser.
The equipment capacity may exceed the design load only when the equipment selected is the
smallest size needed to meet the load.
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.19
16.20
TABLE 16.3 Cooling Capacity of a Typical Rooftop PU, Nominal Capacity 55 Tons; Volume Flow Rate, Standard Air, of 19,250 cfm
Ambient temperature, °F
85 95
Entering wet-bulb temperature, °F
61 67 73 61 67 73
Entering dry-bulb, °F MBH SHR MBH SHR MBH SHR MBH SHR MBH SHR MBH SHR
75 502 81 559 55 621 33 473 83 528 56 587 33
80 506 95 559 69 620 46 479 97 528 71 586 47
85 530 100 559 83 620 59 505 100 528 85 586 60
90 558 100 564 95 619 71 533 100 536 96 585 73
105°F 115°F
61°F 67°F 73°F 61°F 67°F 73°F
Entering dry-bulb, °F MBH SHR MBH SHR MBH SHR MBH SHR MBH SHR MBH SHR
75 443 86 496 57 552 33 413 89 463 58 516 33
80 453 98 495 73 551 47 427 100 462 75 516 48
85 480 100 496 88 551 61 454 100 464 91 515 63
90 508 100 507 98 550 75 481 100 481 100 515 78
MBH: 1000 Btu/h
SHR: Sensible heat ratio
Source: The Trane Company. Reprinted with permission.
16.21
TABLE 16.4 Supply Fan Performance of a Typical Rooftop PU, Nominal Capacity 55 Tons
Total static pressure, in. WC
CFM 2.250 2.500 2.750 3.000 3.250 3.500 3.750 4.000
standard
air rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp rpm bhp
10,000 842 8.09 887 9.18 927 10.28 965 11.37 1,000 12.47 1,034 13.59 1,067 14.74 1,099 15.90
12,000 843 9.17 887 10.31 931 11.50 972 12.75 1,012 14.03 1,050 15.34 1,085 16.65 1,118 17.96
14,000 857 10.72 897 11.85 936 13.01 974 14.25 1,012 15.54 1,050 16.90 1,087 18.31 1,123 19.76
16,000 876 12.59 914 13.81 952 15.05 988 16.31 1,023 17.60 1,057 18.93 1,091 20.30 1,124 21.74
18,000 900 14.80 936 16.08 972 17.40 1,006 18.76 1,040 20.14 1,073 21.54 1,105 22.96 1,137 24.40
20,000 926 17.31 962 18.75 996 20.14 1,029 21.56 1,061 23.01 1,093 24.50 1,124 26.01 1,154 27.54
22,000 946 19.85 986 21.57 1,022 23.22 1,055 24.79 1,086 26.33 1,116 27.88 1,146 29.46
22,500 950 20.50 991 22.28 1,027 23.97 1,061 25.64 1,093 27.24 1,123 28.84
24,000 963 22.44 1,004 24.40
rpm: revolutions per minute
bhp: brake horsepower
Source: The Trane Company. Reprinted with permission.
If the required supply air volume flow rate deviates from the nominal rated value in percentage,
as shown below, then the cooling capacity and sensible cooling capacity can be roughly multiplied
by a multiplier because of the change of the heat-transfer coefficient as follows:
2. Calculate the heating coil load at the winter design condition or at the warm-up period. Determine
the capacity of the gas-fired furnace, electric heater, water or steam heating coil. For a
packaged heat pump, calculate the supplementary heating capacity of the gas-fired or electric
heater.
3. Evaluate the external total pressure loss of the duct system, terminal, and space diffusion devices
(see Chaps. 17 and 18). Determine the speed of the supply fan, relief fan, exhaust fan, or return
fan, such that the volume flow and the fan total pressure of the supply fan, supply fan/ return
fan combination is equal to or greater than the sum of the external total pressure loss and the total
pressure loss in the PU.
For fans for which only a fan static pressure is given, roughly a 0.4-in. WC (100-Pa) of velocity
pressure can be added to the fan static pressure for purposes of rough estimation of fan total
pressure.
Check that the face velocity of the DX coil does not exceed 550 fpm (2.75 m / s) so that condensate
will not carry over.
4. For small PUs, a medium-efficiency air filter of lower pressure drop, such as a final pressure
drop between 0.4 and 0.6 in. WC (100 to 150 Pa), should be selected.
Example 16.1. Select an AHU or a rooftop PU for a typical floor in a commercial building with
the following operating characteristics:
Supply volume flow rate 16,000 cfm (7550 L / s)
Cooling coil load 520,000 Btu /h or 43 tons (152,360 W)
Sensible cooling coil load 364,000 Btu /h (106,650 W)
Outdoor air temperature 95°F (35°C)
Entering coil dry-bulb temperature 80°F (26.7°C)
Entering coil wet-bulb temperature 67°F (19.4°C)
External pressure drop 2.0 in. WC (500 Pa)
Total pressure loss of AHU or PU 2.25 in. WC (563 Pa)
For a 4-row, 12-fins/in. water-colling coil in an AHU, chilled water enters the coil at 45°F
(7.2°C) and is expected to leave the coil at 55°F (12.7°C).
Solution:
1. Divide the supply volume flow rate by 500 fpm, that is, 16,000 /500  32. From Tables 16.1
and 16.2, select an AHU of size 30, which gives a maximum static pressure up to 8 in. WG and a
face area of cooling coil of 29 ft2. The actual face velocity of cooling coil is
vcoil 
16,000
29
 551 fpm (2.75 m / s)
Volume flow 20% 10% 0% 	10% 	20%
Cooling capacity 0.965 0.985 1.0 1.015 1.025
Sensible cooling 0.94 0.97 1.0 1.03 1.06
16.22 CHAPTER SIXTEEN
16.22
There will be no carryover of condensate water droplets; therefore, it is acceptable.
2. From Table 16.2, select a size-30 horizontal draw-through unit with a backward-inclined centrifugal
fan at an impeller diameter of and a fan speed of 1931 rpm. The fan static pressure
is then 4.00 in. WC at a volume flow of 16,000 cfm. From Table 16.2, because the fan outlet velocity
is 2540 fpm, its velocity pressure is then
And the fan total pressure provided by this fan is
4.00 	 0.40  4.40 in. WC (1100 Pa)
This value is greater than the required fan total pressure
2.0 	 2.25  4.25 in. WC (1063 Pa)
From the manufacturer’s catalog, the heights of the fan and coil modules are both 4 ft 6 in. (1.37 m).
For such a height, the headroom of the fan room is usually sufficient.
3. As in Example 15.2, the cooling and dehumidifying capacity per ft2 of coil face area of the
water cooling coil given by the manufacturer’s catalog is 19.1 MBtu /h ft2, and the sensible cooling
capacity is 13.89 MBtu /h ft2. For a coil face area of 29 ft2, the total cooling and dehumidifying
capacity is
Qc  19,100  29  553,900 Btu /h (162,290 W)
and the sensible cooling capacity is
Qcs  13,890  29  402,810 Btu /h (118,020 W)
These capacities both are greater than the required cooling capacity of 520,000 Btu / h (152,360 W)
and the sensible cooling capacity of 364,000 Btu / h (106,650 W).
4. From Table 16.3, select a rooftop PU with a cooling capacity of 55 tons and a sensible heat
ratio SHRc  0.71. When the supply volume flow rate of this PU is 19,250 cfm, at an air entering
coil dry-bulb temperature of 80°F and a wet-bulb temperature of 67°F, and an outdoor air temperature
of 95°F, the cooling capacity is 528 MBtu / h, and the sensible cooling capacity is 0.71 
528  375 MBtu/ h. Both are greater than the required values. The selected 55-ton PU is suitable.
From the manufacturer’s catalog, the face area of this 55-ton DX coil is 37.9 ft2.
If the supply volume flow rate is reduced to 17,625 cfm (19,250  17,625) / 16,000  0.10, or
10 percent greater than the required volume flow rate of 16,000 cfm, then the cooling capacity is
reduced to
528,000  0.985  520,080 Btu /h (152,380 W)
and the sensible cooling capacity will be reduced to
375,000  0.97  363,750 Btu /h (106,580 W)
Only the sensible cooling capacity is slightly less than the required value of 364,000 Btu / h.
5. Assume that the supply fan outlet velocity of the rooftop PU is 2000 fpm (10 m / s); that is,
the velocity pressure at fan outlet is
pv   2000
4005 2
 0.25 in. WC
pv   2540
4005 2
 0.40 in. WC
2214 in.
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.23
From Table 16.4, we see that at a supply volume flow of 17,625 fpm and a fan speed of about 1135
rpm, this rooftop PU can provide a fan total pressure of
pt  4.0 	 0.25  4.25 in. WC (1060 Pa)
This fan total pressure can meet the required value. Check the DX coil’s face velocity
Coil face velocity is less than 550 fpm (2.75 m / s); therefore, there is no condensate droplet
carryover.
16.6 FAN ROOM
Types of Fan Room
A fan room is an enclosure in which an AHU, an air handler, an indoor PU, and other accessories
and air-handling equipment are located. According to ASHRAE Standard 15-1994, for refrigerating
systems of 100 hp (74.6 kW) or less, a fan room may contain refrigeration machinery if
 The fan room is a separated, tight construction with tight fitting doors
 Access by authorized personnel is controlled
 Detectors (refrigerant, oxygen, etc.) are located in refrigerant leaking areas
Refer to Standard 15-1994 for details.
In low-rise buildings of three stories and less, a fan room may be used to serve up to three floors.
In high-rise buildings of four stories and more, a fan room may be used to serve one or more floors,
usually up to 20 floors, depending on the characteristics of the air system, its initial cost, and its operating
cost.
Fan rooms can be classified according to their pressure characteristics as open or isolated.
Open Fan Room. An open fan room is open to the filter end of the AHU or air handler, as shown
in Fig. 16.7. The return ceiling plenum is directly connected to the fan room through an inner-lined
return duct or a return duct with a sound attenuator. Outdoor air is often forced to the fan room by
an outdoor air fan or a makeup AHU, or extracted to the fan room by the supply fan in the AHU.
The fan room becomes the mixing box of the AHU or air handler. Its pressure is often lower than
the static pressure in the return plenum and the outdoor atmosphere. Return air is then extracted to
the fan room through the return duct. An exhaust fan is often installed on the external wall of the
fan room to maintain this pressure difference.
The advantages of this kind of fan room are a positive outdoor air supply and less ductwork in
the fan room. The disadvantages include the following:
 There may be infiltration of uncontrolled outdoor air if the fan room is not airtight.
 The fan room is entirely exposed to outdoor air.
 Sufficient sound attenuation must be provided in the transfer duct or in the air passage that transfers
the return air from the return ceiling plenum to the fan room.
Isolated Fan Room. In this kind of fan room, shown in Fig. 16.8, the outdoor air, the return air,
and the exhaust air are all isolated from the fan room air because of the ductwork. The static pressure
in the fan room depends mainly on the air leakage from or to the AHU or air handler, and the
17,625
37.9
 465 fpm (2.32 m / s)
16.24 CHAPTER SIXTEEN
air passage connecting the fan room and outside atmophere. This kind of fan room is most widely
used in commercial buildings.
Fan rooms located at the perimeter of the building often have direct access from the outside
walls. They are more convenient to provide the outdoor air intake and exhaust. For fan rooms
AIR SYSTEMS: EQUIPMENT—AIR-HANDLING UNITS AND PACKAGED UNITS 16.25
FIGURE 16.7 Open fan room: (a) plan view; (b) sectional view.
located in the interior core, large outdoor and exhaust air risers are required in multistory buildings
when an air economizer cycle is used.
Figure 16.9 shows the plan and sectional views of an isolated fan room for an indoor packaged
unit located in the interior core of the building. The unit is also equipped with a coil module for
water cooling and heating coils and water economizer precooling coils.
16.26 CHAPTER SIXTEEN
FIGURE 16.8 Isolated fan room: (a) plan view; (b) sectional view.
16.27
FIGURE 16.9 Interior core fan room: (a) plan view; (b) sectional view.
Layout Considerations
A satisfactory fan room layout should meet the following requirements:
 It should be compact yet provide sufficient space for the maintenance workers to pull out fan
shafts, coils, and filters. Piping connections on the same side of the AHU or PU are preferable.
 Outdoor intake and exhaust outlets are located either on outside walls perpendicular to each other
or on different walls with a certain distance between them.
 Fire dampers should be installed to separate the fan room and the fire compartment according to
the fire codes. Fan room ventilation and exhaust should be provided to meet the requirements of
the codes.
 Inner-lined square elbows or elbows with 2- or 3-in. -(50- or 75-mm) thick duct liners are used for
better sound attenuation at low frequencies. Sound attenuating devices are required for both the
supply and return sides of the fan.
 A vertical AHU occupies less floor space than a horizontal unit; therefore, it is often the first
choice if the headroom is sufficient.
 An unhoused centrifugal fan located in the exhaust compartment may be the suitable choice of
return fan for its lower noise, and it has the advantage of discharging on both sides.
In Fig. 16.8a, at point rf after the return fan, the pressure is positive. At point m in the mixing box, the
pressure must be negative in order to extract outdoor air. There must be a damper and an appropriate
pressure drop between point rf and m to guarantee such a positive-to-negative pressure conversion.
REFERENCES
ASHRAE, ASHRAE/ IES Standard 90.1-1989, User’s Manual, ASHRAE Inc., Atlanta, GA, 1992.
ASHRAE, Energy Code for Commercial and High-Rise Residential Buildings, Atlanta, 1993.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, Atlanta, 1996.
ASHRAE/ IES Standard 90.1-1999, Energy Standard for Buildings Except New Low Rise Residential Buildings,
Atlanta, 1999.
Bierwirth, H. C., Packaged Heat Pump Primer, Heating / Piping / Air Conditioning, July 1982, pp. 55–59.
Brasch, J. F., Electric Duct Heater Principles, Heating / Piping / Air Conditioning, March 1984, pp. 115–130.
Carrier Corporation, Products and Systems 1992 / 1993 Master Catalog, Carrier Corp., Syracuse, NY, 1992 / 1993.
Gill, K. E., IAQ and Air Handling Unit Design, HPAC, no. 1, 1996, pp. 49–54.
Gill, K. E., Rooftop HVAC, HPAC, no. 7, 1997, pp. 51–55.
Haessig, D. L., A Solution for DX VAV Air Handlers, Heating / Piping / Air Conditioning, no. 5, 1995, pp. 83–86.
Haines, R. W., Stratification, Heating / Piping / Air Conditioning, November 1980, pp. 70–71.
McGuire, A. B., Custom Built HVAC Units, Heating / Piping / Air Conditioning, January 1987, pp. 115–122.
Pannkoke, T., Rooftop HVAC for the 90s, Heating / Piping / Air Conditioning, no. 7, 1993 pp. 33–42.
Riticher, J. J., Low Face Velocity Air Handling Units, Heating / Piping / Air Conditioning, December 1987,
pp. 73–75.
Scolaro, J. F., and Halm, P. E., Application of Combined VAV Air Handlers and DX Cooling HVAC Packages,
Heating / Piping / Air Conditioning, July 1986, pp. 71–82.
The Trane Company, Packaged Rooftop Air Conditioners, The Trane Co., Clarksville, TN, 1997.
Waller, B., Economics of Face Velocities in Air Handling Unit Selection, Heating / Piping / Air Conditioning,
March 1987, pp. 93–94.
Wang, S. K., Air Conditioning, vol. 3, Hong Kong Polytechnic, Hong Kong, 1987.
Weisgerber, J., Custom Built HVAC Penthouses, Heating / Piping / Air Conditioning, November 1986,
pp. 115–117.
16.28 CHAPTER SIXTEEN
CHAPTER 17
AIR SYSTEMS:
AIR DUCT DESIGN
17.1
17.1 BASICS OF AIRFLOW IN
DUCTS 17.2
Bernoulli Equation 17.2
Steady Flow Energy Equation 17.2
Static Pressure, Velocity Pressure, and
Total Pressure 17.3
Stack Effect 17.5
Laminar Flow and Turbulent Flow 17.6
Velocity Distribution 17.7
Equation of Continuity 17.7
17.2 CHARACTERISTICS OF AIRFLOW IN
DUCTS 17.8
Types of Air Duct 17.8
Pressure Characteristics of the
Airflow 17.8
System Pressure Loss 17.10
Criteria of Fan Energy Use 17.10
17.3 DUCT CONSTRUCTION 17.12
Maximum Pressure Difference 17.12
Material 17.12
Rectangular Ducts 17.13
Round Ducts 17.17
Flat Oval Ducts 17.17
Flexible Ducts 17.18
Fiberglass Ducts 17.18
17.4 DUCT HEAT GAIN, HEAT LOSS,
AND DUCT INSULATION 17.19
Temperature Rise or Drop due to Duct
Heat Gain or Loss 17.19
Duct Insulation 17.19
Temperature Rise Curves 17.21
17.5 FRICTIONAL LOSSES 17.22
Darcy-Weisbach Equation 17.22
Friction Factor 17.22
Duct Friction Chart 17.24
Roughness and Temperature
Corrections 17.25
Circular Equivalents 17.27
17.6 DYNAMIC LOSSES 17.31
Elbows 17.31
Converging and Diverging Tees and
Wyes 17.34
Entrances, Exits, Enlargements, and
Contractions 17.38
17.7 FLOW RESISTANCE 17.38
Flow Resistances Connected in
Series 17.40
Flow Resistances Connected in
Parallel 17.41
Flow Resistance of a Y Connection 17.42
Flow Resistance of a Duct System 17.42
17.8 PRINCIPLES AND CONSIDERATIONS
IN AIR DUCT DESIGN 17.43
Optimal Air Duct Design 17.43
Design Velocity 17.45
System Balancing 17.46
Critical Path 17.48
Air Leakage 17.48
Shapes and Material of Air Ducts 17.50
Ductwork Installation 17.50
Fire Protection 17.50
17.9 AIR DUCT DESIGN PROCEDURE
AND DUCT LAYOUT 17.51
Design Procedure 17.51
Duct System Characteristics 17.52
Duct Layout 17.52
17.10 DUCT SIZING METHODS 17.53
Equal-Friction Method 17.53
Constant-Velocity Method 17.53
Static Regain Method 17.54
T Method 17.55
17.11 DUCT SYSTEMS WITH CERTAIN
PRESSURE LOSSES IN BRANCH
TAKEOFFS 17.56
Design Characteristics 17.56
Cost Optimization 17.56
Condensing Two Duct Sections 17.59
Local Loss Coefficients for Diverging Tees
and Wyes 17.60
Return or Exhaust Duct Systems 17.63
17.12 DUCT SYSTEMS WITH NEGLIGIBLE
PRESSURE LOSS AT BRANCH
DUCTS 17.66
Supply Duct Systems 17.66
Pressure Characteristics of Airflow in
Supply Ducts 17.66
Return or Exhaust Duct Systems 17.71
17.13 REQUIREMENTS OF EXHAUST
DUCT SYSTEMS FOR A MINIMUM
VELOCITY 17.72
17.14 COMPUTER-AIDED DUCT DESIGN
AND DRAFTING 17.72
Drafting 17.72
Schedules and Layering 17.72
Design Interface 17.73
Running Processes 17.73
17.15 DUCT LINER AND DUCT
CLEANING 17.74
Duct Liner 17.74
Duct Cleaning 17.74
17.16 PRESSURE AND AIRFLOW
MEASUREMENTS 17.75
Equal-Area versus Log Tchebycheff
Rule 17.77
REFERENCES 17.78
17.1 BASICS OF AIRFLOW IN DUCTS
Bernoulli Equation
The Bernoulli equation relates the mean velocity v, in ft / s (m / s), the pressure p, in lbf / ft2 absolute
(abs.) or psia (Pa abs.), and the elevation z, in ft (m), of a frictionless or ideal fluid at steady state.
When a fluid motion is said to be in steady state, the variables of the fluid at any point along the
fluid flow do not vary with time. Assuming constant density, the Bernoulli equation can be
expressed in the following form:
(17.1)
where p  static pressure, lbf / ft2 abs. (Pa abs.)
 fluid density, lbm/ ft3 (kg /m3)
g  gravitational acceleration, ft / s2 (m/s2)
gc  dimensional constant, 32.2 lbm ft / lbf  s2 (1)
For convenience, lb  lbm (mass).
Steady Flow Energy Equation
For a real fluid flowing between two cross sections in an air duct, pipe, or conduit, energy loss is inevitable
because of the viscosity of the fluid, the presence of the mechanical friction, and eddies.
The energy used to overcome these losses is usually transformed to heat energy. If we ignore the
kinetic energy difference between the value calculated by the mean velocity of the cross section and
the value calculated according to the velocity distribution of the cross section, then the steady flow
energy equation for a unit mass of real fluid is given as
(17.2)
where u  internal energy, Btu / lb (J / kg)
J  Joule’s equivalent, 778 ft lbf / Btu (1)
q  heat supplied, Btu / lb (J / kg)
W  work developed, ft lbf / lb (J /s)
In Eq. (17.2), subscripts 1 and 2 indicate the cross section 1 and 2, respectively, and p1 and p2
denote the absolute static pressure at cross section 1 and 2. Signs of q and W follow the convention
in thermodynamics, i.e., when heat is supplied to the system, q is positive and when heat is released
from the system, q is negative. When work is developed by the system, W is positive; and for work
input to the system, W is negative.
Multiply both sides of Eq. (17.2) by , ignore the difference in densities, and rearrange the
terms. Then each term has the unit of pressure, in lbf / ft2 abs. (Pa abs.), or
(17.3)
For an air duct or piping work without a fan, compressor, and pump, W  0.
Let the pressure loss from viscosity, friction, and eddies between cross sections 1 and 2 be
pf  J(u2 	 u1 	 q); then each term of Eq. (17.3) can be expressed in the form of pressure
p1 
1v1 2
2gc

1gz1
gc
 p2 
2v2 2
2gc

2gz2
gc

W 
J(u2 	 u1 	 q)
p1
1
 u1J 
v1 2
2gc

gz1
gc
 qJ 
p2
r2
 u2J 
v2 2
2gc

gz2
gc
 W
p


v2
2gc

gz
gc
 constant
17.2 CHAPTER SEVENTEEN
(17.4)
If both sides of Eq. (17.2) are multiplied by gc /g and W  0, then each term of the equation is expressed
in the form of head, in ft or in. (m) of fluid column. That is,
(17.5)
Static Pressure, Velocity Pressure, and Total Pressure
Pressure is the force per unit area exerted by a fluid or solid. In an air duct system, a water piping
system, or a refrigerant piping system, fluid pressure including air, water, refrigerant pressure at a
surface or a level, or inside an enclosed vessel, or pressure difference between two surfaces is often
measured under the following conditions:
 Fluid pressure is measured related to a datum of absolute vacuum. Such a measured fluid pressure
is given as absolute pressure and is often represented by the pressure exerted at the bottom surface
of a water column.
 Fluid pressure is often more conveniently measured related to a datum of atmospheric pressure.
Such a measured fluid pressure is given as gauge pressure. The measured gauge pressure that is
greater than the atmosperic pressure is expressed as positive gauge pressure or simply gauge pressure.
That part of measured gauge pressure which is less than the atmospheric pressure is expressed
as negative gauge pressure or vacuum.
 Fluid pressure is measured as a pressure difference, pressure drop, or pressure loss between two
surfaces, two levels, or two cross-sectional surfaces. The involved two measured pressures must
be either both gauge pressure or both absolute pressure.
Consider a supply duct system in a multistory building, as shown in Fig. 17.1. In Eq. (17.4),
since p1  pat1  p1 and p2  pat2  p2, where p1 and p2 represent the gauge static pressure and pat1
and pat2 the atmospheric pressure added on the fluid at cross sections 1 and 2. The relationship of
fluid properties between cross sections 1 and 2 can be expressed as
(17.6)
If the air temperature inside the air duct is equal to the ambient air temperature, and if the stack effect
because of the difference in air densities between the air columns inside the air duct and the
ambient air does not exist, then
Therefore, Eq. (17.6) becomes
(17.7)
Equation (17.7) is one of the primary equations used to determine the pressure characteristics of an
air duct system that does not contain a fan and in which the stack effect is negligible.
Static Pressure. In Eq. (17.7), static pressures p1 and p2 are often represented by ps. In air duct systems,
its unit can be either Pa (pascal, or newtons per square meter) in SI units, or the height of water
p1 
1v1 2
2gc
 p2 
2v2 2
2gc
 pf
pat1 	 pat2  ( 2z 2 	
1z 1)
g
gc
pat1  p1 
1v1 2
2gc

1gz1
gc
 pat2  p2 
2v2 2
2gc

2gz2
gc
 pf
gc p1
g 1

v1 2
2g
 z 1 
gc p2
g 2

v2 2
2g
 z 2 
gc pf
g
p1 
1v1 2
2gc

1gz1
gc
 p2 
2v2 2
2gc

2gz2
gc
 pf
AIR SYSTEMS: AIR DUCT DESIGN 17.3
column, in inches, for I-P units. Either is expressed in gauge pressure or absolute pressure. The relationship
between the static pressure ps, in lbf / ft2, and the height of a water column H, in ft, is
(17.8)
where A  cross-sectional area of water column, ft2
 density of water, lb/ ft3
When the static pressure is expressed as the height of 1 in. of water column absolute pressure (1 in.
WC) or 1 in. of water column gauge pressure (1 in. WG), for a density of water w  62.3 lb / ft3
and numerically g  gc  32.2, we find from Eq. (17.8)
That is 1 in. WC  5.192 lbf/ft2. Because 1 lbf/ft2  47.88 Pa, 1 in. WC  5.192 	 47.88  248.6 Pa.
Velocity Pressure. In Eq. (17.7), the term v2 / (2gc) is called the velocity pressure, or dynamic
pressure, and is represented by the symbol pv, that is,
(17.9) pv 
v2
2gc
ps  1 in. WG and ps  1 in. WC 
wgH
gc

62.3 	 32.2 	 1
32.2 	 12
 5.192 lbf / ft2
ps 
wgHA
gc A

wgH
gc
17.4 CHAPTER SEVENTEEN
FIGURE 17.1 Pressure characteristics of an air duct system.
where  air density, lb/ ft3. For air density  0.075 lb / ft3, if the velocity pressure pv is
expressed in in. WC and the air velocity in fpm or ft / min, according to Eq. (17.9),
and (17.10)
In SI units, if v is expressed in m / s and  in kg /m3, then pv is in Pa, and pv can be calculated by
Eq. (17.9). For SI units, gc  1.
Total Pressure. At any cross-sectional plane perpendicular to the direction of the airflow, the total
pressure of the airstream pt is defined as the sum of the static pressure ps and the velocity pressure
pv, that is,
(17.11)
From Eq. (17.11) velocity pressure pv is also a kind of pressure difference. The units of pt must be
consistent with ps and pv. In I-P units, it is also indicated in inches WC or WG and in SI units in Pa
absolute (abs.) or gauge (g). Substituting Eq. (17.11) into Eq. (17.7), we see that
(17.12)
Equation (17.12) is another primary equation that relates the pressure loss from friction and other
sources, pf, and the total pressure pt1 and pt2 at two cross sections of the air duct system.
Stack Effect
When an air duct system has an elevation difference and the air temperature inside the air duct is
different from the ambient air temperature, the stack effect exists. It affects airflow at different elevations.
During a hot summer day, when the density of the outdoor air is less than the density of the
cold supply air inside the air duct, the pressure exerted by the atmospheric air column between z1
and z2, as shown in Fig. 17.1, is given as
where o  mean density of the ambient air, lb/ ft3 (kg /m3). And the pressure exerted by the air
column inside the air duct between z1 and z2 is
where i  mean density of the supply air inside the air duct, in lb / ft3 (kg /m3). If the differences
between the densities inside the air ducts, 1 and i, and 2 and i, are ignored, substituting these
relationships into Eq. (17.6) yields
(17.13)
and (17.14) pt1 
g( o 	
i)(z 2 	 z 1)
gc
 pt2  pf
p1 
1v1 2
2gc

g( o 	
i)(z 2 	 z 1)
gc
 p2 
2v2 2
2gc
 pf
2gz2 	
1gz1
gc

i g(z 2 	 z 1)
gc
pat1 	 pat2 
o g(z 2 	 z 1)
gc
pt1  pt2  pf
pt  ps  pv pv  pt 	 ps
v  4005 ?pv pv   v
4005 2
 v
60 2

5.192(2pvgc)


5.192 	 2 	 32.2 pv
0.075
 4458pv
AIR SYSTEMS: AIR DUCT DESIGN 17.5
The third term on the left-hand side of Eq. (17.13) and the second term on the left-hand side of
Eq. (17.14) are called the stack effect pst, in lbf / ft2 (Pa),
(17.15)
If pst is expressed in in. WC (1 lbf / ft2  0.1926 in. WC), then
(17.16)
For an upward supply duct system, z2 z1. If cold air is supplied, i o and pst is negative. If
warm air is supplied, o i and pst is positive. For a downward supply duct system with a cold air
supply, pst is positive; if there is a warm air supply, pst is negative.
When supply air is at a temperature of 60°F (15.5°C) and a relative humidity of 80 percent, its
density is 0.075 lb / ft3 (1.205 kg /m3). Also, if the space air has a temperature of 75°F (24°C) and a
relative humidity of 50 percent, its density is 0.073 lb / ft3 (1.171 kg /m3). Numerically, g  gc, for a
difference of z2 	 z1  30 ft (9.14 m), so
pst  0.1926 (0.073 	 0.075)(30)  0.0116 in. WC (2.96 Pa)
For an air-handling unit or a packaged unit that supplies air to the same floor where it is located, the
stack effect is usually ignored.
Laminar Flow and Turbulent Flow
Reynolds identified two types of fluid flow in 1883 by observing the behavior of a stream of dye in
a water flow: laminar flow and turbulent flow. He also discovered that the ratio of inertial to viscous
forces is the criterion that distinguishes these two types of fluid flow. This dimensionless parameter
is now widely known as Reynolds number Re, or
(17.17)
where  density of fluid, lb/ ft3 (kg /m3)
v  velocity of fluid, ft / s (m/s)
L  characteristic length, ft (m)

 viscosity or absolute viscosity, lb/ ft  s (N s/m2)
v  kinematic viscosity, ft2 / s (m2 / s)
Many experiments have shown that laminar flow occurs at Re  2000 in round ducts and pipes. A
transition region exists between 2000 Re  4000. When Re 4000, the fluid flow is probably a
turbulent flow.
At 60°F (15.5°C), the viscosity 
 1.21 	 105 lb / ft  s (18.0 	 106 Ns/m2). For a round duct
of 1-ft (0.305-m) diameter and an airflow through it of 3 ft / s (180 ft /min or 0.915 m / s), the
Reynolds number is
The Re of such an air duct is far greater than 4000. Therefore, the airflow inside the air duct is usually
turbulent except within the boundary layer adjacent to the duct wall.
Re 
vL



0.075 	 3 	 1
1.21 	 105  18,595
Re 
vL



vL
v
pst 
0.1926g( o 	
i)(z 2 	 z 1)
gc
pst 
g( o 	
i)(z 2 	 z 1)
gc
17.6 CHAPTER SEVENTEEN
Velocity Distribution
The velocity distributions of turbulent and laminar flow at a specific cross section in a circular duct
that result from (1) the mechanical friction between the fluid particles and the duct wall and (2) the
shearing stress of the viscous fluid are shown in Fig. 17.2. The difference between these two types
of flow is significant.
For a fully developed turbulent flow, the air velocity v, in fpm (m/s), at various distances from
the duct wall of a circular duct y, in ft (m), varies according to Prantdl’s one-seventh power law as
follows:
(17.18)
where vmax  maximum air velocity on centerline of air duct, fpm (m/ s)
R  radius of duct, ft (m)
The mean air velocity vm lies at a distance about 0.33R from the duct wall.
Equation of Continuity
For one-dimensional fluid flow at steady state, the application of the principle of conservation of
mass gives the following equation of continuity:
(17.19)
where  mass flow rate, lb/s (kg /s)
A  cross-sectional area perpendicular to fluid flow, ft2 (m2)
Subscripts 1 and 2 indicate the cross sections 1 and 2 along the fluid flow.
If the differences in fluid density at various cross sections are negligible, then the equation of
continuity becomes
(17.20)
where  volume flow rate of airflow, cfm (m3 / s)
v  mean air velocity at any specific cross section, fpm (m/ s)
V?
V?
 A1v1  A2v2
m?
m? 
1v1A1 
2v2A2
v
vmax
  v
R 1/7
AIR SYSTEMS: AIR DUCT DESIGN 17.7
FIGURE 17.2 Velocity distribution in a circular duct.
Theoretically, the velocity pressure pvt should be calculated as
(17.21)
Its value is slightly different from pv calculated from Eq. (17.9), which is based on mean velocity v.
For a fully developed turbulent flow, pvt  1.06pv. Since most experimental results of pressure loss
(indicated in terms of velocity pressure) are calculated by pv  v2 / (2gc), for the sake of simplicity
pv is used here instead of pvt.
17.2 CHARACTERISTICS OF AIR FLOW IN DUCTS
Types of Air Duct
Air ducts can be classified into four types according to their transporting functions:
1. Supply duct. Conditioned air is supplied to the conditioned space.
2. Return duct. Space air is returned (1) to the fan room where the air-handling unit is installed or
(2) to the packaged unit.
3. Outdoor air duct. Outdoor air is transported to the air-handling unit, to the fan room, or to the
space directly.
4. Exhaust duct. Space air or contaminated air is exhausted from the space, equipment, fan room,
or localized area.
Each of these four types of duct may also subdivide into headers, main ducts, and branch ducts or
runouts. A header is that part of a duct that connects directly to the supply or exhaust fan before air
is supplied to the main ducts in a large duct system. Main ducts have comparatively greater flow
rates and size, serve a greater conditioned area, and, therefore, allow higher air velocities. Branch
ducts are usually connected to the terminals, hoods, supply outlets, return grilles, and exhaust
hoods. A vertical duct is called a riser. Sometimes, a header or a main duct is also called a trunk.
Pressure Characteristics of the Airflow
During the analysis of the pressure characteristics of airflow in a fan duct system such as the one in
Fig. 17.3, it is assumed that the static pressure of the space air is equal to the static pressure of the
atmospheric air, and the velocity pressure of the space air is equal to zero. Also, for convenient
measurements and presentation, as previously mentioned, the pressure of the atmospheric air is
taken as the datum, that is, pat  0, and pressure is expressed as gauge pressure. When p pat, p is
positive; and if p pat, then p is negative. In a fan duct system, a fan or fans are connected to a
ductwork and equipment.
At cross section R1, as the recirculating air enters the return grille, both the total pressure pt
and static pressure ps decrease as the result of the total pressure loss of the inlet. The velocity pressure
pv, indicated by the shaded section in Fig. 17.3, will gradually increase until it is equal to the
velocity of the branch duct. Both pt and ps are negative so that air will flow from the conditioned
space at a datum of 0 to a negative pressure. Because velocity pressure pv is always positive in the
direction of flow, from Eq. (17.11) pt  ps  pv, so ps is then smaller, or more negative, than pt.
When the recirculating air flows through the branch duct segment R1-11, elbow 11-12, branch
duct segment 12-13, diffuser 13-14, and branch duct segment 14-15, both pt and ps drop because of the
pressure losses. Velocity pressure pv remains the same between cross sections R1 and 13. It gradually
decreases because of the diffuser 13-14 and remains the same between 14 and 15.
As the recirculating air flows through the converging tee 15-1, this straight-through stream meets
with another branch stream of recirculating air from duct section R2-1, at node or junction 1. The
pvt 
[
v2 / (2gc)] vdA
v dA
17.8 CHAPTER SEVENTEEN
combined airstream then becomes the main stream. Total pressure pt usually decreases when
the straight-through stream flows through the converging tee. It may increase if the velocity of the
branch stream is much higher than that of the straight-through stream. However, the sum of the energies
of the straight-through and the branch streams at the upstream side of the converging tee is
always higher than the energy at the downstream side because of the pressure loss of the converging
tee.
In the main duct section 1-21, pt and ps drop further while pv increases because of the higher air
velocity. This is mainly because of the greater volume flow rate in main duct section 1-21. When the
recirculating air enters the air-handling unit, its velocity drops sharply to a value between 400 and
600 fpm (2 and 3 m / s). The air is then mixed with the outdoor airstream from the fresh air duct at
node 2. After the mixing section, pt and ps drop sharply when the mixture of recirculating and outdoor
air flows through the filter and the coil. Total pressure pt and static pressure ps drop to their
minimum values before the inlet of the supply fan Fi.
At the supply fan, pt is raised to its highest value at the fan outlet Fo. Both pt and ps decrease in
duct section Fo-3. At junction 3, the airstream diverges into the main stream or straight-through
stream and the branch stream. Although there is a drop in pt after the main stream passing through
the diverging tee 3-31, ps increases because of the smaller pv after 31. The increase in ps due to the
decrease of pv is known as the static regain ps,r in in. WC (Pa). It can be expressed as
(17.22)
where v3, v31  air velocity at cross section 3 and 31, respectively, fpm (m/ s). In duct sections 3-4
and 4-S3, pt decreases gradually along the direction of airflow. Finally, pt, ps, and pv all drop to 0
after the supply air is discharged to the conditioned space.

(v3 2 	 v31
2 )
2
ps,r   v3
4005 2
	  v31
4005 2
AIR SYSTEMS: AIR DUCT DESIGN 17.9
FIGURE 17.3 Pressure characteristics of a fan duct system.
The pressure characteristics along the airflow can be summarized as follows:
 In most sections, pt of the main airstream decreases along the airflow. However, pt of the main
airstream may increase because of the higher velocity of the combined branch airstream.
 When air flows through the supply fan, mechanical work is done on the air so that pt and ps are
raised from a minimum negative value at the fan inlet to a maximum positive value at the fan
outlet.
 The pressure characteristics between any two cross sections of a duct system are governed by the
change of pt and the pressure loss pf between these two cross sections pt1  pt2  pf. Static
pressure is always calculated as ps  pt 	 pv.
 In a constant-volume air system, the airflow inside an air duct is considered steady and continuous.
Because the change in ps in a fan duct system is small when compared with pat, the airflow is
also considered incompressible.
System Pressure Loss
For an air system (a fan duct system), system pressure loss psy, in in. WC (Pa), is the sum of the
total pressure losses of the return air system pr,s, section R1-21 in Fig. 17.3; the air-handling unit
pAHU, section 21-Fo, or the packaged unit pPU; and the supply air system ps,s, section Fo-S3; all
are expressed in in. WC (Pa). That is,
(17.23)
Both pr,s and ps,s may include the pressure losses of duct sections (including duct segments); duct
fittings such as elbows, diffusers, and converging and diverging tees; components; and equipment.
The sum of pr,s and ps,s is called the external total pressure loss, or external pressure loss
pt,ex, as opposed to the pressure loss in the air-handling unit or packaged unit. The external pressure
loss pt,ex  pr,s  ps,s, in in. WC (Pa).
In commercial buildings, most air systems have a system pressure loss of 2.5 to 6 in. WC (625
to 1500 Pa). Of this, pAHU usually has a value of 1.5 to 3 in. WC (375 to 750 Pa), and ps,s is usually
less than 0.6 psy except in large auditoriums and indoor stadiums.
Criteria of Fan Energy Use
ASHRAE/IESNA Standard 90.1-1999 specifies that each air system having a total fan power
exceeding 5 hp (3.7 kW) shall meet the following specified allowable fan system power (AFSP) in
order to encourage heat recovery, relief fan, and other energy efficient means:
(AFSP)  (fan power limitation)(temperature ratio)  (pressure credit)  (relief fan credit)
(17.24)
Standard 90.1-1999 specifies fan power limitation (including supply, return, relief, and exhaust
fans) as follows:
For constant-volume systems
(17.24a)
Psy
V?
s, d
 0.0011 hp / cfm (0.00174 W s / L) when V?s, d  20,000 cfm
Psy
V?
s, d
 0.0012 hp / cfm (0.0019 W s / L) when V?s,d 20,000 cfm
 pr,s  pPU  ps,s
psy  pr,s  pAHU  ps,s
17.10 CHAPTER SEVENTEEN
where Psy  each air system total power input to the fan motors, hp (kW)
 volume flow rate of air system at design conditions, cfm (L / s)
For variable-air-volume (VAV) system: nameplate fan motor power is 0.0017 hp / cfm (0.0027 W
per L / s); when supply volume flow  20,000 cfm (9440 L / s), allowable nameplate fan motor
power is 0.0015 hp / cfm (W per L / s). That is:
(17.24b)
Psy can be calculated as
(17.25a)
where psy,d  system total pressure loss at design conditions, in. WC (Pa)
f,d, d,d, m,d  fan total efficiency, drive efficiency, and motor efficiency at design conditions
Substituting Eq. (17.25a) into Eq. (17.24), for constant-volume systems, the system total pressure
loss is
(17.25b)
For VAV systems, the system total pressure loss is
(17.25c)
In Eq. (17.24), temperature ratio can be calculated as:
(17.26)
Pressure credit, in hp (kW), can be calculated as:
(17.27a)
Relief fan credit, in hp (kW), can be calculated as:
(17.27b) Relief fan credit  Pr, f 1 	
1 	 V?r, f
V?
s,d 
Pressure credit 

n1
V?
s,d(pfil 	 1.0)
3718


m1
V?
s,dpHR
3718
Temperature ratio 
Tr, set 	 Ts,d
20
psy
f,d d,d m,d
 9.5 in. WC (Pa) when V?s,d  20,000 cfm
psy
f,d d,d m,d
 10.8 in. WC (Pa) when V?s,d 20,000 cfm
psy
f,d d,d m,d
 7.0 in. WC (Pa) when V?s,d  20,000 cfm
psy
f,d d,d m,d
 7.6 in. WC (Pa) when V?s,d 20,000 cfm
Psy 
psy,dV?s,d
6350 f,d d,d m,d
Psy
V?
s, d
 0.0015 hp / cfm (0.0024 W s / L) when V?s,d  20,000 cfm
Psy
V?
s,d
 0.0017 hp / cfm (0.0027 W s / L) when V?s,d 20,000 cfm
V?
s,d
AIR SYSTEMS: AIR DUCT DESIGN 17.11
where Tr, set  room set point temperature, °F (°C)
Ts,d  design supply air temperature for the zone in which the thermostat is located, °F(°C)
pfil, pHR  total pressure loss of the filters and of the heat recovery coils, in WC (Pa)
pr, f  name plate relief fan motor power, hp (kW)
 volume flow rate of relief fan at cooling design operation, cfm (L / s)
Consider a VAV system with a design supply volume flow rate 20,000 cfm (9440 L / s). If a
rooftop packaged system is used with a combined fan, drive, and motor efficiency f,d d,d m,d 
0.45, from Eq. (17.25c), the allowable system pressure loss for this rooftop package system (including
supply and return fans) is
If the design system pressure loss psy 4.3 in. WC (1070 Pa), the HVAC&R system designer is
recommended either to use a relief fan instead of a return fan for relief fan credit, or to take into account
the pressure credit if filter’s total pressure loss is greater than 1 in. WC (250 Pa) or if there is
heat recovery coil, increase the supply temperature differential Ts  (Tr, set 	 Ts) to a value greater
than 20°F (11.1°C) if they are cost effective to meet the fan power limitation in ASHRAE/IESNA
Standard 90.1–1999.
For a VAV reheat central system of  20,000 cfm (9440 L / s), if combined efficiency is 55
percent, from Eq. (17.25c), the allowable system pressure loss is
If the design system total pressure loss 5.2 in. WC, the same as for the rooftop packaged system,
means of relief fan credit, pressure credit due to filters and heat recovery coils, and the increase of
the supply temperature differential greater than 20°F (11.1°C) should be considered. Refer to
ASHRAE Standard 90.1-1999 for details.
17.3 DUCT CONSTRUCTION
Maximum Pressure Difference
Duct systems can be classified according to the maximum pressure difference between the air inside
the duct and the ambient air (also called the static pressure differential) as 0.5 in. WC
(125 Pa), 1 in. WC (250 Pa), 2 in. WC (500 Pa), 3 in. WC (750 Pa), 4 in. WC
(1000 Pa), 6 in. WC (1500 Pa), and 10 in. WC (2500 Pa). In actual practice, the maximum
pressure difference of the supply or return duct system in commercial buildings is usually less
than 3 in. WC (750 Pa).
In commercial buildings, a low-pressure duct system has a static pressure differential of 2 in.
WC (500 Pa) or less, and the maximum air velocity inside the air duct is usually 2400 fpm
(12 m/ s). A medium-pressure duct system has a static pressure differential of 2 to 6 in. WC (500 to
1500 Pa) with a maximum air velocity of about 3500 fpm (17.5 m / s). In industrial duct systems, including
mechanical ventilation, mechanical exhaust, and industrial air pollution control systems,
the pressure difference is often higher. In residential buildings, the static pressure differential of the
duct systems is classified as 0.5 in. WC (125 Pa) or 1 in. WC (250 Pa).
Material
Underwriters’ Laboratory (UL) classifies duct systems according to the flame spread and smoke developed
of the duct material during fire as follows:
psy 
f,d d,d m,d (9.5)  0.55 	 9.5  5.2 in. WC (1300 Pa)
V?
s,d
psy 
f,d d,d m,d (9.5)  0.45 	 9.5  4.3 in. WC (1070 Pa)
V?
s,d
V?
r, f
17.12 CHAPTER SEVENTEEN
Class 0. Zero flame spread, zero smoke developed.
Class 1. A flame spread rating of not more than 25 without evidence of continued progressive
combustion and a smoke developed rating of not more than 50.
Class 2. A flame speed of 50 and a smoke developed rating of 100.
National Fire Protection Association (NFPA) Standard 90A specifies that the material of the ducts
be iron; steel including galvanized sheets, aluminum, concrete, masonary; or clay tile. Ducts fabricated
by these materials are listed as class 0. UL Standard 181 allows class 1 material to be used for
ducts when they do not serve as risers for more than two stories or are not used in temperatures
higher than 250°F (121°C). Fibrous glass and many flexible ducts that are factory-fabricated are
approved by UL as class 1.
Ducts can be classified according to their shapes into rectangular, round, flat oval, and flexible,
as shown in Fig. 17.4.
Rectangular Ducts
For the space available between the structural beam and the ceiling in a building, rectangular ducts
have the greatest cross-sectional area. They are less rigid than round ducts and are more easily
AIR SYSTEMS: AIR DUCT DESIGN 17.13
FIGURE 17.4 Various types of air duct: (a) rectangular duct; (b) round duct with spiral seam; (c) flat oval
duct; (d) flexible duct.
fabricated on-site. The joints of rectangular ducts have a comparatively greater percentage of air
leakage than factory-fabricated spiral-seamed round ducts and flat oval ducts, as well as fiberglass
ducts. Unsealed rectangular ducts may have an air leakage from 15 to 20 percent of the supply volume
flow rate. Rectangular ducts are usually used in low-pressure systems.
The ratio of the long side a to the short side b in a rectangular duct is called the aspect ratio Ras.
The greater Ras, the higher the pressure loss per unit length as well as the heat loss and heat gain per
unit volume flow rate transported. In addition, more labor and material are required.
Galvanized sheet or, more precisely, galvanized coated steel sheet, and aluminum sheet are the materials
most widely used for rectangular ducts. To prevent vibration of the duct wall by the pulsating
airflow, transverse joints and longitudinal seam reinforcements are required in ferrous metal ducts.
The galvanized sheet gauge and thickness for rectangular ducts are listed in Table 17.1.
Table 17.2 gives specifications for rectangular ferrous metal duct construction for commercial systems
based on the publication of the Sheet Metal and Air Conditioning Contractors’ National Association
(SMACNA) titled HVAC Duct Construction Standards—Metal and Flexible. For design and
construction of an economical duct system, it is recommended to select an optimum combination of
minimum galvanized sheet thickness, type of transverse joint reinforcement, and its maximum
spacing for a specific duct dimension at a specific pressure differential between the air inside the
duct and the ambient air.
For rectangular ducts, one uses the same metal thickness for all sides of the duct and evaluates
duct reinforcement on each side separately. In Table 17.2, for a given duct dimension and thickness,
letters indicate the type of duct reinforcement (rigidity class) and numbers indicate maximum spacing,
in ft. Blanks indicate that reinforcement is not required, and dashes denote that such a
combination is not allowed.
Transverse joint reinforcements, abridged from SMACNA’s publication HVAC Duct Construction
Standard—Metal and Flexible and ASHRAE Handbook 1988, Equipment are presented in
Table 17.3. These must be matched with the arrangements in Table 17.2. Duct hangers should be installed
at right angles to the centerline of the duct. Habjan (1984) recommended the following
maximum duct hanger spacing:
17.14 CHAPTER SEVENTEEN
TABLE 17.1 Thickness of Galvanized Sheet for
Rectangular Ducts
Thickness, in. Nominal weight,
Gauge Nominal Minimum lb / ft2
30 0.0157 0.0127 0.656
28 0.0187 0.0157 0.781
26 0.0217 0.0187 0.906
24 0.0276 0.0236 1.156
22 0.0336 0.0296 1.406
20 0.0396 0.0356 1.656
18 0.0516 0.0466 2.156
16 0.0635 0.0575 2.656
14 0.0785 0.0705 3.281
13 0.0934 0.0854 3.906
12 0.1084 0.0994 4.531
11 0.1233 0.1143 5.156
10 0.1382 0.1292 5.781
Note: Minimum thickness is based on thickness tolerances
of hot-dip galvanized sheets in cut lengths and coils (per
ASTM Standard A525). Tolerance is valid for 48- and 60-in.-
wide sheets.
Source: ASHRAE Handbook 1988, Equipment. Reprinted
with permission.
17.15
TABLE 17.2 Rectangular Ferrous Metal Duct Construction for Commercial Buildings
Minimum galvanized steel thickness, in. (gauge)
0.0575 (16) 0.0466 (18) 0.0356 (20) 0.0296 (22) 0.0236 (24) 0.0187 (26)
Duct
Pressure, in. WG
dimensions,
in. 2 3 1 2 3 1 2 3 0.5 1 2 3 0.5 1 2 3 0.5 1 2
Up through 10
12 A-8 A-8
14 A-8 A-8 A-5 A-10 A-5
16 A-8 A-10 A-8 A-10 A-8 A-5 A-8 A-5
18 A-8 A-10 A-8 A-10 A-8 A-5 A-8 A-5
20 B-10 B-10 B-8 A-10 B-8 A-5 A-10 A-5 A-5 A-10 A-8 A-5
22 B-10 C-10 A-10 B-10 B-8 A-10 B-8 A-5 A-10 A-10 A-5 B-5 A-10 A-5 A-5
24 C-10 C-10 B-10 C-10 B-5 B-10 C-8 B-5 A-10 B-10 B-5 B-5 A-10 A-5 B-5
26 C-10 D-10 C-10 D-10 B-10 C-10 C-5 A-10 B-10 C-8 C-5 A-10 B-8 B-5 C-5 A-10 A-5 B-5
28 C-10 D-10 C-10 C-10 D-10 C-10 C-8 C-5 B-10 C-10 C-5 C-5 B-10 C-8 C-5 C-4 B-8 B-5 B-4
30 D-10 D-10 C-10 D-10 D-8 C-10 D-8 C-5 B-10 C-10 C-5 C-5 B-10 C-8 C-5 C-4 B-8 B-5 C-4
36 E-10 E-8 D-10 E-8 E-5 D-10 D-5 E-5 C-10 D-8 D-5 D-4 C-8 C-5 D-4 D-4 C-5 C-5 —
42 E-8 E-5 E-10 E-5 E-5 D-8 E-5 E-5 D-8 D-5 E-5 E-4 D-8 D-5 E-4 E-3 D-5 D-4 —
48 G-8 G-5 E-8 F-5 G-5 E-5 F-5 F-4 D-8 E-5 E-4 E-3 D-5 E-5 E-3 E-2.5 D-5 D-4 —
54 G-5 H-5 F-8 G-5 H-5 E-5 F-4 G-3 D-5 E-5 F-3 G-3 D-5 E-4 F-3 E-2.5 D-5 — —
60 H-5 H-5 G-8 H-5 H-4 F-5 G-4 G-3 E-5 F-5 G-3 G-2.5 E-5 F-4 G-2.5 G-2.5 E-4 — —
72 I-5 I-4 H-5 H-4 H-3 G-4 H-3 H-3 F-5 G-4 H-3 H-2.5 F-4 H-2 H-2 H-2 — — —
84 J-4 J-3 I-5 J-4 J-3 H-4 I-3 J-2.5 H-5 — — — G-4 — — — — — —
96 K-4 L-3 I-4 K-3 K-2.5 I-3 J-2.5 J-2 H-4 — — — — — — — — — —
Over 96 H-2.5 H-2.5 H-2.5 H-2.5 H-2.5 — — — — — — — — — — — — — —
plus plus plus plus plus
rods rods rods rods rods
Note: For a given duct thickness, letters indicate type (rigidity class) of duct reinforcement (see Table 17.3); numbers indicate maximum spacing (ft) between duct reinforcement. Use the same
metal duct thickness on all duct sides.
Source: Adapted with permission from SMACNA, HVAC Duct Construction Standards—Metal and Flexible. Refer to this standard for complete details.
17.16
TABLE 17.3 Transverse Joint Reinforcement
Pressure, 4 in. WG maximum*
Minimum
rigidity Standing Standing S Standing S
class drive slip Standing S Standing S Standing S (bar-reinforced) (angle-reinforced)
Hs 	 T (min), in. W, in. Hs 	 T (min), in. Hs 	 T (min), in. Hs 	 T (min), in. Hs 	 T (min) plus reinforcement (H 	 T), in.
A Use class B
Use class C
B Use class D
C — 1 	 0.0187 1 	 0.0187 Use class F
D — — 1 	 0.0236 1 	 0.0236
E — —
F — —
G — —
H — — — — —
I — — — — — 2 	 0.0356 plus
2 	 2 	 angle
J — — — — — 2 	 0.0356 plus
2 	 2 	 angle
* Acceptable to 36 in. length at 3 in. WG and to 30 in. length at 4 in. WG.
Source: Adapted with permission from SMACNA, HVAC Duct Construction Standards—Metal and Flexible and ASHRAE Handbook 1988, Equipment. Refer to SMACNA
Standard for complete details.
3	16
1	8
11	2 	 11	2 	 3	16 angle
11	2 	 0.0356 plus
11	2 	 1	8 bar
11	2 	 0.0296 plus 11	2 	 0.0466 15	8 	 0.0466 3	16
11	2 	 1	8 bar
11	2 	 0.0236 plus 11	2 	 0.0236 15	8 	 0.0296 3	16
11	8 	 0.0466 11	8 	 0.0356 3	16
11	8 	 0.0187
11	8 	 0.0296
1	2 	 0.0296 11	8 	 0.0.187
1	2 	 0.0187
Round Ducts
For a specified cross-sectional area and mean air velocity, a round duct has less fluid resistance
against airflow than rectangular and flat oval ducts. Round ducts also have better rigidity and
strength. The spiral- and longitude-seamed round ducts used in commercial buildings are usually
factory-fabricated to improve the quality and sealing of the ductwork. The pressure losses can be
calculated more precisely than for rectangular ducts, and result in a better balanced system. Air
leakage can be maintained at about 3 percent as a result of well-sealed seams and joints. Round
ducts have much smaller radiated noise breakout from the duct than rectangular and flat oval ducts.
The main disadvantage of round ducts is the greater space required under the beam for installation.
Factory-fabricated spiral-seamed round ducts are the most widely used air ducts in commercial
buildings. The standard diameters of round ducts range from 4 to 20 in. in 1-in. (100 to 500
mm in 25-mm) increments, from 20 to 36 in. in 2-in. (500 to 900 mm in 50-mm) increments, and
from 36 to 60 in. in 4-in. (900 to 1500 mm in 100-mm) increments. The minimum thickness of galvanized
sheet and fittings for round ducts in duct systems in commercial buildings is listed in
Table 17.4.
Many industrial air pollution control systems often require a velocity around 3000 fpm (15 m / s)
or higher to transport particulates. Round ducts with thicker metal sheets are usually used in such
applications.
Flat Oval Ducts
Flat oval ducts, as shown in Fig. 17.4, have a cross-sectional shape between rectangular and round.
They share the advantages of both the round and the rectangular duct with less large-scale air turbulence
and a small depth of space required during installation. Flat oval ducts are quicker to install
and have lower air leakage because of the factory fabrication.
Flat oval ducts are made in either spiral seam or longitudinal seam. The minimum thickness of
the galvanized sheet and fittings for flat oval duct systems used in commercial buildings is presented
in Table 17.5.
Duct area Maximum spacing, ft (m)
Up to 4 ft2 (0.37 m2) 8 (2.4)
Between 4 and 10 ft2 (0.37 and 0.93 m2) 6 (1.8)
Larger than 10 ft2 (0.93 m2) 4 (1.2)
AIR SYSTEMS: AIR DUCT DESIGN 17.17
TABLE 17.4 Round Ferrous Metal Duct Construction for Duct Systems in Commercial Buildings
Minimum galvanized steel thickness, in.
Pressure, 	2 in. WG Pressure, 2 in. WG
Duct Spiral Longitudinal Spiral Longitudinal Suggested
diameter, in. seam duct seam duct Fittings seam duct seam duct Fittings type of joint
Up through 8 0.0157 0.0236 0.0236 0.0157 0.0157 0.0187 Beaded slip
14 0.0187 0.0236 0.0236 0.0157 0.0187 0.0187 Beaded slip
26 0.0236 0.0296 0.0296 0.0187 0.0236 0.0236 Beaded slip
36 0.0296 0.0356 0.0356 0.0236 0.0296 0.0296 Beaded slip
50 0.0356 0.0466 0.0466 0.0296 0.0356 0.0356 Flange
60 0.0466 0.0575 0.0575 0.0356 0.0466 0.0466 Flange
84 0.0575 0.0705 0.0705 0.0466 0.0575 0.0575 Flange
Source: Adapted with permission from SMACNA, HVAC Duct Construction Standards—Metal and Flexible. Refer to SMACNA
Standard for complete details.
Flexible Ducts
Flexible ducts are often used to connect the main duct or the diffusers to the terminal box. Their
flexibility and ease of removal allow allocation and relocation of the terminal devices. Flexible
ducts are usually made of multiple-ply polyester film reinforced by a helical steel wire core or corrugated
aluminum spiral strips. The duct is often insulated by a fiberglass blanket 1 or 2 in. (25 to
50 mm) thick. The outer surface of the flexible duct is usually covered with aluminum foil or other
types of vapor barriers to prevent the permeation of water vapor into the insulation layer.
The inside diameter of flexible ducts may range from 2 to 10 in. in 1-in. (50 to 250 mm in
25-mm) increments and from 12 to 20 in. in 2-in. (300 to 500 mm in 50-mm) increments. The flexible
duct should be as short as possible, and its length should be fully extended to minimize flow
resistance.
Fiberglass Ducts
Fiberglass duct boards are usually made in 1-in. (25-mm) thickness. They are fabricated into rectangular
ducts by closures. A fiberglass duct with a 1.5-in. (38-mm) thickness may be used in the Gulf
area of the United States where the climate is hot and humid in summer, to minimize duct heat gain.
Round molded fiberglass ducts are sometimes used.
Fiberglass ducts have a good thermal performance. For a 1-in. (25-mm) thickness of duct board,
the U value is 0.21 Btu /h ft2 °F at 2000 fpm (1.192 W/m2 °C at 10 m / s) air velocity, which is
better than a galvanized sheet metal duct with a 1-in. (25-mm) inner liner. Fiberglass duct has good
sound attenuation characteristics. Its air leakage is usually 5 percent or less, which is far less than
that of a sheet-metal rectangular duct that is not well sealed. Another important advantage of fiberglass
duct is its lower cost.
The closures, also called taping systems, are tapes used to form rectangular duct sections from
duct boards and to join the sections and fittings into an integrated duct system. The improved acrylic
pressure-sensitive tapes provide a better bond than before. Heat-sensitive solid polymer adhesive
closures show themselves to be good sealing tapes even if dust, oil, or water is present on the surface
of the duct board. Mastic and glass fabric closures are also used in many fiberglass duct systems.
Tests show that the ongoing emission of fiberglass from the duct board was less than that contained
in outdoor air. Fiberglass ducts have a slightly higher friction loss than galvanized sheet duct
(0.03 in. WC or 7.5 Pa greater for a length of 100 ft or 30.5 m). They are also not as strong as metal
sheet duct. They must be handled carefully to prevent damage during installation.
Fiberglass ducts are used in duct systems with a pressure differential of 2 in. WC (500 Pa)
or less. Many codes restrict the use of fiberglass in sensitive areas such as operating rooms and
maternity wards.
17.18 CHAPTER SEVENTEEN
TABLE 17.5 Flat Oval Duct Construction for Positive-Pressure Duct
Systems in Commercial Buildings
Minimum galvanized steel thickness, in.
Major axis, Spiral Longitudinal Suggested
in. seam duct seam duct Fittings type of joint
Up through 24 0.0236 0.0356 0.0356 Beaded slip
36 0.0296 0.0356 0.0356 Beaded slip
48 0.0296 0.0466 0.0466 Flange
60 0.0356 0.0466 0.0466 Flange
70 0.0356 0.0575 0.0575 Flange
Over 70 0.0466 0.0575 0.0575 Flange
Source: Adapted with permission from SMACNA, HVAC Duct Construction Standards—
Metal and Flexible. Refer to SMACNA Standard for complete details.
17.4 DUCT HEAT GAIN, HEAT LOSS, AND DUCT INSULATION
Temperature Rise or Drop due to Duct Heat Gain or Loss
The temperature rise or drop from duct heat gain or loss is one of the parameters that affect the supply
air temperature as well as the supply volume flow rate in the air conditioning system design.
Heat gain or loss through the duct wall of a rectangular duct section with a constant-volume flow
rate qd, in Btu / h (W), can be calculated as
(17.28)
For a round duct section
17.29)
where U  overall heat transfer coefficient of duct wall, Btu/h ft2 °F (W/m2 °K)
P, L  perimeter and length of duct, ft (m)
Dd  diameter of round duct, ft (m)
Tam  temperature of ambient air, °F (°C)
Ten, Tlv  temperature of air entering and leaving duct section, °F (°C)
The temperature increase or drop of the air flowing through a duct section is given as
(17.30)
where Ad  cross-sectional area of duct, ft2 (m2). In Eq. (17.30), the mean air velocity v is
expressed in fpm [m / (60 s)].
We substitute Eq. (17.28) into Eq. (17.30). For rectangular duct let
(17.31)
For round duct, let
(17.32)
Then the temperature of air leaving the duct section is
(17.33)
Duct Insulation
Duct insulation is mounted or inner-lined to reduce heat loss and heat gain as well as to prevent the
condensation on the outer surface of the duct. It is usually in the form of duct wrap (outer surface),
duct inner liner, or fiberglass duct boards. Duct liner provides both thermal insulation and sound
attenuation. The thickness of an insulation layer is based on economical analysis.
ASHRAE/IESNA Standard 90.1-1999 mandates that all supply and return ducts and plenums
installed as part of an HVAC&R air distribution system shall be thermally insulated as listed in
Table 17.6. The insulated R-values in h  ft2 °F/ Btu (m2 °C/W), are for the insulation installed and
T1v 
2Tam  Ten( y 	 1)
y  1
y 
30Ddv scpa
UL
y 
120 Adv scpa
UPL
T1v 	 Ten 
qd
60Adv scpa
qd  UDdL Tam 	
Ten 	 T1v
2 
qd  UPL Tam 	
Ten  T1v
2 
AIR SYSTEMS: AIR DUCT DESIGN 17.19
do not include air film resistances. The required minimum thickness does not consider water vapor
transmission and possible surface condensation. The recommended thickness of insulation layer, or
duct wrap, for R-3.5 h ft2 °F/ Btu (0.62 m2 °C/ W) is 1 to 1.5 in. (25 to 38 mm), and for R-6
h ft2 °F/ Btu (1.06 m2 °C/ W) is 2 to 3 in. (50 to 75 mm).
Exceptions of duct and plenum insulation include
 Factory-installed plenums, casings, and ductwork as part of the equipment
 Ducts or plenums located in heated spaces, semiheated spaces, or cooled spaces
 For runout less than 10 ft (3 m) in length to air terminals or air outlets, R-value of insulation need
not exceed R-3.5.
 Backs of air outlets and outlet plenums exposed to unconditioned or indirectly conditioned space
with face area exceeding 5 ft2 (0.5 m2) need not exceed R-2, for those 5 ft2 (0.5 m2) or smaller
need not be insulated.
17.20 CHAPTER SEVENTEEN
TABLE 17.6 Minimum Duct Insulation R-Value,* Cooling and Heating Supply Ducts and Return Ducts
Duct location
Climate zone
Unvented Unvented
attic with attic with Indirectly
Envelop Ventilated backloaded roof Unconditioned conditioned
criteria table HDD65 CDD50 Exterior attic ceiling insulation space† space§ Buried
Heating ducts only
5-1 to 5-7 0-1800 All None None None None None None None
5-8 to 5-12 1801-3600 All R-3.5 None None None None None None
5-13 to 5-15 3601-5400 All R-3.5 None None None None None None
5-16 to 5-18 5401-7200 All R-6 R-3.5 None None None None R-3.5
Cooling only ducts
5-15, 18, 20, All 0-1800 R-1.9 R-1.9 R-1.9 R-1.9 R-1.9 None None
22 to 26
5-12, 14, 17, All 1801-3600 R-3.5 R-1.9 R-3.5 R-1.9 R-1.9 None None
19, 21
5-7, 9, 11, All 3601-5400 R-3.5 R-3.5 R-6 R-1.9 R-1.9 None None
13, 16
5-4, 6, 8, 10 All 5401-7200 R-6 R-6 R-6 R-3.5 R-1.9 None None
Combined heating and cooling ducts
5-9 1801-2700 0-5400 R-6 R-3.5 R-6 R-1.9 R-1.9 None
5-10 2701-3600 5401  R-6 R-6 R-6 R-3.5 R-3.5 None
5-11 2701-3600 3601-5400 R-6 R-6 R-6 R-3.5 R-3.5 None
5-12 2701-3600 0-3600 R-3.5 R-3.5 R-3.5 R-1.9 R-1.9 None
5-13 3601-5400 3601  R-6 R-6 R-6 R-3.5 R-3.5 None
5-14 3601-5400 1801-3600 R-6 R-3.5 R-6 R-1.9 R-3.5 None
5-15 3601-5400 0-1800 R-3.5 R-3.5 R-3.5 R-1.9 R-1.9 None
Return ducts
5-1 to 5-26 All climates R-3.5 R-3.5 R-3.5 None None None None
*Insulation R-values, measured in (h ft2 °F) / Btu, are for the insulation as installed and do not include film resistance. The required minimum
thicknesses do not consider water vapor transmission and possible surface condensation. Where exterior walls are used as plenum walls, wall insulation
shall be as required by the most restrictive condition of 6.2.4.2 or Section 5. Insulation resistance measured on a horizontal plane in accordance
with ASTM C518 at a mean temperature of 75°F at the installed thickness.
†Includes crawl spaces, both ventilated and nonventilated.
§Includes return air plenums with or without exposed roofs above.
Source: ASHRAE/IESNA Standard 90.1-1999. Reprinted by permission.
If the temperature of the ambient air is 80°F (26.7°C) with a relative humidity of 50 percent, its dew
point is 59°F (18.3°C). Only when the outer surface temperature of the duct Tsd 59°F (15°C) will
condensation not occur. Refer to ASHRAE Standard 90.1-1999 for details.
Temperature Rise Curves
Duct heat gain or loss and the temperature rise or drop of the air inside the duct depend on air
velocity, duct dimensions, and duct insulation. Figure 17.5 shows curves for the temperature rise in
round ducts. These curves are calculated according to Eqs. (17.30) to (17.33) under these conditions:
 Thickness of the insulation layer of duct wrap is 1.5 in. (38 mm), and the thermal conductivity of
the insulating material k  0.30 Btu in /h ft2 °F (0.043 W/m°C).
 Heat transfer coefficient of the outer surface of the duct
h  1.6 Btu /h ft2 °F (9.1 W/m2 °C) (17.34)
 Convective heat-transfer coefficient of the inside surface hc, in Btu /h ft2 °F, can be calculated by
(17.35)
where ReD  Reynolds number based on duct diameter as characteristic length
Pr  Prandtl number
 Air temperature inside the air duct is assumed to be 55°F (12.8°C).
 Temperature difference between the air inside the duct and the ambient air surrounding the duct is
25°F (13.9°C).
hc  0.023  k
Dd
ReD
0.8
Pr0.4
AIR SYSTEMS: AIR DUCT DESIGN 17.21
FIGURE 17.5 Temperature rise curves from duct heat gain.
If the temperature rise or drop has been determined, duct heat gain or heat loss can be either
expressed in percentage of supply temperature differential or calculated by Eqs. (17.28) and (17.29).
17.5 FRICTIONAL LOSSES
In an air duct system, there are two types of resistance against the airflow: frictional losses and
dynamic losses.
Darcey-Weisbach Equation
Frictional losses result mainly from the shearing stress between the fluid layers of the laminar sublayer,
which is adjacent to the surface of the duct wall. Friction also exists when the fluid particles
in the turbulent flow bump against the protuberances of the duct wall. These lead to the production
of eddies and energy loss. Friction losses occur along the entire length of the air duct.
For a steady, incompressible fluid flow in a round duct or a circular pipe, friction head loss Hf, in
ft (m) of air column, can be calculated by the Darcy-Weisbach equation in the following form:
(17.36)
If the friction loss is presented in the form of pressure loss pf, in lbf / ft2 (Pa), then
(17.37)
where f  friction factor
L  length of duct or pipe, ft (m)
D  diameter of duct or pipe, ft (m)
v  mean air velocity in duct, ft / s (m / s)
If air velocity v is expressed in fpm and pf is expressed in in. WC, then a conversion factor
Fcv  (1 / 5.19)(1 / 60)2  0.0000535 should be used. That is,
(17.38)
In Eqs. (17.36), (17.37), and (17.38), strictly speaking, D should be replaced by hydraulic diameter
Dh. But for round ducts and pipes, D  Dh.
Friction Factor
Friction loss is directly proportional to the friction factor f, which is dimensionless. The relationship
between f and the parameters that influence the magnitude of the friction factor is shown in
Fig. 17.6, which is called a Moody diagram. In Fig. 17.6, the term represents the absolute roughness
of the surface protuberances, expressed in ft (m), as shown in Fig. 17.7. The ratio /D indicates
the relative roughness of the duct or pipe.
For laminar flow in an air duct when ReD 2000, f is affected mainly by the viscous force of
the fluid flow and, therefore, is a function of ReD only. That is,
(17.39) f 
64
ReD
pf  0.0000535f  L
D 
v2
2gc 
pf  f  L
D 
v2
2gc 
Hf  f  L
D  v2
2g 
17.22 CHAPTER SEVENTEEN
In an ideal smooth tube or duct, that is, /D 10	5, if ReD 4000, the surface roughness is submerged
in the laminar sublayer with a thickness of , and the fluid moves smoothly, passing over
the protuberances. In this case, f decreases with an increase of ReD. The relationship between f and
ReD may be expressed by the Blasius empirical formula
(17.40) f 
0.316
ReD
0.25
AIR SYSTEMS: AIR DUCT DESIGN 17.23
FIGURE 17.6 Moody diagram. (Source: Moody, L.F., Transactions A.S.M.E., vol. 66, 1994. Reprinted with permission.)
With a further increase of ReD, the laminar sublayer becomes thinner, even thinner than the height
of the irregularities , that is,  . The protuberances form the separation of fluid flow, enhance
the formation of vortices, and, therefore, increase the pressure loss as well as the value of f at
greater ReD.
If ReD is beyond a limit called the Rouse limit, f depends mainly on the relative roughness of the
duct wall. The Rouse limit line can be determined by ReD  200/ , as shown in Fig. 17.6.
Duct Friction Chart
In most air ducts, ReD ranges from 1 	 104 to 2 	 106, and /D may vary from 0.005 to 0.00015.
This covers a transition zone between hydraulic smooth pipes and the Rouse limit line. Within this
region, f is a function of both ReD and /D. Colebrook (1939) recommended the following empirical
formula to relate ReD, /D, and f for air ducts:
(17.41)
In Eq. (17.41), and D must be expressed in the same units. Swamee and Jain suggested the use of
an explicit expression that can give approximately the same value of f as Colebrook’s formula:
(17.42)
In air duct calculations, a rough estimate of f can be determined from the Moody diagram based on
knowing the relative roughness /D and ReD.
For practical calculations, a friction chart for round ducts, developed by Wright, in the form
shown in Fig. 17.8, is widely used. In this chart, air volume flow rate , in cfm, and the frictional V?
f 
0.25
{log [ / (3.7D)  5.74 / (0.9 ReD)]}2
1
?f
 	2 log  
3.7D

2.51
ReD?f 
?f(/D)
17.24 CHAPTER SEVENTEEN
FIGURE 17.7 Modes of airflow when air passes over and around surface
protuberances of the duct wall: (a)  ; (b)  .
loss per unit length pf, in in. WC per 100 ft, are used as coordinates. The mean air velocity v, in
fpm, and the duct diameter are shown in inclined lines in this chart.
The duct friction chart can be applied to the following conditions without corrections:
 A temperature from 41 to 95°F (5 to 35°C)
 Up to an elevation of 1600 ft (488 m)
 Duct pressure difference of 20 in. WC (5000 Pa) with respect to ambient air pressure
 Duct material of medium-smooth roughness
Roughness and Temperature Corrections
The absolute roughness , in ft, given in ASHRAE Handbook 1989, Fundamentals, is listed in
Table 17.7. When duct material differs from medium-smooth roughness, or when the air duct is
installed at an elevation above 1600 ft (488 m), or when warm air is supplied at a temperature
higher than 95°F (35°C), then corrections should be made to pf as follows:
(17.43)
where pf,c  friction loss found from the duct friction chart, in. WC (Pa). In Eq. (17.43), Ksr indicates
the correction factor for surface roughness, which is dimensionless, and can be calculated as
(17.44)
Here fc denotes the friction factor of duct material of surface roughness that is specified by the duct
friction chart, that is,  0.0005 ft (0.09 mm). The symbol fa represents the actual friction factor of
duct material with surface roughness differing from fc. Both fa and fc can be calculated from
Eq. (17.41) or (17.42).
Ksr 
fa
fc
pf  KsrKTKe1 pf,c
AIR SYSTEMS: AIR DUCT DESIGN 17.25
TABLE 17.7 Duct Roughness
Duct material (roughness, ft) Roughness category Absolute roughness , ft
Uncoated carbon steel, clean (0.00015 ft) Smooth 0.0001
PVC plastic pipe (0.00003–0.00015 ft)
Aluminum (0.00015–0.0002 ft)
Galvanized steel, longitudinal seams, 4-ft joints Medium smooth 0.0003
(0.00016–0.00032 ft)
Galvanized steel, spiral seams, with 1, 2, and 3
ribs, 12-ft joints (0.00018–0.00038 ft)
Galvanized steel, longitudinal seams, 2.5-ft joints Average 0.0005
(0.0005 ft)
Fibrous glass duct, rigid Medium rough 0.003
Fibrous glass duct liner, air side with facing
material (0.0005 ft)
Fibrous glass duct liner, air side spray-coated (0.015 ft) Rough 0.01
Flexible duct, metallic (0.004–0.007 ft when fully
extended)
Flexible duct, all types of fabric and wire
(0.0035–0.015 ft when fully extended)
Concrete (0.001–0.01 ft)
Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with permission.
The term KT indicates the correction factor for air temperature inside the duct, which affects the
density of the air; KT is dimensionless and can be calculated as
(17.45)
where Ta  actual air temperature inside duct, °F. The term Kel indicates the correction factor for
elevation, which also affects the density of the air; it is dimensionless. When the elevation is greater
than 1600 ft (488 m), Kel can be evaluated as
(17.46)
where pat  actual atmospheric or barometric pressure, in. Hg
pat,mm  actual atmospheric or barometric pressure, mm Hg

pat,mm
760
Kel 
pat
29.92
KT   530
Ta  460 0.825
17.26 CHAPTER SEVENTEEN
FIGURE 17.8 Friction chart for round ducts. (Source: ASHRAE Handbook 1989. Reprinted with permission.)
Example 17.1. A fabric and wire flexible duct of 8-in. (200 mm) diameter installed in a commercial
building at sea level has a surface roughness  0.12 in. (3 mm). The mean air velocity inside
the air duct is 800 fpm (4 m / s). Find the correction factor of surface roughness. The viscosity of air
at 60°F is 1.21 	 10	5 lb / ft  s (2.0 	 10	5 Pa s).
Solution. For the galvanized sheet duct specified in the duct friction chart,  0.0005 ft, or
0.006 in. (0.15 mm). Then
From Eq. (17.42),
For the fabric and wire flexible duct, ReD is the same as for the galvanized sheet duct. From the data
given,  0.12 in., so the actual friction factor is
From Eq. (17.44), the correction factor of surface roughness is
Circular Equivalents
In Eq. (17.37), if D is replaced by Dh, then
Apparently, for circular or noncircular air ducts with different cross-sectional shapes and the same hydraulic
diameter, the pressure loss is the same for an equal length of air duct at equal mean air velocities.
Circular equivalents are used to convert the dimension of a noncircular duct into an equivalent
diameter De, in in. (mm), of a round duct when their volume flow rates and frictional losses per unit
length pf,u are equal. A noncircular duct must be converted to a circular equivalent first before determining
its pf,u from the duct friction chart. The hydraulic diameter Dh, in in. (mm), is defined as
(17.47)
where A  area, in.2 (mm2) and P  perimeter, in. (mm). Based on experimental results, Heubscher
(1948) recommended the following formula to calculate De for rectangular duct at equal and pf,u:
(17.48)
where a, b  dimensions of two sides of rectangular duct, in. (mm). The circular equivalents of
rectangular ducts at various dimensions calculated by Eq. (17.48) are listed in Table 17.8.
De 
1.30(ab)0.625
(a  b)0.25
V?
Dh 
4A
P
V?
pf  f  L
Dh 
v2
2gc 
Ksr 
fa
fc

0.044
0.0204
 2.16
fa 
0.25
{log [0.12 / (3.7 	 8)  5.74 / (0.9 	 5.5 	 104)]}2  0.044

0.25
{log [0.006 / (3.7 	 8)  5.74 / (0.9 	 5.5 	 104)]}2  0.0204
fc 
0.25
{log [ / (3.7D)  5.74 / (0.9 ReD)]}2
ReD 
vD



0.075 	 1000 	 8
60 	 12 	 1.21 	 10	5  5.5 	 104
AIR SYSTEMS: AIR DUCT DESIGN 17.27
17.28
TABLE 17.8 Circular Equivalents of Rectangular Ducts for Equal Friction and Capacity
One side,
rectangular
Adjacent side, in.
duct, in. 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0 9.0 10.0 11.0 12.0 13.0 14.0 15.0 16.0
3.0 3.8 4.0 4.2 4.4 4.6 4.7 4.9 5.1 5.2 5.5 5.7 6.0 6.2 6.4 6.6 6.8 7.0
3.5 4.1 4.3 4.6 4.8 5.0 5.2 5.3 5.5 5.7 6.0 6.3 6.5 6.8 7.0 7.2 7.5 7.7
4.0 4.4 4.6 4.9 5.1 5.3 5.5 5.7 5.9 6.1 6.4 6.7 7.0 7.3 7.6 7.8 8.0 8.3
4.5 4.6 4.9 5.2 5.4 5.7 5.9 6.1 6.3 6.5 6.9 7.2 7.5 7.8 8.1 8.4 8.6 8.8
5.0 4.9 5.2 5.5 5.7 6.0 6.2 6.4 6.7 6.9 7.3 7.6 8.0 8.3 8.6 8.9 9.1 9.4
5.5 5.1 5.4 5.7 6.0 6.3 6.5 6.8 7.0 7.2 7.6 8.0 8.4 8.7 9.0 9.3 9.6 9.9
One side, Side
rectan- rectangular
Adjacent side, in.
gular
duct, in. 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 22 24 26 28 30 duct
6 6.6 6
7 7.1 7.7 7
8 7.6 8.2 8.7 8
9 8.0 8.7 9.3 9.8 9
10 8.4 9.1 9.8 10.4 10.9 10
11 8.8 9.5 10.2 10.9 11.5 12.0 11
12 9.1 9.9 10.7 11.3 12.0 12.6 13.1 12
13 9.5 10.3 11.1 11.8 12.4 13.1 13.7 14.2 13
14 9.8 10.7 11.5 12.2 12.9 13.5 14.2 14.7 15.3 14
15 10.1 11.0 11.8 12.6 13.3 14.0 14.6 15.3 15.8 16.4 15
16 10.4 11.3 12.2 13.0 13.7 14.4 15.1 15.7 16.4 16.9 17.5 16
17 10.7 11.6 12.5 13.4 14.1 14.9 15.6 16.2 16.8 17.4 18.0 18.6 17
18 11.0 11.9 12.9 13.7 14.5 15.3 16.0 16.7 17.3 17.9 18.5 19.1 19.7 18
19 11.2 12.2 13.2 14.1 14.9 15.7 16.4 17.1 17.8 18.4 19.0 19.6 20.2 20.8 19
20 11.5 12.5 13.5 14.4 15.2 16.0 16.8 17.5 18.2 18.9 19.5 20.1 20.7 21.3 21.9 20
22 12.0 13.0 14.1 15.0 15.9 16.8 17.6 18.3 19.1 19.8 20.4 21.1 21.7 22.3 22.9 24.0 22
24 12.4 13.5 14.6 15.6 16.5 17.4 18.3 19.1 19.9 20.6 21.3 22.0 22.7 23.3 23.9 25.1 26.2 24
26 12.8 14.0 15.1 16.2 17.1 18.1 19.0 19.8 20.6 21.4 22.1 22.9 23.5 24.2 24.9 26.1 27.3 28.4 26
28 13.2 14.5 15.6 16.7 17.7 18.7 19.6 20.5 21.3 22.1 22.9 23.7 24.4 25.1 25.8 27.1 28.3 29.5 30.6 28
30 13.6 14.9 16.1 17.2 18.3 19.3 20.2 21.1 22.0 22.9 23.7 24.4 25.2 25.9 26.6 28.0 29.3 30.5 31.7 32.8 30
32 14.0 15.3 16.5 17.7 18.8 19.8 20.8 21.8 22.7 23.5 24.4 25.2 26.0 26.7 27.5 28.9 30.2 31.5 32.7 33.9 32
34 14.4 15.7 17.0 18.2 19.3 20.4 21.4 22.4 23.3 24.2 25.1 25.9 26.7 27.5 28.3 29.7 31.0 32.4 33.7 34.9 34
36 14.7 16.1 17.4 18.6 19.8 20.9 21.9 22.9 23.9 24.8 25.7 26.6 27.4 28.2 29.0 30.5 32.0 33.3 34.6 35.9 36
38 15.0 16.5 17.8 19.0 20.2 21.4 22.4 23.5 24.5 25.4 26.4 27.2 28.1 28.9 29.8 31.3 32.8 34.2 35.6 36.8 38
40 15.3 16.8 18.2 19.5 20.7 21.8 22.9 24.0 25.0 26.0 27.0 27.9 28.8 29.6 30.5 32.1 33.6 35.1 36.4 37.8 40
17.29
42 15.6 17.1 18.5 19.9 21.1 22.3 23.4 24.5 25.6 26.6 27.6 28.5 29.4 30.3 31.2 32.8 34.4 35.9 37.3 38.7 42
44 15.9 17.5 18.9 20.3 21.5 22.7 23.9 25.0 26.1 27.1 28.1 29.1 30.0 30.9 31.8 33.5 35.1 36.7 38.1 39.5 44
46 16.2 17.8 19.3 20.6 21.9 23.2 24.4 25.5 26.6 27.7 28.7 29.7 30.6 31.6 32.5 34.2 35.9 37.4 38.9 40.4 46
48 16.5 18.1 19.6 21.0 22.3 23.6 24.8 26.0 27.1 28.2 29.2 30.2 31.2 32.2 33.1 34.9 36.6 38.2 39.7 41.2 48
50 16.8 18.4 19.9 21.4 22.7 24.0 25.2 26.4 27.6 28.7 29.8 30.8 31.8 32.8 33.7 35.5 37.2 38.9 40.5 42.0 50
52 17.1 18.7 20.2 21.7 23.1 24.4 25.7 26.9 28.0 29.2 30.3 31.3 32.3 33.3 34.3 36.2 37.9 39.6 41.2 42.8 52
54 17.3 19.0 20.6 22.0 23.5 24.8 26.1 27.3 28.5 29.7 30.8 31.8 32.9 33.9 34.9 36.8 38.6 40.3 41.9 43.5 54
56 17.6 19.3 20.9 22.4 23.8 25.2 26.5 27.7 28.9 30.1 31.2 32.3 33.4 34.4 35.4 37.4 39.2 41.0 42.7 44.3 56
58 17.8 19.5 21.2 22.7 24.2 25.5 26.9 28.2 29.4 30.6 31.7 32.8 33.9 35.0 36.0 38.0 39.8 41.6 43.3 45.0 58
60 18.1 19.8 21.5 23.0 24.5 25.9 27.3 28.6 29.8 31.0 32.2 33.3 34.4 35.5 36.5 38.5 40.4 42.3 44.0 45.7 60
62 20.1 21.7 23.3 24.8 26.3 27.6 28.9 30.2 31.5 32.6 33.8 34.9 36.0 37.1 39.1 41.0 42.9 44.7 46.4 62
64 20.3 22.0 23.6 25.1 26.6 28.0 29.3 30.6 31.9 33.1 34.3 35.4 36.5 37.6 39.6 41.6 43.5 45.3 47.1 64
66 20.6 22.3 23.9 25.5 26.9 28.4 29.7 31.0 32.3 33.5 34.7 35.9 37.0 38.1 40.2 42.2 44.1 46.0 47.7 66
68 20.8 22.6 24.2 25.8 27.3 28.7 30.1 31.4 32.7 33.9 35.2 36.3 37.5 38.6 40.7 42.8 44.7 46.6 48.4 68
70 21.0 22.8 24.5 26.1 27.6 29.1 30.4 31.8 33.1 34.4 35.6 36.8 37.9 39.1 41.2 43.3 45.3 47.2 49.0 70
72 23.1 24.8 26.4 27.9 29.4 30.8 32.2 33.5 34.8 36.0 37.2 38.4 39.5 41.7 43.8 45.8 47.8 49.6 72
74 23.3 25.1 26.7 28.2 29.7 31.2 32.5 33.9 35.2 36.4 37.7 38.8 40.0 42.2 44.4 46.4 48.4 50.3 74
76 23.6 25.3 27.0 28.5 30.0 31.5 32.9 34.3 35.6 36.8 38.1 39.3 40.5 42.7 44.9 47.0 48.9 50.9 76
78 23.8 25.6 27.3 28.8 30.4 31.8 33.3 34.6 36.0 37.2 38.5 39.7 40.9 43.2 45.4 47.5 49.5 51.4 78
80 24.1 25.8 27.5 29.1 30.7 32.2 33.6 35.0 36.3 37.6 38.9 40.2 41.4 43.7 45.9 48.0 50.1 52.0 80
82 26.1 27.8 29.4 31.0 32.5 34.0 35.4 36.1 38.0 39.3 40.6 41.8 44.1 46.4 48.5 50.6 52.6 82
84 26.4 28.1 29.7 31.3 32.8 34.3 35.7 37.1 38.4 39.7 41.0 42.2 44.6 46.9 49.0 51.1 53.2 84
86 26.6 28.3 30.0 31.6 33.1 34.6 36.1 37.4 38.8 40.1 41.4 42.6 45.0 47.3 49.6 51.6 53.7 86
88 26.9 28.6 30.3 31.9 33.4 34.9 36.4 37.8 39.2 40.5 41.8 43.1 45.5 47.8 50.0 52.2 54.3 88
90 27.1 28.9 30.6 32.2 33.8 35.3 36.7 38.2 39.5 40.9 42.2 43.5 45.9 48.3 50.5 52.7 54.8 90
92 29.1 30.8 32.5 34.1 35.6 37.1 38.5 39.9 41.3 42.6 43.9 46.4 48.7 51.0 53.2 55.3 92
96 29.6 31.4 33.0 34.7 36.2 37.7 39.2 40.6 42.0 43.3 44.7 47.2 49.6 52.0 54.2 56.4 96
One side, Side
rectan- rectangular
Adjacent side, in.
gular
duct, in. 32 34 36 38 40 42 44 46 48 50 52 56 60 64 68 72 76 80 84 88 duct
32 35.0 32
34 36.1 37.2 34
36 37.1 38.2 39.4 36
38 38.1 39.3 40.4 41.5 38
40 39.0 40.3 41.5 42.6 43.7 40
42 40.0 41.2 42.5 43.7 44.8 45.9 42
44 40.9 42.2 43.5 44.7 45.8 47.0 48.1 44
46 41.8 43.1 44.4 45.7 46.9 48.0 49.2 50.3 46
48 42.6 44.0 45.3 46.6 47.9 46.1 50.2 51.4 52.5 48
50 43.6 44.9 46.2 47.5 48.8 50.0 51.2 52.4 53.6 54.7 50
17.30
TABLE 17.8 (Continued )
One side, Side
rectan- rectangular
Adjacent side, in.
gular
duct, in. 32 34 36 38 40 42 44 46 48 50 52 56 60 64 68 72 76 80 84 88 duct
52 44.3 45.7 47.1 48.4 49.7 51.0 52.2 53.4 54.6 55.7 56.8 52
54 45.1 46.5 48.0 49.3 50.7 52.0 53.2 54.4 55.6 56.8 57.9 54
56 45.8 47.3 48.8 50.2 51.6 52.9 54.2 55.4 56.6 57.8 59.0 61.2 56
58 46.6 48.1 49.6 51.0 52.4 53.8 55.1 56.4 57.6 58.8 60.0 62.3 58
60 47.3 48.9 50.4 51.9 53.3 54.7 60.0 57.3 58.6 59.8 61.0 63.4 65.6 60
62 48.0 49.6 51.2 52.7 54.1 55.5 56.9 58.2 59.5 60.8 62.0 64.4 66.7 62
64 48.7 50.4 51.9 53.5 54.9 56.4 57.8 59.1 60.4 61.7 63.0 65.4 67.7 70.0 64
66 49.4 51.1 52.7 54.2 55.7 57.2 58.6 60.0 61.3 62.6 63.9 66.4 68.8 71.0 66
68 50.1 51.8 53.4 55.0 56.6 58.0 59.4 60.8 62.2 63.6 64.9 67.4 69.8 72.1 74.3 68
70 50.8 52.5 54.1 55.7 57.3 58.8 60.3 61.7 63.1 64.4 65.8 68.3 70.8 73.2 75.4 70
72 51.4 53.2 54.8 56.6 58.0 59.6 61.1 62.5 63.9 65.3 66.7 69.3 71.8 74.2 76.5 78.7 72
74 52.1 53.8 55.5 57.2 58.8 60.3 61.9 63.3 64.8 66.2 67.5 70.2 72.7 75.2 77.5 79.8 74
76 52.7 54.5 56.2 57.9 59.5 61.1 62.6 64.1 65.6 67.0 68.4 71.1 73.7 76.2 78.6 80.9 83.1 76
78 53.3 55.1 56.9 58.6 60.2 61.8 63.4 64.9 66.4 67.9 69.3 72.0 74.6 77.1 79.6 81.9 84.2 78
80 53.9 55.8 57.5 59.3 60.9 62.6 64.1 65.7 67.2 68.7 70.1 72.9 75.4 78.1 80.6 82.9 85.2 87.5 80
82 54.5 56.4 58.2 59.9 61.6 63.3 64.9 66.5 68.0 69.5 70.9 73.7 76.4 79.0 81.5 84.0 86.3 88.5 82
84 55.1 57.0 58.8 60.6 62.3 64.0 65.6 67.2 68.7 70.3 71.7 74.6 77.3 80.0 82.5 85.0 87.3 89.6 91.8 84
86 55.7 57.6 59.4 61.2 63.0 64.7 66.3 67.9 69.5 71.0 72.5 75.4 78.2 80.9 83.5 85.9 88.3 90.7 92.9 86
88 56.3 58.2 60.1 61.9 63.6 65.4 67.0 68.7 70.2 71.8 73.3 76.3 79.1 81.8 84.4 86.9 89.3 91.7 94.0 96.2 88
90 56.8 58.8 60.7 62.5 64.3 66.0 67.7 69.4 71.0 72.6 74.1 77.1 79.9 82.7 85.3 87.9 90.3 92.7 95.0 97.3 90
92 57.4 59.3 61.3 63.1 64.9 66.7 68.4 70.1 71.7 73.3 74.9 77.9 80.8 83.5 86.2 88.8 91.3 93.7 96.1 98.4 92
94 57.9 59.9 61.9 63.7 65.6 67.4 69.1 70.8 72.4 74.0 75.6 78.7 81.6 84.4 87.1 89.7 92.3 94.7 97.1 99.4 94
96 58.4 60.5 62.4 64.3 66.2 68.0 69.7 71.5 73.1 74.8 76.3 79.4 82.4 85.3 88.0 90.7 93.2 95.7 98.1 100.5 96
Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with permission.
For galvanized steel flat oval ducts with spiral seams, Heyt and Diaz (1975) proposed the
following formula to calculate the circular equivalent for use of the duct friction chart:
(17.49)
Here A is the cross-sectional area of the flat oval duct, in.2 (mm2), and is given as
(17.50)
and the perimeter P, in in. (mm), is calculated as
P  b  2(a  b) (17.51)
The dimensions a and b of the flat oval duct are shown in Fig. 17.4c.
17.6 DYNAMIC LOSSES
When air flows through duct fittings, such as, elbows, tees, diffusers, contractions, entrances and exits,
or certain equipment, a change in velocity or direction of flow may occur. Such a change leads to flow
separation and the formation of eddies and disturbances in that area. The energy loss resulting from
these eddies and disturbances is called dynamic loss pdy, in in. WC (Pa). Although a duct fitting is
fairly short, the disturbances it produces may persist over a considerable distance downstream.
In addition to the presence of dynamic loss pdy, frictional loss pf occurs when an airstream
flows through a duct fitting. For convenience in calculation, the length of the duct fitting is usually
added to the adjacent duct sections connected with this duct fitting of the same mean air velocity.
When airstreams of the same Reynolds number flow through geometrically similar duct fittings,
that is, in dynamic similarity, the dynamic loss is proportional to their velocity pressure pv. Dynamic
loss may be calculated as
(17.52)
where Co  local loss coefficient or dynamic loss coefficient
 air density, lb/ ft3 (kg /m3)
vo  mean air velocity of airstream at reference cross section o, fpm (m/ s)
gc  dimensional constant, 32.2 lbmft / lbf  s2, for SI units, gc  1
cf  conversion factor, for SI units, cf  1
Because the mean velocity of the airstream may vary at different ends of a duct fitting, Co is always
specified with respect to a velocity of a reference cross section o in the duct fitting.
Elbows
An elbow is a duct fitting in which the airflow changes direction. Elbows are shown in Figs. 17.9
and 17.10. Consider an elbow that makes a 90° turn in a round duct, as shown in Fig. 17.9a.
Because of the change of airstream direction, centrifugal force is created and acts toward the outer
wall of the duct. When the airstream flows from the straight part of the duct to the curved part, it is
accompanied by an increase in pressure and a decrease in air velocity at the outer wall. At the same
time, a decrease in pressure and an increase in air velocity take place at the inner wall. Therefore, a
diffuser effect occurs near the outer wall, and a bell mouth forms near the inner wall. After turning,
the opposite effect takes place as the airstream flows from the curved part to the straight part of the
pdy  Copv 
Co vo 2
2gccf
 Co vo
4005 2
A 
b2
4
 b(a 	 b)
De 
1.55A0.625
P0.25
AIR SYSTEMS: AIR DUCT DESIGN 17.31
duct. The diffuser effects lead to flow separations from both walls, and eddies and large-scale turbulence
form in regions AB and CE, as shown in Fig. 17.9a.
In a rectangular elbow, a radial pressure gradient is also formed by the centrifugal force along its
centerline NR, as shown in Fig. 17.10c. A secondary circulation is formed along with the main
forward airstream.
The magnitude of the local loss coefficient of an elbow is influenced by the following factors:
 Turning angle of the elbow
 Ratio of centerline radius (CLR) to diameter  Rc /D or Rc /W, where Rc represents the throat radius
and W the width of the duct, both in in. (mm), as shown in Fig. 17.9a
 A three-gore (number of pieces), five-gore, or pleated seven-gore 90° elbow (Fig. 17.9b)
 Installation of splitter vanes, which reduce the eddies and turbulence in an elbow
 Shape of cross-sectional area of the duct
As the CLR becomes greater, the flow resistance of the airstream becomes smaller. However, a
greater CLR requires more duct material, a higher labor cost, and a larger allocated space. A value
of Rc /D (CLR)  1 or 1.5 is often used if the space is available.
The installation of splitter vanes, or turning vanes, in rectangular ducts can effectively reduce
the pressure loss at the elbow. For a rectangular 90° elbow with Rc /W  0.75 and two splitter
17.32 CHAPTER SEVENTEEN
FIGURE 17.9 Round elbows: (a) region of eddies and
turbulences in a round elbow; (b) five-piece 90° round
elbow.
vanes, as shown in Fig. 17.10a, the local loss coefficient Co is only 0.04 to 0.05. The values of
Ro, R1, and R2 can be found from Table 17.11. For details, refer to ASHRAE Handbook 1997,
Fundamentals.
For a mitered rectangular 90° elbow without turning vanes, Co is about 1.1. If turning vanes are
installed, Co drops to only 0.12 to 0.18. Installation of splitter vanes in mitered elbows is also
AIR SYSTEMS: AIR DUCT DESIGN 17.33
FIGURE 17.10 Rectangular elbows: (a) rectangular elbow, smooth
radius, two splitter vanes; (b) mitered elbow; (c) secondary flow in a
mitered elbow.
advantageous for noise control because of the smaller fan total pressure and, therefore, lower fan
noise as well as lower airflow noise at the elbow.
Converging and Diverging Tees and Wyes
A branch duct that combines with or diverges from the main duct at an angle of 90° is called a
tee. However, if the angle lies between 15° and 75°, it is called a wye (Figs. 17.11 and 17.12). Tees
and wyes can be round, flat oval, or rectangular. Various types of converging and diverging tees and
wyes for round and flat oval ducts appear in Fig. 17.12.
The function of a converging tee or wye is to combine the airstream from the branch duct with
the airstream from the main duct. The function of a diverging tee or wye is to diverge part of the
airflow from the main duct into the branch takeoff. For airstreams flowing through a converging or
diverging tee or wye, the dynamic losses for the main stream can be calculated as
(17.53)
ps,c 
Cs,c vc
2
2gccf
 Cs,c  vc
4005 2
pc,s 
Cc,s vc
2
2gccf
 Cc,s vc
4005 2
17.34 CHAPTER SEVENTEEN
FIGURE 17.11 Converging and diverging wyes: (a) converging
wye, rectangular,  90°; (b) diverging wye, rectangular,
 90°.
17.35
FIGURE 17.12 Round and flat oval tees and wyes: (a) round tees, wyes, and cross; (b) diverging
tees and wyes with elbows; (c) flat oval tees and wyes.
For the branch stream,
(17.54)
In Eqs. (17.53) and (17.54), subscript c represents the common end, s the straight-through end, and
b the branch takeoff. The subscript c,s indicates the flow of main stream from the common end to
the straight-through end, and s,c the flow from the straight-through end to the common end. Similarly,
c,b denotes the flow of the branch stream from the common end to the branch takeoff, and b,c
the airflow from the branch duct to the common end. The airstreams flowing through a rectangular
converging or diverging tee are shown in Fig. 17.13.
As mentioned in Sec. 17.2, the total pressure of the main stream may increase when it flows
through a converging or diverging wye or tee. This occurs because energy is received from the
airstream of higher velocity or the diverging of the slowly moving boundary layer into branch
takeoff from the main airstream. However, the sum of the energies of the main and branch streams
leaving the duct fitting is always smaller than that entering the duct fitting because of the energy
losses.
The magnitude of local loss coefficients Cc,s, Cs,c, Cb,c, and Cc,b is affected by the shape and construction
of the tee or wye, the velocity ratios vs /vc and vb /vc, the volume flow ratios and
, and area ratios As /Ac and Ab /Ac.
For rectangular ducts, the converging or diverging wye of the configuration shown in Fig. 17.13
gives the lower Co and, therefore, less energy loss. For round ducts, the Co values for various types
of diverging tees and wyes with nearly the same outlet direction and the same velocity ratio
vb /vc  0.6, as shown in Fig. 17.12b, are quite different.
V?
b /V?c
V?
s /V?c
pb,c 
Cb,c vc
2
2gccf
 Cb,c  vc
4005 2
pc,b 
Cc,b vc
2
2gccf
 Cc,b  vc
4005 2
17.36 CHAPTER SEVENTEEN
FIGURE 17.13 Airflows through a rectangular converging or diverging wye.
Tee, diverging, round, with 45° elbow, branch
90° to main
Tee, diverging, round, with 90° elbow, branch
90° to main
Tee, diverging, round, with 45° elbow, conical
branch 90° to main
Co  1.60
Co  1.18
Co  0.84
The different Co values help in choosing the optimum duct fitting and balancing the total pressure
between different paths of airflow in an air duct system.
In Table 17.11 are listed the local loss coefficients of diverging tees and wyes. For details, refer
to ASHRAE Handbook Fundamentals (1997) and Idelchik (1986).
AIR SYSTEMS: AIR DUCT DESIGN 17.37
FIGURE 17.14 Openings mounted on a duct or a duct wall: (a) entrances; (b) exits.
Wye, 45°, diverging, round, with 60° elbow,
branch 90° to main
Wye, 45°, diverging, round, with 60° elbow,
conical branch 90° to main
Co  0.68
Co  0.52
Entrances, Exits, Enlargements, and Contractions
Entrances and exits are the end openings mounted on a duct or a duct wall, as shown in Fig. 17.14.
Because of the change in direction of the streamlines at the entrance, eddies and large-scale
turbulences develop along the duct wall when the airstream passes through the entrance. Generally,
the total pressure drop pt of the airstream before it enters the entrance is negligible. A sharp-edge
entrance may have a Co  0.9, whereas for an entrance flush-mounted with the wall, Co reduces to
0.5. An entrance with a conical converging bellmouth may further reduce Co to 0.4. If an entrance
with a smooth converging bell mouth is installed, Co could be as low as 0.1.
When the air flows through a wall outlet or an exit, flow separation occurs along the surface of
the vanes, and wakes are formed downstream, so there is a drop in total pressure. The velocity of
the airstream reaches its maximum value at the vena contracta, where the cross section of airflow is
minimum and the static pressure is negative. The total pressure loss at the outlet always includes the
velocity pressure of the discharge airstream.
Various types of return inlets, such as grilles and louvers, and supply outlets, such as diffusers,
are discussed in Chapter 18.
When air flows through an enlargement or a contraction, flow separation occurs and produces eddies
and large-scale turbulences after the enlargement, or before and after the contraction. Both cause
a total pressure loss pt, as shown in Fig. 17.15. To reduce the energy loss, a gradual expansion, often
called a diffuser, or converging transition is preferred. An expansion with an including angle of enlargement
 14°, as shown in Fig. 17.15a, is ideal. In actual practice, may be from 14° to 45° because
of limited space. For converging transitions, an including angle of 30° to 60° is usually used.
17.7 FLOW RESISTANCE
Flow resistance is a property of fluid flow that measures the characteristics of a flow passage which
resist the fluid flow in that passage with a corresponding total pressure loss at a specific volume
17.38 CHAPTER SEVENTEEN
FIGURE 17.15 Enlargements and contractions: (a) abrupt enlargement; (b) sudden contraction.
flow rate. Consider a round duct section (Fig. 17.16), between two cross-sectional planes 1 and 2,
with a diameter D, in ft (mm), and length L, in ft (m). The fluid flow has a velocity of v, in fpm
(m/ s), and a volume flow rate of , in cfm (m3 / s). From Eq. (17.37), the total pressure loss of the
fluid flow pt, in in. WC (Pa), between planes 1 and 2 can be calculated as
(17.55)
where f  friction factor
 density of fluid, lb/ ft3 (kg /m3)
gc  dimensional constant, 32.2 lbmft / lbf  s2 (for SI unit, gc  1)
Cv  conversion constant
Let R  8Cv fL / (gc 2D5), so that total pressure loss becomes
(17.56) R 
pt
V?
2
pt  RV?
2
pt  Cv f  L
D 
v2
2gc  
8Cv f L 
gc 2D5
V?
2
V?
AIR SYSTEMS: AIR DUCT DESIGN 17.39
FIGURE 17.16 Total pressure loss pt and flow resistance R of a round duct section.
In Eq. (17.56), R is defined as flow resistance, which indicates the resistance to fluid flow of this
duct section, and is characterized by its specific total pressure loss and volume flow rate. Flow resistance
in I-P units is expressed as in. WC/ (cfm)2 (in SI units, Pa s2/m6).
For a given duct section, D and L are constants. In addition, the difference in the mean values of f and
at different volume flow rates is small, so R can be considered a constant. The relationship between pt
and of this section can be represented by a parabola whose vertex coincides with the origin.
Differentiating Eq. (17.56) with respect to , then, gives
(17.57)
This means that the slope of any segment of the curve pt  R 2 is 2R . Consequently, the
greater the flow resistance R, the steeper the slope of the characteristic curve of this duct
section.
Because the flow resistance R of a specific duct section is related to by means of Eq. (17.56),
the total pressure loss of the fluid flow in this duct section at any can be calculated. Therefore, the
total pressure loss pt curve of this duct section can be plotted at different values, and forms the
pt- diagram. On the pt- diagram, each characteristic curve represents a unique value of flow
resistance R. If a duct section contains additional dynamic losses Cdy v2 / (2gc), as mentioned in Sec.
17.6, the relationship between pt, R, and defined by Eq. (17.56) still holds, but the flow resistance
R is changed to
(17.58)
where Cdy  local loss coefficient.
Flow Resistances Connected in Series
Consider a duct system that consists of several duct sections connected in series with air flowing
inside the duct, as shown in Fig. 17.17a. When these duct sections are connected in series, the volume
flow rate of air that flows through each section must be the same. The total pressure loss
between planes 1 and 4 is the sum of the total pressure losses of each duct section, i.e.,
(17.59)
where R12, R23, . . . , Rn  flow resistances of duct sections 1, 2, . . . , n, in. WC/ (cfm)2 (Pas2/m6)
p12, p23, . . . , pn  total pressure losses across flow resistances R12, R23, . . . , Rn, in. WC
(Pa)
12, 23, . . . , n  volume flow rates of air flowing through resistances R12, R23, . . . ,
Rn, cfm (m3 / s)
The flow resistance of the duct system Rs is equal to the sum of individual flow resistances of duct
sections connected in series, or
(17.60)
If the characteristic curves of each individual duct section with flow resistances of R12,
R23, . . . , Rn are plotted on a pt- diagram, then the characteristic curve of the entire duct can
be plotted either by using the relationship Rs  pt / 2 or by graphical method. Plotting by graphical
method is done by drawing a constant line. The points of pt on the characteristic curve of
the duct can be found by summing the individual total pressure losses p12, p23, . . . , pn of
individual duct sections 1, 2, . . . , n, as shown in Fig. 17.17a.
V?
V?
V?
Rs  R12  R23      Rn
V?
V?
V?
 (R12  R23      Rn)V? 2  RsV? 2
pt  p12  p23      pn
R 
8 (Cv fL /D  Cdy)
gc 2D4
V?
V?
V?
V?
V?
V?
V?
V?
dpt
dV?  2RV?
V?
V?
17.40 CHAPTER SEVENTEEN
Flow Resistances Connected in Parallel
When a duct system has several sections connected in parallel, as shown in Fig. 17.17b, the total
volume flow rate of this duct that flows through cross-sectional planes 1 and 2 is the sum of the
individual volume flow rates , , . . . , of each duct section, i.e.,
(17.61)
where p12  total pressure loss between planes 1 and 2, in. WC (Pa)
RA, RB, . . . , Rn  flow resistance of duct sections A, B, . . . , n, in. WC/ (cfm)2
As for the entire duct, . Here Rp represents the flow resistance of the entire duct
whose sections are in parallel connection. Then
(17.62)
1
?Rp

1
?RA

1
?RB
    
1
?Rn
V?
 ?p12 /Rp
 ?p12
RA
 ?p12
RB
     ?p12
Rn
V?
 V?A  V?B      V?n
V?
n V?B V?A
V?
AIR SYSTEMS: AIR DUCT DESIGN 17.41
FIGURE 17.17 Combination of flow resistances: (a) connected in series; (b) connected in parallel.
If duct sections are connected in parallel, the total pressure loss across the common junctions 1 and
2 must be the same, or
pt  p12 (17.63)
For two sections with flow resistances RA and RB connected in parallel, the flow resistance of the
combination is then
(17.64)
As in series connection, the characteristic curve of the duct with segments connected in parallel
can be plotted on the pt- diagram either by the relationship pt  Rp
2 or by graphical method.
In the graphical method, draw a constant-pt line; the total volume flow rate on the characteristic
curve of the duct can be found by adding the volume flow rates of the individual sections, as shown
in Fig. 17.17b.
Flow Resistance of a Y Connection
When a round duct consists of a main duct section and two branches, as shown in Fig. 17.18, a
Y-connection flow circuit is formed. In a Y-connection flow circuit, the volume flow of the main
duct section 01, denoted by 01, is the sum of the volume flow rates at branches 11 and 12, that
is, 01  11  12. If the ambient air that surrounds the duct is at atmospheric pressure, then
p11  p12 and p01  p02.
Because p11  R11 11
2 ,  p12  R12 12
2 ,
(17.65)
Therefore, total pressure loss at ductwork 01 and 02 is calculated as
(17.66)
Similarly, (17.67)
For a given duct, Cdy, L01, L11, L12, D01, D11, and D12 are constants. Air density  can be considered
a constant. Although the friction factor f is a function of air velocity v and the volume flow rate
, at high Reynolds numbers f is least affected and can also be taken as a constant. Therefore, R01
and R02 can be considered constants.
Figure 17.18 also shows the characteristic curve of a Y connection plotted on a ptdiagram.
Flow Resistance of a Duct System
Supply and return duct systems are composed of Y connections. If the ambient air is at atmospheric
pressure, the total pressure loss of a supply duct system and its flow characteristics, as shown in
V?
V?
p02'  R02'V? 01
2
 R01V? 01
2
 R01V? 01
2  R11V? 11
2  (R01  R11K11
2 )V? 01
2
p01  p01  p11
V?
12 
1
1  ?R12/R11
V?
01  K12V?01
V?
11 
1
1  ?R11/R12
V?
01  K11V?01
V?
V?
V?
V?
V?
V?
V?
V?
Rp 
RARB
RA  RB  2?RARB
17.42 CHAPTER SEVENTEEN
Fig. 17.19a, are given as
(17.68)
and (17.69)
The characteristic curve of a duct system on a pt- diagram is called the system curve. It is shown
in Fig. 17.19b.
17.8 PRINCIPLES AND CONSIDERATIONS IN AIR DUCT DESIGN
Optimal Air Duct Design
An optimal air duct system transports the required amount of conditioned, recirculated, or exhaust
air to the specific space and meets the following requirements:
V?
R0n  R01  R12K12
2      R(n	1)nK(n	1)n
 R0nV? 01
2
 R01V? 01
2  R12K12
2 V? 01
2      R(n	1)nK(n	1) nV? 01
2
p0n'  p01  p12      p(n	1)n
AIR SYSTEMS: AIR DUCT DESIGN 17.43
FIGURE 17.18 Flow resistance for a Y connection.
17.44 CHAPTER SEVENTEEN
FIGURE 17.19 Flow characteristics of a supply duct system: (a) schematic diagram; (b) system curve.
 An optimal duct system layout within the allocated space
 A satisfactory system balance, achieved through the pressure balance of various paths by changing
duct sizes or using different configurations of duct fittings
 Space sound level lower than the allowable limits
 Optimum energy loss and initial cost
 Installation with only necessary balancing devices such as air dampers and nozzle plates
 National, ASHRAE, and local codes of fire protection, duct construction, and duct insulation met
These requirements result in the design of optimal duct layout, duct size, and total pressure loss of
the duct system.
Air duct system design often requires comprehensive analysis and computer-aided calculating
programs. Different air duct systems have different transport functions and thus have their own
characteristics. It is difficult to combine the influences of cost, system balance, and noise together
with duct characteristics into one or two representative indices.
Design Velocity
For any air duct system, the nearer the duct section is to the fan outlet or inlet, the greater its volume
flow rate. The fan outlet or the main duct connected to the fan outlet is often the location
where maximum air velocity occurs. The maximum design air velocity vd,max is determined mainly
according to the space available, noise control, energy use, and cost considerations. Return and
branch ducts are nearer to the conditioned space and have comparatively lower volume flow rates
than the supply ducts; therefore, supply main ducts often allow a higher vd,max than return and
branch ducts, except when there is a surplus pressure in a branch duct or a higher branch velocity is
required to produce a negative local loss coefficient.
For supply air duct systems in high-rise commercial buildings, the maximum design air velocity
in the supply duct vd,max is often determined by the space available between the bottom of the beam
and the suspended ceiling, as allocated by the architect, where the main duct traverses under the
beam. Because of the impact in recent years of energy-efficient design of HVAC&R systems in
commercial buildings (as influenced by ASHRAE Standards, local codes, and regulations) the following
are true:
 In supply main ducts vd,max usually does not exceeds 3000 fpm (15 m / s). Airflow noise must be
checked at dampers, elbows, and branch takeoffs to satisfy the indoor NC range.
 In buildings with more demanding noise control criteria, such as hotels, apartments, and hospital
wards, in supply main ducts usually vd,max  2000 to 2500 fpm (10 to 12.5 m / s), in return main
ducts vd,max  1600 fpm (8 m / s), and in branch ducts vd,max  1200 fpm (6 m / s).
Higher air velocity results in a higher energy cost, and lower air velocity increases the material
and labor costs of the installation. If a commercial building has sufficient headroom in its ceiling
plenum, or if an industrial application has enough space at a higher level, an optimization procedure
can then be used to reach a compromise between energy and installation costs. A check of airflow
noise is still necessary.
For a particulate-transporting duct system, the air velocity must be higher than a specific value at
any section of the duct system to float and transport the particulates. Design velocities for the components
in an air duct system are listed in Table 17.9. The face velocity vfc, in fpm (m / s), is defined as
(17.70)
where  air volume flow rate flowing through component, cfm (m3 / s)
Ag, fc  gross face area of component perpendicular to airflow, width 	 height, ft2 (m2)
V?
vfc 
V?
Ag, fc
AIR SYSTEMS: AIR DUCT DESIGN 17.45
These face velocities are recommended based on the effectiveness in operation of the system component
and its optimum pressure loss.
System Balancing
For an air duct system, system balancing means that the volume flow rate that comes from each
outlet or flows into each inlet should be (1) equal or nearly equal to the design value for constant-
volume systems and (2) equal to the predetermined values at maximum and minimum flow
for variable-air-volume systems. System balancing is one of the primary requirements in air
duct design. For supply duct systems installed in commercial buildings, using dampers only to
provide design airflow often causes additional air leakage, as well as an increase in installation
cost, and in some cases objectional noise. Therefore, system balancing using dampers only is
not recommended.
A typical small supply duct system in which pt of the conditioned space is equal to zero is
shown in Fig. 17.20. For such a supply duct system, the pressure and volume flow characteristics
are as follows:
17.46 CHAPTER SEVENTEEN
TABLE 17.9 Design Velocities for Air-Handling System
Components
Face velocity,
Duct element fpm
Louvers
Intake
1. 7000 cfm and greater 400
2. 2000 to 7000 cfm 250 to 400
Exhaust
1. 5000 cfm and greater 500
2. Less than 5000 cfm 300 to 400
Filters
Panel filters
1. Viscous impingement 200 to 800
2. Dry-type extended-surface
Flat (low efficiency) Duct velocity
Pleated media (medium efficiency) up to 750
HEPA 250
Renewable media filters
1. Moving-curtain viscous impingement 500
2. Moving-curtain dry-media 200
Electronic air cleaners, ionizing-type 150 to 350
Heating coils
Steam and hot water 400 to 600
(200 min.,
1500 max.)
Dehumidifying coils 500 to 600
Air washers
Spray type 300 to 700
High-velocity spray type 1200 to 1800
Source: Adapted with permission from ASHRAE Handbook 1989,
Fundamentals.
 From node 1, the total pressure loss of the airstreams flowing through various paths 1-3, 1-2-4,
and 1-2-5 to the conditioned space is always equal, i.e.,
(17.71)
 Volume flow rates , , and supplied from the branch takeoffs, 1-3, 2-4, and 2-5 depend on
the size and the configuration of the duct fittings and duct sections in the main duct and in the
branch takeoffs as well as the characteritics of the supply outlet.
The relationship between the total pressure loss of any duct section pt, the volume flow rate of
the duct section , and the flow resistance R can be expressed as follows:
(17.72)
For duct paths 1-3 and 1-2-45,
Here, path 1-2-45 indicates the airflow having volume flow rate and flowing through nodes
1, 2 and a combined parallel path 2-4 and 2-5.
At design conditions, the flow resistance R1-2-45 is determined in such a manner that the total
pressure loss pt,1-2-45 along path 1-2-45 at a volume flow rate of is balanced with the total
pressure loss pt,1-3.
 If the diverging wye or tee, terminal, duct fittings, and duct section used in path 1-3 are similar to
those of paths 2-4 and 2-5, most probably will be greater than the required design value as the
result of lower R1-3. To have required flowing through path 1-3 at design conditions to provide
a system balance, the following means are needed: (1) Decrease R1-2-45, including an increase in
the size of duct section 1-2; (2) increase R1-3, mainly by using a smaller duct in section 1-3 and a
diverging wye or tee of greater local loss coefficient Cc,b.
 In a duct system, if the flow resistance of each branch duct Rb is greater, the duct system is more
easily adjusted to achieve an equal amount of supply volume flow from each of the branch supply
outlets. For the branch takeoff or connecting duct having a length less than 1 to 2 ft (0.3 to 0.6 m)
long, it is often difficult to increase Rb by reducing its size. Adding a volume damper directly on the
supply outlet may alter space airflow patterns. A better remedy is to vary the sizes of the successive
main duct sections with outlets of greater flow resistance to achieve a better system balance.
 A variable-air-volume duct system installed with a VAV box in each branch takeoff does provide
system balancing automatically. However, only a part of the modulating capacity of a VAV box is
allowed to be used to provide system balance of a specific branch takeoff (such as less than 20
percent), so that the quality of its modulation control at part-load operation is not impaired.
V?
1
V?
1
V?
2  V?3
V?
2  V?3
 R1-3V ?1
2  R1-2-45(V?2  V?3)2
pt,1-3  pt, 1-2-45
pt  RV?
2
V?
V?
3 V?2 V?1
pt,1-3  pt, 1-2-4  pt, 1-2-5
AIR SYSTEMS: AIR DUCT DESIGN 17.47
FIGURE 17.20 System balancing and critical path of a supply duct
system.
Critical Path
For any air duct system there exists a critical path, or design path of airflow, whose total flow resistance
Rdgn is maximum compared with other airflow paths when the volume flow rate of the critical
path is equal to the design value at design conditions.
A critical path is usually a duct path with more duct fittings and comparatively higher volume
flow rate; additionally, it may be the longest one. In Fig. 17.20, path FO-1-2-5 may be the critical
path. For an energy-efficient air duct design, the total pressure loss, including the local loss coeffi-
cients of the duct fittings along the critical path, should always be minimized, especially at the fan
inlet and outlet or their vicinity (see Chap. 21).
To reduce the dynamic losses of the critical path, the following are recommended:
 Maintain an optimum air velocity for airflow through the duct fittings.
 Emphasize the reduction of dynamic losses of duct fittings that are nearer to the fan outlet or inlet
with a high air velocity, especially the fan system effects. Fan system effects are discussed in
Chap. 21.
 Evans and Tsal (1996) recommend use of 90° elbows with an Rc /D or Rc/W of 1.5. If space is
not available, use one or two splitter vanes in the elbows with throat radius or Rc/W  0.1; the
throat radius should not be smaller than 4 in. (100 mm). Turning vanes should not be used in a
transitional elbow or an elbow other than 90°. Proper turning vane installation is critical to performance,
which favors a factory-made unit.
 Set two duct fittings as far apart as allowable; if they are too close together, the eddies and largescale
turbulences of the first duct fitting often affect the velocity distribution in the second duct
fitting and considerably increase the pressure loss in the second fitting.
For other duct paths, if the total pressure loss at the design flow rate is smaller than the pressure
loss available, a smaller duct size and duct fittings of greater local loss coefficients may provide a
better balance.
Air Leakage
Conditioned air leakage from the joints and seams of the air duct to the space, which is not air conditioned,
is always a waste of refrigeration or heating energy as well as fan power. Based on their
tested results, Swim and Griggs (1995) reported that the joints were the major leakage sites and accounted
for 62 to 90 percent of the total air leakage from joints and seams. Air leakage depends
mainly on the use of sealant on joints and seams, the quality of fabrication, and the shape of the
ducts. Heat-sensitive tapes, mastic and glass fabric, and many other materials are used as the
sealants for the joints and seams.
Duct leakage classifications, based on tests conducted by AISI, SMACNA, ASHRAE, and the
Thermal Insulation Manufacturers Association (TIMA), are presented in Table 17.10. The air leakage
rate , in cfm/100 ft2 (L/ s per m2) of duct surface area can be calculated by the following formula:
(17.73)
The leakage class can then be calculated as
(17.73a)
where CL  leakage class, cfm per 100-ft2 duct surface area at a static pressure difference of 1 in.
WC [based on Eq. (17.73a)]
C  constant affected by area characteristics of leakage path
psd  static pressure differential between air inside and outside duct, in. WC (Pa)
CL 
V?
L
psd
0.65
V?
L  CL psd
0.65
V?
L
17.48 CHAPTER SEVENTEEN
ASHRAE/IESNA Standard 90.1-1999 mandates that ductwork and plenum shall be sealed in accordance
with the following requirements:
The sealing requirement of sealing levels A, B, and C are as follows:
Supply duct
Location 2 in. WC 2 in. WC Exhaust duct Return duct
Outdoors A A C A
Unconditioned spaces B A C B
Conditioned spaces C B B C
AIR SYSTEMS: AIR DUCT DESIGN 17.49
TABLE 17.10 Duct Leakage Classification
Predicted leakage
class CL
Type of duct Sealed Unsealed
Metal (flexible excluded)
Round and oval 3 30
(6 to 70)
Rectangular
2 in. WC 12 48
(both positive and (12 to 110)
negative pressures)
2 and 10 in. WC 6 48
(both positive and (12 to 110)
negative pressures)
Flexible
Metal, aluminum 8 30
(12 to 54)
Nonmetal 12 30
(4 to 54)
Fibrous glass
Rectangular 6 NA
Round 3 NA
The leakage classes listed in this table are averages based on
tests conducted by AISA/SMACNA 1972, ASHRAE/SMACNA/
TIMA 1985, and ASHRAE 1988. Leakage classes listed are not
necessarily recommendations on allowable leakage. The designer
should determine allowable leakage and specify acceptable duct
leakage classifications.
Source: ASHRAE Handbook 1989, Fundamentals. Reprinted
with permission.
A  All transverse joints, longitudinal
seams, and duct wall penetrations.
Pressure-sensitive tape shall not be
used as the primary sealant.
B  All transverse joints, longitudinal
seams. Pressure-sensitive tape shall
not be used as the primary sealant.
C  Transverse joints only.
Longitudinal seams are joints oriented in the direction of airflow. Transverse joints are connections
of two duct sections oriented perpendicular to air flow. Duct wall penetrations are openings made
by any screw fastener, pipe, rode, or wire. Spiral lock seams in round and flat oval duct need not be
sealed. All other connections are considered transverse joints.
ASHRAE/IESNA Standard 90.1-1999 also mandates that:
 Metal round and flat oval ducts should have a leaking class CL3, and metal rectangular ducts, rectangular
fibrous ducts, and round flexible ducts have a CL6.
 Ducts that are designed to operate at a static pressure exceeding 3 in. WC (750 Pa) shall be leak
tested according to industry accepted test procedures. Representative sections totaling no less
than 25 percent of total installed duct area for the designed pressure class shall be tested.
 The maximum permitted duct leakage , in cfm / 100 ft2 (L/sm2), shall be calculated from
Eq. (17.73).
Shapes and Material of Air Ducts
When a designer chooses the shape (round, rectangular, or flat oval duct) or material (galvanized
sheet, aluminum, fiberglass, or other materials) of an air duct, the choices depend mainly on the space
available, noise, cost, local customs and union agreements, experience, quality, and the requirements
of the project. In many high-rise commercial buildings, factory-fabricated round ducts and sometimes
flat oval ducts with spiral seams are used because they have fewer sound problems, lower air leakage,
and many configurations of wyes and tees available for easier pressure balance. Round ducts also have
the advantage of high breakout transmission loss at low frequencies (see Chap. 19).
For ducts running inside the air conditioned space in industrial applications, metal rectangular
ducts are often chosen for their large cross-sectional areas and convenient fabrication. Round ducts
are often used for more demanding projects. In projects designed for lower cost, adequate duct insulation,
and sound attenuation, fiberglass ducts may sometimes be the optimum selection.
Ductwork Installation
Ductwork installation, workmanship, materials, and methods must be monitored at all stages of the
design and construction process to ensure that they meet the design intent.
Fire Protection
The design of air duct systems must meet the requirements of National Fire Codes NFPA 90A,
Standards for the Installation of Air Conditioning and Ventilating Systems, Warm Air Heating and
Air Conditioning Systems, and Blower and Exhaust Systems as well as local codes. Refer to these
standards for details. The following are some of the requirements:
 The duct material discussed in Sec. 17.3 must be made of class 0 or class 1 material. Also the
duct coverings and linings—including adhesive, insulation, banding, coating, and film covering
the outside surface and material lining the inside surface of the duct—must have a flame spread
rating not over 25 and a smoke development rating not over 50 except for ducts outside buildings.
Supply ducts that are completely encased in a concrete floor slab not less than 2 in. (50 mm) thick
need not meet the class 0 or class 1 requirement.
 Vertical ducts more than two stories high must be constructed of masonry, concrete, or clay tile.
 When ducts pass through the floors of buildings, the vertical openings must be enclosed with partitions
and walls with a fire protection rating of not less than 1 h in buildings less than four stories
high and greater than 2 h in buildings four stories and higher.
 Clearances between the ducts and combustible construction and material must be made as speci-
fied in NFPA 90A.
 The opening through a fire wall by the duct system must be protected by (1) a fire damper closing
automatically within the fire wall and having a fire protection rating of not less than 3 h or (2) fire
doors on the two sides of the fire wall. A service opening must be provided in ducts adjacent to
V?
L
17.50 CHAPTER SEVENTEEN
each fire damper. Many regulatory agencies have very rigid requirements for fire dampers, smoke
dampers, fire /smoke dampers in combination, and smoke venting.
 When a duct penetrates through walls, floors, and partitions, the gap between the ducts and the
walls, floors, and partitions must be filled with noncombustible material to prevent the spread of
flames and smoke.
 Duct systems for transporting products, vapor, or dust in industrial applications must be constructed
entirely of metal or noncombustible material. Longitudinal seams must be continuously
welded, lapped and riveted, or spot-welded on maximum centers of 3 in. (75 mm). Transitions
must be 5 in. (125 mm) long for every 1-in. (25-mm) change in diameter. Rectangular ducts may
be used only when the space is not available and must be made as square as possible.
17.9 AIR DUCT DESIGN PROCEDURE AND DUCT LAYOUT
Design Procedure
Before an air duct system is designed, the supply volume flow rate for each conditioned space,
room, or zone should be calculated, and the locations of the supply outlets and return inlets should
also be settled according to the requirements of space air diffusion (see Chap. 18). For an air duct
system, the supply volume flow rate of cold supply air in summer is usually greater than the warm
volume flow rate needed in winter. If an air duct system conditions the space with cold air supply in
summer, it often also conditions the space with warm air supply in winter.
Computer-aided duct design and sizing programs are widely used for more precise calculation
and optimum sizing of large and more complicated duct systems. Manual air duct design and sizing
are often limited to small and simple duct systems. Computer-aided duct design and sizing
programs are discussed in a later section in this chapter.
The design procedure for an air duct system is as follows:
1. Designer should verify local customs, local codes, local union agreements, and material availability
constraints before proceeding with duct design.
2. The designer proposes a preliminary duct layout to connect the supply outlets and return inlets
with the fan(s) and other system components through the main ducts and branch takeoffs. The
shape of the air duct is selected. Space available under the beam often determines the shape of the
duct and affects the layout in high-rise buildings.
3. The duct layout is divided into consecutive duct sections, which converge and diverge at nodes
or junctions. In a duct layout, a node or junction is represented by a cross-sectional plane perpendicular
to airflow. The volume flow rate of any of the cross sections perpendicular to airflow in a duct
section remains constant. A duct section may contain one or more duct segments (including duct fittings).
A duct system should be divided at a node or junction where the airflow rate changes.
4. The local loss coefficients of the duct fittings along the tentative critical path should be minimized,
especially adjacent to fan inlets and outlets.
5. Duct sizing methods should be selected according to the characteristics of the air duct system.
The maximum design air velocity is determined based on the space available, noise, energy use,
and initial cost of the duct system. Various duct sections along the tentative critical path are sized.
6. The total pressure loss of the tentative critical path as well as the air duct system is calculated.
7. The designer sizes the branch ducts and balances the total pressure at each junction of the
duct system by varying the duct and component sizes, and the configuration of the duct fittings.
8. The supply volume flow rates are adjusted according to the duct heat gain at each supply outlet.
9. The designer resizes the duct sections, recalculates the total pressure loss, and balances the
parallel paths from each junction.
AIR SYSTEMS: AIR DUCT DESIGN 17.51
10. The airborne and breakout sound level from various paths should be checked and the necessary
attenuation added to meet requirements.
Duct System Characteristics
Air duct systems can be classified into the following three categories according to their system
characteristics:
1. Supply duct, return duct, or exhaust duct systems with a certain pressure loss in branch takeoffs
2. Supply duct, return duct, or exhaust duct systems in which supply outlets or return grilles either
are mounted directly on the duct or have only very short connecting duct between the outlet or
inlet and the main duct
3. Industrial exhaust duct systems to transport dust particulates or other particulate products
Duct Layout
When a designer starts to sketch a preliminary duct layout using computer-aided design and drafting
(CADD) or manually, the size of the air duct system (the conditioned space served by the air
duct system) must be decided first. The size of an air duct system must be consistent with the size
of the air system or even the air conditioning system. From the point of view of the air duct system
itself, a smaller and shorter system requires less power consumption by the fan and shows a smaller
duct heat gain or loss. The air duct system is also comparatively easier to balance and to operate.
If the designer uses a more symmetric layout, as shown in Fig. 17.21, a more direct and simpler
form for the critical path can generally be derived. A symmetric layout usually has a smaller main
duct and a shorter design path; it is easier to provide system balance for a symmetric than a nonsymmetric
layout. A more direct and simpler form of critical path usually means a lower total pressure
loss of the duct system.
For variable-air-volume air duct systems, the ends of the main ducts are connected to each other to
form one or more duct loops, as shown by the dotted lines in Fig. 17.21. Duct looping(s) allow some
of the duct sections to be fed from the opposition direction. Balance points exist where the total pressure
of the opposite flowing airstreams is zero. The positions of the balance points often follow the
sun’s position and the induced cooling load at various zones during the operating period. Duct looping
optimizes transporting capacity and results in a smaller main duct than without looping.
17.52 CHAPTER SEVENTEEN
FIGURE 17.21 A typical supply duct system with symmetric duct layout (bold line) and duct loopings (connected by dotted lines) for a
typical floor in a high-rise building.
The designer then compares various alternative layouts and reduces the number of duct fittings,
especially the fittings with higher velocity and high local loss coefficients along the critical path, in
order to achieve a duct system with lower pressure loss.
When duct systems are installed in commercial and public buildings without suspended ceiling, duct
runs should be closely matched with the building structures and give a neat and harmonious appearance.
17.10 DUCT SIZING METHODS
Duct sizing determines the dimensions of each duct section in the air duct system. After the duct
sections have been sized, the total pressure loss of the air duct system can then be calculated, and
the supply, return or relief fan total pressure can be calculated from the total pressure losses of the
supply and return duct systems and the pressure loss in the air-handling unit or packaged unit.
Four duct-sizing methods are currently used:
1. Equal-friction method with maximum velocity
2. Constant-velocity method
3. Static regain method
4. T method
Equal-Friction Method
This method sizes the air duct so that the duct friction loss per unit length pf,u at various duct sections
always remains constant. The final dimensions of sized ducts should be rounded to standard
size. The total pressure loss of the duct system pt, in in. WC (Pa), equals the sum of the frictional
losses and dynamic losses at various duct sections along the critical path:
(17.74)
where L1, L2, . . . , Ln  length of duct sections 1, 2, . . . , n, ft (m)
Le1, Le2, . . . , Len  equivalent length of duct fittings in duct sections 1, 2, . . . , n, ft (m)
If the dynamic loss of a duct fitting is equal to the friction loss of a duct section of length Le, in ft
(m), then
and the equivalent length is
(17.75)
The selection of pf,u is usually based on experience, such as 0.1 in. WC per 100 ft (0.82 Pa / m) for
low-pressure systems. A maximum velocity is often used as the upper limit.
The equal-friction method does not aim at an optimal cost. Dampers are sometimes necessary
for a system balance. Because of its simple calculations, the equal-friction method is still used in
many low-pressure systems in which airborne noise due to higher air velocity is not a problem or
for small duct systems.
Constant-Velocity Method
The constant-velocity method is often used for exhaust systems that convey dust particles in industrial
applications. This method first determines the minimum air velocity at various duct sections
Le 
CoD
f
Co v2
2gc

f(Le / D) v2
2gc
pt  pf,u[(L1  L2      Ln)  (Le1  Le2      Len)]
AIR SYSTEMS: AIR DUCT DESIGN 17.53
according to the requirement to float the particles, either by calculation or by experience. On the basis
of the determined air velocity, the cross-sectional area and, therefore, the dimension of the duct
can be estimated and then rounded to a standard size. The total pressure loss of the duct system pt,
in in. WC (Pa), along the critical path can be calculated as
(17.76)
where v1, v2, . . . , vn  mean air velocity at duct sections 1, 2, . . . , n, respectively, fpm (m/ s)
C1, C2, . . . , Cn  local loss coefficients at duct sections 1, 2, . . . , n, respectively
K  5.35 	 10	5 for I-P unit (1 for SI unit)
Static Regain Method
This method sizes the air duct so that the increase of static pressure (static regain) due to the reduction
of air velocity in the supply main duct after each branch takeoff nearly offsets the pressure loss
of the succeeding duct section along the main duct. As a consequence, the static pressure at the
common end of the diverging tee or wye of the sized duct section remains approximately the same
as that of the preceding section.
A rectangular duct section 1-2 between the cross-sectional planes 1 and 2 is illustrated in
Fig. 17.22. The size of this duct section is to be determined. Let v1 and v2 be the mean velocities at
      fnLn
Dn
 Cnvn 2
pt 
K
2gc
	 f1L1
D1
 C1v1 2   f2L2
D2
 C2v2 2
17.54 CHAPTER SEVENTEEN
FIGURE 17.22 Pressure characteristics of a main duct section.
planes 1 and 2, and the volume flow rates, and A1 and A2 the cross-sectional areas. The total
pressure loss in duct section 1-2 consists of the duct friction loss pf1-2 and the dynamic loss of the
main airstream flowing through the diverging tee p1c,s. The relationship between the total pressure
at planes 1 and 2 can be expressed as
(17.77)
because pt  ps  pv. Ignore the difference between air densities 1 and 2. Let pf1-2  pf,uL1-2.
Here L1-2 represents the length of the duct section 1-2, in ft (m). If the static pressures at planes 1
and 2 are equal, that is, ps1  ps2, then
(17.78)
If v is expressed in fpm and pf,u in in. WC per 100 ft,  0.075 lb / ft3, and gc  32.2
lbmft / lbf  s2, the mean air velocity of the sized duct section is
(17.79)
In SI units,
(17.79a)
with v1 and v2 both in m / s, pf,u in Pa /m, L1-2 in m, and in kg /m3.
For any duct section between cross-sectional planes n 	 1 and n, if the total local loss coeffi-
cient of the duct fittings is Cn, excluding the local loss coefficient C(n	1)c,s, the mean air velocity of
the sized duct section is
(17.80)
Because vn	1, n, Ln, and Cn are known values, by using iteration methods, vn can be determined.
The dimension of the duct section and its rounded standard size can also be determined.
The static regain method can be applied only to supply duct systems. It tends to produce a more
even static pressure at the common end of each diverging tee or wye leading to the corresponding
branch takeoff, which is helpful to the system balance. It does not consider cost optimization. The
main duct sections remote from the fan discharge often have larger dimensions than those in the
equal-friction method. Sound level and space required should be checked against determined air
velocity and dimension.
When one is using the static regain method to size air ducts, it is not recommended to allow only
part of static regain to be used in the calculation.
T Method
The T method, first introduced by Tsal et al. (1988), is an optimizing procedure to size air ducts by
minimizing their life-cycle cost. It is based on the tree-staging idea and is therefore called the T
method. The goal of this method is to optimize the ratio between the velocities in all sections of the
duct system. The T method consists of the following procedures:
1. System condensing—condensing various duct sections of a duct system into a single imaginary
duct section having the same hydraulic characteristics and installation costs as the duct system
vn  [(1 	 C(n	1) c,s)vn	1
2 	 1.6 	 105 pf,uLn]
1  Cn 
0.5
v2  	 (1 	 C1 c,s)v1 2 	 2pf,uL1-2

0.5
v2  [(1 	 C1 c,s) v1 2 	 1.6 	 105pf,uL1-2]0.5
1(v1 2 	 v2 2)
2gc
 pf,uL1-2 
C1 c,s 1v1 2
2gc
pt1  pt2  pf1-2  p1 c,s
V?
2 V?1
AIR SYSTEMS: AIR DUCT DESIGN 17.55
2. Fan selection—selecting a fan that provides the optimum system pressure loss
3. System expansion—expanding the imaginary duct section into the original duct system before
condensing with the optimum distribution of total pressure loss between various duct
sections
During optimization, local loss coefficients are considered constant at various stages of iteration.
For details, refer to Tsal et al. (1988).
The T method can be used for sizing duct systems with certain total pressure losses in the
branch ducts. However, the local loss coefficients are actually varied at various stages of the iteration
and should be taken into consideration during optimization.
17.11 DUCT SYSTEMS WITH CERTAIN PRESSURE LOSSES
IN BRANCH TAKEOFFS
Design Characteristics
Supply, return, or exhaust duct systems with certain pressure losses in branch takeoffs have the following
characteristics:
 Duct is sized based on the optimization of the life-cycle cost of various duct sections of the duct
system as well as the space available in the building.
 System balancing is achieved mainly through pressure balancing of various duct paths by changing
of duct sizes and the use of various configurations of duct fittings and terminals instead of
dampers or other devices.
 Sound level will be checked and analyzed. Excess pressure at each inlet of VAV box at design
conditions is to be avoided. Sound attenuation arrangements are added if necessary.
 Local loss coefficients of the duct fittings and equipment along the critical path are minimized. It
may be beneficial to use the surplus pressure available in the branch takeoff to produce a higher
branch duct velocity and a smaller straight-through local loss coefficient Cs,c.
 Supply volume flow rates are adjusted according to the duct heat gain. For VAV systems, diversity
factors are used to determine the volume flow rate of various duct sections along the critical path
so that the volume flow rate nearly matches the block load at the fan discharge.
Cost Optimization
For any duct section in an air duct system, the total life-cycle cost Cto, as shown in Fig. 17.23, in
dollars, can be calculated as
(17.81)
In Eq. (17.81), CRF indicates the capital recovery factor and can be calculated as follows:
(17.82)
where i  interest rate and n  number of years under consideration.
The first-year energy cost Ce, in dollars, can be calculated as the product of electric energy
consumed at the fan and the unit energy cost Er, in $ /kWh, times the annual operating hours tan, in
CRF 
i(1  i)n
(1  i)n 	 1
Cto  Ce
1
CRF
 Cdi
17.56 CHAPTER SEVENTEEN
h, or in I-P units
(17.83)
For SI units,
(17.84)
where pt  total pressure loss of duct section, in. WC (Pa)
 volume flow rate of duct section, cfm (m3 / s)
f, m  total efficiency of fan and efficiency of motor
Let z1  Ke(1 / CRF). Then we can write
(17.85)
Installation cost for round duct Cdi, in dollars, can be calculated as
(17.86)
where D  diameter of round duct, ft (m)
L  length of duct section, ft (m)
Ciu  unit cost of duct installation, $/ ft2 ($ /m2)
Cdi 
DLCiu
1
CRF
Ce  z 1 pt
V?
Ke 
V?
Ertan
1000 f m
 pt 1.175 	 10	4V?Ertan
f m
  Ke pt
Ce  pt
0.746 V?Ertan
6350 f m
AIR SYSTEMS: AIR DUCT DESIGN 17.57
FIGURE 17.23 Cost analysis for a duct system.
The surface area of rectangular duct Arec  2(W  H)L. Here W is the width of the duct, and H is
the height of the duct, both in ft (m). Then for rectangular duct
(17.87)
where Rrec  ratio of surface area of rectangular duct to surface area of round duct. For any duct
section, the total pressure loss pt, in in. WC (Pa), is
(17.88)
where C  sum of local loss coefficients in duct section, based on air velocity of sized duct
section
v  mean air velocity in sized duct section, fpm (m/ s)
For frictional loss D  f (pt 	0.2), and for dynamic loss D  f (pt 	0.25). However, for simplicity, let
us take D  f (pt 	0.22) and consider a constant. Then
(17.89)
Because gc  32.2 lbmft / lbf  s2,
(17.90)
Let z2  0.1097( /gc)0.22Ciu. By substituting into Eq. (17.86), for a round duct, we have
(17.91)
For a rectangular duct
(17.92)
Seeking a minimum cost by taking derivatives of Eq. (17.81) with respect to pt, making it equal to
zero, we have
Because gc  32.2 lbmft / lbf  s2, then for a round duct
(17.93)
For a rectangular duct,
(17.94) pt  108Rrec	 LCiu f m
(1 / CRF)Ertan 0.82	 fL
?D
 C ?D0.18 1
V?
0.46 
pt  108	 LCiu f m
(1 / CRF)Ertan 0.82	 fL
?D
 C?D0.18 1
V?
0.46 
dCto
dpt
 z 1 	 0.22z 2K pt
	1.22  0
Cdi  Rrecz 2K pt
	0.22
Cdi 
DLCiu  z 2K pt
	0.22
where K  ( fL/?D  C?D)0.22V? 0.44L.
D  0.0511	 fL
?D
 C?D0.22
V?
0.44
pt
	0.22
 0.1097

gc 0.22
K
pt
	0.22
L
D  0.1097 	 fL / ?D  C?D) 
gc 0.22
V?
0.44
pt
	0.22
( fL / ?D)  C?D
pt 
5.35 	 10	5( fL/D  C) v2
2gc
Cdi  2(W  H )LCiu 
2(W  H )
D
DLCiu  Rrec DLCiu
17.58 CHAPTER SEVENTEEN
When the total pressure loss of the duct section that has the minimum cost is calculated from
Eq. (17.93) or (17.94), the diameter or the circular equivalent of the duct section of minimum cost
can then be calculated by Eq. (17.90).
Condensing Two Duct Sections
According to the T method, if two duct sections 1 and 2 connected in series are condensed into an
imaginary section 1-2, as shown in Fig. 17.24a, then the volume flow rate at each duct section must
be equal, i.e.,
(17.95)
Here subscripts 1 and 2 and 1-2 represent the duct sections 1 and 2 and the imaginary duct section
1-2, respectively.
The total pressure loss of the condensed duct section must equal the sum of the total pressure
losses of duct sections 1 and 2, or
(17.96)
and the installation cost of the condensed section is
(17.97)
When two duct sections 1 and 2 connected in parallel are condensed into an imaginary duct section
1-2, as shown in Fig. 17.24b, the following relationships hold:
(17.98)
Cdi1-2  Cdi1  Cdi2
pt1-2  pt1  pt2
V?
1-2  V?1  V?2
 z 2(K1 pt1 	 0.22  K2 pt2 	0.22)  z 2K1-2 pt1-2
	0.22
Cdi1-2  Cdi1  Cdi2
pt1  pt2  pt1-2
V?
1  V?2  V?1-2
AIR SYSTEMS: AIR DUCT DESIGN 17.59
FIGURE 17.24 Two duct sections 1 and 2 condensed into an imaginary
duct section 1-2: (a) two duct sections 1 and 2 connected in series;
(b) two duct sections 1 and 2 connected in parallel.
If the whole duct system is condensed into one imaginary duct section, its minimum cost can be
found by taking derivatives with respect to pt of the imaginary duct section. In using such a procedure,
the minimum cost includes both the main and branch ducts. However, if the local loss coefficients
of the terminals and converging and diverging wyes are assumed constant, the benefits
of using smaller terminals and different configuration of wyes to balance the branch duct paths
are lost, as demonstrated by Dean et al. (1985). A more flexible and often economical alternative
is that only the sizes of duct sections along the critical path of a duct system are determined according
to the cost optimization procedure—that is, Eqs. (17.93), (17.94), and (17.90)—and
rounded to standard size. The sizes of the duct sections of other duct paths, such as branch takeoffs,
should be determined according to the difference between the total pressure at the junction
and the space pressure for system or pressure balance using optimum terminals, wyes, and fittings.
Local Loss Coefficients for Diverging Tees and Wyes
For a supply duct system, there are often more diverging tees or wyes than elbows along the design
path. Selecting diverging tees or wyes with smaller local loss coefficients Cc,s has a definite influence
on the design of a duct system with minimizing total pressure loss of critical path.
Table 17.11 shows Cc,s values for diverging wyes and tees for round ducts and diverging
wyes for rectangular ducts. The following are recommendations for proper selection of wyes
and tees:
 Select a diverging wye instead of a diverging tee except when there is surplus total pressure at the
branch takeoffs. A 30° angle between the branch and the main duct has a smaller Cc,s value than a
45° or a 60° angle.
17.60 CHAPTER SEVENTEEN
TABLE 17.11 Local Loss Coefficients for Elbows, Diverging Tees, and Diverging Wyes
(1) Elbow; 3, 4, and 5 pieces, round
Coefficients for 90° elbows Co
No. of Rc/D
pieces 0.75 1.0 1.5 2.0
5 0.46 0.33 0.24 0.19
4 0.50 0.37 0.27 0.24
3 0.54 0.42 0.34 0.33
Angle correction factors K (Idelchik, 1986; diagram 6-1)
 0 20 30 45 60 75 90 110 130 150 180
K 0 0.31 0.45 0.60 0.78 0.90 1.00 1.13 1.20 1.28 1.40
17.61
TABLE 17.11 Local Loss Coefficients for Elbows, Diverging Tees, and Diverging Wyes
(Continued )
(2) Elbow, smooth radius with splitter vanes, rectangular
Coefficients for elbows with two splitter vanes Co
H/W
R/W Rc/W CR 0.25 0.5 1.0 1.5 2.0 3.0 4.0 5.0 6.0 7.0 8.0
0.05 0.55 0.362 0.26 0.20 0.22 0.25 0.28 0.33 0.37 0.41 0.45 0.48 0.51
0.10 0.60 0.450 0.17 0.13 0.11 0.12 0.13 0.15 0.16 0.17 0.19 0.20 0.21
0.15 0.65 0.507 0.12 0.09 0.08 0.08 0.08 0.09 0.10 0.10 0.11 0.11 0.11
0.20 0.70 0.550 0.09 0.07 0.06 0.05 0.06 0.06 0.06 0.06 0.07 0.07 0.07
0.25 0.75 0.585 0.08 0.05 0.04 0.04 0.04 0.04 0.05 0.05 0.05 0.05 0.05
0.30 0.80 0.613 0.06 0.04 0.03 0.03 0.03 0.03 0.03 0.03 0.04 0.04 0.04
(3) Wye, 45°, round, with 60° elbow, branch 90° to main
Branch
b / c 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Cc,b 1.0 0.88 0.77 0.68 0.65 0.69 0.73 0.88 1.14 1.54 2.2
Main
s / c 0 0.1 0.2 0.3 0.4 0.5 0.6 0.8 1.0
Cc,s 0.40 0.32 0.26 0.20 0.14 0.10 0.06 0.02 0
(4) Tee, diverging, round, with 45° elbow, branch 90° to main
 The smaller the difference in air velocity between the common end and the straight-through end,
the lower the Cc,s value.
 At a greater ratio of branch duct velocity to main duct velocity vb /vc, especially when vb /vc 1.2,
the Cc,s value for the diverging wye of rectangular duct with smooth elbow takeoff is negative.
17.62
TABLE 17.11 Local Loss Coefficients for Elbows, Diverging Tees, and Diverging Wyes
(Continued )
Branch
Vh /Vc 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Cc,b 1.0 1.32 1.51 1.60 1.65 1.74 1.87 2.0 2.2 2.5 2.7
For main local loss coefficient Cc,s, see values in (3)
(5) Tee, diverging, rectangular (Idelchik, 1986; diagram 7-21)
Branch, Cc,b
Ab /As Ab /Ac 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0.25 0.25 0.55 0.50 0.60 0.85 1.2 1.8 3.1 4.4 6.0
0.33 0.25 0.35 0.35 0.50 0.80 1.3 2.0 2.8 3.8 5.0
0.5 0.5 0.62 0.48 0.40 0.40 0.48 0.60 0.78 1.1 1.5
0.67 0.5 0.52 0.40 0.32 0.30 0.34 0.44 0.62 0.92 1.4
1.0 0.5 0.44 0.38 0.38 0.41 0.52 0.68 0.92 1.2 1.6
1.0 1.0 0.67 0.55 0.46 0.37 0.32 0.29 0.29 0.30 0.37
1.33 1.0 0.70 0.60 0.51 0.42 0.34 0.28 0.26 0.26 0.29
2.0 1.0 0.60 0.52 0.43 0.33 0.24 0.17 0.15 0.17 0.21
Main, Cc,s
Ab /As Ab /Ac 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
0.25 0.25 	0.01 	 0.03 	 0.01 0.05 0.13 0.21 0.29 0.38 0.46
0.33 0.25 0.08 0 	 0.02 	 0.01 0.02 0.08 0.16 0.24 0.34
0.5 0.5 	 0.03 	 0.06 	 0.05 0 0.06 0.12 0.19 0.27 0.35
0.67 0.5 0.04 	 0.02 	 0.04 	 0.03 	 0.01 0.04 0.12 0.23 0.37
1.0 0.5 0.72 0.48 0.28 0.13 0.05 0.04 0.09 0.18 0.30
1.0 1.0 	 0.02 	 0.04 	 0.04 	 0.01 0.06 0.13 0.22 0.30 0.38
1.33 1.0 0.10 0 0.01 	 0.03 	 0.01 0.03 0.10 0.20 0.30
2.0 1.0 0.62 0.38 0.23 0.23 0.08 0.05 0.06 0.10 0.20
Source: ASHRAE Handbook 1989, Fundamentals. Reprinted with permission. For details, refer to ASHRAE
Handbook.
V?
b / V?c
V?
b / V?c
Return or Exhaust Duct Systems
Behls (1978) compared four designs of a bunker ventilation duct system with 10 risers connected to
a main exhaust duct. The results of these four designs are as follows:
Using a greater branch duct velocity than the main duct, and varying the sizes of the risers, select
proper duct fittings to decrease system total pressure loss and system unbalance.
Example 17.2. For a round supply duct system made of galvanized steel with spiral seams, as
shown in Fig. 17.20, its operational and constructional characteristics are as follows:
System total
pressure loss, Riser unbalance,
in. WC (Pa) in. WC (Pa)
Fixed-diameter riser, tee / diffuser 	7.39 (	1837) 6.55 (1628)
Variable-diameter riser, tee / diffuser 	5.89 (	1464) 1.77 (440)
Fixed-diameter riser, conical tee 	5.29 (	1315) 4.06 (1009)
Variable-diameter riser conical tee 	4.03 (	1002) 1.62 (403)
AIR SYSTEMS: AIR DUCT DESIGN 17.63
Supply air temperature
Kinematic viscosity of air
Density of supply air
Absolute roughness
Local loss coefficients in section FO-1
Fan total efficiency, average
Motor efficiency
Electrical energy cost
Installation cost of duct
CRF
Annual operating hours
60°F (15.6°C)
1.59 	 10	4 ft2 / s (1.46 	 10	5 m2 / s)
0.075 lb / ft3 (1.20 kg /m3)
0.0003 ft (0.09 mm)
0.5
0.7
0.8
$0.08 /kWh
$3.5 /ft2
0.10
3000
The total supply volume flow rate at the fan discharge is 3000 cfm (1.415 m3 / s), and the adjusted
volume flow rates because of duct heat gain for each of the branch takeoffs are illustrated in
Fig. 17.20. First, size this supply duct system. Then if each branch takeoff is connected to a VAV
box that needs a total pressure loss of 0.75 in. WC (186 Pa) at its inlet, calculate the total pressure
loss required for this supply duct system excluding the VAV box and downstream flexible duct and
diffuser.
Solution:
1. For the first iteration, start with the round duct section 2-5, which may be the last section of
the design critical path. Assume a diameter of 12 in., or 1 ft, and an air velocity of 1200 fpm, or 20
ft / s. Assume also that the local loss coefficient of the straight-through stream of the diverging tee
Cc,s  0.15, and for the elbow Co  0.22. The Reynolds number based on diameter D is
ReD 
vD
n

20 	 1
1.59 	 10	4  125,786
From Eq. (17.42), the friction factor is
From Eq. (17.93), the optimum pressure loss for the round duct section 2-5 is
And from Eq. (17.90), the diameter of the duct section 2-5 when pt  0.13 in. WC is
2. Duct section 1-2 is one of the sections of the critical path. For this section, if we assume that
Cc,s  0.05 and the diameter is 1.2 ft, then the calculated friction factor f  0.0155 and the optimal
total pressure loss is
and the sized diameter is
3. Duct section FO-1 is one of the sections of the critical path. For this section, C  0.5. Let
us assume that the diameter is 1.67 ft and the calculated f  0.0147. Then
		 0.147 	 25
?1.67
 0.5?1.670.0750.18 1
30000.46  0.0692 in. WG
ptFO-1  108 25 	 3.5 	 0.7 	 0.8
10 	 0.08 	 3000 0.82
 1.229 ft or 14.75 in. (375 mm)
D1-2  0.051	 0.0155 	 20
?1.2
 0.05?1.20.0750.22
20250.440.0575	0.22
		 0.155 	 20
?1.2
 0.05?1.20.0750.18 1
20250.46   0.0575 in. WC
pt1-2  108 20 	 3.5 	 0.7 	 0.8
10 	 0.08 	 3000 0.82
 0.927 ft or 11.12 in. (282 mm)
 0.0511	 0.0166 	 30
?1.0
 0.37?1.00.0750.22
10250.44
1
0.130.22
D2-5  0.0511	 fL
?D
 C?D0.22
V?
0.44
pt
	0.22
	 	 0.0166 	 30
?1.0
 0.37?10.0750.18 1
10250.46   0.13 in. WC
 108 30 	 3.5 	 0.7 	 0.8
10 	 0.08 	 3000 0.82
pt2-5  108	 LCiu f m
(1/CRF)Ertan 0.82	 fL
?D
 C?D0.18 1
V?
0.46 

0.25
{log [0.0003 / (3.7 	 1)  5.74 / (0.9 	 125,786)]}2  0.0166
f 
0.25
{log [ / (3.7D)  5.74 / (0.9ReD)]}2
17.64 CHAPTER SEVENTEEN
and the sized diameter
4. Duct section 2-4 is another leg from junction 2. It must have the same total pressure loss as
duct section 2-5. Assume that Cc,b  1 and the diameter is 11 in., or 0.917 ft. Then
 0.978 ft or 11.78 in. (299 mm)
5. Duct section 1-3 is another leg from junction 1. It must have the sum of the total pressure loss
of duct sections 1-2 and 2-5, that is, pt1-3  0.0575  0.13  0.1875 in. WC. Assume that Cc,b 
0.8 and the diameter is also 11 in. Then
 0.858 ft or 10.3 in. (262 mm)
6. The results of the first iteration are listed in Table 17.12. These results are rounded to standard
sizes and provide the information for the selection of the diverging tee and wye and the determination
of the local loss coefficient.
7. For branch takeoffs 2-4 and 1-3, select proper diverging tees and wyes and, therefore, the Cc,b
values based on the air velocity of the sized duct sections. Vary the size of the duct sections if necessary
so that their pt values are approximately equal to the pt of section 2-5 for branch duct 2-4,
and the pt of the sum of sections 2-5 and 1-2 for branch 1-3. Use Eq. (17.88) to calculate the total
pressure loss of sections 2-4 and 1-3 according to the rounded diameters.
8. After the diverging tee or wye is selected, recalculate the optimum total pressure losses and
diameters for duct sections 2-5, 1-2, and FO-1 from Eqs. (17.93) and (17.90).
9. After two iterations, the final sizes of the duct sections, as listed in Table 17.12, are the following:
Section 2-5, 11 in.; wye, 45°, diverging, round, 30° elbow
Section 1-2, 14 in.; tee, diverging, round, 90° elbow
Section FO-1, 20 in.
Section 2-4, 11 in.
Section 1-3, 12 in.
For duct section 1-2, at vs /vc  1.2, Cc,s is about 0.07 according to Idelchik (1986), diagram 7-17.
D1-3  0.051	 0.0166 	 12
?0.917
 0.8?0.9170.0750.22
9750.440.1875	0.22
D2-4  0.051	 0.0166 	 12
?0.917
 1?0.9170.0750.22
10000.440.13	0.22
 1.734 ft or 20.81 in. (529 mm)
DFO-1  0.051	 0.0147 	 25
?1.67   0.5?1.670.0750.22
30000.440.0692	0.22
AIR SYSTEMS: AIR DUCT DESIGN 17.65
TABLE 17.12 Results of Computations of Duct Sizes and Total Pressure Loss in Example 17.2
First iteration Final results
Volume
Air
Friction
Duct flow V,
Diameter
Rounded
velocity
factor
Velocity ratio
pt. in.
section cfm ft in. in. fpm fps f vs /vc vb /vc Cc,s Cc,b C Diameter WC
2-5 1025 0.927 11.12 11 1553 25.88 0.0166 0.94 0.011 0.22 11 0.116
1-2 2025 1.229 14.75 15 1894 31.97 0.0155 1.20 0.07 14 0.0645
FO-1 3000 1.734 20.81 20 1374 22.92 0.0147 0.5 20 0.0847
2-4 1000 0.978 11.78 11 1515 25.25 0.80 0.65 11
1-3 975 0.858 10.3 11 1241 20.69 0.90 1.70 12
10. Because the local loss coefficient Cc,b for the branch stream from the 45° diverging wye at
junction 2 can be reduced to 0.45 if a conical branch is used, the critical path consists of duct sections
FO-1, 1-2, and 2-5. The total pressure loss of this supply duct system excluding the VAV box
and downstream flexible duct and diffuser, from Table 17.12, is
pt  0.116  0.0645  0.0847  0.2652 in. WC (66 Pa)
17.12 DUCT SYSTEMS WITH NEGLIGIBLE PRESSURE LOSS
AT BRANCH DUCTS
Supply Duct Systems
When supply outlets are directly mounted on the main duct without branch takeoffs, or the connecting
duct between the supply outlet and the main duct is about 2 ft or less, the total pressure loss of
the branch duct, excluding the supply outlet, is often very small or negligible. In such circumstances,
system balancing of the supply duct system depends mainly on the sizes of the successive
main duct sections. If volume dampers are installed just before the outlet or in the connecting duct,
the damper modulation will also vary the space diffusion airflow pattern.
Pressure Characteristics of Airflow in Supply Ducts
The rectangular supply duct with transversal slots shown in Fig. 17.25 is an example of a supply
duct system with a supply outlet directly mounted on the main duct. The volume flow per ft2 (m2) of
floor area of this type of duct system may sometimes exceed 8 cfm / ft2 (40.6 L / s m2). This type of
supply duct system is often used in many industrial applications.
Consider two planes n and n  1, with a distance of 1 unit length, say, 1 ft (0.305 m), between
them. If the space air is at atmospheric pressure, the total pressure loss of the supply air that flows
from plane n, turns a 90°, and discharges through the slots is
(17.99)
where pc,bn, pton  dynamic loss of branch stream when it makes a 90° turn and when it discharges
from slots, lb / ft2 or in. WC (Pa)
Cc,bn  local loss coefficient of diverging branch stream with reference to velocity at
plane n
Co  local loss coefficient of transversal slots with reference to velocity at slot
vn, von  air velocity at plane n and at slot from plane n, ft / s (m/s)
Similarly, at plane n  1 the total pressure loss of the branch stream when it discharges from the
slots is
(17.100)
The relationship of the total pressure between planes n and n  1 along the airflow is given as
(17.101)
In Eq. (17.101), fn, Ln, Dn, and Cs,cn indicate the friction factor, length, circular equivalent, and Cc,s
value between plane n and n  1, respectively. Substituting Eqs. (17.99) and (17.100) into
ptn  pt(n1) 
( fnLn/Dn  Cc,sn) vn 2
2gc
pt(n1) 
(Cc,b(n1)v(n1)
2  Covo(n1)
2 ) 
2gc
ptn  ptn  pc,bn  pton 
(Cc,bnvn 2  Covon
2 ) 
2gc
17.66 CHAPTER SEVENTEEN
Eq. (17.101), Wang et al. (1984) recommended the following formula to calculate vn1, in ft / s
(m/ s), based on the balanced total pressure before the transversal slots:
(17.102)
In Eq. (17.102), von and vo(n1), in ft / s (m / s), are the supply air velocities at the slots. For a cold air
supply, usually vo(n1) von is desirable because of the effect of the duct heat gain.
For a free area ratio of slot area to duct wall area of 0.5, Co for a perforated plate including discharged
velocity pressure can be taken as 1.5. The vn, Ln, and Dn are known values during the calculation
of vn1. Local loss coefficients Cc,bn and Cc,sn can be determined from the experimental curves
shown in Figs. 17.26 and 17.27. The coefficient Cc,b(n1) can be assumed to have a value similar to
Cc,bn, and f can be calculated from Eq. (17.42). The mean air velocity in the duct at plane n  1, denoted
by vn1, can then be calculated from Eq. (17.102).
In Figs. 17.26 and 17.27, on indicates the volume flow rate of supply air discharged from the
slots per ft length of the duct, in cfm / ft (L / s m); n represents the volume flow rate of the supply
air inside the duct at plane n, in cfm (m3 / s).
Sizing of this kind of supply duct system starts from the duct section immediately after the
fan discharge. This section can be sized from Eqs. (17.90) and (17.93) based on life-cycle cost
V?
V? v(n

1) 

	
(
von
2 	 vo(n1)
2 )Co  Cc,bn 	 Cc,sn 	
fnLn
Dn
vn 2  1
Cc,b(n1) 
0.5
AIR SYSTEMS: AIR DUCT DESIGN 17.67
FIGURE 17.25 Rectangular supply duct with transversal slots.
optimization. The designer should check for space available and noise, if sound control is required.
The successive duct sections can then be sized by calculating from Eq. (17.102).
If supply duct systems with negligible pressure loss at branch ducts are installed with outlets
whose Cc,bn and Cc,sn are not known, the static regain method is recommended for sizing the supply
main duct for a large duct system and the equal-friction method for a small system.
Example 17.3. A galvanized steel rectangular duct section with transversal slots has the following
construction and operational characteristics at plane n
17.68 CHAPTER SEVENTEEN
FIGURE 17.26 Local loss coefficient Cc,s versus volume flow ratio . (Source: ASHRAE Transactions,
1984, Part II A. Reprinted with permission.)
V?on / V?n
Dimension:
Absolute roughness
Supply air temperature
Volume flow rate n V?

4-ft width 	 3-ft height (1219 mm 	 914 mm)
0.0005 ft (0.15 mm)
65°F (18.3°C)
23,760 cfm (11.21 m3 / s)
If the height of the rectangular duct remains the same, size the dimension of this duct at plane
n  1, 10 ft (3.05 m) from plane n.
Solution
1. From the information given, the air velocity at plane n is
vn 
23,760
60 	 4 	 3
 33 ft / s (1980 fpm)
AIR SYSTEMS: AIR DUCT DESIGN 17.69
FIGURE 17.27 Local loss coefficient Cc,b versus volume flow ratio . (Source: ASHRAE Transactions,
1984, Part II A. Reprinted with permission.)
V?on / V?n
Volume flow discharged from slots related to
plane n
Discharged velocity at slots related to plane n
Required discharged velocity at plane n  1
270 cfm / ft (417.9 L / s m)
7.5 ft / s or 450 fpm (2.29 m / s)
7.6 ft / s or 456 fpm (2.32 m / s)
17.70 CHAPTER SEVENTEEN
And from Eq. (17.48), the circular equivalent is
The Reynolds number of the supply air
The friction factor
2. From the information given, the volume flow ratio is
and the velocity ratio is
From Figs. 17.26 and 17.27, Cc,sn  	 0.0025 per ft and Cc,bn  1.22.
3. Assume that Cc,b(n1)  1.23. Then from Eq. (17.102),
4. The required volume flow rate from the slots related to plane n  1 is
Then the volume flow rate of air at plane n  1 is
The duct area required at plane n  1 is
An1 
21,024
60 	 29.6
 11.83 ft2 (1.00 m2)
V?
o(n1)  23,760 	 273.6 	 10  21,024 cfm
V?
o(n1) 
9270 	 7.6
7.5
 273.6 cfm / ft
 29.6 ft / s (9.02 m / s)
 (7.52 	 7.62)1.5  	 1.22 	 (0.0139 	 10 / 2.9) 	 0.0025 	 10)]332
1.23 0.5
v(n1)  	(von
2 	 vo(n1)
2 )Co  Cc,bn 	 Cc,sn 	
fL
D vn 2
1
Cc,b(n1) 
0.5
von
vn

7.5
33
 0.227
V?
on
V?
n

270
23,760
 0.0113

0.25
{log [0.0005 / (3.7 	 2.9)  5.74 / (0.9 	 596,700)]}2  0.0139
f 
0.25
{log [ / (3.7D)  5.74 / (0.9 ReD)]}2
ReD 
33 	 2.9
1.62 	 10	4  596,700
De 
1.3(ab)0.625
(a  b)0.25 
1.3(3 	 4)0.625
(3  4)0.25  2.9 ft
The width of the duct is
Return or Exhaust Duct Systems
Return duct systems with very short connecting ducts between the return inlet and the main duct
are often used in industrial applications. They are often used in large workshops and need nearly
equal return volume flow rate for each branch intake. To provide a better system balance, it is necessary
to reduce the total pressure difference between the branch inlets at the two ends of the
main duct.
For a conditioned space where noise is not a major problem, one effective means of reducing the
total pressure loss of a long return main duct is to have a higher branch duct velocity than the velocity
in the main duct. Because a total pressure difference between the branch inlets and the main
duct at two ends is inevitable, the area of the return grille can be varied along the main duct to provide
a more even return volume flow rate. Unlike the supply outlets, the variation of the area of the
return grille has a minor effect on the space air diffusion.
A return duct system with short connecting ducts is shown in Fig. 17.28 for a textile mill.
Twelve floor grilles are connected to the return main duct at the lower floor. The connecting duct
has a floor to main return duct difference of 2.6 ft (0.79 m). The sizes of the branch ducts vary from
16 in. 	 16 in. (406 mm 	 406 mm) at the remote end to 12 in. 	 8 in. (305 mm 	 203 mm) near
the fan end. The main duct varies from 16 in. 	 32 in. (406 mm 	 813 mm) to 60 in. 	 32 in.
(1524 mm 	 813 mm). In Fig. 17.28, the variations in total pressure pt and static pressure ps along
the main duct are shown by the upper and middle curves, in in. WG (Pa); and the velocity in main
return duct v, in fpm (m / s), is shown by the lower curve. The velocity in the branches varies from
700 fpm (3.56 m / s) at the remote end to 2000 fpm (10.16 m / s) at the fan end. Note that pt at grille
6 is 	0.111 in. WG (	27.6 Pa), 	0.105 in. WG (	26.1 Pa) at grille 5, and 	0.114 in. WG
W 
10.78
3
 3.59 ft (1.095 m)
AIR SYSTEMS: AIR DUCT DESIGN 17.71
FIGURE 17.28 A return duct system with short connecting ducts for a textile mill.
(	28.3 Pa) at grille 2. Such a satisfactory pressure balance along the return duct is mainly due to
the higher branch velocity in the grilles near the fan end.
17.13 REQUIREMENTS OF EXHAUST DUCT SYSTEMS FOR A
MINIMUM VELOCITY
To transport dust particles or particulate products contained in the air, exhaust duct systems require a
minimum velocity in all duct sections of the system. These systems are used in many industrial applications
and usually have air velocity ranging between 2400 and 4000 fpm (12.2 and 20.3 m/ s). In
addition to the variation of the sizes of the branches’ ducts, it is essential to select the proper configuration
of duct fittings to provide a better system balance and to reduce the total pressure loss of the
system (critical path). The following are recommendations for exhaust duct system designs:
 Round ducts usually produce smaller pressure losses; they are more rigid in construction.
 Air velocity inside the duct must not exceed the minimum velocity too much, in order not to
waste energy.
 Well-sealed joints and seams are important for reducing air leakage at higher pressure differentials.
17.14 COMPUTER-AIDED DUCT DESIGN AND DRAFTING
According to Amistadi (1993) and “Scientific Computing” (1998), the following computer-aided
duct design and drafting (CADD) programs are widely used in air duct design: E20-II Duct Design,
Carrier Corp.; Varitrane Duct Designer, The Trane Co.; and Softdesk HVAC. Carrier’s E20-II requires
a disk space of installation of 4 MBytes and a RAM of 1 MByte, Trane’s Varitrane needs a
disk space of 13 MBytes and a RAM of 12 MBytes, and Softdesk HVAC needs a disk space of
7 MBytes and a RAM of 8 MBytes. All these three programs will run in Windows 95.
An effective and intelligent computer-aided duct design and drafting program provides the
following functions:
Drafting
The computer program provides full, three-dimensional (3D) capabilities and draws round, rectangular,
and flat oval ducts and fittings in any cross-sectional plane. The program offers two types of
single-line layouts: (1) duct and fitting, which automatically produce elbows, transitions, and reducers
from plan or multistory views, and (2) isometric, which allows the designer to produce a 3D
drawing in an easy way. A single-line layout can be converted to double-line or 3D ducts and
fittings layouts, or the designer can construct the 2D double-line or 3D ducts and fittings drawings
directly. Duct layout can be produced in 2D or 3D in different layers simultaneously.
Schedules and Layering
Schedules of a computer-aided duct design program include diffusers, ducts, fittings, and related
equipment, such as terminals and air-handling units or packaged units. Duct schedules include the
tag number, quantity, size, type, length, gauge, and bill of material. Fitting schedules include the tag
number, quantity, type, name, gauge, fitting junction sizes, and cfm (L / s). Values can be totaled and
stored for each duct or fitting type. Tag editing capacity includes: creating, moving, and renumbering.
The computer program prefers to have a nested-layer hierachy of five levels deep including
building information, plan types, layer name and level (elevation), and construction modifier. Build-
17.72 CHAPTER SEVENTEEN
ing information layers are grouped by supply, return, and exhaust ducts and subdivided into ducts,
diffusers, and fittings. For each of these duct components, layers can be single-line, 2D or 3D construction
with tags, labels, dimensions, and accessories. Plan types similar to the drawing information
layers include 2D or 3D supply, return, and exhaust ducts. Besides, there are also floor, interior,
exterior, and bird’s-eye view plans.
A level layer indicates its attribute characteristic and the vertical elevation. Each layer can be
edited, modified, and displayed.
Design Interface
Computer-aided duct design software either must provide interfaces with available duct sizing programs
or itself must have the software for duct sizing. It is preferable that the duct design software
size the ducts by using equal friction, static regain, constant-velocity, and even cost optimiztion T
methods. The duct design software extracts system geometry and flow rates from the AutoCAD
drawing, exports the information to the selected duct sizing program, and uses the sizes calculated
to produce a double-line or 3D drawing to replace the centerline drawing.
Running Processes
Drawing a Centerline Layout. Designers are required to enter a fan and centerline layout in AutoCAD.
They have the option to add terminals, diffusers, necessary dampers, fittings, duct sections,
and related data to the drawing. Terminal text, line color, and fitting text are interpreted as the air-
flow, duct shape, and fitting configurations. An ASHRAE Duct Fittings Database with local loss coefficients
of 225 types of fittings provides default fitting types and coefficients.
Analyzing the Duct System. With a completed centerline layout, designers are asked to choose
the duct sizing software. Default values that are used for the centerline layout, such as duct shape,
and airflow rates are passed to the duct-sizing software.
The designer is asked to select the fan location. The computer program connects all lines to the
the fan in an attempt to connect each line to the end of the preceding line. The computer program
tests the continuity of the duct system by checking that all selected components are connected and
all junctions can be found. All lines must have a fan, terminal, or diffuser, or another centerline at
each end. The computer program generates nodes and inserts a node tag for the fan, main duct section,
and branch takeoff endpoints. Node shape is determined according to the duct shape and class
of node. Duct section information is stored in the upstream node. Each main duct node and branch
duct node are set with default attributes required by the duct-sizing program.
Editing Default Attributes and Running Duct-Sizing Program. If the designer is satified with all
default values, he or she can proceed to run the duct-sizing computer programs. If the designer needs to
change any of the default values, she or he can use AutoCAD to edit duct-sizing input default attributes
and visually pick the duct sections from the layout for editing before running the duct-sizing program.
When the editing work is done, modified node data are used to update input data for the ductsizing
program. The designer uses the command to run the chosen duct-sizing program.
Converting from Centerline to Double-Line or 3D. Centerline layout or centerline 3D layout can
be input into sizing calculations, and the computer program will produce a double-line layout or a
3D duct system layout. The conversion begins with the fan and produces the layout in the following
order: fittings, transitions, reducers, duct segments, and dampers.
Total Pressure Loss and Fan-Sizing Calculations. Some computer-aided duct design computer
programs have the ability to determine an existing duct layout shape and size in the CAD environment
and load it into the duct system total pressure loss and fan sizing calculations.
AIR SYSTEMS: AIR DUCT DESIGN 17.73
17.15 DUCT LINER AND DUCT CLEANING
Duct Liner
Duct liners are lined internally on the inner surface of the duct wall. They are mainly used for noise
attenuation and form part of the duct insulation layer that helps to reduce the heat transfer between
the conditioned air inside the duct and the ambient air surrounding the duct. Fiberglass blanket or
boards with a thickness of in. (13 mm), 1 in. (25 mm), in. (38 mm), and 2 in. (50 mm) have
been used as duct liner for decades and proved to be cost-effective.
In 1994, the U.S. Department of Health and Human Services (DHHS) announced that fiberglass
will be listed as a material “reasonably anticipated to be a carcinogen.” Although there are many valid
arguments concerning mold growth and fiber erosion from fiberglass duct liners, the use of fiberglass
in several institutional, educational, and medical projects has been banned or severely limited. Recommendations
concerning the use of fiberglass duct liner and alternatives in HVAC&R systems are discussed
in Secs. 19.3 and 19.5.
Duct Cleaning
The accumulation of dirt and debris as well as the growth of mold and fungi inside the ductwork
normally will not occur in a properly designed (such as equipped with medium- or high-efficiency
air filters), installed, operated, and maintained air system including ductwork. When the
cause of indoor air pollution is dirty ductwork—accumulation of dirt and debris or mold growth,
or both inside the ductwork—source removal or duct cleaning is often the most effective treatment.
Spielmann (1997) discussed the most widely used duct cleaning equipment and techniques
as follows:
Planning and System Inspection. Drawings are used as a valuable planning tool. Duct cleaning is
usually performed in zones of about 25 ft (7.5 m) or less. Contaminated ducts and uncleanable
types of ducts, such as flexible ducts and certain lined ducts, that cannot be cleaned without damaging
their duct liner should be identified. Inspection should be performed when the area is unoccupied
and the air system is turned off.
Starting the Job. Turn off the air system. The return side is often 5 to 10 times dirtier than the
supply side and is always cleaned first, before the supply side. Lay drop cloths to protect occupied
areas. Duct cleaning is started from return grilles and outdoor air intakes. Existing openings should
be used for access whenever possible. Additional access openings often must be cut. Cleaning
zones are isolated by using inflatable bladders inserted into the ducts. Duct sections are cleaned
along the ductwork toward the air-handling unit or air handler.
Agitation Devices. Contaminants are agitated by cleaning devices so that dirt and debris are loosened
drom the duct wall and extracted by the airstream. There are three currently used agitation
methods:
 Manual agitation using contact vacuuming requires physical access to all surfaces.
 Use portable compressed-air skipper nozzles to dislodge debris from the duct surfaces.
 Adopt a rotary brush or other cleaning head spinning on the end of a motor-driven flexible shaft.
When using a rotary brush in a lined duct, do not allow the brush to remain in the same spot for
too long.
Most of the duct cleaning work requires a combination of these agitation methods.
Duct Vacuum. A duct vacuum (negative air machine) is used to extract loosened debris into the
airstream of negative pressure so that contaminants will not pollute the conditioned space. A duct
11	2
1	2
17.74 CHAPTER SEVENTEEN
vacuum used in duct cleaning is often portable and consists of a blower, a HEPA filter to collect
debris, an outer casing, and connecting flexible duct. A large duct vacuum often has an extracting
volume flow rate of 4000 cfm (1.89 m3 / s) and more. For each cleaning zone, there are two openings
: one is the access to the agitation device, and the other opening connects the duct under cleaning
directly to the portable duct vacuum. A smaller portable duct vacuum is used to extract debris
from the fan blades, coils, and condensate drip pans.
Sealing of Access Openings. After the duct cleaning is completed, the cleaned ducts should be
isolated from the cleaning zone in which duct cleaning is progressing by means of inflatable bladder.
The access opening should be patched using reusable doors, sheet metal, duct tape, and
sealants.
17.16 PRESSURE AND AIRFLOW MEASUREMENTS
The total pressure of the airflow is the sum of the static pressure, which acts in all directions, and
the velocity pressure, which results from the impact and the inertia of the airflow. The Pitot tube
and manometers, shown in Figs. 17.29 and 17.30, are widely used to measure total pressure, static
pressure, velocity pressure, and thus airflow inside air ducts.
A standard Pitot tube consists of two concentric tubes having an outside diameter D  in.
(8 mm) and an inner tube of diameter of in. (3.2 mm), as shown in Fig. 17.29. The inner tube
opens to the airflow at the nose, and the other end is used as the total pressure tap. Eight equally
spaced holes of 0.04-in. (1-mm) diameter allow the air to flow into the hollow space between the
outer and inner tubes.
The hollow space is connected to the static pressure tap. Because the small holes are located
perpendicular to the centerline of the head, they are able to measure the static pressure when the
head of the Pitot tube is placed in a position opposite to the direction of the airflow. The centerlines
of the small holes are located at a distance of 8D from the nose and 16D from the stem. The negative
pressure produced at the nose is nearly balanced by the positive pressure at the stem.
The U tube shown in Fig. 17.30a is the simplest type of manometer used to measure pressure in
air ducts. The vertical difference in the liquid column indicates the pressure reading. For more
1	8
5	16
AIR SYSTEMS: AIR DUCT DESIGN 17.75
FIGURE 17.29 Pitot tube.
17.76 CHAPTER SEVENTEEN
FIGURE 17.30 Manometers: (a) U tube; (b) inclined
manometer.
FIGURE 17.31 Pressure measurements in air ducts.
accurate measurement, an inclined manometer (see Fig. 17.30b) is often used. The relationship between
the height of the liquid column H and the length of the magnified inclined scale is
H  L sin  (17.103)
where L  length of inclined scale with liquid column, ft (m)
 angle of inclination of inclined tube, deg
Because the cross-sectional area of the inclined tube is very small compared with that of the reservoir,
the change in the liquid level of the reservoir can be ignored.
The following precautions should be taken when one takes pressure measurements in an air duct
such as that shown in Fig. 17.31:
 The nose of the Pitot tube must always be placed opposite to the direction of airflow whether in
the suction side or discharge side of the fan.
 When the total pressure or static pressure is measured, one leg of the U tube or inclined manometer
must be open to the atmospheric air as the reference datum. The smaller pressure is always
connected to the open end of the inclined tube of the inclined manometer.
 For a velocity pressure measurement, the total and static pressure taps must be connected to the
two ends of the manometer. The total pressure tap is connected to the reservoir, and the static
pressure tap is connected to the inclined tube. Velocity pressure is always positive in the direction
of airflow.
Equal-Area Versus Log Tchebycheff Rule
Because velocity is usually not uniform across the cross-sectional area, a transverse is often used to
determine the average velocity. The equal-area method was widely used before. Recently ASHRAE
AIR SYSTEMS: AIR DUCT DESIGN 17.77
0.032D
0.135D
(a) (b)
0.939W
0.765W
0.563W
0.437W
0.500H
0.712H
0.926H
H
0.235W
0.061W
0.074H
0.288H
W
0.968D
0.865D
0.679D
0.321D
D
60
60
60
60
60
60
FIGURE 17.32 Measuring points in rectangular and round duct transverse: (a) log Tchebycheff rule for rectangular
duct transverse; (b) log-linear rule for round duct transverse.
Standard 111-88, the Associated Air Balance Council (AABC), and the National Environmental
Balancing Bureau (NEBB) all adopted the log Tchebycheff rule for rectangular duct and round duct
transverse, as shown in Fig. 17.32a and b, because this rule provided greatest accuracy. The reason
is that the location of transverse points of these rules has taken into account the effect of the duct
wall friction and the reduction in velocity near the duct walls. For round duct transverse, the log
Tchebycheff rule and log-linear transverse method are similar. Refer to ASHRAE Standard 111-88
for details. The log Tchebycheff rule was developed by a mathematician named Tchebycheff in
1977.
In MacFerran (1999), airflow in ducts measured by the equal-area method was often higher than
correct results, from 7 percent higher maybe up to 25 percent higher.
When the velocity pressure is measured, in inches WC (Pa), the air velocity, in fpm, can be calculated
from Eq. (17.10). For SI units, the air velocity, in m / s, can be calculated from Eq. (17.9), in
which gc  1.
REFERENCES
Amistadi, H., Design and Drawing Software Review, Engineered Systems, no. 6, 1993, pp. 18–29.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, 1977a.
ASHRAE, Duct Fittings Loss Coefficient Tables, Atlanta, 1997b.
ASHRAE, ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, ASHRAE Inc., Atlanta, 1999.
Behls, H., Balanced Duct Systems, ASHRAE Transactions, 1978, Part I, pp. 624–646.
Brooks, P. J., Duct Design Fundamentals, ASHRAE Journal, no. 4, 1995, pp. 69–76.
Coe, P. E., The Economics of VAV Duct Looping, Heating /Piping /Air Conditioning, August 1983, pp. 61–64.
Colebrook, C. F., Turbulent Flow in Pipes with Particular Reference to the Transition Region Between the
Smooth and Rough Pipe Laws, Journal of Institute of Civil Engineers, February 1939, p. 133.
Dean, R. H., Dean, F. J., and Ratzenberger, J., Importance of Duct Design for VAV Systems, Heating / Piping /
Air Conditioning, August 1985, pp. 91–104.
Evans, R. A., and Tsal, R. J., Basic Tips for Duct Design, ASHRAE Journal, no. 7, 1996, pp. 37–42.
Habjan, J., Medium Pressure Duct Sizing and Design, Heating / Piping / Air Conditioning, December 1984,
pp. 95–100.
Heyt, H. W., and Diaz, M. J., Pressure Drop in Flat Oval Spiral Air Duct, ASHRAE Transactions, 1975, Part II,
pp. 221–232.
Heubscher, R. G., Friction Equivalents for Round, Square and Rectangular Ducts, ASHVE Transactions, 1948,
pp. 101–144.
Idelchik, I. E., Handbook of Hydraulic Resistance, 2d ed., Hemisphere Publishing Corp.,Washington, 1986.
MacFerran, E. L., Equal Area vs. Log-Tchebycheff, Heating / Piping / Air Conditioning, no. 12, 1999, pp.
26–31.
Miller, E. B., and Weaver, R. D., Computer-Aided Duct Design in a Small Engineering Office, ASHRAE Transactions,
1978, Part I, pp. 647–664.
NFPA 90A, Standard for the Installation of Air Conditioning and Ventilating Systems, National Fire Protection
Association, Quincy, MA, 1985 ed.
North American Insulation Manufacturers Association, New Duct Liner Standard Addresses HVAC System
Challenges, Heating / Piping / Air Conditioning, no. 2, 1994, pp. 61–67.
Scientific Computing, Software Review: Up for Review (Again), Engineered Systems, no. 1, 1998, pp. 76–84.
Shitzer, A., and Arkin, H., Study of Economic and Engineering Parameters Related to the Cost of an Optimal
Air Supply Duct System, ASHRAE Transactions, 1979, Part II, pp. 363–374.
Spielmann, S., Check Your Basic Skill on Duct and Ventilation System Cleaning, Air Conditioning, Heating &
Refrigeration News, Feb. 24, 1997, pp. 3–5.
17.78 CHAPTER SEVENTEEN
Stoecker,W. F., and Bertschi, R. I., Design Duct System for Minimum Life-Cycle Cost, Proceedings of Improving
Efficiency and Performance of HVAC Equipment and Systems for Commercial and Industrial Buildings,
vol. 1, 1976, pp. 200–207.
Swim,W. B., and Griggs, E. I., Duct Leakage Measurement and Analysis, ASHRAE Transactions, 1995, Part I,
pp. 274–291.
TIMA/HPAC, Fiber Glass Duct Systems, Heating / Piping / Air Conditioning, October 1986 Supplement.
Tsal, R. J., and Behls, H. F., Evaluation of Duct Design Methods, ASHRAE Transactions, 1986, Part I A,
pp. 347–361.
Tsal, R. J., Behls, H. F., and Mangel, R., T-Method Duct Design, Part I: Optimization Theory, ASHRAE Transactions,
1988, Part II, pp. 90–111.
Tsal, R. J., Behls, H. F., and Mangel, R., T-Method Duct Design, Part II: Calculation Procedure and Economical
Analysis, ASHRAE Transactions, 1988, Part II, pp. 112–150.
Wang, S. K., Air Conditioning, vol. 2, Hong Kong Polytechnic, Hong Kong, 1987.
Wang, S. K., Leung, K. L., and Wong,W. K., Sizing a Rectangular Supply Duct with Transversal Slots by Using
Optimal Cost and Balanced Total Pressure Principle, ASHRAE Transactions, 1984, Part II A, pp. 414–429.
AIR SYSTEMS: AIR DUCT DESIGN 17.79
CHAPTER 18
AIR SYSTEMS:
SPACE AIR DIFFUSION
18.1
18.6 DESIGN PROCEDURE OF
MIXING FLOW AIR DIFFUSION
SYSTEM 18.31
Select the Type of Supply Outlet 18.31
Volume Flow Rate per Outlet or per Unit
Length 18.32
Choose an Optimum Throw–
Characteristic Length Ratio 18.33
Determine the Design Characteristics
of Slot Diffusers in Perimeter
Zone 18.33
Select the Specific Outlet from
Manufacturer’s Catalog 18.34
Determination of Final Layout of Supply
Outlets and Return Inlets 18.34
18.7 DISPLACEMENT FLOW AND
UNIDIRECTIONAL FLOW 18.38
Displacement Flow 18.38
Unidirectional Flow 18.38
Unidirectional Flow for Clean
Rooms 18.39
Ventilating Ceiling 18.40
Ceiling Plenum and Supply Air
Velocity 18.41
18.8 STRATIFIED DISPLACEMENT
FLOW 18.42
Two-Zone Stratified Model 18.42
Operating Characteristics 18.42
Comparison of Stratified Displacement
Flow and Mixing Flow 18.43
18.9 PROJECTING FLOW—SPOT
COOLING/HEATING 18.44
Benefits of Projecting Flow 18.44
Industrial Spot Cooling
Systems 18.44
Recommendations in Spot Cooling
Design 18.46
Desktop Task Conditioning Systems
18.46
Performance of Desktop Task
Conditioning Systems 18.47
Application of Desktop Task Conditioning
Systems 18.48
18.10 UPWARD FLOW UNDERFLOOR
AIR DISTRIBUTION 18.48
Upward Flow from Floor Plenum 18.48
Thermal Storage of Floor Plenum 18.49
Heat Unneutralized 18.50
Maintaining a Consistent Access Plenum
Temperature 18.50
Floor Plenum Master Zone Air
Temperature Control 18.50
18.1 PRINCIPLES OF SPACE AIR
DIFFUSION 18.2
Draft and Effective Draft
Temperature 18.2
Air Diffusion Performance
Index 18.3
Space Diffusion Effectiveness
Factor 18.4
Ventilation Effectiveness and Air
Exchange Effectiveness 18.4
18.2 AIR JETS 18.5
Free Isothermal Jets 18.5
Throw, Entrainment Ratio, and
Characteristic Length 18.7
Confined Air Jets 18.8
Free Nonisothermal Jets 18.10
18.3 SUPPLY OUTLETS AND RETURN
INLETS 18.11
Grilles and Registers 18.11
Ceiling Diffusers 18.12
Slot Diffusers 18.14
Nozzles 18.16
Accessories for Supply Outlets 18.17
Return and Exhaust Inlets 18.17
Light Troffer Diffuser and Troffer-Diffuser
Slot 18.19
18.4 MIXING FLOW 18.20
Airflow Pattern and Space Air
Diffusion 18.20
Principles and Characteristics of Mixing
Flow 18.20
Mixing Flow Using High Side
Outlets 18.21
Mixing Flow Using Ceiling
Diffusers 18.23
Mixing Flow Using Slot
Diffusers 18.24
Mixing Flow Using Sill or Floor
Outlets 18.24
Stratified Mixing Flow 18.25
18.5 COLD AIR DISTRIBUTION 18.28
Cold Air Distribution versus Conventional
Air Distribution 18.28
High Induction Nozzle Diffusers 18.28
Characteristics of Cold Air Distribution
Systems 18.29
Case Study—Florida Elementary
School 18.29
Performance of Ceiling and Slot
Diffusers 18.29
Cold Air Distribution with Fan-Powered
VAV Boxes 18.30
Surface Condensation 18.30
18.1 PRINCIPLES OF SPACE AIR DIFFUSION
Space air diffusion distributes the conditioned air containing outdoor air to the occupied zone (or a
given enclosure) in a conditioned space according to the occupants’ requirements. An occupied
zone is a space with the following dimensions: (width of room  1 ft)  (depth of room  1 ft) 
(a height of 6 ft). A satisfactory space air diffusion evenly distributes the conditioned and outdoor
air to provide a healthful and comfortable indoor environment for the occupants or the appropriate
environment for a specific manufacturing process, at optimum cost.
Because space air diffusion is the last process of air conditioning and takes place entirely within
the conditioned space, it directly affects the effectiveness of air conditioning. Because the diffused
air and ambient air are transparent, space air diffusion is also difficult to trace.
Draft and Effective Draft Temperature
One of the most common complaints about air conditioned space is draft. Draft is defined as an unwanted
local cooling of the human body caused by air movement and lower space air temperature.
Recent developments by Fanger et al. (1988, 1989) demonstrate that air velocity fluctuations also
create drafts, and they investigated the impact of turbulence intensity on the sensation of draft. Figure
18.1 depicts velocity fluctuations in a typical conditioned space. A new dimensionless parameter
called the turbulence intensity Itur is introduced and expressed as
(18.1) Itur 
v
vm
18.2 CHAPTER EIGHTEEN
Design Considerations 18.50
Applications of Upward Flow Underfloor
Air Distribution 18.51
18.11 COMPUTATIONAL FLUID
DYNAMICS 18.51
CFD Becomes More Popular 18.52
Reynolds-Averaged Navier-Stokes
Equations 18.52
Numerical Methods 18.52
Conducting CFD
Experiments 18.54
REFERENCES 18.54
FIGURE 18.1 Fluctuations of air velocity in a typical air conditioned
space. (Adapted with permission from ASHRAE Journal, April 1989, p. 30.)
where v  standard deviation of air velocity fluctuations, fpm (m/ s)
vm  mean air velocity, fpm (m/ s)
Experiments showed that if the space temperature Tx  73.4°F (23°C) and vm  30 fpm (0.15
m/ s), the percentage of dissatisfied occupants may increase from about 10 to 15 percent when Itur
increases from 0.1 to 0.5. A parameter called the effective draft temperature , in °F (°C), which
combines the effects of uneven space air temperature and air movement, is often used to assess the
deviations of local magnitudes from the mean value, and it is defined as
(18.2)
where Tx, vx  space air temperature and velocity at specific location, °F (°C) and fpm (m / s)
Tr  mean space air temperature or set point, °F (°C)
In Eq. (18.2), a is a conversion constant to combine the effects of space air temperature and air
movements; its value is 0.07 when T is expressed in °F and v is expressed in fpm, and 8 when T is
expressed in °C and v in m / s.
The desirable mean space air velocity vrm, in fpm (m/ s), is closely related to the space air
temperature Tr to be maintained, the metabolic rate, and the clothing insulation of the occupant.
According to Fig. 4.8, for a space relative humidity 	 50 percent, Trad  Tr , a metabolic rate of
400 Btu /h (117.2 W), and a clothing insulation of 0.6 clo (0.093 m2  °C/W), their relationship may
be given as follows:
Air Diffusion Performance Index
To an extent, the magnitude of effective draft temperature also reflects the degree of thermal comfort
that can be provided by comfort air conditioning systems. During cooling mode in commercial
and public buildings, if the space temperature is maintained between 75 and 78°F (23.9 and
25.6°C), Trad  Tr, space air velocity vr 	 55 fpm (0.28 m / s), and the space relative humidity is between
30 and 70 percent, then most sedentary occupants feel comfortable when 3°F 	 	
2°F (1.7°C 	 	 1.1°C). The air diffusion performance index (ADPI), in percent, which
evaluates the performance of space air diffusion, is calculated as
(18.3)
where N number of points measured in occupied zone in which  3°F 	 	 2°F
( 1.7°C 	 	 1.1°C)
N  total number of points measured in occupied zone
The higher the ADPI, the higher the percentage of occupants who feel comfortable. Maximum
ADPI approaches 100 percent.
Because air velocity fluctuations create drafts, therefore, they also affect ADPI. Abu-Ei-Hassan
et al. (1996), based on the impact of turbulence intensity Tu recommended by Fanger et al. (1988),
conducted tests to determine the effect of turbulence intensity Tu (velocity fluctuations) on the
ADPI. Abu-Ei-Hassan et al. took the centerline velocity of the test room, which has nearly the same
flow characteristic as the centerline velocity of the air jet, as the air velocities of the local points,
and the ADPIs thus calculated as Air Diffusion Performance Index that take into consideration the
Tu impact. The results showed that for a sidewall outlet space air diffusion system with a room
ADPI 
N
 100
N
Tr, °F 70 72 74 76 78 80
vrm, fpm 16 20 25 32 40 55

 Tx  Tr  a(vx  vrm )
AIR SYSTEMS: SPACE AIR DIFFUSION 18.3
cooling load intensity of 10 to 20 Btu /h ft2 (31.5 to 63 W/m2) the Air Diffusion Performance Index
considering the turbulence effect were about 10 percent less than those that did not take the impact
of Tu into account. At higher cooling loads and airflow rates, the effect of Tu became larger.
Actually, the drop in ADPI is less than 10 percent for cooling load intensity between 10 and 20
Btu/h ft2 (31.5 and 63 W/m2) due to the impact of Tu because the points of air velocity located on
the centerline of the test room are only a small portion of the total number of measured points.
In cooling mode operation, ADPI is an appropriate index to use to evaluate a space air diffusion
system. For heating mode operation, the temperature gradient between two points in the occupied
zone may be a better indicator of the thermal comfort of the occupants and the effectiveness of a
space air diffusion system. Usually, the temperature gradient is less than 5°F (2.8°C) within an
occupied zone.
Space Diffusion Effectiveness Factor
The effectiveness of a space air diffusion system can also be assessed by using a space diffusion effectiveness
factor for air temperature 
T or for air contamination 
C. Both factors are dimensionless.
Effectiveness factor 
T compares temperature differentials and 
C compares contamination differentials
as follows:
(18.4)
where T  temperature, °F (°C)
C  concentration of air contamination, g/m3
In Eq. (18.4), subscript re represents the recirculating air, ex the exhaust air, r the space air or air at
the measuring point, and s the supply air. When 
 1, the space air diffusion is considered effective.
If 
	 1, a portion of supply air has failed to supply the occupied zone and exhaust through
the return or exhaust inlets directly. Parameters in both the numerator and denominator must be in
the same units.
Ventilation Effectiveness and Air Exchange Effectiveness
ASHRAE Handbook 1997, Fundamentals, defines ventilation effectiveness as a description of an air
distribution system’s ability to remove internally generated contaminants from a zone, space, or
building. Air change effectiveness 
ex is defined as air system ability to deliver ventilation air to a
zone, space, or building. The age of air age, in minutes or hours, is the time period that outdoor
ventilation air has been in a zone, space, or building, and age can be evaluated for an existing building
using the trace gas method. Chamberlin et al. (1999) proposed the following when growth
(step-up) and decay (step-down) tracer methods are used:
(18.5a)
where Cavg  time-averaged concentration of tracer gas measured in test zone from when tracer
gas injections began to when last sample is taken, g/m3 or ppm
Cend, Cstart  concentration of tracer gas when last sample was taken at time end and when first
sample was taken at start, g/m3 or ppm
 ( stop 
start)(Cavg  Cstart)

age  ( end 
initial)1 
Cavg
Cend 

C 
Cex  Cs
Cr  Cs

T 
Tre  Ts
Tr  Ts

Tex  Ts
Tr  Ts
18.4 CHAPTER EIGHTEEN
end  initial  time elapsed from when tracer gas injection began to when last sample was taken,
min or h
stop  start  time elapsed from when first sample was taken to when tracer gas injection stopped,
min or h
The youngest air is the freshest air through infiltration. A longer age means a poorer outdoor air delivery.
Another measure of ventilation is the nominal time constant N, in min or h, and it can be calculated
as
(18.5b)
where  supply volume flow rate to zone, space, or room, cfm (m3 /s)
Vr  volume of zone, space, or room, ft3 (m3)
Air exchange effectiveness is a relative index to measure how well the outdoor air is diffused to various
locations in the occupied zone. Nominal air change effectiveness 
N can be calculated as
(18.6)
and age,N is evaluated through tracer gas measurement. Normally, for a properly designed, installed,
operated, and maintained air distribution system, 
N  1.
18.2 AIR JETS
An air jet is an airstream that discharges from an outlet with a significantly higher velocity than that
of the surrounding air, and moves along its centerline until its terminal velocity reduces to a value
that equals or approximately equals the velocity of the ambient air. Because of the turbulence of air
particles, air jets tend to spread. They also rise or fall depending on the buoyancy of the airstream.
The outer boundary of an air jet where air is moving at a perceptible velocity, such as 150 fpm
(0.75 m / s), 100 fpm (0.5 m / s), or 50 fpm (0.25 m / s), is called the envelope. In general, air jets can
be classified as free or confined, isothermal or non isothermal, and axial or radial. A free air jet is an
ideal air jet whose envelope (outer boundary) is not confined by the enclosure of the conditioned
space. A confined air jet’s envelope is confined by the ceiling, floor, walls, windows, and furniture
of the conditioned space.
According to experimental results, the characteristics of an air jet approach those of a free air jet
when the relationship  50 is satisfied. Here Ar represents the cross-sectional area of the
enclosure perpendicular to the centerline of the air jet, in ft2 (m2), and Do indicates the diameter or
the circular equivalent of the supply outlet, in ft (m).
An isothermal air jet is one whose temperature is equal or nearly equal to the temperature of the
ambient air. A nonisothermal air jet is one whose temperature is different from that of the ambient
air in the conditioned space. An axial air jet projects in one direction, and a radial air jet projects radially
in all directions.
Free Isothermal Jets
Along the centerline of a free isothermal jet, four zones can be identified, as shown in Fig. 18.2:
1. Core zone. In the core zone, the centerline velocity remains unchanged. This zone extends about
4Do from the surface of the outlet.
?Ar /Do

N 
N
age,N
V?
s
N 
V?
s
Vr
AIR SYSTEMS: SPACE AIR DIFFUSION 18.5
2. Transition zone. In the transition zone, the centerline velocity decreases inversely with the
square root of the distance from the surface of the outlet. This zone extends about 8Do.
3. Main zone. In the main zone, turbulent flow is fully developed, and the maximum velocity
decreases inversely with the distance from the surface of the outlet. Even when the air jet is
discharged from a rectangular outlet, the cross section of the airstream becomes circular in the
main zone. This zone extends about 25 to 100Do in length.
4. Terminal zone. In the terminal zone, the maximum air velocity decreases rapidly to a value less
than 50 fpm (0.25 m / s) within a distance of a few outlet diameters.
In core, transition, and main zones, the fluctuating velocity components transport momentum across
the boundary. Therefore, ambient air is induced into the air jet, and the airstream diverges to a
greater spread either vertically or horizontally. The angle of divergence 2 of a free isothermal jet
discharged from a nozzle is about 22°. For an air jet discharged from a slot, its angle of divergence
2  perpendicular to the slot is about 33°. If the guide vanes at the outlet are deflected at an angle
from the straight position, the spread of the air jet is greater.
The velocity profiles at different cross-sectional planes perpendicular to the airflow in the main
zone of a free isothermal jet are similar to one another. At any specific cross-sectional plane, the
velocity profile can be approximated by the formula
(18.7)
where v  air velocity at distance R from centerline of air jet, fpm (m/ s)
vc  centerline velocity, fpm (m/ s)
R  radial distance from centerline of air jet, ft (m)
R0.5  radial distance from centerline of air jet to a point where velocity equals 0.5vc, ft (m)
Research shows that the law of conservation of momentum can be assumed in the main zone of
a free air jet. The application of this law to determine the centerline velocity of a free air jet gives
the following results:
(18.8)
where K, K  centerline velocity constants, depending mainly on type of outlets; normally,
K  1.13K
Ho  width of radial air jet at outlet or vena contracta, ft (m)
vo  mean velocity at vena contracta, fpm (m/ s)
vc
vo

K
x/Do

K
x/?Ao

K
x/Ho
 R
R0.5 2
 3.3 log
vc
v
18.6 CHAPTER EIGHTEEN
FIGURE 18.2 Four zones of a free, isothermal, axial air jet.
x  distance from surface of outlet to a cross-sectional plane having centerline velocity
vc, ft (m)
In Eq. (18.8), Ao, in ft2 (m2), represents the effective area of the airstream, i.e., the minimum area at
the vena contracta. It can be calculated as
(18.9)
where Ac  core area of outlet, i.e., surface area of opening, ft2 (m2)
vcore  face velocity at core of outlet, fpm (m/ s)
Cd  discharge coefficient, usually between 0.65 and 0.9
Rfa  ratio of free area to gross area (free area is net area of opening through which air can
pass)
When vo is 500 to 1000 fpm (2.5 to 5 m / s), for round free openings, K  5; for rectangular openings
and linear slot diffusers, K  4.9. When vo is 2000 to 5000 fpm (10 to 25 m / s), for round free
openings, K  6.2.
When multiple air jets are discharged into a conditioned space at the same level, each air jet
behaves independently until the jets meet. From the point where the jets meet, the velocities between
the centerlines of the air jets increase until they are equal to the centerline velocities of the
air jets.
Throw, Entrainment Ratio, and Characteristic Length
Throw in ft (m), is defined as the horizontal or vertical axial distance from the outlet to a cross-sectional
plane where the maximum velocity of the airstream at the terminal zone has been reduced to
50 fpm (0.25 m / s), 100 fpm (0.5 m / s), or 150 fpm (0.75 m / s). The throw is indicated by Tv, and
the subscript denotes the terminal velocity for which the throw is measured. For instance, T50 (T0.25)
indicates the throw with a terminal velocity of 50 fpm (0.25 m / s).
From Eqs. (18.8) and (18.9), throw can be calculated as
(18.10)
where vt,max  maximum velocity of airstream at terminal zone, fpm (m/ s)
 supply volume flow rate at outlet, cfm (m3 /s)
For a specific configuration of supply outlet, throw Tv depends on both the supply volume flow rate
s and supply outlet velocity vcore.
The entrainment ratio Ren is the ratio of the volume flow rate of the total air at a specific crosssectional
plane of the air jet x to the volume flow rate of the supply air o discharged from the outlet
,which is sometimes called primary air. Total air is the sum of supply air and induced air.
The entrainment ratio is proportional to the distance or square root of the distance from the outlet.
For circular jets in the main zone, the entrainment ratio can be calculated as
For a long slot, the entrainment ratio is
(18.11) Ren 
V?
x
V?
o
 ? 2x
KHo
 ?2vo
vc
Ren 
V?
x
V?
o

2x
K?Ao

2vo
vc
V?
V? V?
V?
s
Tv 
KV?s
vt, max?AcCd Rfa
Ao  AcCd Rfa
vo  vcoreCd Rfa
AIR SYSTEMS: SPACE AIR DIFFUSION 18.7
The characteristic length L, in ft (m), is the horizontal distance from the surface of the outlet to
the nearest vertical opposite wall or the horizontal distance from the surface of the outlet to the
midplane between two outlets in the direction of airflow, or the distance to the closest intersection
of air jets. The ratio of throw to characteristic length Tv /L is related to the ADPI of various supply
outlets and has been used as a parameter in space diffusion design.
Confined Air Jets
In actual practice, most air jets are confined by the boundary of the room or the conditioned space.
For a confined air jet, the total momentum of the fluid elements decreases gradually as distance
from the outlet increases because of the friction between the airstream and the boundary.
Surface Effect. When a primary airstream discharged from a supply outlet flows along a surface,
the velocity of the primary airstream is significantly higher than that of the ambient air, and a lowerpressure
region is formed near the surface along the airflow, as shown in Fig. 18.3. Consequently, in-
18.8 CHAPTER EIGHTEEN
FIGURE 18.3 Surface effect.
FIGURE 18.4 Centerline velocities of confined air jets.
duced ambient air at a comparatively higher pressure presses the air jet against the surface, even
when it is a curved surface. Such a phenomenon is called the surface effect or the Conda effect.
Friction between the air jet and the boundary decreases the centerline velocity of confined air
jets, as shown in Fig. 18.4. However, because of the surface effect, the throw of a confined air jet is
longer, and the drop from the horizontal axis smaller, than that of a free air jet.
Airflow Pattern. The flow pattern and characteristics of a confined air jet using a sidewall outlet
were introduced by Russian scientists during the 1950s. In an air conditioned room whose supply
outlet is located above the occupied zone and whose exhaust opening is on the same side of the supply
outlet, the supply air jet clings to the surface of the ceiling and mixes with the room air. An induced
reverse airstream, with more even velocity and temperature distribution than that of the air
jet, covers the occupied zone.
Figure. 18.5 shows the airflow pattern of a typical confined isothermal jet. When supply air is discharged
from the circular outlet and moves along the surface of the ceiling, the fluctuating velocity
components continue to transport momentum across the boundary of the air jet. Therefore, ambient air
is induced into the air jet, and the induced circulating airflow occupies most of the enclosed space.
As the air jet moves forward, its mass flow rate increases and mean air velocity decreases until it
arrives at a cross-sectional plane where dimensionless distance s  0.22. Term s is defined as
(18.12)
where a  turbulence factor; for a circular nozzle a  0.076, and for a rectangular outlet without
guide vanes a  0.15.
s 
ax
?0.5Ar
AIR SYSTEMS: SPACE AIR DIFFUSION 18.9
FIGURE 18.5 Airflow pattern of a typical confined air jet.
The air jet terminates beyond s  0.22, and the airstream makes a 180° turn, forming a reverse
airstream flowing in the opposite direction. The majority of the reverse airstream turns upward and
is induced by the air jet again. Only a portion of it is exhausted outside the room.
The angle of divergence, the velocity profile, and the calculation of the entrainment ratio of con-
fined jets are similar to those of free jets. Based on the principle of continuity of mass at steady
state, other characteristics of confined air jets can be summarized as follows:
 If there is no infiltration into or exfiltration from the room, the mass flow rate of supply air is exactly
equal to the exhaust air.
 If there are no obstructions in the room, the ideal stream lines of the induced air are closed curves
and form recirculating flow.
 If the supply outlet and exhaust inlet are located on the same side of the room, at any crosssectional
plane perpendicular to the horizontal airflow, the mass flow rates of airstreams flowing
at opposite directions must be equal.
 The volume flow rate of induced air is equal to or several times greater than that of supply air at
the cross-sectional plane where s  0.22. The characteristics of airstreams in the occupied zone
depend mainly on the induced reverse airstream. The volume flow rate and the air velocity of the
reverse airstream are highest at the cross-sectional plane where s  0.22.
Free Nonisothermal Jets
When the temperature of the conditioned air discharged from a supply outlet is different from that
of the ambient air in a conditioned space, the buoyancy of the fluid particles causes the trajectory of
the air jet to deviate from the axis of the free isothermal jet. A cold air jet will descend, and a warm
air jet will ascend. Figure 18.6 shows a free cold air jet discharged horizontally from a nozzle.
18.10 CHAPTER EIGHTEEN
FIGURE 18.6 Path of a cold air jet discharged horizontally from a nozzle.
In Fig. 18.6, x is the horizontal distance between a cross-sectional plane in a cold air jet and the
surface of the outlet, and y is the drop of the cold air jet, i.e., the vertical distance between the horizontal
axis of the nozzle and the center of the specific cross-sectional plane. The velocity profile at
the specific cross-sectional plane is indicated by the solid line, and the temperature profile is indicated
by the dotted line.
According to the experimental results of Kostel (1955), the following empirical formula can be
used to determine the vertical drop of a cold air jet and the vertical rise of the warm air jet discharged
from a nozzle:
(18.13)
where  angle between centerline of nozzle and horizontal axis of nozzle, deg
K  constant
Also Ar, known as Archimedes’ number, indicates the buoyant forces of the air jet and is given as
(18.14)
where g  gravitational acceleration, ft / s2 (m/s2)
To  temperature difference between supply air at outlet and ambient air in conditioned
space, °F (°C)
TrR  absolute temperature of space air, °R (K)
vo  air velocity at supply outlet, f/s (m/s)
For free air jets, Kostel (1955) determined that K  0.065.
For a nonisothermal jet, the relationship between the decay of the centerline velocity and the
centerline temperature difference Tc, in °F (°C), can be shown as
(18.15)
where subscript c represents centerline, o the outlet, and r the space. The characteristics of isothermal
radial air jets are similar to those of axial air jets. When determining the centerline velocity, use
x/Ho for radial jets instead of x / for axial jets.
18.3 SUPPLY OUTLETS AND RETURN INLETS
The proper type of supply outlet for a conditioned space is largely determined by the architectural
setup of the room, the airflow pattern needed, the indoor environmental requirements, and the load
conditions. Five types of supply outlets are currently used: grilles and registers, ceiling diffusers,
slot diffusers, light troffer diffusers, and nozzles. A window sill outlet is actually a type of grille
mounted at the top of a fan coil.
Grilles and Registers
A grille or grill, as shown in Fig. 18.7a, is an outlet for supply air or an inlet for return air or exhaust
air. A register is a grille with a volume control damper. Figure 18.7b shows a single-deflection
grille. It consists of a frame and one set of adjustable vanes. Either vertical or horizontal vanes may
be used as face vanes to deflect the airstream.
?Ao
Tc
To

Tc  Tr
Ts  Tr

0.8vc
vo
Ar 
g?Ao
vo 2
To
TrR
y
?Ao

x
?Ao
tan 
 K Ar  x
?Ao
cos 3
AIR SYSTEMS: SPACE AIR DIFFUSION 18.11
Figure 18.7c shows a double-deflection register. It is able to deflect the airstream both horizontally
and vertically. A volume damper is also used to adjust the volume flow through the register.
Usually extruded aluminum vanes, aluminum or steel frames, and steel dampers are used. A baked
enamel finish gives the grille an attractive appearance.
Grilles have a comparatively lower entrainment ratio, greater drop, longer throw, and higher air
velocities in the occupied zone than slot and ceiling diffusers.
In the manufacturer’s catalog, the performance of a grille or register is defined by the following
parameters:
 Core size or core area Ac, which indicates the total plane area of an opening, ft2 (m2)
 Volume flow rate , cfm (m3 /s, or L / s)
 Air velocity v  /Ak, fpm (m/ s)
 Total pressure loss pt, in. WC (Pa)
 Throw at terminal velocities of 50 fpm (0.25 m / s), 100 fpm (0.5 m / s), and 150 fpm (0.75 m / s)
with horizontal vanes deflected at 0°, 22.5°, and 45°
 Noise criteria curve
Here Ak is the area factor representing the net or corrected area, in ft2 (m2), which is the unobstructed
area of a grille, register, or diffuser measured at the vena contracta.
For Air Diffusion Council–certified performance data, throw is tested on isothermal jets in order
to avoid the influence of buoyancy forces at different supply temperature differential Ts. During
testing, the difference between supply and room air temperatures cannot exceed 2 °F (1.11°C).
Ceiling Diffusers
A ceiling diffuser consists of a series of concentric rings or inner cones made up of vanes arranged in
fixed directions and an outer shell or frame, as shown in Fig. 18.8a. Ceiling diffusers can be round,
square, or rectangular. Square diffusers are most widely used. Supply air is discharged through the
concentric air passages or directional passages in one, two, three, or in all directions by using different
V?
V?
18.12 CHAPTER EIGHTEEN
FIGURE 18.7 Supply grille and register: (a) grille, front view; (b) single-deflection, vertical vanes; (c) double-de-
flection register, with vertical and horizontal vanes.
18.13 AIR SYSTEMS: SPACE AIR DIFFUSION
..(>
(b)
FIGURE 18.8 Ceiling diffusers: (a) square and rectangular
ceiling diffusers; (b) removable inner-core ceiling diffuser;
(c) perforated ceiling diffuser. (c)
types of inner cone and vanes. The air discharge pattern of many ceiling diffusers may be changed
from horizontal to vertical by means of adjustable inner cones or special deflecting vanes. Ceiling diffusers
are often mounted at the center of the conditioned space and discharge air in all directions.
An important characteristic of a ceiling diffuser is its induction effect. Induction is the volume
flow rate of space air induced into the outlet by the primary airstream. It is also called aspiration.
The ratio of the volume flow rate of induced air to the volume flow rate of supply air ind / s, both
in cfm (L / s), is called the induction ratio. A ceiling diffuser with a higher induction ratio (see Fig.
18.8b) is beneficial for a higher supply temperature differential Ts for cold air distribution in an
ice storage system, which is discussed later in Sec. 18.5 and in Chap. 31.
Figure 18.8c shows a perforated ceiling diffuser. A perforated surface increases the induction effect.
It is also used to match the appearance of the ceiling tile. The diameter of the holes on the perforated
panel is typically in. (4.8 mm). Some ceiling diffusers are designed to discharge air
evenly in all directions so that the surrounding ceiling surfaces are less likely to be smudged.
The size of a ceiling diffuser is determined by its neck size and is therefore closely related to the
neck velocity of the diffuser. Ceiling diffusers are made of an outer shell and a removable inner
core. The opposed-blade damper with an equalizing device of parallel blades mounted within a
frame is often used to adjust the volume flow and to provide a more uniform airflow.
Ceiling diffusers are usually made of aluminum-coated heavy-gauge steel, extruded aluminum,
or teak wood. In addition to high induction ratios, ceiling diffusers have a good surface effect and a
shorter throw and can supply air in all directions. Horizontally projected ceiling diffusers are suitable
for conditioned spaces with low headroom.
The performance data of a ceiling diffuser are defined by the following parameters:
 Neck velocity vneck, that is, mean air velocity at the neck of the ceiling diffuser, fpm (m/ s)
 Volume flow rate s  Ak vs, in cfm (L / s); area factor Ak, in ft2 (m2), is determined at the testing
laboratory, and the average supply velocity vs is usually measured at a specified position at the
outlet by a deflecting vane anemometer, often called a velometer
 Total pressure loss pt, in in. WC (Pa)
 Throw, ft (m)
 Noise criteria curve
Slot Diffusers
A slot diffuser consists of a plenum box with single or multiple slots and air deflecting vanes, as
shown in Fig. 18.9. The slots are available in 0.5-in. (13-mm), 0.75-in. (19-mm), 1-in. (25-mm),
and 1.5-in. (38-mm) widths. The 0.75- and 1-in. slot widths are most commonly used. Slots are typically
2, 3, 4, 5, or 6 ft (600, 900, 1200, 1500, or 1800 mm) long. Air discharged from a slot diffuser
can be projected horizontally or vertically. With a single-slot diffuser, air is always discharged
in one direction. With multiple slots, air can be horizontally discharged either left or right, or a
combination of both, or one slot can discharge vertically while another discharges horizontally. Slot
diffusers are often called T-bar diffusers, as they are compatible with standard ceiling T bars.
Air enters the plenum box from the round inlet at one side. The function of the plenum box is to
distribute the air more evenly at the slot. The plenum box is sometimes insulated internally. After
passing the deflecting vanes, the air is discharged from the slot horizontally (left or right) or vertically
according to the required airflow pattern. Figure 18.9 shows the airstreams discharged from
slot diffusers in various directions. Different manufacturers use their own patented air deflecting
vanes. Deflecting vanes can also be used to block the airflow entirely if necessary.
Extruded aluminum is widely used as the material for the structural member of the slot, and a
galvanized sheet is used for the plenum box. The performance of a slot diffuser is defined by the
following parameters:
 Width of slot, number of slots, and length of slots
V?
3
16
V?
V?
18.14 CHAPTER EIGHTEEN
AIR SYSTEMS: SPACE AIR DIFFUSION 18.15
Air
deflecting
vanes
Slot
width
Inlet
(a)
(b)
(c)
(d)
Innerlined
material
Plenum
box
Ceiling
FIGURE 18.9 Slot diffusers: (a) horizontal left; (b)
horizontal right; (c) vertical; (d ) two-slot, two-way.
 The volume flow rate per ft (m) of slot, or slot intensity, in cfm / ft (L / sm)
 Throw, total pressure loss, and noise criteria curve
Table 18.1 lists the performance data of a typical slot diffuser of 0.75-in. (19-mm) slot width,
taken from a manufacturer’s catalog.
Nozzles
A nozzle is a round supply outlet (see Fig. 18.10). The airstream discharged from a nozzle is contracted
just before the outlet, which results in a higher flow velocity and more even distribution. The
purpose of using a nozzle instead of other types of supply outlets is to provide a longer throw and a
smaller spread. Round nozzles (Fig. 18.10a) are usually used in air diffusion systems with large
18.16 CHAPTER EIGHTEEN
TABLE 18.1 Performance Data for Typical 0.75-in. (19-mn) Width Slot Diffuser with Plenum
Total H .004 .015 .032 .058 .091 .125 .175 .230
Slots pressure,
in.WC V .003 .011 .024 .044 .067 .095 .125 .170
Cfm/Ft 6 12 18 24 30 36 42 48
H 1,1,3 1,3,9 3,6,11 5,9,13 8,10,15 9,11,16 10,12,17 11,13,19
1 Throw, ft
V 2 6 10 12 14 15 16 17
NC — — — 18 24 30 35 39
Cfm/Ft 12 24 36 48 60 72 84 96
H 1,1,5 2,5,13 5,10,16 9,13,19 11,15,21 13,16,23 14,17,25 15,19,27
2 Throw, ft
V 4 9 14 17 19 21 23 24
NC — — 13 23 29 35 40 44
Cfm/Ft 18 36 54 72 90 108 126 144
H 1,2,8 4,8,16 7,12,20 11,16,23 14,18,26 16,20,28 18,22,30 19,23,32
3 Throw, ft
V 6 11 17 21 23 26 28 30
NC — — 16 26 32 38 43 47
Source: TITUS. Reprinted by permission.
6 in. (150 mm) 6 in. (150 mm)
Ceiling
Supply
air 40 F
(4.4 C)
Induced air
(b)
FIGURE 18.10 Round nozzle and nozzle diffuser (a) Round nozzle and (b) nozzle diffuser.
space volume or in spot cooling / heating, which is covered in detail in Sec. 18.9. Nozzle diffusers
(Fig. 18.10b) have a high induction ratio and are used in cold air distribution, which is discussed in
Sec. 18.5.
Accessories for Supply Outlets
Grilles, registers, and diffusers may require certain accessories to modulate the volume flow rate of
each supply outlet and to distribute the air more evenly at the outlet. Currently used accessories include
 Opposed-blade dampers. Usually they are attached to the register upstream from the vanes, or
installed at the outlet collar leading to a ceiling diffuser, as shown in Fig. 18.11a.
 Split dampers. A piece of sheet metal is hinged at one end to the duct wall, as shown in
Fig. 18.11b. An equalizing grid is often used at the same time to distribute the supply air more
evenly to the outlets. A split damper is also often installed at the junction of two main ducts, to
adjust the volume flow rates of the supply air between these ducts.
 Gang-operated turning vanes. These vanes are always installed before the takeoff to the outlet, as
shown in Fig. 18.11c. The tips of the turning vanes through which air enters the takeoff remain parallel
to the direction of the airflow. Gang-operated turning vanes are especially suitable for
adjusting the airflow rate for registers that are mounted on the air duct with a very short outlet
collar.
Return and Exhaust Inlets
Various types of return inlets are used to return space air to the air-handling unit or packaged unit in
the fan room. Exhaust air inlets are usually connected to the exhaust duct. Before contaminated air
is exhausted outside, it is usually cleaned and treated with special air cleaning equipment.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.17
FIGURE 18.11 Accessories for supply outlets: (a) opposed-blade damper;
(b) split damper with equalizing device; (c) gang-operated turning vanes.
Return and exhaust inlets can be classified as return or exhaust grilles, return or exhaust registers,
return slots, and ventilated light troffers.
Return grilles and registers (see Fig. 18.12) are similar in shape and construction to supply
grilles and registers. In Fig. 18.12a, only one set of vanes is required to direct the air into the return
duct, so vision-proofed curved vanes are often used instead of the straight vanes. Vanes can be fixed
or adjustable. An opposed-blade damper is sometimes installed behind the curved vanes in a return
register, as shown in Fig. 18.12b. Wire-mesh grilles, as shown in Fig. 18.12c, are simple and inexpensive.
They are often installed in places where the appearance of the grille is not important. In
many industrial buildings, floor grilles are used when the return duct is below the floor level, as
shown in Fig. 18.12d. When room air is required to return to the fan room through a corridor and a
return duct, a door louver is often used to transport return air through the door. Vision-proofed,
right-angle louvers are generally used for this purpose, as shown in Fig. 18.12e. Local codes should
be consulted before the door louvers are installed.
When a ceiling plenum is used as a return plenum (as in many commercial buildings), return
slots are used to draw the return air through the ceiling. Return slots are similar to slot diffusers.
They are usually made with single or double slots, as shown in Fig. 18.13. The width of the slot
18.18 CHAPTER EIGHTEEN
FIGURE 18.12 Various types of return grilles and registers: (a) return grille with curved vanes; (b) return
register; (c) wire-mesh grille; (d ) floor-type return grille; (e) door louvers.
varies from 0.5 to 1.75 in. (13 to 45 mm). Slots are usually available in lengths of 2, 2.5, 3, 4, and
5 ft (600, 750, 900, 1200, and 1500 mm).
The capacity of return slots is also expressed in cfm / ft (L / s m) of slot length. If a return
slot has the same configuration, slot width, and slot intensity as the supply slot diffuser, then its total
pressure loss, in in. WC (Pa), is nearly the same as the vertical projection total pressure loss of
the supply slot diffuser. Return slots are usually inner-lined with sound absorption material. This
material prevents sound transmission between two rooms via the return slots and the ceiling
plenum.
Exhaust inlets are similar to the return inlets but they often have a simpler design.
In return and exhaust inlets, air velocity decreases very sharply as the distance from the surface
of the inlet increases. This phenomenon is the major difference in airflow characteristics between a
return or an exhaust inlet and a supply outlet.
Light Troffer Diffuser and Troffer-Diffuser Slot
Light troffer diffuser combines a fluorescent light troffer and slot diffuser, sometimes in saddle
type, as shown in Fig. 18.14a. The slot can be used as a supply outlet or return inlet. A troffer-diffuser
slot is a combination of light troffer, slot diffuser, and return slot, as shown in Fig. 18.14b.
Light troffer diffusers and troffer-diffuser slots serve two main purposes:
1. To maintain a lower air temperature in the light troffer, which increases the luminous efficiency
of fluorescent lamps
2. To form an integrated layout of light troffer, diffuser, and return slots on suspended ceilings
In addition to these benefits, a combination of a light troffer and return slot also reduces the space
cooling load because return air absorbs part of the heat released by the fluorescent lights.
In a light troffer diffuser, surface slots are often used. Air can be discharged horizontally or vertically.
Troffers and slots are available in 2-, 3-, and 4-ft ( 600-, 900-, and 1200-mm) lengths.
To prevent the deposit of dust particles on the surface of the fluorescent tube, the light troffer
should be designed so that return air does not come in direct contact with the fluorescent tube. The
result is a high ventilation efficiency and a neat appearance.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.19
FIGURE 18.13 Return slots: (a) single-slot; (b) double-slot.
18.4 MIXING FLOW
Airflow Pattern and Space Air Diffusion
Airflow pattern determines the performance of space air diffusion in the occupied zone of commercial
buildings or in the working area of a factory. The optimum airflow pattern for an occupied zone
depends mainly on the indoor temperature, relative humidity, and indoor air quality requirements;
the outdoor air supply; and the characteristics of the building. Four types of airflow patterns are currently
used in air conditioned space:
1. Mixing flow
2. Displacement flow
3. Projecting flow
4. Upward flow
Principles and Characteristics of Mixing Flow
The airflow pattern of an air conditioned space is said to be in mixing flow only when the supply air
is thoroughly, or nearly thoroughly, mixed with the ambient air, and the occupied zone or working
area is dominated by the induced recirculating flow.
18.20 CHAPTER EIGHTEEN
FIGURE 18.14 Light troffer, slot diffuser, and return slot combination:
(a) light troffer diffuser; (b) troffer-diffuser slot.
In many commercial and industrial buildings, cold supply air at 50 to 60°F (10 to 15.6°C) and
a velocity of 400 to 800 fpm (2 to 4 m / s) must be nearly thoroughly mixed with the ambient
air first. The temperature of the mixture is raised to 72 to 74°F (22.2 to 23.3°C), and its velocity
drops to less than 70 fpm (0.35 m / s) before it enters the occupied zone and working area, or drafts
occur. Mixing flow is often the best choice for comfort air conditioning systems when the supply
temperature differential is greater than 15°F (8.3°C) and the supply air velocity exceeds 300 fpm
(1.5 m / s).
Straub and Chen (1957) tested mixing flow patterns by using five types of supply outlets. These
airflow patterns have the following characteristics:
Induction of Space Air to the Air Jet. Mixing of induced air and the air jet reduces the differences
in air velocity and temperature between the air jet and the ambient air to acceptable limits.
Reverse Airstream in the Occupied Zone. The induced reverse airstream covers the occupied
zone with a more even temperature and velocity distribution.
Minimize the Stagnant Area in the Occupied Zone. A stagnant area is a zone with natural convective
currents. Air velocities in most of the stagnant zones are lower than 20 fpm (0.1 m / s). Air
stratifies into layers, with a significant temperature gradient from the bottom to the top of the stagnant
zone. If the recirculating flow fills the entire occupied zone, no space remains for stagnant air.
Types and Locations of Return and Exhaust Inlets. Because the return or exhaust airflow rate is
only a fraction of the maximum total airflow of the supply air jet or the maximum airflow of the reverse
airstream, the location of the return or exhaust inlet does not significantly affect the airflow
pattern of mixing flow in the occupied zone. However, the location of the return and exhaust inlets
does affect the thermal effectiveness factor 
T of space air diffusion. Therefore, for an induced
recirculating flow pattern, the return intake should be located near the end of the reverse airstream,
in the vicinity of a heat source, or within the stagnant zone.
For a conditioned space in which the space air is contaminated by dust particles, toxic gases, or
odors, several evenly distributed return or exhaust grilles should be installed. Collective return
grilles, i.e., one or two large return inlets instead of many inlets, impair indoor air quality. They are
only suitable under the following circumstances:
 When the conditioned space is small
 When it is difficult to provide a return duct
 When lower initial cost is of prime importance
Types and Locations of Supply Outlets. The type and the location of supply outlets affect the
mixing flow pattern and its performance. Currently, mixing airflow incorporates the following types
of supply outlets:
 High side outlets
 Ceiling diffusers
 Slot diffusers
 Sill and floor outlets
 Outlets from stratified mixing flow
Mixing Flow Using High Side Outlets
A high side outlet may be a sidewall outlet mounted at the high level of a conditioned space, or an
outlet mounted directly on the supply duct. Figure 18.15 shows a mixing flow using a high sidewall
AIR SYSTEMS: SPACE AIR DIFFUSION 18.21
outlet. In Fig. 18.15, the shaded black envelope shows the primary air envelope of the air jet, and
the dotted envelope shows the total air envelope.
As the air jet discharges from the high sidewall outlet, the surface effect tends to keep the air jet
in contact with the ceiling. During cooling, the air jet induces the ambient air and deflects downward
when it strikes on the opposite wall. As soon as the maximum air velocity of the air jet drops
to about 50 fpm (0.25 m / s), the air jet terminates as it flows along the opposite wall. The induction
of space air from the occupied zone into the air jet forms the reverse airstream and fills the occupied
zone.
If the throw of a high sidewall outlet is longer than the sum of the length of the room and the
height of the opposite wall, then the air jet is deflected by the opposite wall and floor and enters the
occupied zone with excessive air velocity. Both the air jet and the reverse airstream fill the occupied
zone with a higher air velocity and greater temperature difference.
If the throw is too short, the air jet drops and may enter the occupied zone directly. This causes
higher air velocity and lower temperature (cold draft) in the occupied zone.
During heating, warm air tends to rise and results in a shorter throw. As the induced airstream
rises, a stagnant zone may form between the floor and the induced airstream. A warm air jet with
sufficient velocity and a longer throw can reduce or even eliminate the stagnant zone.
If the airstream at the terminal zone of the warm air jet comes in contact with the cold inner surface
of the window, it converts to a cold airstream and forms a cold draft when it flows downward and
enters the occupied zone. The installation of a baseboard heater under the window sill can prevent the
formation of this cold draft and reduce the vertical temperature gradient in the occupied zone.
The most suitable location for a return inlet is on the ceiling outside the air jet. Return air flows
through the return inlets and often uses the ceiling plenum as a return plenum. Such an arrangement
often results in a higher effectiveness factor 
T and a lower installation cost for the return system.
For a room without a suspended ceiling, a cold or warm air jet has a shorter throw because of
lack of surface effect. The buoyancy force causes a cold air jet to drop sooner, and it may enter the
occupied zone directly. The higher rise of a warm air jet forms a greater stagnant zone above the
floor level.
18.22 CHAPTER EIGHTEEN
FIGURE 18.15 Mixing flow using high side outlets. (Adapted with permission from ASHRAE Handbook 1989,
Fundamentals.)
Mixing flow using high side outlets usually gives a longer throw and a higher air velocity in the
occupied zone than other supply outlets. Drop must be checked to prevent the air jet from entering
the occupied zone directly in a large conditioned space.
Mixing Flow Using Ceiling Diffusers
The airflow pattern of mixing flow using ceiling diffusers is similar to that of high side outlets.
Ceiling diffusers produce a better surface effect, a shorter throw, a lower and more even distribution
of air velocity, and a more even temperature in the occupied zone.
Figure 18.16 shows mixing flow using ceiling diffusers. During cooling, the cold air jet usually
maintains contact with the ceiling, producing a better surface effect than sidewall outlets. The reverse
airstream fills the occupied zone. During heating, thinner vertical spread and shorter throw
create a larger, higher stagnant zone. The best location for return inlets is on the ceiling within the
terminal zone of the air jet.
When two air jets discharge horizontally from ceiling diffusers or other outlets in opposite directions
along the ceiling, with a ceiling height of 10 ft (3.05 m), the air velocity where two air jets
collide should be less than 120 fpm (0.6 m / s) in order to maintain an air velocity below 50 fpm
(0.25 m / s) 5 ft (1.5 m) above the floor. Mixing flow using ceiling diffusers is widely used for conditioned
spaces with limited ceiling height. Many ceiling diffusers are designed to have a high
AIR SYSTEMS: SPACE AIR DIFFUSION 18.23
FIGURE 18.16 Mixing flow using ceiling diffusers. (Adapted with permission from ASHRAE Handbook 1989,
Fundamentals.)
induction ratio and to produce a very good surface effect. These ceiling diffusers are often used in
variable-air-volume systems.
Mixing Flow Using Slot Diffusers
Slot diffusers are narrower and longer than ceiling diffusers and grilles. When supply air is discharged
horizontally from a ceiling slot diffuser, it has a thinner vertical spread and an excellent
surface effect. Figure 18.17 shows mixing flow using slot diffusers discharging vertically downward
into the perimeter zone and also discharging horizontally along the ceiling in both the perimeter and
interior zones of a large office.
Because of its superior surface effect, the horizontally discharged cold air jet remains in contact
with the suspended ceiling even when the supply air volume flow rate is reduced in a variable-airvolume
system. The occupied zone is then filled with the reverse airstream at a more uniform temperature
and air velocity.
In order to counteract natural convection along the inner surface of a window in the perimeter
zone, the air jet is often projected downward toward the window. A sufficient throw is important to
produce a downward air jet that offsets the cold draft at the inner surface of the window during winter
heating. The occupied zone fills with induced airflow at an air velocity lower than 50 fpm (0.25
m/ s).
For a mixing flow using slot diffusers, the location of return slot is preferably aligned with the
supply slots.
In addition to their excellent surface effect in mixing flow, slot diffusers have a linear appearance
that can be coordinated easily with electric lights and ceiling modular arrangements. Therefore,
they are widely used in buildings using VAV systems with moderate loads and normal ceiling
height.
Mixing Flow Using Sill or Floor Outlets
The purpose of using a sill or floor outlet is to counteract the cold draft flowing downward along the
inner surface of the window when the outdoor temperature is below 30°F ( 1.1°C). Figure 18.18
shows a mixing flow using a sill outlet.
During cooling, a stagnant zone may form above the eddies beyond the terminal zone of the air
jet if the supply air velocity is not high enough. Meanwhile the space under the cold air jet is filled
18.24 CHAPTER EIGHTEEN
FIGURE 18.17 Mixing flow using slot diffusers: (a) perimeter and interior zone, cooling; (b) perimeter zone,
heating.
with reverse airstreams and cooled. During heating, a stagnant zone may form below the ascending
induced airstream.
Mixing flow with sill or floor outlets has been widely used in buildings with large window areas
or in raised-floor offices. The direction and amount of airflow from floor diffusers usually can be
adjusted according to the requirements of the occupants.
Stratified Mixing Flow
In a building with a high ceiling, it is more economical to stratify the building vertically into two
zones (the stratified upper zone and the cooled lower zone) or three zones (upper, transition, and
lower zones) during cooling. The upper boundary of the lower zone is at the level of the supply outlet
where the air jet projects horizontally.
Gorton and Sassi (1982) and Bagheri and Gorton (1987) performed a series of model studies
and experiments about stratified space air diffusion. Figure 18.19a shows the elevation view of the
nuclear reactor facility in which the experiments were conducted. The ceiling height was 41 ft
(12.5 m). An air-handling unit was located at floor level, and two diffusers were mounted on
the supply riser 16 ft (4.9 m) above the floor level. The supply velocity at the outlet was about
1000 fpm (5 m/ s), and the supply temperature was around 60°F (15.6°C).
Figure 18.19b shows the space air temperature profile at 4 p.m. at various height levels during
cooling. This temperature profile could be divided into
 An upper zone, from the ceiling down to about 22 ft (6.7 m) from the floor level, in which air
temperature varied from 82.5 to 79°F (28 to 26.1°C)
 A transition zone, between 16 and 22 ft (4.9 and 6.7 m) from the floor, in which air temperature
varied from 79 to 74.5°F (26.1 to 23.6°C)
AIR SYSTEMS: SPACE AIR DIFFUSION 18.25
FIGURE 18.18 Mixing flow using sill outlet.(Adapted with permission from
ASHRAE Handbook 1989, Fundamentals.)
18.26 CHAPTER EIGHTEEN
FIGURE 18.19 Stratified mixing flow in a nuclear reactor
facility: (a) elevation view of a nuclear reactor facility;
(b) temperature profile during cooling; (c) temperature pro-
file during heating.
 A lower zone below 16 ft (4.9 m), in which air temperature varied between 72 and 72.5°F (22.2
and 22.5°C)
The induced recirculating flow under the cold air jet formed the lower zone. Most of the lower
zone was filled with reversed airstreams. The upper zone and the transition zone were formed because
of the entered heat gains at the upper level and a weaker recirculating induced flow above the
cold air jet. Because of the drop of the cold air jet, a greater amount of recirculating flow was induced
in the cold air jet in the lower zone than in the upper zone. Consequently, most of the cooling
occurred in the lower zone. The formation of a stagnant layer near the ceiling also retarded airflow
and heat transfer between outdoor and indoor air in the upper zone.
Figure 18.19c shows the space temperature profile at various height levels at 4 p.m. during heating.
The supply air temperature was about 85°F (29.5°C) and the outside temperature 40°F (4.4°C). The
air temperature varied from 80 to 81.5°F (26.7 to 27.5°C) from a height of 0 to 36 ft (0 to 11.0 m)
from floor level. There was no evidence of significant thermal stratification.
Figure 18.20 shows a stratified mixing flow in a large, high-ceiling indoor stadium using supply
nozzles. During summer cooling, the cold air jet creates two induced recirculating airflows: a
cooled lower occupied zone and a stratified upper zone. If a higher air velocity is required in the
lower occupied zone (the spectator’s area), the supply nozzles should project at an inclined angle.
Otherwise, the cold air jet should project horizontally.
During heating, if the throw of the warm air jet is long enough to arrive at the lowest spectator
seat, the whole spectator area is filled with the reversed airstream. No stagnant zone is formed in
the lower occupied zone. Return inlets should be located in the lower cooling zone. They should be
evenly distributed under the spectator seats.
Characteristics of the stratified mixing flow during summer cooling mode operation are as follows:
 Convective heat transfer from the hot roof is effectively blocked by the higher-temperature or
stagnant air layer in the upper zone.
 Cooling loads that occur in the lower zone (windows, walls, lights, occupants, and equipment)
should be offset by cold supply air.
 Radiant heat from the roof, upper external walls, and electric lights in the upper zone enters the
occupied zone and converts to cooling load.
 Although supply airflow rate and supply temperature affect the throw and drop of the air jet, the
induced recirculating airflow pattern and the stratified upper and lower zones remain the same.
 The height of the supply air jet determines the upper boundary of the lower zone.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.27
FIGURE 18.20 Stratified mixing flow in a large indoor stadium using supply nozzles.
 The location of return inlets influences the cooling load only when they are located in the upper
or transition zone. A portion of exhaust air should be extracted at a higher level.
18.5 COLD AIR DISTRIBUTION
Cold Air Distribution Versus Conventional Air Distribution
Because of the lower chilled water discharge temperature of 34 to 38°F (1.1 to 2.2°C) from ice
storage systems, the supply air temperature of the air system can often be reduced to 40 to 45°F
(4.4 to 7.2°F) and results in a cold air distribution (low-temperature distribution) system instead of
a conventional system of 55°F (12.8°F) supply temperature. Dorgan and Elleson (1993) and
Scofield (1993) summarized that a cold air distribution system has the following advantages when
compared with a conventional space air diffusion system:
 Due to the low dew point of supply air, space relative humidity can be maintained between 35 and
45 percent. Cold air distribution can increase the space temperature slightly without causing discomfort.
 For a cold air distribution system using fan-powered VAV boxes, Ts is equal to the air temperature
discharged from the fan-powered unit during summer cooling mode operation if the duct heat
gain between the fan-powered VAV box and the supply outlet is ignored. If Ts  44°F (6.7°C) and
space air temperature Tr  78°F (25.5°C), then the supply temperature differential for cold air
distribution Tr  Ts  78  44  34°F (18.9°C). Compared with a supply temperature differential
Tr  Ts  20°F for a conventional system, cold air distribution systems reduce the design
supply volume flow rate by 44 percent.
 With cold air distribution, fans, ducts, and the associated air system components can be downsized.
Smaller ducts require less space clearance between the ceiling and structural members, and
thus a reduction of floor-to-floor height in high-rise buildings. Also there is a significant increase
in rental space because of the reduction of the mechanical rooms and duct shafts.
 Fan energy use can save as much as 40 percent.
 Savings are seen in construction cost because of the reduction of the structural height and smaller
ductwork and air system components (sometimes it is the primary consideration).
 There are reduced fan sound levels and, therefore, less sound attenuation for a specified space NC
rating.
 Cold air distribution is often beneficial for a retrofit project in which the existing air distribution
system may be used for increased space cooling loads.
The major reservation of using cold air distribution, especially for VAV systems in large buildings,
is the justification of a reduction of 40 to 45 percent of supply air at a time of increased concern
over IAQ and the dumping of the cold air jet at reduced volume flow.
High Induction Nozzle Diffusers
Two types of space diffusion systems are used in cold air distribution: (1) Cold supply air from the
air-handling unit or packaged unit is supplied directly to the conditioned space by means of high
induction diffusers; (2) fan-powered VAV boxes are used to mix low-temperature supply air with return
air, and the mixture at a temperature between 55 and 60°F (13 and 16°C) is then supplied to
the conditioned space.
18.28 CHAPTER EIGHTEEN
Fields and Knebel (1991) and Knebel and John (1993) introduced a high induction nozzle
diffuser used in cold air distribution, as shown in Fig. 18.10b. A typical 4-ft- (1.22-m-) long high
induction nozzle diffuser has 57 0.75-in. (19-mm) nozzles on each side of the diffuser. At an airflow
rate of 350 cfm (165 L / s), it has a supply air velocity of about 1000 fpm (5 m / s), a total pressure
loss at the supply outlet of 0.213 in. WC (53 Pa), a throw T50  33 ft (T0.25  10 m), and a drop of
about 0.8 ft (0.24 m).
Characteristics of Cold Air Distribution Systems
A satisfactory cold air distribution system often has the following characteristics:
 A higher value, as defined in Sec. 18.2, to provide adequate surrounding spaces for the
supply air jets to induce sufficient ambient air and fully mix the ambient air with the 40 to 45°F
(4.4 to 7.2°C) low-temperature supply air
 A higher supply air velocity and jet turbulence (depending on the configuration of the air diffusion
system) to provide momentum for the induced airstreams
 An excellent surface effect with adequate throw and small drop to prevent supply air jet from entering
the occupied zone directly
 An ADPI 80 percent at both design and reduced airflow
Case Study—Florida Elementary School
According to Knebel and John (1993), in a Florida elementary school, the HVAC&R system consists
of a 40,000 cfm (18,880 L / s) blow-through AHU with airfoil fan and an ice storage system.
The DDC controlled VAV terminals are used to throttle the airflow and reset the supply air temperature.
The classroom that uses cold air distribution is 27 ft (8.3 m) wide by 36 ft (11.0 m) long.
Three 4-ft nozzle-diffusers, each with 114 nozzles, were set in one row, designed to produce a
throw of 18 ft (5.4 m) on each side. Each nozzle-diffuser had a design airflow rate of 200 cfm (94
L/ s) and a supply velocity of 500 fpm (2.5 m / s) at the nozzle outlet. With this velocity, the 50 fpm
(0.25 m / s) air jet envelope was 30 in. (0.76 m) from the nozzle. The air motion in the space was between
0 and 10 fpm (0 and 0.05 m / s).
The diffuser supplier modified the number of active nozzles in each diffuser. The supply velocity
at the nozzle outlets was raised to 2000 fpm (10 m / s) at the design airflow and to 1000 fpm
(5 m/ s) at 50 percent part-load operation. For an outlet velocity of 2000 fpm (10 m / s) and a supply
air temperature of 40°F (4.4°C), space air velocity had been raised to 15 to 45 fpm (0.075 to
0.23 m/ s) and the measured ADPI was 100 percent. At part-load operation, when the outlet velocity
dropped to 1000 fpm (5 m / s) with an supply air temperature of 41°F (5°C), the space air velocity
varied from 10 to 40 fpm (0.05 to 0.20 m / s) and the measured ADPI was 96.2 percent. The
estimated value for this cold air distribution system in this classroom is about 37.
Performance of Ceiling and Slot Diffusers
Zhang et al. (1994) conducted ventilation performance tests for two square diffusers (three concentric
inner cones discharged in all directions, which is similar to the ceiling diffuser shown in Fig. 18.8b)
and one-slot diffusers (horizontally projected) using 38°F (3.3°C) supply air for cold air distribution
and 55°F (12.8°C) supply air for conventional air distribution. Test results were as follows:
 Cold and conventional air distribution systems performed equally well in delivering the outdoor
ventilation air into the occupied zone.
?Ar /Do
?Ar /Do
AIR SYSTEMS: SPACE AIR DIFFUSION 18.29
 The average air exchange effectiveness (AEE) of the small square diffuser using cold air distribution
was 1.01, and using conventional air distribution was 0.99. The average AEE for the slot
diffuser using cold air distribution was 0.99, and using conventional air distribution was 1.00. The
high magnitudes of AEE (  0.90) showed that almost all the supply air was delivered into the
occupied zone.
 The ADPI for the large square diffuser using cold air distribution was 92.5 percent, for the small
square diffuser was 96.7 percent, and for the slot diffuser was 96.7 percent. The cold air distribution
system for the slot diffuser has a value of 22. The ADPI for small square diffuser using
conventional air distribution was 96.7 percent and for the slot diffuser was 90.0 percent.
 Cold air distribution with these three diffusers provided satisfactory ventilation performance.
There was no appreciable difference between the cold and conventional air distribution
systems.
Cold Air Distribution with Fan-Powered VAV Boxes
There are two types of fan-powered VAV box used in cold air distribution: parallel fan-powered box
and series fan-powered box. In a parallel fan-powered box, the cold primary air from the AHU does
not flow through the VAV box fan but instead mixes with the recirculating air after the box fan outlet,
whereas in a series fan-powered box, the cold primary air flows through the box fan. Parallel
and series fan-powered boxes are discussed in Sec. 21.5. For 44°F (6.7°C) cold primary air and
78°F (25.6°C) ceiling plenum air, 40 percent of 78°F (25.6°C) return air is often extracted to the
fan-powered box and is mixed with 60 percent of 44°F (6.7°C) cold primary air from the AHU, resulting
at a 58°F (14.4°C) mixture, the same as in the conventional air distribution system supplied
to the space. A fan-powered box has the benefits of providing space air motion.
Elleson (1993) compared the energy use with and without the fan-powered mixing boxes in cold
air distribution. In general, the energy required for continuous mixing box fan operation is greater
than the supply fan energy savings resulting from the reduced supply temperature.
Box fan energy use can be reduced by operating the box fan only when space air motion falls
below a minimum level. Supply air temperature reset can further reduce fan energy consumption.
Direct supply of cold primary air to the conditioned space entails the lowest fan energy use. Furthermore,
compared to the direct supply of cold primary air, cold air distribution using fan-powered
boxes has a higher noise level and initial cost and requires more site maintenance. Direct supply of
low-temperature cold primary air to the conditioned space is recommended.
Surface Condensation
For a space air temperature of 75°F (23.9°C) and a relative humidity of 50 percent, the dew point of
space air is 55°F (12.8°C). During cold air distribution, the space relative humidity can be reduced
to 40 percent, and the dew point is then about 49.5°F (9.7°C). For a cold air distribution with a supply
air of 40°F (4.4°C), the temperature differential between the supply air inside the ducts and the
ambient air increased to 75  40  35°F (19.4°C) instead of 75  55  20°F (11.1°C) in a conventional
air distribution system. Harmon and Yu (1993) reported that the actual space relative
humidity levels measured were 3 to 5 percent above those projected by calculations. In cold air distribution
systems, in order to prevent surface condensation, the AHU or PU, ducts, duct access
doors, VAV boxes, flexible ducts, and diffusers should be insulated with a layer of fiberglass protected
by a vapor barrier, or a closed-cell elastomeric or cellular glass insulation, with a thickness of
2 in. (50 mm). It is preferable to select VAV boxes with thermally isolated inlet connections and
properly insulated diffusers.
Plenum space must be properly sealed to prevent outside air infiltration. Gradually cool the supply
air and the space air during each early morning in summer to form a soft start, so that the outside
surface will not suddenly be cooled below the dew point of the ambient air.
?Ar /Do
18.30 CHAPTER EIGHTEEN
18.6 DESIGN PROCEDURE OF MIXING FLOW AIR
DIFFUSION SYSTEM
The supply volume flow rates of various control zones at summer and winter design conditions are
usually determined before a mixing flow air diffusion system is designed. Because cooling mode
operations are used in both perimeter and interior zones in summer and often in interior zones even
in winter, heating mode operations often occur only in the perimeter zone during winter. Therefore,
cooling mode operation is the basic consideration in space air diffusion design. In the perimeter
zone, heating mode space air diffusion must be considered, and a compromise may be necessary.
The design procedure for such an air diffusion system follows.
Select the Type of Supply Outlet
Selection of the supply outlet depends on the following:
 Requirements of indoor environmental control. If the conditioned space needs controlled air
movements, such as for a badminton and table tennis tournament, or precise air temperature control,
a high side outlet is not the right choice.
 Shape, size, and ceiling height of the building. For buildings with limited ceiling height, ceiling
and slot diffusers are often the best choice. For large buildings with high ceilings, high side outlets
mounted at high levels to form stratified induced recirculating flow patterns are recommended.
In a perimeter zone, an overhead two-way slot diffuser projected down toward the window
and horizontally projected to the room, a ceiling diffuser with a throw to the inner surface of
the window glass, or a sill outlet should be used.
 Surface effect. A good surface effect is especially important to the VAV system because it allows
the supply volume flow rate to be reduced to half or even 30 percent of the design flow.
 Volume flow per ft2 of floor area. Sidewall outlets are limited to a lower volume flow per ft2 of
floor area, in cfm / ft2 (L/sm2), because of the higher air velocity in the occupied zone. The slot
diffuser has a narrower slot width and can only project in one or two directions. Therefore, the
volume flow per ft2 for a slot diffuser is smaller than that of a ceiling diffuser. Table 18.2 lists the
volume flow per ft2 for various types of supply outlet recommended by ASHRAE Handbook 1992,
HVAC Systems and Equipment.
 Appearance. The shape and configuration of outlets and inlets are closely related to the interior
appearance of the building and should be coordinated with inlets and lighting troffers.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.31
TABLE 18.2 Volume Flow per ft2 Floor Area
Approximate
maximum
air changes
Supply air density, per hour for
Type of outlet cfm/ ft2 of floor space 10-ft (3-m) ceiling
Grille 0.6–1.2 7
Slot diffuser 0.8–2.0 12
Perforated panel 0.9–3.0 18
Ceiling diffuser 0.9–5.0 30
Ventilating ceiling 1.0–10.0 60
Source: ASHRAE Handbook 1992, HVAC Systems and Equipment.
Reprinted with permission.
 Cost. In many commercial buildings, cost is often an important factor in determining the type of
supply outlet.
Volume Flow Rate per Outlet or per Unit Length
For a specific supply volume flow rate in a conditioned space, the volume flow rate per supply
outlet s,out, in cfm (L / s), determines the number of outlets in the conditioned space. As defined in
Eq. (18.10), both s,out and supply outlet velocity vs,o or vcore, in fpm (m / s), affect the throw of the
supply airstream. However, s,out has a significantly greater influence than vs,o, especially for slot
diffusers.
The volume flow rate per supply outlet depends mainly on the throw required to provide a satisfactory
space air diffusion system design. Load density, space air diffusion system characteristics,
sound control requirements, and cost considerations are also factors that determine s,out. In a VAV
system in a high-rise office building, avoiding the discharge of cold air into the space directly at a
reduced volume flow rate is one of the primary considerations.
For a space air diffusion system using slot diffusers, the supply volume flow per ft (m) length of
slot diffuser (the slot intensity), in cfm / ft (L / s m), is often an important index, especially in the
return slots. It is usually 15 to 40 cfm / ft (23 to 62 L / s m). All return slots mounted on the same
V?
V?
V?
V?
18.32 CHAPTER EIGHTEEN
TABLE 18.3 Relationship between ADPI and T50 /L and T100 /L
Terminal Load density, T50 /L Maximum For ADPI Range of
device Btu/h ft2 for max. ADPI ADPI greater than T50 /L
High sidewall grilles 80 1.8 68 — —
60 1.8 72 70 1.5–2.2
40 1.6 78 70 1.2–2.3
20 1.5 85 80 1.0–1.9
Circular ceiling diffusers 80 0.8 76 70 0.7–1.3
60 0.8 83 80 0.7–1.2
40 0.8 88 80 0.5–1.5
20 0.8 93 90 0.7–1.3
Sill grille straight vanes 80 1.7 61 60 1.5–1.7
60 1.7 72 70 1.4–1.7
40 1.3 86 80 1.2–1.8
20 0.9 95 90 0.8–1.3
Sill grille spread vanes 80 0.7 94 90 0.8–1.5
60 0.7 94 80 0.6–1.7
40 0.7 94 — —
20 0.7 94 — —
Slot diffusers (for T100 /L) 80 0.3 85 80 0.3–0.7
60 0.3 88 80 0.3–0.8
40 0.3 91 80 0.3–1.1
20 0.3 92 80 0.3–1.5
Light troffer diffusers 60 2.5 86 80 	 3.8
40 1.0 92 90 	 3.0
20 1.0 95 90 	 4.5
Perforated and louvered
ceiling diffusers 35–160 2.0 96 90 1.4–2.7
80 1.0–3.4
Source: Adapted with permission from ASHRAE Handbook 1989, Fundamentals. Reprinted with permission.
return ceiling plenum must have the same total pressure loss, or using same type of return slot, the
same slot intensity.
In a closed office of an area of 150 ft2 (14 m2) or less with only one side of external wall, usually
one ceiling diffuser is sufficient.
Choose an Optimum Throw–Characteristic Length Ratio
For most types of supply outlets, the selection of an optimum throw–characteristic length ratio
T50 /L (T0.25 /L) or T100 /L (T0.5 /L) determines the layout and grid of the supply outlets. An optimum
T/L ratio should meet the following requirements:
 Selected T50 /L or T100 /L values should have an ADPI greater than 70 and 80 percent, as listed in
Table 18.3.
 The spread at the end of the supply air jet covers or almost covers the width of the conditioned
space before the air jet enters the occupied zone as shown in Fig. 18.15.
Miller (1976) and others conducted many experiments and determined the relationship between
the ADPI and T50 /L or T100 /L. They found that the cooling load density is also a factor. Table 18.3
lists the relationship between ADPI and T50 /L or T100 /L at various load densities. The characteristic
length of a sill grille is defined as the length of the room or space in the direction of jet flow. In
Table 18.3, T50 and T100 (T0.25 and T0.5) are throws of isothermal jets given in most of the manufacturers’
catalogs for selection convenience.
Generally, high sidewall outlets have a longer throw and therefore a higher T50 /L (T0.25 /L) range
than do ceiling diffusers. Slot diffusers have a wider T100 /L (T0.5 /L) range in which ADPI exceeds
80 percent than do high sidewall outlets. This characteristic makes them suitable for use at reduced
volume flows. Square ceiling diffusers have a T50 /L (T0.25 /L) range similar to that of the circular
ones. Higher cooling load density usually results in a lower ADPI.
For VAV systems, Tv/L must be selected within the satisfactory range, that is, ADPI  80 percent,
for both maximum and minimum airflow.
Determine the Design Characteristics of Slot Diffusers in Perimeter Zone
In the perimeter zone, overhead slot diffusers and ceiling diffusers are widely used. The characteristic
length of an overhead slot diffuser in the perimeter zone has not been clearly defined. Based on
tests performed by Straub and Cooper (1991), Lorch and Straub (1983), and Int-Hout (1998), the
design characteristics of overhead slot diffusers in the perimeter zone are as follows:
 Two-slot, two-way overhead slot diffusers should be used. The suitable distance between the
overhead slot diffuser and the external wall is about 1 ft (0.3 m). One of the slots near the window
is set to blow down, and the other slot blows horizontally into the room. Such an arrangement offsets
the cold draft along the window in winter heating.
 During cooling mode operation, ADPI exceeding 80 percent occurs within a wider range of slot
intensities, load densities, and spacing between the diffuser and external wall than during heating
mode. Heating mode operation requires careful consideration.
 According to ASHRAE Standard 55-1992, the criteria for an acceptable space air diffusion
in heating mode operation are based on the requirement that the vertical temperature gradient in
the occupied zone measured between 4-in. (0.1-m) and 67-in. (1.7-m) levels not exceed 5°F
(3°C).
 According to ASHRAE Handbook 1997, Fundamentals, the supply temperature differential for
warm air during heating mode Ts  Tr should not be greater than 15°F (8.3°C), to prevent excessive
buoyancy.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.33
 A terminal velocity of 150 fpm (0.75 m / s) is recommended at the 5-ft (1.5-m) level near the window,
to offset cold drafts in winter.
 An evenly distributed air volume flow rate between the vertically and horizontally discharged
slots is recommended.
Select the Specific Supply Outlet from Manufacturer’s Catalog
After an optimum T50 /L (T0.25 /L) or T100 /L (T0.5 /L) and characteristic length L are determined from
the preliminary layout, the supply outlet can be selected from the manufacturer’s catalog from
known T50 (T0.25) or T100 (T0.5) and supply volume flow rates, after the following parameters have
been checked:
Sound Level. The combined sound level of terminal and outlet should be at least 3 dB lower than
the recommended NC criteria in the conditioned space. For optimum noise control, the recommended
air velocities at the supply outlet are as follows:
Residences, apartments, churches, 500 to 750 fpm (2.5 to 3.75 m / s)
hotel guest rooms, theaters, private offices
General offices 500 to 1250 fpm (2.5 to 6.25 m / s)
The outlet velocity for the ceiling diffuser can be calculated by dividing the volume flow by area
factor Ak, given in the manufacturer’s catalog.
Drop of Cold Air Jet. Drop of a cold air jet should be checked if the cold jet enters the occupied
zone directly. Figure 18.21a shows the drop and other data of typical sidewall outlets mounted
within 1 ft (0.3 m) of the ceiling with an adjustable vane deflection of vertical 5° up and horizontal
0°. Figure 18.21b shows the drop for sidewall outlets without surface effects and a vane deflection
of vertical 15° up and horizontal 0°. Both are cold air jets based on a supply temperature differential
of 20°F (11.1°C).
Total Pressure Loss of the Supply Outlet. The total pressure loss of supply air when it flows
through a slot diffuser with a slot width of 0.75 in. (19 mm) is usually between 0.05 and 0.20 in.
WC (12 and 50 Pa). For a ceiling diffuser, it is between 0.02 and 0.2 in. WC (5 and 50 Pa). A total
pressure loss higher than 0.20 in. WC (50 Pa) is unsatisfactory.
Determination of Final Layout of Supply Outlets and Return Inlets
The determination of the volume flow rate per outlet, the number of outlets, T50 /L (T0.25 /L), and the
selection of specific outlets from the manufacturer’s catalog is often an iteration procedure. After
that, final layout of supply outlets and return inlets can be determined.
In the perimeter zone, the high side wall outlet in recently constructed hotel guest rooms are often
located on the partition wall adjacent to the window. In the interior zone, ceiling and slot diffusers
are usually located directly above the center of the conditioned space.
As mentioned in Sec. 18.4, if a ceiling plenum is used as return plenum, the return inlets should
be located outside the supply air jet, above the return airstream, or near a concentrated heat source
for a better effectiveness factor 
.
The recommended face velocities for return inlets are as follows:
Above the occupied zone 800 to 1000 fpm (4 to 5 m / s)
Within the occupied zone 400 to 600 fpm (2 to 3 m / s)
Door louvers 300 to 500 fpm (1.5 to 2.5 m / s)
18.34 CHAPTER EIGHTEEN
Example 18.1. A large, open office has a perimeter length of external wall of 30 ft (9.1 m) and a
depth of 35 ft (10.7 m) of which 15 ft (4.5 m) is the perimeter zone, as shown in Fig. 18.22. The supply
air volume flow rates and load densities for both perimeter and interior zones are as listed below:
Design supply volume flow rate, cfm (L / s)
Perimeter zone 960 (453)
Interior zone 500 (236)
AIR SYSTEMS: SPACE AIR DIFFUSION 18.35
FIGURE 18.21 Performance of cold air jets based on a supply temperature differential of 20°F. (a) Performance
of cold air jets from typical grilles and registers mounted within 1 ft below ceiling; deflection: vertical 5°
up, horizontal, 0°. (b) Performance of free cold air jets from typical grilles and registers; deflection: vertical 15°
up, horizontal, 0°. (Source: Titus Products. Reprinted with permission.)
Supply temperature differential 15°F (8.3°C)
Sound level NC 40
Cooling load density in perimeter zone 50 Btu /h ft2 (158 W/m2)
Linear density of heating load 300 Btu /h ft (288 W/m)
Design a space air diffusion system for this office for year-round operation using a VAV system.
Its minimum supply volume flow rate is 50 percent of the design volume flow at part-load operation.
The ceiling plenum is used as a return plenum.
Solution
1. For a large, open office using a VAV system, a space air diffusion system of mixing airflow
pattern using ceiling slot diffusers is a suitable choice because of its good surface effect and wider
T100 /L (T0.5 /L) range.
2. For the perimeter zone, the overhead two-way two-slot diffuser should be located on the suspended
ceiling parallel to the external wall and window glass. One of the slots should discharge
downward toward the window glass. The space between the slot diffusers and the window glass is
1 ft (0.3 m). Another slot discharges horizontally inward from the window glass.
3. In Table 18.3 the range of T100 /L for slot diffusers having ADPI  80 percent at a load density
of 50 Btu /h ft2 is between 0.3 and 0.95. Considering the reduction of supply volume flow rate
at 50 percent load, select a T100 /L ratio around 0.7. Because the characteristic length L in the
perimeter zone is L  15  1  14 ft, the required throw for the slot diffuser in the perimeter zone
18.36 CHAPTER EIGHTEEN
FIGURE 18.22 Layout of slot diffusers for a large open office in Example 18.1. All dimensions in feet.
is T100 /L  0.7, or
14  9.8 ft
From Table 18.1, two-way two-slot diffusers with a length of 4 ft (1.2 m) and a slot width of 0.75
in. (19 mm) are selected. At a slot intensity of 30 cfm / ft (46 L / s m), its performance is as follows:
Throw
Horizontal projection, ft (m) 8-10-15 (2.4-3-4.6)
Vertical projection T50 (T0.25), ft(m) 14 (4.3)
Total pressure loss
Horizontal, in. WC (Pa) 0.091 (23)
Vertical, in. WC (Pa) 0.067 (17)
Sound level, NC curve 29
4. If the supply volume flow rate in the perimeter zone is evenly split between horizontal and
downward airstreams, the number of slot diffusers is 0.5(960) / (4  30)  4, so four slot diffusers
should be used. For the downward discharge slot, the vertical throw T150  0.5 T50  0.5  14 
7 ft, which is greater than the required throw T150 (ceiling height 9  distance from floor level 5) 
4 ft. The terminal velocity is higher than 150 fpm (0.75 m / s) near the window glass at a level of
5 ft from the floor level.
For the horizontally discharged slot, T100 is 10 ft. Because the slot is installed 1 ft away from the
window, T100 /L  (10 1)/15  0.73. From Table 18.3, at a load density of 50 Btu /h ft2, the
range of T100 /L ratio for ADPI  80 percent is between 0.3 and 0.95; T100 /L  0.73 is a suitable
setup.
5. Heating load in the perimeter zone is
If the supply volume flow in heating mode operation is still 960 cfm, and if the density of supply air
is 0.072 lb / ft3, then the supply differential can be calculated as
This value is far smaller than 15°F (8.3°C).
As the terminal velocity is approximately equal to 150 fpm (0.75 m / s) near the window glass at
a level 5 ft (1.5 m) from the floor level, and a far lower Ts  Ts, cold draft does not occur at the inner
glass surface at a heating load linear density of 300 Btu /h ft (288 W/m2).
6. Four 4-ft two-slot, of 0.75-in. slot width, return slots of the same configuration as the supply
slot diffusers with a slot intensity of 900 / (2  4  4)  28 cfm / ft (43 L / s m) are installed in the
perimeter zone. Their locations are arranged on two sides of the room in a direction perpendicular
to the supply outlets shown in Fig. 18.22 for a better effectiveness factor 
.
Because the total pressure TP loss of return slots is nearly equal to the TP of the vertical projection
of the same type of supply slot diffusers, from Table 18.1, for a slot intensity of 28 cfm / ft, TP
for selected return slots is 0.06 in. WC (15 Pa).
7. The depth of the interior zone is 20 ft. It is better to select two 4-ft two-way two-slot supply
slot diffusers with 0.75-in. slot width and located at the midpoint of the interior zone. Each one-way
single-slot diffuser needs a throw of 20/2  10 ft. From Table 18.1, for a slot intensity of
500/ (2  4  2)  31 cfm / ft, horizontal throw T100 is about 10 ft and the ratio T100 /L  10/10 
1.0. Such a setup is beneficial during part-load operation.

9000
60  960  0.072  0.243
 8.9F (5C)
Ts  Tr 
Qrh
60V?s s Cpa
Qrh  30  300  9000 Btu / h
T100  0.7L  0.7 
AIR SYSTEMS: SPACE AIR DIFFUSION 18.37
The return slots in the interior zone must have the same total pressure loss TP  0.06 in. WC,
i.e., the same slot intensity, as in the perimeter zone. This is because both perimeter and interior
zone inlets are installed on the same ceiling and use the ceiling plenum as the return plenum.
For the interior zone, return slots with a slot intensity of 28 cfm / ft  500/28  18 ft (5.4 m)
must be installed. Select two 5-ft two-way two-slot return slots with a slot width of 0.75 in. The diffusers
will be located on the ceiling at a position midway from the depth boundaries of the interior
zone, as shown in Fig. 18.22.
8. In the perimeter zone, during cooling mode operation of 50 percent part-load operation, the
throw–characteristic length ratio may reduce to
In the interior zone,
Both are still within 0.3 to 1.4 at a load density 25 Btu /h ft2.
9. During heating mode part-load operation, the supply air temperature differential, the buoyancy
effect, and the vertical temperature gradients will be smaller than those at design load.
18.7 DISPLACEMENT FLOW AND UNIDIRECTIONAL FLOW
Displacement Flow
Displacement flow is a flow pattern in an air conditioned space in which cold supply air, at a velocity
nearly equal to the required velocity in the space, enters the occupied zone or working area and
displaces the original space air with a pistonlike airflow without mixing the supply air and the original
space air.
Compared with mixing flow, displacement flow provides a better indoor air quality in the occupied
zone. If cold air is supplied at a velocity nearly equal to the mean air velocity in the occupied
zone with a small supply temperature differential for comfort air conditioning systems, there will be
low turbulence intensities and fewer draft problems. On the other hand, displacement flow usually
requires a greater supply volume flow rate and a higher construction cost.
There are three types of displacement flow: unidirectional flow, downward uniform flow, and
stratified displacement flow.
Unidirectional Flow
In a unidirectional flow, the conditioned supply airstream flows in the same direction as a uniform
airflow and showers the entire working area or occupied zone. Supply airstreams occupy the entire
working area or occupied zone of the conditioned space.
Types of unidirectional flow currently used in clean rooms include downward unidirectional
flow and horizontal unidirectional flow, as shown in Fig. 18.23. Unidirectional flow refers to airflow
patterns in which the stream lines of airflow are uniform and move in the same direction. Because
of the unidirectional flow, contaminants generated in the space cannot move laterally against the
downward airflow, and dust particles will not be carried to higher levels by recirculating flow or
large eddies.
Such an airflow pattern in clean rooms was formerly known as laminar flow. Laminar flow is often
confused with the types of fluid flow distinguished by Reynolds number in fluid mechanics.
T100
L

4.5
0.5  20
 0.45
T100
L

4.5
14
 0.32
18.38 CHAPTER EIGHTEEN
Most forced airflows driven by fans in air conditioned space are turbulent flow except the laminar
flow in the boundary layers adjacent to the surface of the building envelope. The Reynolds number
of forced airflow, even at very low air velocity, is usually greater than 1  104. As of 1991,
ASHRAE Handbook 1991, HVAC Applications, refers to this flow pattern as unidirectional flow.
Unidirectional Flow for Clean Rooms
Figure 18.23a shows a downward unidirectional flow for clean rooms. After flowing through the
high-efficiency particulate air (HEPA) filters or ultralow penetration (ULPA) filters, ultraclean
air discharges uniformly downward in parallel stream lines and enters the working area. It then
AIR SYSTEMS: SPACE AIR DIFFUSION 18.39
FIGURE 18.23 Unidirectional flow for clean rooms: (a) downward unidirectional flow for clean room; (b) horizontal
unidirectional flow for clean room; (c) unidirectional flow for clean workstation.
flows through the raised floor grating and is returned to the recirculating air unit at the top of the
clean room.
Figure 18.23b shows a horizontal unidirectional flow for a clean room. Instead of flowing downward,
the clean airstream discharges horizontally from the HEPA filters on one side of the room and
flows through the working area. The contamination level near the return inlets of a horizontal unidirectional
flow may be higher than those of a downward unidirectional flow.
Figure 18.23c shows a clean workstation that provides a small noncontaminated working area
for a single worker by means of horizontal or downward unidirectional flow and HEPA filters.
Clean workstations can achieve a high degree of contamination control over a limited area for many
practical applications.
To provide parallel stream lines, air velocity of 60 to 90 fpm (0.3 to 0.45 m / s) is required. Unidirectional
flow provides a direct and predictable path of submicrometer-size dust particles and
minimizes the opportunities for these particles to contaminate working parts. It also captures internally
generated dust particles and carries them away. Most dust particles in unidirectional flow
reestablish their parallel stream lines after the downstream eddies of an obstruction.
The supply temperture differential Tr-Ts for unidirectional flow for clean rooms depends mainly
on the required space velocity and the cooling loads to be removed within the working area. A case
study of a class 10 clean room is discussed in Chap. 30.
Ventilating Ceiling
A ventilating ceiling is sometimes called a perforated ceiling. It creates a downward uniform flow
similar to the downward unidirectional flow for clean rooms. In most cases, ventilating ceilings discharge
conditioned air through the entire ceiling to form a downward uniform flow, except in the
area occupied by light troffers. The primary differences between unidirectional flow for clean
rooms and downward uniform flow from ventilating ceilings are as follows:
 Unidirectional flow requires a 60 to 90 fpm (0.3 to 0.45 m / s) air velocity and ultraclean air in the
working area, and ventilating ceilings usually have a mean air velocity of less than 15 fpm (0.075
m/ s) of conditioned air.
 There is no mixing of supply and space air in unidirectional flow; whereas just below the perforated
ceiling, supply air is mixed with the ambient air at a vertical distance of less than 1 ft (0.3
m) in downward uniform flow from the ventilating ceiling.
18.40 CHAPTER EIGHTEEN
FIGURE 18.24 Ventilating ceiling.
Figure 18.24 shows a downward uniform airflow. In Fig. 18.24, conditioned air is first supplied
to the ceiling plenum through supply outlets inside the plenum. It is then squeezed through the
holes or slots of the ventilating ceiling and discharged to the conditioned space in a downward uniform
flow. When parallel airstreams combine and flow through the working area, they arrive at the
floor level, turn to the side return grilles at low levels, and then return to the air-handling unit or
packaged unit in the fan room.
Downward uniform airflow patterns from ventilating ceilings have the following characteristics:
 There is no mixing or induced recirculating flow in the working area or occupied zone, so dust
particles contained in the space air at a lower level are not carried to a high level to enter the
working area again.
 A very low air velocity can be maintained in the working area or the occupied zone even when
the volume flow per ft2 /Afl, in cfm / ft2 (L/sm2) is very high. Here represents the supply volume
flow rate, in cfm (m3/s), and Afl represents the floor area, in ft2 (m2).
 When /Afl  3 cfm / ft2 (15 L / sm2) the downward uniform flow becomes prominent.
Ceiling Plenum and Supply Air Velocity
To create a more uniform supply air velocity vo, in fpm (m / s), at the perforated holes or slots, the
maximum velocity of the airstream crossing the ventilating ceiling inside the ceiling plenum vp, in
fpm (m/ s), should be low. Usually, vp  0.5vo. Moreover, the clearance between the beam and the
ventilating ceiling Hip, in ft (m), should be always greater than 8 in. (200 mm ) to prevent a greater
pressure loss against the airflow crossing the ventilating ceiling inside the plenum. The greater the
supply volume flow rate, the higher Hip. If the ceiling plenum is of sufficient height Hp and few
small obstructions other than the beams are present, distributing ductwork inside the ceiling plenum
is not necessary to provide a uniform air supply.
Supply air discharged from the outlets into the ceiling plenum always moves in an upward
direction so that the velocity pressure of the discharged airstream does not affect the supply air
velocity at the perforated openings. The construction of the ceiling plenum must meet the requirements
of the National Fire Protection Association (NFPA) and local fire protection codes.
Ventilating ceilings are always made of noncombustible materials such as metal strips or
mineral acoustic tiles. As a supply plenum, the ceiling plenum must be insulated against heat
gain and loss, and condensation on its inner surface if it is adjacent to an area that is not air conditioned.
Supply air velocity at the perforated openings vo must be optimal. When vo exceeds 1000 fpm
(5 m/ s), objectionable noise is generated. A higher vo also means higher pressure at the ceiling
plenum and, therefore, greater air leakage. On the other hand, a lower vo may cause unevenness in
the supply airflow rate because of a higher vp / vo ratio. Supply air velocity vo is usually between 200
and 700 fpm (1 and 3.5 m / s). When vo has been determined, the perforated area Ao, in ft2, (m2) can
be calculated as
(18.16)
where s  supply volume flow rate, cfm (m3 /s). Generally, the diameter of the circular holes or
the slot width is less than 0.25 in. (6.5 mm). The number of holes or slots can then be determined
accordingly. Based on test results, the local loss coefficient of the perforated holes and slots Co including
discharge air velocity pressure at the perforated openings is about 2.75.
Ventilated ceilings are often used in industrial applications where very low air movements or
precise control of space temperature up to 68  0.1°F (20  0.056°C) is needed. They are also
used in indoor sport stadiums for badminton and table tennis tournaments, where an air velocity of
less than 40 fpm (0.2 m / s) should be maintained in the arena.
V?
Ao 
V?
s
vo
V?
V?
V?
AIR SYSTEMS: SPACE AIR DIFFUSION 18.41
18.8 STRATIFIED DISPLACEMENT FLOW
Two-Zone Stratified Model
Stratified displacement flow supplies conditioned cold air (in Scandinavian countries usually 100
percent outdoor air at lower velocity) at a low-level supply outlet, as shown in Fig. 18.25. The cold
supply air, with a volume flow rate of s, in cfm (L / s), flows in a thin layer along the floor. Above
the heat and contaminant sources, heated air containing contaminants rises upward because of its
buoyancy effect. Supply air is then entrained into the upward convective flow with a volume flow
rate of conv, in cfm (L / s). When upward convective flow arrives at a height where its volume flow
rate is equal to the supply volume flow rate, that is, conv  s, this height is recognized as the stationary
level zstat, in ft (m). Above zstat, ambient air is induced into the upward convective flow until
it reaches the ceiling. Upward convective flow with induced air spreads laterally along the ceiling.
An amount nearly equal to the supply volume flow rate exhausts or returns to the fan room through
the exhaust or return inlet near the ceiling. The remaining portion containing the contaminant descends
to the stationary level to be induced into the upward convective flow and recirculated.
The stratified level divides the room vertically into two zones: an upper zone and a lower zone.
In the lower zone, only the supply air is induced into the upward convective flow, as its volume flow
rate is smaller than the supply volume flow rate conv 	 s. In the upper zone, the portion of air and
contaminants that is greater than the supply flow rate conv  s recirculates. For cold air supply,
the mean temperature of the upper zone is usually 1 to 2°F (0.56 to 1.1°C) higher than that of the
lower zone.
Operating Characteristics
Stratified displacement flow (or displacement ventilation) was introduced to Scandinavian countries
in the early 1970s as a means of improving general ventilation in industrial applications. For comfort
systems in public and office buildings, it was adopted in the early 1980s.
V?
V?
V?
V?
V?
V?
V?
V?
18.42 CHAPTER EIGHTEEN
FIGURE 18.25 Stratified displacement flow in a typical room.
According to Svensson (1989), Mathisen (1989), Sandberg and Blomqvist (1989), and Sepp?nen
et al. (1989), stratified displacement flow has the following characteristics:
 Cold air supply of usually 100 percent outdoor air is used to remove cooling loads in conditioned
space. Heating is usually provided by radiant panels or a baseboard heater under the windows or
on the walls.
 Air must be supplied at low velocity, usually less than 60 fpm (0.3 m / s) and at a height often less
than 1.8 ft (0.54 m) from the floor level.
 Cold air is generally supplied at a temperature 5 to 9°F (2.8 to 5°C) lower than that of the air in
the occupied zone or working area.
 The height of the lower zone, or zstat, generally should be higher than the breath line of
a seated occupant (4.5 ft or 1.4 m). Stationary level depends mainly on the supply volume flow
rate s.
 In the lower zone, all air is supply air, theoretically, except the downward cold drafts. Supply air
is supplied to the occupied zone directly without mixing with the ambient air.
 Because of the small supply temperature differential, the maximum cooling load that can be
removed from the room is considerably less than mixing flow. Some scientists recommend a
maximum cooling load of 13 Btu /h ft2 (41 W/m2) at a ceiling height of 9 ft (2.7 m). For greater
cooling loads, additional cooling panels mounted on the ceiling or part of space air mixing should
be used.
 Return or exhaust inlets are located near the ceiling level.
 According to the required thermal comfort of the occupant, the vertical temperature difference between
the 0.3- and 5.5-ft (0.09- and 1.7-m) level as specified in Sec. 4.8 should not exceed 5°F
(2.8°C).
Busweiler (1993) reported the use of a combination of radiant cooling, displacement ventilation,
and desiccant cooling during the retrofit of the HVAC&R system of a conference room in a hotel in
Bremen, Germany. Because of the limited space in the ceiling plenum, air ducts for a conventional
air conditioning system could not be installed. The cooled ceiling was used to remove the sensible
heat and controlled the room temperature. Displacement ventilation supplied 100 percent outdoor
air from wall outlets near to the floor level at a velocity of 40 fpm (0.2 m / s). Desiccant-based cooling
was sometimes used to cool and dehumidify the outdoor air and maintain a room temperature
of 68°F (20°C) and a relative humidity of about 60 percent. The first-year operation in the Bremen
hotel reveals that these systems are a good and practical combination.
Comparison of Stratified Displacement Flow and Mixing Flow
Compared with mixing flow space diffusion systems, stratified displacement flow has the following
advantages:
 Contamination level in the occupied zone is lower, and therefore indoor air quality is better.
 Space diffusion effectiveness factors 
T and 
C are higher.
 Turbulence intensities Itur are lower, and therefore fewer draft problems occur even at higher mean
air velocities.
The disadvantages of the stratified displacement flow are as follows:
 Initial cost is significantly higher if cooling load density is greater than the recommended maximum
13 Btu/h ft2 (41 W/m2) and if cooling panels are added.
 Energy cost is comparatively higher.
 Stratified displacement flow is for space cooling only.
V?
AIR SYSTEMS: SPACE AIR DIFFUSION 18.43
18.9 PROJECTING FLOW—SPOT COOLING/HEATING
Benefits of Projecting Flow
In a projecting flow air pattern, the cold or warm air jet is deliberately projected into part of the occupied
zone or working area, which is often called the target zone. The result is control of the environment
in a small or localized area, or microenvironmental control. This projecting airflow pattern
is used in spot cooling / heating, task air conditioning, or controlling personal environments.
Spot cooling / heating using projecting flow has many advantages over conventional space air
diffusion:
 It provides better control of temperature, air cleanliness, and air movements in a localized environment.
Spot cooling improves the occupants’ thermal conditions significantly, and their heat
strain decreases.
 There is greater direct outdoor air supply.
 There is direct and efficient handling of local loads.
 Occupants can have greater control of their own microenvironment or personalized environment.
The main disadvantages of spot cooling / heating include draft discomfort or discomfort due to the
pressure of the air jet, limited area of environmental control, and a more complicated space air diffusion
system design.
Air jets in projecting flow are usually free jets with high entrainment ratios because of their
large contact area with ambient air. Long-throat round nozzles are often used as supply outlets for
spot cooling in large spaces.
Two types of projecting flow are currently used today: industrial spot cooling systems and desktop
task air conditioning systems.
Industrial Spot Cooling Systems
In an industrial spot cooling system, as shown in Fig. 18.26a, the temperature differential between
the target zone and the room air (ambient air) is often 5°F (2.8°C) or greater. Melikov et al. (1994)
summarized the following based on their and other research workers’ experimental results:
Distance Between Target Zone and Supply Outlet. The core zone of a free air jet (from surface
of outlet to x/Do  4) and transition zone (x/Do between 4 and 12) had a stronger cooling capacity
than full developed main zone. This was due to the intensive mixing of the cold jet air with the
room air in the main zone. The difference between the target temperature Tj and the outlet temperature
To increased as the distance between the target zone and the outlet x and the ambient air temperature
Ta were increased.
For a target temperature of 77°F (25°C) and an ambient (room) temperature of 86°F (30°C), if
x/Do  2, the required outlet temperature To was about 71.6°F (22°C). If x/Do  8, To was about
64.4°F (18°C). The size of the target zone can be estimated since the angle of divergence of the air
jet  22° and the distance from the outlet x are known. High turbulence intensity caused intensive
mixing of the cold air jet and, therefore, decreased the cooling capacity of the air jet.
Vertical Versus Horizontal Jet. Many reseachers preferred a vertical air jet because a horizontal
air jet required a large diameter of the jet’s outlet and the worker’s body was not exposed symmetrically
to the cold air shower. In Melikov’s experiments, when the room temperature was 82.4°F
(28°C), the target velocity vj preferred by the subjects participating in the experiments was almost
identical with vertical and horizontal cooling jets. However, when the room temperatures were
91.4°F (33°C) and 100.4°F (38°C), experimental results confirmed that a vertical jet cooled the subjects
more than a horizontal jet.
18.44 CHAPTER EIGHTEEN
AIR SYSTEMS: SPACE AIR DIFFUSION 18.45
Air
damper
Vertical jet
0.5 m
1 m
(a)
(b) Access floor panel
Mixing
box
Supply
nozzles
(outlets)
Control
panel
FIGURE 18.26 Projecting flow: (a) an industrial spot cooling system; (b) a desktop task conditioning
system.
Target Velocities. The preferred target velocities vj of participating subjects at various conditions
were as follows:
Study Ta, °F(°C) Tj, °F(°C) vj, fpm (m/ s) Thermal sensation Jet type
Melikov et al. (1994) 91.4 (33) 82.4 (28) 280 (1.4) Slightly warm Vertical
Horizontal
Olesen and Nielsen 86 (30) 75.2 (24) 320 (1.6) Slightly warm Vertical
(1983) 86 (30) 75.2 (24) 400 (2.0) Slightly warm Horizontal
Melikov et al. (1994) 86 (30) 75.2 (24) 440 (2.2) Vertical
Horizontal
Fanger et al. (1974) 82.4 (28) 82.4 (28) 160 (0.8) Neutral Horizotal
Whole body
Melikov et al. (1994) 82.4 (28) 82.4 (28) 300 (1.5) Vertical
82.4 (28) 82.4 (28) 300 (1.5) Horizontal
The preferred target velocities show substantial differences from different studies. The amount of
the body exposed to the air jet is also an important factor that affects vj.
Thermal Sensation. Large differences occurred in the local thermal sensation because of the
velocity and temperature distribution of the airflow around the human body. Under these circumstances,
an occupant’s acceptance of the thermal environment showed a compromise between the
thermal sensation of the whole body (general thermal sensation) and the thermal sensation for individual
parts of the body (local thermal sensation).
When air velocities were above 200 fpm (1 m / s), the airflow exerted a physical pressure on the
skin that produced a discomfort. This pressure is the sum of the static and velocity pressure, the total
pressure which must take into account the direction of airflow. At high local air temperatures
when T 82.4°F (28°C), a subject participating in the tests preferred the slight warmth instead
neutral sensation because of the annoying effect of the pressure at the high-velocity air jet. At low
jet temperatures, local discomfort created by draft was mostly at the head and neck for a vertical jet
and at the neck, face, and left shoulder for a left horizontal jet.
Recommendations in Spot Cooling Design
Melikov et al. (1994) and Brown (1988) recommend the following for spot cooling design:
 Allow the occupant to have individual control of the jet velocity and temperature. Spot cooling
systems that permit the occupant to adjust the distance between the target zone and the outlet x
and the direction of the air jet are preferrable.
 The target zone should be located in the core or transition zone of a cooling jet; typically, x/Do 
4 to 5 from the nozzle outlet.
 The maximum target velocity (mean value) should not exceed 440 fpm (2.2 m / s), and the maximum
local velocity should not exceed 840 fpm (4.2 m / s).
 When ambient air temperature is above 95°F (35°C), use of a combination of a vertical jet, a
helmet, and a clothing ensemble to protect the neck from draft is required.
Desktop Task Conditioning Systems
System Description. Bauman et al. (1993) reported the results of laboratory and field measurements
investigating the performance of a desktop task conditioning system. A typical desktop task
conditioning system, as shown in Fig. 18.26b, consists of a self-powered mixing box, two small
18.46 CHAPTER EIGHTEEN
fans, two desktop supply outlets, flexible ducts, and a control panel. Inside the mixing box, there is
a variable-speed fan to extract primary air (cold conditioned air from the AHUs) either from the
low-pressure floor plenum or from the supply duct in the ceiling plenum by means of connecting
flexible ducts. Another fan pulls recirculating air from the knee space under the desk. The primary
air and recirculating air are mixed together according to the required fractions by regulating the
opening of their dampers in each of the transporting ducts. The mixture is then drawn through an
electrostatic air filter. The primary air damper is never allowed to close completely, to ensure that
the minimum amount of outdoor ventilation air is supplied all the time. The mixing box is hung on
the back or on the corner of the desk and is connected to the two supply nozzles (outlets) on the top
of the desk by flexible ducts.
Two supply nozzles may be rotated 360° on the horizontal plane. They are installed with horizontal
guide vanes which are adjustable  30° in the vertical plane.
Control Panel. A desktop DDC unit controller uses adjustable sliders to control the variablespeed
fan and modulate the volume flow of the supply air from the nozzles. The supply air temperature
is adjusted by varying the opening of the primary and recirculating dampers, and thus the ratio
of primary to recirculating air. A 200-W electric radiant heating panel in the knee space is employed
during winter heating. The dimming of the occupant’s task light controls the lighting luminance
and a white noise generator which has a continuous frequency spectrum with equal
energy/Hz over a specified frequency range.
A motion detector–based occupancy sensor sends a signal to the DDC unit which shuts the
desktop system off when the workstation is unoccupied for a specified time.
Operating Characteristic. Each desktop unit can provide 40 to 250 cfm (20 to 70 L / s) of supply
air from the supply nozzles. Even when the two fans in the desktop unit are shut off, 40 cfm (20
L/ s) of primary air will still be delivered to the conditioned space for the minimum outdoor ventilation
air requirement.
The maximum outlet velocity measured at the face of the 2.3- by 4-in. (58- by 100-mm) supply
outlet during laboratory tests varied between 6.5 and 24.5 fps (2 and 7.5 m / s). For a typical desktop
system, the primary air temperature supplied from a variable-air-volume (VAV) AHU was 55°F
(12.8°C). After mixing with recirculating air, the supply air temperature from the desktop outlet
was 65°F (18.3°C).
The supply nozzles are 5 ft (1.52 m) wide by 30 in. (0.76 m) deep at the back corners of the
desk. Air is supplied from these two nozzles toward a focus point near the center of the front edge
of the desk. For isothermal air jets of horizontal supply with a volume flow of about 90 cfm
(43 L/ s), the average speed at the work location in front of the desk was 3.3 fps (1 m / s).
Performance of Desktop Task Conditioning Systems
Based on the laboratory investigation and field-measured results, the performance of desktop task
conditioning systems was summarized in Bauman et al. (1993) as follows:
 The desktop unit can be controlled to produce a wide range of thermal conditions, allowing occupants
the opportunity to fine-tune the local environment to their individual comfort preferences.
 Under the influence of different load density of the adjacent workstation, or under a warm average
room air temperature, a desktop unit is capable of being fine-tuned to maintain a nearly comfortable
environment, or to maintain a temperature of 1 to 3°F (0.56 to 1.7°C) lower than that of the
workstation without the desktop unit, even at relatively low supply air volume flow rates.
 At the same supply volume flow rate, a larger supply nozzle was shown to deliver lower air
velocity than the smaller nozzle. A larger nozzle reduces the potential to produce draft while
maintaining improved ventilation performance at moderate to high supply volume flow rates.
 Turning off the desktop unit whenever the workstation is unoccupied significantly reduces its energy
use.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.47
Application of Desktop Task Conditioning Systems
The first large installation of the desktop system was recently completed in a new office building for
an insurance company in West Bend, Wisconsin. This building has 370 desktop units and was fully
occupied in July 1991. Reseachers had tracked the productivity of more than 100 employees before
and after they moved to the new building. The study has concluded that the desktop system does
have a positive impact on worker productivity.
The desktop systems provide excellent individual thermal control for occupants and a direct
supply of primary air including outdoor ventilation air to occupants. However, desktop systems often
need a floor plenum for primary air supply and a ceiling plenum for return air.
Desktop units and task air conditioning for microenvironmental (personalized environmental)
control have great potential because of the smaller controlled environment and therefore lower energy
use. Air conditioning integration with furniture may be important in the future.
18.10 UPWARD FLOW UNDERFLOOR AIR DISTRIBUTION
Upward Flow from Floor Plenum
Upward flow from a raised floor has been successfully used for space air diffusion in computer
rooms and other industrial applications with high cooling load density, such as 60 to 300 Btu /h ft2
(189 to 946 W/m2). Figure 18.27 shows an upward flow space air diffusion from a raised floor
plenum—an underfloor air distribution system (an air distribution system consists of ductwork distribution,
terminals, and space air diffusion) with the following characteristics:
18.48 CHAPTER EIGHTEEN
Concrete
structural
slab
Ceiling
Return
air
Access
floor
panel
80 F
55 F
65 F
Ceiling plenum
Floor
diffuser
Fan
diffuser
Desktop
supply
outlet
Mixing
box
Floor
plenum
Cold primary
air supply duct
Conditioned
space
Return air
FIGURE 18.27 Upward flow underfloor air distribution system.
 Conditioned air is supplied to an access floor plenum. The floor is usually raised from the structural
floor 12 to 18 in. (0.30 to 0.45 m) depending on whether fan-coil units (FCUs) or water
source heat pumps (WSHPs) are installed inside the floor plenum. Cable services are usually provided
below the access floor.
 Types of supply outlets include floor diffusers, fan-driven floor diffusers, desktop units, and supply
outlets from FCUs and WSHPs. The floor plenum can be pressurized at a positive pressure of
usually less than 0.08 in. WG (20 Pag) if conditioned air is squeezed out from the floor plenum
through floor diffusers directly. For fan diffusers, desktop units, grilles from FCUs, and WSHPs,
conditioned air is forced to the occupied zone directly by individually installed small fans. Under
these conditions, the floor plenum pressure should be maintained at slightly negative pressure between
 0.02 and  0.05 in. WG ( 5 and  12.5 Pag).
 In the occupied zone, the airflow patterns of the supply air discharged from the floor diffusers,
supply outlets of the FCUs, and WSHPs are mixing flow, and the air discharged from the desktop
nozzles is projecting flow. However, from the floor level up to the ceiling plenum, the upward
flow is the dominating airflow pattern.
 Cool primary air (including outdoor ventilation air) from the air-handling unit is supplied at neutral
pressure to the floor plenum at a temperature of 45°F (7.2°C) for cold air distribution and
55°F (12.8°C) for conventional air distribution. Because all the supply outlets of the underfloor
air distribution system are located in the occupied zone, the cool primary air must be mixed with
the recirculating air by using fan diffusers, desktop units, FCUs, and WSHPs. The temperature of
the mixture at the supply outlet (supply air temperature) is often between 63 and 65°F (17 and
18.5°C).
 After the space cooling load is absorbed by the supply air, the bouyant space air rises. The temperature
of the return air entering the light troffers is usually between 80 and 82°F (26.7 and
27.8°C). Return air then enters the ceiling plenum. A part of it is returned to the floor plenum
through the return air shafts along the columns, and the remaining part is returned to the fan room
through a return system.
 FCUs or WSHPs are often installed inside the access floor plenum and are used to offset the
building envelope heat losses and heat gains.
In recent years, there have been rapid developments in office automation. Underfloor air distribution
is a new air distribution system that intends to keep in pace with these advances. Shute (1995)
summarized the experience of an integrated access floor plenum prototype project which has been
in full operation for 4 years as follows:
Thermal Storage of Floor Plenum
The distribution of cold primary air in the floor plenum is not fully ducted. If the floor
plenum is composed of a 4-in. (100-mm) concrete structural slab and a concrete-filled access
floor, the cold primary air within the floor plenum is in direct contact with a density of 50 lb / ft2
(244 kg /m2) thermally active concrete and the floor panel which are potentially thermal storage
media.
During a winter night, the mass may provide after-hours low-grade night setback space
heating, and creates a cooled reservoir to assist in offsetting a part of the cooling load the next
day. On a winter day, part of the internal heat gain is stored, and the recovered heat is used to
assist the following night’s setback space heating. During the summer, from early morning
(4.00 a.m.), the system is started to subcool the mass, which helps to reduce the electric peak demand.
Observations with a laboratory experiment have indicated that a 4-in. (100-mm) structural slab
plus the total mass of the access floor panel could participate in a 6 to 7°F (3.3 to 3.9°C) temperature
variation cycle within a 24-h period. This allows storage of up to 30 percent of a typical daily
cooling load.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.49
Heat Unneutralized
When the supply air from the floor level is returned at the ceiling level, upward airflow lifts the heat
gains in the upward flow direction and returns some heat unneutralized to the ceiling. Furthermore,
the upward airflow captures most of the lighting heat gain in the return airflow. Research has
demonstrated that for a typical office, 15 percent of space heat gain and most of ceiling-level lighting
heat gain can be considered unneutralized because of the localized thermal buoyancy. In offices
where equipment is concentrated, convective plumes, upward warm expanding airflow above the
heat sources, can capture an even greater percentage of space load by the floor-to-ceiling airflow,
and results in greater capability to exhaust heat through ceiling return than the conventional space
air diffusion system.
Maintaining a Consistent Access Plenum Temperature
Because of the smaller supply air temperature differential Tr  Ts in underfloor air distribution, if the
underfloor plenum has hot or cold areas, the occupants’ comfort will not be satisfied by pulling air from
these localized areas of temperature inconsistency. The following rules are based on test results:
 Supply air colder than 63°F (17.2°C) is perceived as too cold by occupants. Blend air at the floor
outlets and at other supply units.
 Primary air degradation becomes unacceptable when it travels a distance exceeding 30 ft (9 m)
from the point of air introduction within the access floor plenum. Provide sufficient insulated
ductwork to distribute primary air.
 A variation of more than 3°F (1.7°C) across a master control zone may exceed the air transfer fan
adjustment range to compensate for this variation. Practical experience limits the master zoning (a
zone corresponds to an area served by a VAV box or other terminals) to about 3000 ft2 (279 m2) if
primary air is supplied from a central riser with VAV control, or a maximum of 10,000 ft2
(929 m2) if primary air is supplied from an on-floor fan room.
Floor Plenum Master Zone Air Temperature Control
For a VAV system, as space load varies, the supply volume flow rate supply to each master control
zone through an individually controlled VAV box will be varied accordingly. Because of the thermal
storage effect, the floor plenum air temperature does not respond as quickly as required. It takes approximately
20 min before the plenum temperature falls sufficiently to call for proper action.
As the space load reduces, more recirculating air is pulled to mix with the cold primary air. The
primary air rate becomes proportionally greater in the floor plenum. This effect increases the pressure
on the floor plenum, and its pressure becomes positive. The pressurized cold primary air overpowers
the normal downward return airflow. If a temperature sensor is located in the downward air-
flow path, this cold air will be sensed by the sensor, with a resulting rapid reponse to cut the
primary air supply flow rate.
Design Considerations
 A structural concrete slab of 4-in. (100-mm) thickness that separates the supply raised floor
plenum and the return ceiling plenum is recommended. An attempt to baffle supply and return
sections and integrate them in the plenum of the same level has inherent difficulties due to air
leakages through plenum baffles, short-circuiting of air in the space, and limiting of the supply air
temperature range.
18.50 CHAPTER EIGHTEEN
 The diameter of the floor diffuser is usually 8 to 15 in. (200 to 375 mm). The mean air velocity at
the outlet is often 200 to 400 fpm (1 to 2 m / s). For a 12-in. (300-mm) floor diffuser, the maximum
supply volume flow rate max  150 cfm (71 L / s).
 The mean air velocity of a radial air jet of a floor diffuser in which the supply air is discharged in
a radial direction along the floor decays more rapidly than a vertical air jet of the same size and
supply volume flow rate.
 According to Sodec and Craig (1990), the attenuation of the ductwork between two floor diffusers
in adjacent rooms is usually greater than 29 dB. Therefore, conversation transmission through
floor diffusers is negligible.
Applications of Upward Flow Underfloor Air Distribution
Compared with a conventional ceiling supply mixing flow air diffusion system, the primary advantages
of an upward flow underfloor air distribution system are as follows:
 It keeps pace with the requirements of the raised floor plenum in office buildings of high automation.
 Conditioned air (including outdoor ventilation air) is supplied to the occupied zone directly,
which results in a better indoor air quality.
 When air is supplied from the floor diffusers and discharged and returned from the ceiling
plenum, usually there is no stagnant air in the occupied zone.
 Upward airflow lifts some unneutralized heat and carries most of the lighting heat gains at high
level to the ceiling plenum; these result in a higher return air temperature and, therefore, a higher
space diffusion effectiveness factor and a low space cooling load.
 It utilizes the mass heat storage of the access floor panels, and part of the structural concrete slab
subcools the floor plenum and reduces the summer peak electric demand charge.
Disadvantages of upward flow underfloor air distribution include the following:
 Valuable floor space may be occupied by the floor diffusers and the buffer zones between the air
jets and the workstation.
 Both floor and ceiling plenums are required, and the initial investment is higher than that for conventional
ceiling supply mixing flow systems.
For offices and applications other than computer rooms, upward flow underfloor air distribution
for commercial buildings was first developed in Germany, and many underfloor air distribution systems
have recently been adopted in South Africa, Hong Kong, and Japan. In the 1990s, some modern
upward flow underfloor air distribution systems were completed in Canada and the United
States. One, an underfloor air distribution system using 370 desktop units for an office building in
West Bend,Wisconsin, was fully occupied in July 1991 (Bauman et al., 1993).
Upward flow underfloor air distribution systems are still in the developmental stage. Where a
floor plenum is needed, an underfloor air distribution system might be a suitable choice.
18.11 COMPUTATIONAL FLUID DYNAMICS
Computational fluid dynamics (CFD) including heat transfer is a kind of computing technique for
the quantitative prediction of air motion and heat transfer, and it presumes that the physical behavior
of airflow and thermal systems can be predicted by approximated Navier-Stokes and thermal
equations.
V?
AIR SYSTEMS: SPACE AIR DIFFUSION 18.51
CFD Becomes More Popular
HVAC&R engineers traditionally conducted scale-model physical experiments to study air motion
and temperature distribution by maintaining similarity through a high Re number and an appropriate
Ar corresponding to the adopted scale. CFD was first developed in Europe and Japan. In the
late 1980s, ASHRAE organized a research project which thoroughly investigated many issues associated
with CFD simulation. In the late 1990s, CFD became more popular in the HVAC&R industry
for the determination of locations of supply outlets and return inlets, airflow patterns, air velocity
and temperature distributions, and contaminant concentration and removal in an HVAC&R system.
Baker et al. (1997) and Ladeinde and Neavon (1997) summarized the reasons for CFD’s popularity:
 CFD has the ability to establish firm quantitative data regarding air motion and can predict fluid
characteristics and pressure differentials to very low levels that are essentially impossible during
experiments.
 The cost of CFD is lower than that of scale-model experiments.
 Earlier CFD codes were developed on supercomputers. Today computing software is available for
simulations to be performed on personal computers.
 People are slowly getting over the fear of using CFD.
Reynolds-Averaged Navier-Stokes Equations
The mathematical expressions describing the basic space-time relationship between mass, velocity,
and temperature are expressed in partial differential equations called Navier-Stokes (NS) equations.
These equations may be only directly applied to laminar flow fields. Baker et al. (1997) pointed out
that CFD modeling based on NS equations requires a large number of assumptions and approximations.
These approximations are closely related to the CFD computing results.
For predicting turbulent flow, an approximation called Reynolds averaging (Ra) is used to
convert this time-unsteady flow to a mean velocity presentation. A principal assumption for CFD is
the turbulence closure model to govern the Reynolds-averaged Navier-Stokes partial differential
equations. The selected CFD has been a two-equation turbulent kinetic energy (TKE) model. The
TKE model assumes a fully turbulent flow existing everywhere for room airflow prediction. Actually,
at normal air change rates per hour, fully turbulent flow occurs only in supply ducts, air jets,
downstream of the edge of the obstacles. Elsewhere, the flow is more likely weakly turbulent and
actually time-unsteady. A quantitative measurement of the degree of turbulence of an airflow is the
turbulence Reynolds number (Ret) which is defined as the ratio of the turbulence eddy viscosity to
the kinematic viscosity: Ret  vt/v. For laminar flow, Ret  0, and for fully turbulent flow 50 	
Ret 	 103.
Numerical Methods
Because these Reynolds-averaged partial differential equations using the turbulence Reynold
number Ret are highly nonlinear, they are not solvable by explicit, analytical methods. In these
partial differential equations, the velocities u, v, w; and pressure p; temperature T; and some
scalar 	 are dependent variables to be calculated. Space displacements x, y, z and time t are independent
variables. Initial and boundary conditions must be specified. Numerical solution (approximate
methods) such as via the finite element and finite volume methods is more popular in CFD
simulation.
Ladeinde and Neavon (1997) illustrated the finite element procedure during the calculation
of airflow in a duct section. Figure 18.28a shows this duct section that contains an inlet and an
outlet. The interior of the duct that constitutes the region of airflow (domain of the CFD model)
18.52 CHAPTER EIGHTEEN
is discretized into nodes and cells as shown in Fig. 18.28b. The integration of the equations using
the finite element method gives the dependent variables at each node of the CFD model.
The graphic of the vectors representing the velocities using the finite element method is shown in
Fig. 18.29.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.53
Flow outlet
Flow inlet
Solid wall
Solid wall
cell node
(a)
(b)
D C
B A
A
F E
D C
B
F E
FIGURE 18.28 Calculation of airflow using CFD in a duct section: (a) a
duct section with sudden expansion; (b) finite element model for airflow in a
duct section. (Source: ASHRAE Journal, January 1997. Authors Ladeinde
and Nearon. Reprinted by permission.)
FIGURE 18.29 Velocity vectors of the airflow in a duct section. (Source: ASHRAE Journal, January 1997. Authors Ladeinde and
Nearon. Reprinted by permission.)
Conducting CFD Experiments
CFD will not replace the physical experiments for airflow and heat transfer in the HVAC&R industry
completely; however, it will significantly reduce the physical experiments to study the airflow
patterns and the space diffusion effectiveness factor of various space air diffusion systems, the removal
of contaminants in clean rooms, and the effectiveness of the introduction of outdoor air to
the occupied zone.
Baker et al. (1997) asked, How accurate can CFD be? This depends on the form of Reynolds-averaged
Navier-Stokes partial differential (ReNSPD) equations selected as well as the boundary conditions.
Therefore, a range of CFD simulations or CFD experiments must be conducted to verify solution
sensitivity to model parameters and boundary conditions embedded in the selected ReNSPD
equations and their numerical method calculation.
REFERENCES
Abu-Ei-Hassan, M. B., Hosni, M. H., and Miller, P. L., Evaluation of Turbulence Effect on Air Distribution
Performance Index (ADPI), ASHRAE Transactions, 1996, Part II, pp. 322–331.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, 1997.
Bagheri, H. M., and Gorton, R. L., Verification of Stratified Air-Conditioning Design, ASHRAE Transactions,
1987a, Part II, pp. 211–227.
Bagheri, H. M., and Gorton, R. L., Performance Characteristics of a System Designed for Stratified Cooling Operating
during the Heating Season, ASHRAE Transactions, 1987b, Part II, pp. 367–381.
Baker, A. J., Kelso, R. M., Gordon, E. B., Roy, S., and Schaub, E. G., Computational Fluid Dynamics: A Two-
Edged Sword, ASHRAE Journal, no. 8, 1997, pp. 51–58.
Bauman, F. S., Zhang, H., Arens, E. A., and Benton, C. C., Localized Comfort Control with a Desktop
Task Conditioning System: Laboratory and Field Measurements, ASHRAE Transactions, 1993, Part II,
pp. 733–749.
Brockmeyer, I. H. P., Air Flow Pattern and Its Influence on the Economy of Air Conditioning, ASHRAE Transactions,
1981, Part I, pp. 1127–1142.
Brown, C. E., Spot Cooling / Heating and Ventilation Effectiveness, ASHRAE Transactions, 1988, Part I,
pp. 678–684.
Busweiler, U., Air Conditioning with a Combination of Radiant Cooling, Displacement Ventilation, and Desiccant
Cooling, ASHRAE Transactions, 1993, Part II, pp. 503–510.
Chamberlin, G. A., Schwenk, D. M., Maki, K. S., Li, Z., and Christianson, L. L., VAV Systems and Outdoor
Air, ASHRAE Journal, no. 10, 1999, pp. 39–47.
Dorgan, C. E., and Elleson, J. S., Cold Air Distribution, ASHRAE Transactions, 1988, Part I, pp. 2008–2025.
Dorgan, C. E., and Elleson, J. S., Design of Cold Air Distribution System with Ice Storage, ASHRAE Transactions,
1989, Part I, pp. 1317–1322.
Dorgan, C. E., and Elleson, J. S., Design Guide for Thermal Storgage, ASHRAE Inc., Atlanta, 1993.
Elleson, J. S., Energy Use of Fan-Powered Mixing Box with Cold Air Distribution, ASHRAE Transactions,
1993, Part I, pp. 1349–1358.
Fanger, P. O., Melikov, A. K., Hanzawa, H., and Ring, J., Air Turbulence and Sensation of Draught, Energy and
Buildings, vol. 12, 1988, pp. 21–29.
Fanger, P. O., Melikow, A. K., Hanzawa, H., and Ring, J., Turbulence and Draft, ASHRAE Journal, April 1989,
pp. 18–25.
Fields,W. G., and Knebel, D. E., Cost Effective Thermal Energy Storage, Heating / Piping / Air Conditioning,
no. 7, 1991, pp. 59–72.
Fish,W. J., and Faulkner, D., Air Exchange Effectiveness in Office Buildings: Measurement Techniques and
Results, Reprints of International Symposium on Room Air Convection and Ventilation Effectiveness, July
22–24, 1992, University of Tokyo.
18.54 CHAPTER EIGHTEEN
Genter, R. E., Air Distribution for Raised Floor Offices, ASHRAE Transactions, 1989, Part II, pp. 141–146.
Gorton, R. L., and Sassi, M. M., Determination of Temperature Profiles and Loads in a Thermally Stratified Air
Conditioning System: Part 1-Model Studies, ASHRAE Transactions, 1982, Part II, pp. 14–32.
Hanzawa, H., Melikov, A. K., and Fanger, P. O., Air Flow Characteristics in the Occupied Zone of Ventilated
Spaces, ASHRAE Transactions 1987, Part I, pp. 524–539.
Harmon, J. L., and Yu, H. C., Cold Air Distribution and Concern about Condensation, ASHRAE Journal, no. 5,
1993, pp. 34–40.
Hart, G. H., and Int-Hout, D., The Performance of a Continuous Linear Diffuser in the Interior Zone of an Open
Office Environment, ASHRAE Transactions, 1981, Part II, pp. 311–320.
Heinemeier, K. E., Schiller, G. E., and Benton, C. C., Task Conditioning for the Workplace: Issues and Challenges,
ASHRAE Transactions, 1990, Part II, pp. 678–689.
Int-Hout, D., Air Distribution for Comfort and IAQ, HPAC, no. 3, 1998, pp. 59–70.
Int-Hout, D., and Weed, J. B., Throw: The Air Distribution Quantifier, ASHRAE Transactions, 1988, Part I,
pp. 667–677.
Knebel, D. E., and John, D. A., Cold Air Distribution, Application, and Field Evaluation of a Nozzle-Type
Diffuser, ASHRAE Transactions, 1993, Part I, pp. 1337–1348.
Kostel, A., Path of Horizontally Projected Heated and Chilled Air Jets, ASHRAE Transactions, 1955, p. 213.
Ladeinde, F., and Nearon, M. D., CFD Applications in the HVAC&R Industry, ASHRAE Journal, no. 1, 1997,
pp. 44–48.
Lorch, F. A., and Straub, H. E., Performance of Overhead Slot Diffusers with Simulated Heating and Cooling
Conditions, ASHRAE Transactions, 1983, Part I B, pp. 200–211.
Mathisen, H. M., Case Studies of Displacement Ventilation in Public Halls, ASHRAE Transactions, 1989,
Part II, pp. 1018–1027.
Mayer, E., Physical Causes for Draft: Some New Findings, ASHRAE Transactions, 1987, Part I,
pp. 540 – 548.
McCarry, B., Innovative Underfloor System, ASHRAE Journal, no. 3, 1998, pp. 76–79.
Melikov, A. K., Arakelian, R. S., Halkjaer, L., and Fanger, P. O., Spot Cooling—Part 2: Recommendations for
Design of Spot-Cooling Systems, ASHRAE Transactions, 1994, Part II, pp. 500–510.
Miller, Jr., P. L., Application Criteria for the Air Diffusion Performance Index (ADPI), ASHRAE Transactions,
1976, Part II, pp. 206–218.
Miller, Jr., P. L., Room Air Diffusion Systems: A Re-evaluation of Design Data, ASHRAE Transactions, 1979,
Part II, pp. 375–384.
Sandberg, M., and Blomqvist, C., Displacement Ventilation Systems in Office Rooms, ASHRAE Transactions,
1989, Part II, pp. 1041–1049.
Sepp?nen, O. A., Fisk,W. J., Eto, J., and Grimsrud, D. T., Comparison of Conventional Mixing and Displacement
Air-Conditioning and Ventilating Systems in U.S. Commercial Buildings, ASHRAE Transactions, 1989,
Part II, pp. 1028–1040.
Scofield, C. M., Low Temperature Air with High IAQ for Tropical Climates, ASHRAE Journal, no. 3, 1993,
pp. 52–59.
Shute, R. W., Integrated Access Floor HVAC: Lessons Learned, ASHRAE Transactions, 1995, Part II,
pp. 877–886.
Sodec, F., and Craig, R., The Underfloor Air Supply System—The European Experience, ASHRAE Transactions,
1990, Part II, pp. 690–695.
Straub, H. E., and Chen, M. M., Distribution of Air within a Room, for Year-Round Air Conditioning—Part II,
University of Illinois, Engineering Experiment Station Bulletin, No. 442, 1957.
Straub, H. E., and Cooper, J. G., Space Heating with Ceiling Diffusers, Heating / Piping / Air Conditioning, May
1991, pp. 49–55.
Svensson, A. G. L., Nordic Experiences of Displacement Ventilation Systems, ASHRAE Transactions, 1989,
Part II, pp. 1013–1017.
Wang, S. K., Air Conditioning, vol. 2, Hong Kong Polytechnic, Hong Kong, 1987.
Wang, S. K., Kwok, K. W., and Watt, S. F., Characteristics of a Space Diffusion System in an Indoor Sport
Stadium, ASHRAE Transactions, 1985, Part II B, pp. 416–435.
AIR SYSTEMS: SPACE AIR DIFFUSION 18.55
Wendes, H., Supply Outlets for VAV Systems, Heating / Piping / Air Conditioning, February 1989,
pp. 67 – 71.
Williams, P. T., Baker, A. K., and Kelso, R. M., Numerical Calculation of Room Air Motion—Part 2: The
Continuity Constraint Finite Element Method for Three-Dimensional Incompressible Thermal Flows, ASHRAE
Transactions, 1994, Part I, pp. 531–548.
Yuan, X,. Chen, Q., and Clickman, L. R., Performance Evaluation and Design Guidelines for Displacement Ventilation,
ASHRAE Transactions, 1999, Part I, pp. 298–309.
Zhang, J. S., Zhang, R., Li, Z., Shaw, C. Y., Christianson, L. L., and Sparks, L. H., An Experimental Study of the
Ventilation Performance of Cold-Air Distribution Systems, ASHRAE Transactions, 1994, Part II, pp. 360–367.
18.56 CHAPTER EIGHTEEN
CHAPTER 19
SOUND CONTROL
19.1
19.1 SOUND CONTROL AND SOUND
PATHS 19.1
Sound Control 19.1
Sound Paths 19.2
Control at Design Stage 19.3
Recommended Procedure for
Noise Control 19.3
19.2 FAN, COMPRESSOR, PUMP, AND
AIRFLOW NOISE 19.4
Fan Noise 19.4
Noise from Chillers and Pumps 19.4
Airflow Noise 19.5
19.3 SOUND ATTENUATION ALONG
DUCT-BORNE PATH 19.6
Sound Attenuation in Ducts 19.6
Sound Attenuation at Elbows and Branch
Takeoffs 19.9
End Reflection Loss 19.10
Duct-Borne Crosstalk 19.11
Attenuation along Duct-Borne
Path 19.11
19.4 SILENCERS 19.12
Types of Silencers 19.13
Characteristics of Silencers 19.14
Location of Silencers 19.15
Active Silencers 19.16
Selection of Silencers 19.17
19.5 FIBERGLASS IN HVAC&R
SYSTEMS 19.17
Problems 19.17
Recommendations 19.18
19.6 RADIATED NOISE AND
TRANSMISSION LOSSES 19.18
Breakout and Break-in 19.18
Duct Rumble 19.19
Transmission Losses 19.19
Breakout and Break-in Sound
Power Level 19.19
Transmission Loss for Selected Building
Structures 19.23
19.7 RELATIONSHIP BETWEEN ROOM
SOUND POWER LEVEL AND ROOM
SOUND PRESSURE LEVEL 19.23
Single or Multiple Sound
Sources 19.23
Array of Ceiling Diffusers 19.24
19.8 NOISE CONTROL FOR A TYPICAL
AIR SYSTEM 19.25
Combination of Supply Fan Noise and
Terminal Noise 19.25
Estimate Sound Pressure Level
for Spaces Served by Terminal
Units 19.25
Environmental Adjustment Factor 19.26
Plenum Ceiling Effect 19.26
19.9 ROOFTOP PACKAGED UNITS 19.29
Basics 19.29
Sound Sources and Paths 19.30
Discharge Side Duct Breakout 19.31
Sound Path on Return Side 19.31
Structure-Borne Noise 19.32
REFERENCES 19.32
19.1 SOUND CONTROL AND SOUND PATHS
Sound Control
Sound control for an air conditioned space or, more accurately, an occupied zone is provided to attenuate
the HVAC&R equipment sound to an acceptable background level and to provide a suitable acoustic
environment for people in the occupied zone. A suitable acoustic environment is as important as a comfortable
thermal environment to the occupants. Noise, or any unwanted sound, is always annoying.
The objective of sound control should also be to provide an unobtrusive background sound at a
level low enough that it does not interfere with human activities. Sound quality depends on the relative
intensities of the sound levels in various octave bands of the audible spectrum. Unobtrusive
sound quality means the following: a balanced distribution over a wide frequency range; no tonal
characteristics such as hiss, whistle, or rumble; and a steady sound level.
The background sound level to be provided must be optimized. An NC or RC design criterion
that is too low always means an unnecessary higher cost. The recommended indoor design RC or
NC criteria range is listed in Table 4.8.
Sound Paths
When HVAC&R equipment or a component generates sound, it is often called a sound generator or
sound source. Sound created at the source is received by the occupant through various sound transmission
paths. HVAC&R equipment that acts as noise generators includes fans, compressors,
pumps, and dampers; this noise is in addition to other airflow noise in ducts. Of these, the fan is the
major noise source in HVAC&R equipment to be controlled and attenuated because fans are widely
used and have more sound transmission paths than compressors and pumps.
Sound transmits from the source to the receiver via the following paths:
 Duct-borne paths. These are noise transmission paths through ducts, duct fittings, and other air
system components, at either the supply air side or the return air side, as shown in Fig. 19.1. The
noise thus transmitted is often called system noise.
 Radiated sound path. This is the path along which noise radiates through the duct walls or casings
into the ceiling plenum or into an adjacent area and then transmits to the receiver.
 Airborne path. This path includes the equipment noise transmitted to adjacent areas through the
air.
 Structure-borne path. These are the paths along which equipment vibration and noise are transmitted
through the building structure.
19.2 CHAPTER NINETEEN
FIGURE 19.1 Fan room and sound paths.
Structure-borne transmission is actually a combination of sound and vibration effects. This problem
is usually solved through the combined effort of the structural and HVAC&R designers.
Control at Design Stage
The evaluation and control of all noise transmission paths should take place at the design stage, and
it should be a cooperative effort of the architect and the structural and mechanical engineers. If an
excessive noise level occurs due to improper design—the designer fails to analyze the potential
noise problem, there are mistakes in analysis, or data are used improperly—remedial measures to
reduce the noise level are usually expensive and are often less effective.
One effective measure to reduce noise in buildings is to locate the fan room or other sound
source away from critical areas, such as conference rooms and executive offices. For purposes of
noise control, the fan room may be located near rest rooms, stairwells, or copy rooms.
For noise control, theoretical estimation and prediction schemes are mainly guidelines. Previous
experience and field performance are important. Especially for projects with strict acoustic requirements,
field/laboratory tests and checkouts are often necessary.
Recommended Procedure for Noise Control
Before the analysis and evaluation of sound levels, the indoor design NC or RC criteria are usually
determined, and the necessary data to evaluate sound levels at various points of interest are collected
and investigated. The recommended procedure for noise control uses the following approach:
Source : Path (attenuation) : Receiver
The basic procedure is as follows:
1. Determine the sources of the noise from the fan, compressor, and pump, and the sound power
level generated. Use certified manufacturer’s data.
2. Carefully analyze all the possible sound paths that can transmit noise from the source to the
occupied zone. An overlooked sound path may affect the final results.
3. Calculate all the sound attenuation and transmission losses in each sound path during transmission.
4. Determine the airflow noise Laf, in duct-borne path due to dampers, elbow, or junctions and
the attenuated fan noise near the damper Lat, fan, both in dB. If Lat, fan  Laf  8 dB, Laf can be
ignored. Only when Lat, fan  Laf  8 dB should Laf be added to Lat, fan for duct-borne transmission
calculations.
5. For duct-borne paths, determine the attenuated sound power level of fan noise at the supply
outlet (system noise or room source Lwr), in dB re 1012W, that affects the receiver. Convert Lwr to a
sound pressure level received by the receiver at a chosen point in a room Lpr or at a plane 5 ft (1.5
m) from the floor Lpt, both in dB re 20 Pa.
6. For radiated sound transmissions or airborne transmissions, the sound power level of the
sound radiated into the receiving room should be converted to a room sound pressure level.
7. The resultant sound pressure level Lp at the center frequency of each octave band, in dB, that
is perceived by the receiver at a chosen point or at a plane 5 ft (1.5 m) from the floor level is the
sum of sound pressure levels from all sound paths. Determine the NC or RC LNC, in dB, from the
calculated Lp in each octave band.
8. Compare LNC with the design criteria NC or RC Lp, NC. If LNC  Lp, NC, then one should add
a silencer (or an active noise control device, which is discussed in later sections) to the duct-borne
sound path, change the configuration of the duct system, or build structures to bring LNC down to
or slightly below Lp, NC.
SOUND CONTROL 19.3
9. Check this noise control prediction scheme with the actual field results of projects of similar
noise control arrangements, if possible.
19.2 FAN, COMPRESSOR, PUMP, AND AIRFLOW NOISE
Fan Noise
Use the fan noise data provided by the manufacturer. If the manufacturer’s data are not available,
estimate the sound power level for the fan by Eq. (15.21). The specific sound power levels at fullload
operation are listed in Table 15.1.
Fans should be selected for maximum efficiency so as to produce less fan noise. When a fan is
operated at off-peak efficiency conditions, according to ASHRAE Handbook 1999, HVAC Applications,
an additional amount up to 6 dB may be added to a plenum fan sound power level operating
at the right-hand side of the peak efficiency along the fan curve. Because the calculated system
pressure loss is often less than the actual system pressure loss, a centrifugal fan is often
selected to operate just at the right-hand side of its rated peak efficiency. When a smaller fan is
operated at higher volume flow rates, a noise up to 5 dB will often be produced. Fans should not
be operated in surge or stall region at any time because fan noise and vibrations are substantially
increased.
According to ASHRAE Handbook 1999, HVAC Applications, backward-inclined (BI) or air foil
(AF) centrifugal fans often have higher sound power level in the mid- and upper-frequency range
especially at the blade passing frequency. They are more sensitive to inlet flow distortion than forward-
curved (FC) centrifugal fans. When inlet vanes are installed inside the fan inlet, the sound
level at the blade passing frequency increases 2 to 8 dB depending on the percentage of the volume
flow rate reduced. Externally mounted inlet vanes increase about 2 to 3 dB. BI and AF fans show a
narrower operating range and a clearly defined surging region than FC fans. Inside the operating
range, BI and AF fans are usually perceived to have lower sound power level at lower frequency
(below 100 Hz) than FC fans. For centrifugal fans, air volume flow control by varying fan speed
with PWM ac inverter is the acoustically preferred method.
Forward-curved centrifugal fans have a wide operating range without a clearly defined surge
region. They are comparatively insensitive to flow distortion at their inlet. In FC fans, modulation of
inlet vanes in a VAV system does not cause an increase in sound level at a frequency of 63 Hz and
above. FC fans are considered having 31.5- and 63-Hz rumbles.
Plug and plenum fans need a slightly greater energy use, however, they can have a significantly
lower sound power level if the fan plenum is properly sized and acoustically treated.
Axial fans are considered to have lower sound power level in lower frequency than centrifugal
fans. They are often used in applications where the high-frequency noise can be attenuated by
silencers. Axial fans are most sensitive to inlet flow obstructions. Varying blade angle to control
volume flow rate is predominately used in vane-axial fans for a VAV system when the system load
is reduced. When the air volume flow reduces from 80 to 40 percent, there is a corresponding noise
reduction of 2 to 5 dB between 125 and 4000 Hz.
Noise from Chillers and Pumps
If manufacturer’s data are not available, according to ASHRAE Handbook 1991, HVAC
Applications, the sound pressure level LpA, in dBA, for centrifugal chillers at a distance 3.3 ft (1 m)
from the chiller can be calculated as
(19.1)
where TR 	 refrigeration capacity, in tons. For reciprocating chillers, LpA in dBA at a distance of
LpA 	 60  11 log TR
19.4 CHAPTER NINETEEN
3.3 ft (1 m) is
(19.2)
The sound pressure level Lp, in dB, at the center frequency of various octave bands can be obtained
by adding the following values at each octave band to the calculated LpA:
The centrifugal chiller previously mentioned is constructed hermetically, has internal gears, and
operates at medium or full load. At light load operation, Lp may increase 10 to 13 dB for the shaft
frequency in the band, 8 to 10 dB in the blade pass frequency band, and 5 dB for the remaining
octave bands. For circulating pumps, LpA at a distance of 3.3 ft (1 m) is
(19.3)
where hp 	 power input to the pump, hp.
Airflow Noise
Airflow noise is an aerodynamic noise generated in flow passages or at duct fittings. Airflow noise
is the result of vortices passing around obstacles, such as elbows, dampers, branch takeoffs, terminals,
and diffusers, that cause local acceleration and deformation and thereby produce local dynamic
compression of the air. Airflow noise can also be generated when vortices pass through solid
discontinuities.
Airflow noise is mainly determined by the velocity of air flowing through the air passage (duct
velocities) and the constrictions due to duct fittings. The geometry of the duct fittings also affects
the airflow turbulence and vortices and, therefore, the airflow noise.
Maximum Duct Velocities. Airflow noise can be reduced by using low duct velocities, preventing
abrupt changes, and providing sufficient sound attenuation. According to ASHRAE Handbook 1999,
HVAC Applications, the maximum recommended airflow velocities in ducts or in free openings at
various specified RC (NC) ratings are as follows:
Maximum airflow velocity, fpm
Rectangular Round Supply Return
RC (NC) duct duct outlet inlet
In shaft or above drywall ceiling 45 3500 5000
35 2500 3500
25 1700 2500
Above suspended acoustic ceiling 45 2500 4500
35 1750 3000
25 1200 2000
Within occupied space 45 2000 3900
35 1450 2600
25 950 1700
Free openings 45 625 750
35 500 600
25 350 425
LpA 	 77  10 log hp
Center frequency of octave bands, Hz
63 125 250 500 1k 2k 4k
Centrifugal chiller 8 5 6 7 8 5 8
Reciprocating chiller 19 11 7 1 4 9 14
LpA 	 71  9 log TR
SOUND CONTROL 19.5
Branch ducts should have air velocities about 80 percent of the listed value. Airflow velocity of the
final runout to outlets should be 50 percent of the listed value. The presence of the elbow, diffuser,
and other fittings can increase airflow noise substantially, and therefore the duct velocity should
reduce accordingly.
Diffusers and Grilles. The manufacturer’s sound rating of a diffuser is obtained with a uniform
velocity distribution throughout the neck. If a volume damper is installed immediately before the
diffuser, the turbulent airflow at the neck will be significantly increased compared to the manufacturer’s
laboratory-tested data. This turbulence can be considerably reduced by adding an equalizing
grid in the neck of the diffuser. Volume dampers should not be located closer than 5 ft (1.5 m) from
a supply outlet.
The sound power level added to the diffuser sound rating is proportional to the damper pressure
ratio (DPR). The DPR is equal to the throttled pressure loss across the damper, divided by the minimum
pressure loss across the damper when its blades are fully opened. According to the data
provided by ASHRAE Handbook 1999, HVAC Applications, if a volume damper is installed in the
neck of a linear diffuser, when the DPR varies from 1.5 to 6, a corresponding 5 to 24 dB should be
added to the diffuser sound rating. If the volume damper is installed in the inlet of the plenum of a
linear diffuser, when DPR varies from 1.5 to 6, a corresponding of 2 to 9 dB should be added. If the
volume damper is installed at least 5 ft (1.5 m) from the inlet plenum of a linear diffuser, only a corresponding
of 0 to 5 dB should be added.
Airflow noise generated at the diffusers or grilles at the end of duct-borne paths is difficult to
attenuate except by reducing their neck or face velocities. Imbalance of volume flow between various
branch runouts may create a greater branch duct velocity and grille face velocity in one of the
branch runouts which has the lowest flow resistance.
Poor Fan Entry and Discharge Conditions. Noise is generated because of the abrupt air passage
constrictions and sudden changes in airflow direction. Both cause flow turbulence and flow separation
as well as energy losses. Such low-frequency noise results in a duct rumble and is very difficult
to attenuate. The designer should carefully design the fan intake and discharge connections. Doing
so is the best way to control this kind of airflow noise as well as to provide an energy-efficient air
system.
19.3 SOUND ATTENUATION ALONG DUCT-BORNE PATH
Sound Attenuation in Ducts
Sound attenuation is the reduction in the intensity of sound, expressed in watts per unit area, as it
travels along a sound transmission path from a source to a receiver. Sound attenuation is achieved
by (1) the absorption of sound energy by the absorptive material, (2) spherical spreading and
scattering, and (3) reflection of sound waves incident upon a surface.
Sound attenuation in duct sections, duct fittings, silencers, and other sound-reducing elements
can be indicated by insertion loss. Insertion loss (IL) at a specific frequency is the reduction in
sound power level, in dB re 1012 W, measured at the receiver when a sound attenuation element is
inserted in the transmission path between the sound source and the receiver. Noise reduction is the
difference in sound pressure levels between any two points along the sound transmission path.
Sound absorptivity is the ability of a material to absorb sound energy. When a sound wave
impinges on the surface of a porous sound-absorbing material, air vibrates within the small pores.
The flow resistance of air and its vibration converts a portion of the absorbed sound energy to heat.
The fraction of the incident sound power that is absorbed is called the sound absorption coefficient
	. Most sound-absorbing materials have a low 	at low frequencies and a higher 	at high frequencies.
For a typical sound absorption material, 	 might be equal to 0.15 in the octave band whose
center frequency is 63 Hz, and 		 0.9 at 1000 Hz.
19.6 CHAPTER NINETEEN
Ducts can be unlined or inner-lined with foil-coated encapsulated fiberglass or other soundabsorptive
materials, usually at a thickness of 1 in. (25 mm), sometimes 2 in. (50 mm), to absorb
fan noise. Inner-lined ducts have better sound attenuation than unlined ducts, and the inner-lined
sound-absorbing layer can also serve as thermal insulation.
In unlined round ducts, the sound attenuation, called natural attenuation, is only about one-tenth
that in unlined rectangular ducts as follows:
Natural attenuation of unlined rectangular ducts is listed in Table 19.1.
For inner-lined ducts, the sound attenuation, or insertion loss, of a given length of duct section
depends on the cross-sectional area and the properties of the absorptive material. Sound attenuation
in inner-lined round duct with a 1-in. (25-mm) thickness of duct liner is listed in Table 19.2, and for
1-in-. (25-mm-) thick inner-lined straight rectangular ducts it is listed in Table 19.3.
Round duct attenuation, dB/ft
Octave band center frequency, Hz
Diameter, in. 125 250 500 1000
D  7 0.03 0.05 0.05 0.10
7 D  15 0.03 0.03 0.05 0.07
15 D  30 0.02 0.02 0.03 0.05
30 D  60 0.01 0.01 0.02 0.02
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted
with permission.
SOUND CONTROL 19.7
TABLE 19.1 Approximate Natural Attenuation of Sound in
Unlined Rectangular Sheet-Metal Ducts
Attenuation, dB/ ft
Duct size, P/A, Octave band center frequency, Hz
in. 
 in. 1/ ft 63 125 250 250
6 
 6 8.0 0.30 0.20 0.10 0.10
12 
 12 4.0 0.35 0.20 0.10 0.06
12 
 24 3.0 0.40 0.20 0.10 0.05
24 
 24 2.0 0.25 0.20 0.10 0.03
48 
 48 1.0 0.15 0.10 0.07 0.02
72 
 72 0.7 0.10 0.10 0.05 0.02
P/ A ratio indicates perimeter / area ratio.
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted with
permission.
TABLE 19.2 Round Duct Attenuation (Insertion Loss)
Octave band center frequency, Hz
Duct 63 125 250 500 1000 2000 4000 8000
diameter, in. Approximate attenuation for 1-in. duct liner, dB/ ft
6 0.38 0.59 0.93 1.53 2.17 2.31 2.04 1.26
12 0.23 0.46 0.81 1.45 2.18 1.91 1.48 1.05
24 0.07 0.25 0.57 1.28 1.71 1.24 0.85 0.80
48 0 0 0.18 0.63 0.26 0.34 0.45 0.44
Source: Abridged with permission from ASHRAE Handbook 1995, HVAC Applications.
19.8 CHAPTER NINETEEN
TABLE 19.3 Sound Attenuation (Insertion Loss) for Straight Lined Sheet-Metal Rectangular Ducts, Lining
Thickness 1 in.
Internal
Octave and center frequency, Hz
cross-sectional 63 125 250 500 1000 2000 4000 8000
dimensions, in. Attenuation, dB/ ft
8 
 8 0.77 0.90 1.29 2.57 6.23 5.94 3.49 2.42
8 
 18 0.60 0.67 1.04 2.22 5.10 4.72 3.06 2.26
12 
 12 0.56 0.60 0.96 2.09 4.70 4.29 2.90 2.20
12 
 24 0.50 0.48 0.79 1.81 3.85 3.41 2.54 2.06
18 
 18 0.40 0.43 0.74 1.70 3.54 3.10 2.41 2.00
18 
 36 0.30 0.32 0.59 1.47 2.90 2.46 2.11 1.87
24 
 24 0.30 0.32 0.59 1.47 2.90 2.46 2.11 1.87
24 
 48 0.23 0.24 0.47 1.27 2.37 1.95 1.85 1.76
36 
 36 0.21 0.21 0.43 1.20 2.19 1.78 1.75 1.71
36 
 72 0.16 0.16 0.34 1.03 1.79 1.41 1.54 1.60
48 
 48 0.16 0.16 0.34 1.03 1.79 1.41 1.54 1.60
48 
 72 0.14 0.13 0.29 0.94 1.58 1.22 1.42 1.54
1. Based on measurements of surface-coated duct liners of 1.5 to 3 lb / ft3 density. For the specific materials tested, liner density
had a minor effect over the nominal range of 1.5 to 3 lb / ft3.
2. Add natural attenuation (Table 19.1) to obtain total attenuation.
Source: Abridged with permission from ASHRAE Handbook 1995, HVAC Applications.
TABLE 19.4 Lined Flexible Duct Insertion Loss
Insertion loss dB
Diameter, Length,
Octave band center frequency, Hz
in. ft 63 125 250 500 1000 2000 4000
8 12 8 11 21 33 37 37 24
9 6 8 16 25 28 28 18
6 4 6 11 17 19 19 12
3 2 3 5 8 9 9 6
9 12 8 11 22 33 37 36 22
9 6 8 17 25 28 27 17
6 4 6 11 17 19 18 11
3 2 3 6 8 9 9 6
10 12 8 10 22 32 36 34 21
9 6 8 17 24 27 26 16
6 4 5 11 16 18 17 11
3 2 3 6 8 9 9 5
12 12 7 9 20 30 34 31 18
9 5 7 15 23 26 23 14
6 3 5 10 15 17 16 9
3 2 2 5 8 9 8 5
14 12 5 7 16 27 31 27 14
9 4 5 12 20 23 20 11
6 3 4 8 14 16 14 7
3 1 2 4 7 8 7 4
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted with permission.
Lined flexible ducts reduce duct-borne noise significantly. Insertion losses for lined flexible ducts
at various diameters and duct lengths are listed in Table 19.4. The length of the flexible duct should
normally be from 3 to 6 ft (0.9 to 1.8 m). An abrupt bend may create unacceptably high noise.
The insertion loss (sound attenuation) of lined round, flat oval, rectangular, and flexible ducts is
a function of their inside diameter (or the smaller inside dimension of rectangular and flat oval
ducts). Generally, as the inside diameter increases, the insertion loss decreases. Surprisingly, the 3 : 1
aspect ratio flat oval duct has approximately the same IL as a round duct with the same diameter as
the minor axis.
The thickness of the duct liner has a definite effect on the IL of lined ducts. For duct liner of rectangular
duct with a thickness of 2 in. (50 mm), the sound attenuation is nearly double the value of
a 1-in. (25-mm) duct liner for octave band center frequencies between 250 and 500 Hz. The ILs are
nearly the same for both round and rectangular duct liners with a thickness of 1 in. (25 mm) for
octave band center frequencies of 1000 Hz and greater.
According to Bodley (1981), the density of the duct liner material has a minor effect on the IL
of inner-lined ducts. The IL is affected by the absorption coefficient of the lining material and its
thickness.
Sound Attenuation at Elbows and Branch Takeoffs
Sound is reflected by elbows. Lined or unlined elbows provide sound attenuation to reduce the
duct-borne noise. Sound attenuation depends on the size of duct, whether the elbow has a duct lining
before and / or after the elbow, and whether turning vanes are provided. The estimated sound
attenuation for unlined round elbows is indicated in Table 19.5.
A mitered 90° elbow, often called a square elbow, provides effective noise attenuation by means
of reflection loss in the octave bands whose center frequencies equal or exceed 125 Hz. Rumble
noise is difficult to attenuate even for a square elbow. In Table 19.6 are listed the insertion losses for
elbows that are followed or proceeded by duct lining. Square elbows with turning vanes are slightly
less effective in sound attenuation than elbows without turning vanes.
For any branch takeoff from a diverging wye in a main duct, the dominating sound attenuation is
the acoustic energy distribution taking place between the main duct and the branch ducts. That part
of sound attenuation because of the branch power division Lw,b, in dB, is proportional to the ratio
of the area of branch Abr, in ft2 (m2), to the sum of the area of branches Abr, in ft2 (m2), and can be
calculated as
(19.4) Lw,b 	 10 log
Abr
Abr
SOUND CONTROL 19.9
TABLE 19.5 Insertion Loss of Unlined
Round Elbows
Insertion loss,
dB
fw 1.9 0
1.9 fw 3.8 1
3.8 fw 7.5 2
fw  7.5 3
fw 	 f 
 w
f 	 frequency, kHz
w 	 diameter for round duct, in.
Source: ASHRAE Handbook 1995, HVAC Applications.
Reprinted with permission.
The sound attenuation of branch power division Lw,b should be subtracted from the sound power
level of duct-borne noise at the common end of the diverging wye in the main duct just before the
branch takeoff Lw,c, in dB, to obtain the sound power level of the branch takeoff Lw,b, in dB,
(19.5)
If the minor sound attenuations in the diverging wye are ignored, based on the principle of energy
conservation, the sum of the sound power levels of the duct-borne noise at the straight-through end
Lw,s, in dB, and the branch takeoff Lw,b just after the diverging wye Lw,s  Lw,b 	 Lw,c, from
Eq. (19.5); therefore
(19.6)
End Reflection Loss
When a sound wave propagates through the end opening of a duct section into a room, a significant
amount of sound is reflected back to the duct, and thus substantially reduces the low-frequency
sound that reaches the receiver in the occupied zone. According to ASHRAE Handbook 1995,
HVAC Applications, the sound attenuation of end reflection loss for ducts terminated in free space
and for ducts terminated flush with a wall is listed in Table 19.7. Diffusers that terminate in a
suspended acoustic ceiling can be considered as terminating in free space. In Table 19.7, for a diffuser
with a rectangular connecting duct, its diameter D 	 , in in. (mm), where A 	 area of
the rectangular duct, in.2 (mm2).
Values listed in Table 19.7 were tested based on straight round connecting ducts leading to the
diffusers. In actual installation, there may be dampers and guide vanes. The connecting duct
between a difffuser and a main duct or terminal may not be straight and smooth. All these factors
may affect the listed values in Table 19.7.
4A /
Lw,s 	 Lw,b
Lw,b 	 Lw,c  Lw,b
19.10 CHAPTER NINETEEN
TABLE 19.6 Insertion Loss of Square Elbows
Insertion loss, dB
Unlined Lined
Without turning vanes
fw 1.9 0 0
1.9 fw 3.8 1 1
3.8 fw 7.5 5 6
7.5 fw 15 8 11
15 fw 30 4 10
fw  30 3 10
With turning vanes
fw 1.9 0 0
1.9 fw 3.8 1 1
3.8 fw 7.5 4 4
7.5 fw 15 6 7
fw  15 4 7
fw 	 f 
 w
f 	 frequency, kHz
w 	 duct width or depth of square duct or diameter
for round duct, in.
Source: ASHRAE Handbook 1995, HVAC Applications.
Reprinted with permission.
Duct-Borne Crosstalk
Duct-borne crosstalk or plenum crosstalk is the sound transmitted between two rooms through the
duct systems or ceiling plenum. Usually, the sound attenuation in the duct system or in the ceiling
plenum should be 5 dB greater than the transmission loss of the architectural partition between two
rooms. Transmission loss is discussed in detail in the following sections.
Duct lining, devious duct runs, flexible ducts, inner lining in the slot diffusers, and return slots
are often used to attenuate the sound transmitted through duct-borne or plenum-connected paths to
avoid crosstalk. A well-sealed architectural partition in the ceiling plenum with an inner-lined return
air passage effectively prevents crosstalk through the ceiling plenum for applications with strict
NC or RC criteria in the occupied zone.
Attenuation along Duct-Borne Path
For a typical VAV system with a supply fan installed in the AHU inside a fan room, as shown in
Fig. 19.1, the sound attenuation along the duct-borne paths—the supply duct side and the return
duct side—are as follows:
Supply duct-borne path D1 Return duct-borne path D2
Supply side silencer Return side silencer
Supply duct attenuation Return duct attenuation
Elbows attenuation Elbows attenuation
SOUND CONTROL 19.11
TABLE 19.7 Duct End Reflection Loss
End reflection loss, dB
Duct diameter,
Octave band center frequency, Hz
in. 63 125 250 500 1000 2000
Duct terminated in free space
6 20 14 9 5 2 1
8 18 12 7 3 1 0
10 16 11 6 2 1 0
12 14 9 5 2 1 0
16 12 7 3 1 0 0
20 10 6 2 1 0 0
24 9 5 2 1 0 0
28 8 4 1 0 0 0
32 7 3 1 0 0 0
Duct terminated flush with wall
6 18 13 8 4 1
8 16 11 6 2 1
10 14 9 5 2 1
12 13 8 4 1 0
16 10 6 2 1 0
20 9 5 2 1 0
24 8 4 1 0 0
28 7 3 1 0 0
32 6 2 1 0 0
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted by permission.
On the return air side, an air filter made of a fibrous material may have a certain degree of sound attenuation.
This could be considered an added safe factor in the return duct-borne path. Ceiling
plenum attenuation is discussed in the following section.
All sound attenuation should be subtracted from the sound power level of the supply fan Lw,f, in
dB re 1012 W, and converted to the sound pressure level at the receiver Lp,r, in dB re 20 Pa,
through the room effect.
19.4 SILENCERS
The purpose of a silencer is to reduce the sound power level of a fan, an airflow noise, or other
sound source transmitted along a duct-borne path or airborne path to a required level.
Supply duct-borne path D1 Return duct-borne path D2
Branch power division Ceiling plenum attenuation or branch power division
Flexible duct attenuation Return slot attenuation
Room effect Room effect
19.12 CHAPTER NINETEEN
FIGURE 19.2 Rectangular and cylindrical silencers: (a) rectangular silencer and (b) cylindrical silencer.
Types of Silencer
Silencers can be classified into rectanguar, cylindrical, and sound-attenuating plenum according to
their configuration. A typical rectangular splitter silencer is shown in Fig. 19.2a. Inside the retangular
casing are a number of flat splitters, depending on the width of the silencer. These splitters direct
the airflow into small sound-attenuating passages. The splitter is made from an envelope containing
sound-attenuating material, such as fiberglass or mineral wool, with protected noneroding facing.
The thickness of a splitter is often between 1 and 4 in. (25 and 100 mm). Splitters often have a
round instead of a flat nose, to reduce their airflow resistance. A rectangular silencer is often connected
with retangular ducts or sometimes with rectangular fan intakes and discharges.
A cylindrical silencer has an outer cylindrical jacket and an inner concentric center body, as
shown in Fig. 19.2b. Both the cylindrical jacket and the center body contain sound-attenuating material
and noneroding facing. A cylindrical silencer is often used in conjunction with vane-axial fans
and in round duct systems.
A sound-attenuating plenum consists of several splitters in the form of two, three, or four
successive square elbows, shown in Fig. 19.3, filled with sound-absorbing material with noneroding
SOUND CONTROL 19.13
FIGURE 19.3 Sound-attenuating plenum.
facing. A sound plenum effectively attenuates low-frequency noise and is often located between the
collective return grille and the return air intake of a packaged unit. A low velocity is important in a
sound-attenuating plenum because of the high local loss coefficient of square elbows.
Silencers used in HVAC&R systems can again be classified according to the means of sound
attenuation into following types:
1. Dissipative silencers. These silencers often use face-covered or encapsulated acoustic material,
such as fiberglass, mineral wool, and acrylic polymers, to attenuate noise over a broad range of
frequencies. The facing material can be made of
 Galvanized or aluminum sheet with a perforated area not exceeding 22 percent of the face
area
 Acrylic polymers or polymer sheets
2. Packless silencers. There is no fibrous fill. Noise is attenuated by means of acoustically resistive
perforations in the splitters. They are often made of sintered aluminum or acrylic plastics.
3. Reflection-dissipative silencers. These silencers use the combined effect of sound reflection and
dissipation in airflow passages of successive square elbows.
4. Active silencers. These silencers produce low-frequency inverse sound waves to cancel the unwanted
noise.
Characteristics of Silencers
The acoustic and aerodynamic characteristics of a silencer are mainly indicated by four parameters.
19.14 CHAPTER NINETEEN
FIGURE 19.4 Difference in insertion loss of a rectangular silencer between
forward and reverse flow (at 2000 fpm or 10 m / s face velocity).
(Source: Handbook of HVAC Design, 1990. Reprinted by permission.)
Insertion Loss. This is the capacity of a silencer to reduce the sound power level of fan noise or
other noise at various octave band center frequencies. The IL of a silencer is affected by the air-
flow direction, especially when air velocity is greater than 200 fpm (1 m / s). A sound wave that
propagates in the same direction as the airflow is said to be in forward flow, and one that propagates
opposite to the airflow is said to be in reverse flow. At low frequencies, reverse flow has a
longer contact time and, therefore, a higher IL. At high frequencies, sound waves tend to refract
toward the absorptive surface in a silencer under reverse flow. Therefore, these effects increase the
high-frequency attenuation in the forward flow and decrease the attenuation in reverse flow. The
difference in IL between forward and reverse flow of a rectangular silencer at 2000 fpm (10 m / s)
face velocity is presented in Fig. 19.4.
Operating Parameters. The operating parameters for a silencer are as follows:
 Volume flow rate of air through the silencer sil, in cfm (m3 /s)
 Cross-sectional area of the silencer, including the free area Afree, ft2 (m2), i,e., area of the passages
through which air flows, and the cross-sectional area of the silencer Asil, ft2 (m2)
 Free area ratio Rfree, which is defined as Rfree 	 Afree /Asil
 Face velocity vsil, in fpm (m / s), defined as vsil 	 sil /Asil
Many silencers have a free air ratio between 0.3 and 0.8. The actual mean air velocity inside
the airflow passages in a silencer vfree, in fpm (m / s), may be 1.25 to 3.3 times the face velocity
vsil. The face velocity of a silencer and sound-attenuating plenum is 500 to 2000 fpm (2.5 to
10 m/ s).
Self-Noise. This is the lower limit of sound power level, in dB, that a specific silencer can approach
at various octave band center frequencies.
Pressure Drop psil. This is the total pressure drop of airstream when it flows through a
silencer. Pressure drop psil is a function of its face velocity, free area ratio, length, and the con-
figuration of the splitter or the center body. In general, ASHRAE Handbook 1995, HVAC Applications,
recommends that if psil  0.35 in. WC (87 Pa), both vfree and airflow noise should be
investigated.
Location of Silencers
The location of a silencer to attenuate fan noise should meet the following two requirements:
1. There must be a minimum distance Lsil between the upstream fan discharge outlet or other duct
fittings and the silencer, to ensure a uniform approach velocity at silencer inlet or an undisturbed
discharging velocity at the silencer exit. From the fan discharge, Lsil must be equal to or greater
than the distance of one duct diameter for every 1000 fpm (5 m / s) average duct velocity. From
the fan intake, Lsil should be equal to or greater than 0.75 duct diameter for every 1000 fpm
(5 m/ s) average duct velocity.
2. The silencer should straddle or be adjacent to the solid wall of the fan room (or mechanical
room), as shown in Figs. 16.7a and 16.8a. Such a setup can prevent or reduce the following fan
noise transmissions:
 The fan noise that has not been attenuated by the silencer will not break out into the room adjacent
to the fan room.
 The noise in the fan room may have the chance to break in to the supply main duct after the
silencer.
V?
V?
SOUND CONTROL 19.15
Active Silencers
Active silencers use ducted enclosure to cancel duct-borne, low-frequency fan noise (including
rumbles) by producing sound waves of equal amplitude and opposite phase. The primary sound
source is the unwanted fan noise. The secondary sound source which cancels the unwanted source
comprises the inverse sound waves from a loudspeaker.
Operating-Characteristics. An active silencer consists of two or more microphones, a microprocessor-
based controller, a loudspeaker, and a ducted enclosure. When a fan noise is propagated
down along a duct, as shown in Fig. 19.5, an input microphone located in the airstream measures
the noise and sends an electric signal proportional to the sound wave to the controller. The microprocessor-
based controller calculates the amplitude, frequency, and phase of the propagating sound
and sends a cancel signal to the loudspeaker. The loudspeaker broadcasts sound waves of the same
amplitude and frequency as the unwanted noise, but 180° out of phase. The destructive interference
between these two sound sources results in the cancellation of the incident fan noise by the secondary
source broadcasted by the loudspeaker. The amplitude of the fan noise is reduced downstream
of the loudspeaker. An error microphone measures the residual noise. This feedback is used
to optimize the performance of the active silencer.
Frequency Limits. Gelin (1997) noted that in an active silencer, the ducted enclosure acts as a
waveguide for the sound waves. At lower frequencies, when the wavelength of the sound is at least
2 times longer than the greatest dimension of the waveguide, sound waves may propagate in plain
waves in which sound pressure and the phase of the sound wave are uniform across any duct cross
section. At intermediate and high frequencies, the wavelength is smaller than a waveguide dimension,
and the sound waves may propagate as plain waves as well as more complicated sound waves.
A low-frequency fan noise is easily measured, defined, and reproduced 180° out of phase, whereas
for a high-frequency noise it is very difficult to do these things. Therefore, active silencers work
best at an octave band center frequency range of 31 to 250 Hz.
System Characteristics. The following are the system characteristics of active silencers:
 In an active silencer, sound energy is added to the fan noise to cancel it, whereas in a traditional
passive silencer (such as dissipative and packless silencers), sound energy is converted to heat and
removed from the system.
 Because the microphone in an active silencer cannot distinguish the fan noise and the turbulent air-
flow noise, it is acceptable practice that the air velocity in the ducted enclosure of an active silencer
19.16 CHAPTER NINETEEN
Fan
noise
DDC
controller
Input
microphone
Error
microphone
Loudspeaker
FIGURE 19.5 An active silencer.
not exceed 1500 fpm (7.5 m / s). When the duct velocity exceeds 2500 fpm (12.5 m / s), the turbulent
noise may completely mask fan noise, and the effect of an active silencer is reduced to nil.
 Pressure drop of a passive silencer is often 0.2 to 0.35 in. WC (50 to 63 Pa). The electric energy
required to produce the canceling sound is only 40 W, a substantial saving compared to a passive
silencer.
 Due to the development of the microprocessor-based technologies, in 1997, the price of the controller
used in an active silencer dropped substantially compared to that in 1987.
Performance. Gelin (1997) compared the insertion loss, in dB, of an active silencer, a passive silencer,
and a duct liner for a large fan of 5000 cfm (2360 L / s) volume flow rate as follows:
Octave bands, Hz 31 63 125 250 500 1000 2000 4000
Active silencer 5–8 10–15 10–15 2–4 — — — —
1-in. Lined, 84-in.- (2.1-m-) long duct 2 2 4 10 20 17 15
84-in. (2.1-m) Prefabricated silencer 8 17 26 43 43 26 19
The pressure drop of the 84-in. (2.1-m) prefabricated silencer is 0.24 in. WC (60 Pa).
The active silencer has a good sound attenuation in frequencies between 31 and 125 Hz, the duct
liner provides effective sound attenuation in frequencies of 500 Hz and more, and the 84-in. (2.1-m)
prefabricated silencer is effective in frequencies between 63 and 4000 Hz.
Selection of Silencers
During the selection of silencers, the following points should be considered:
 Dominating frequencies of noise. For a low-frequency fan “rumble” noise, an active silencer or a
reflective dissipative silencer should be selected. For a broadband noise, a combination of active
and dissipative silencers or a dissipative silencer with thick splitters, of 2- to 4-in. (50- to 100-
mm) thickness, is preferable.
 Design requirements. For clean rooms, fiber-free all-metal silencers are often selected. For health
care facilities, fiber-free packless silencers should be selected.
 Wet surfaces. If there are wet surfaces in the air system, fiber-free all-metal silencers should be selected.
For a specific type of silencer, variations in its length and free area ratios often meet most of the requirements
for IL and pressure drops. Silencers are usually made in lengths of 3, 5, 7, and 10 ft
(0.9, 1.5, 2.1, and 3 m). The IL and psil of a silencer are proportional to its length.
A lower face velocity and greater free air ratio for a specific silencer at a given length always result
in a lower pressure drop and therefore lower energy consumption. If space is allowed, a lower
vsil is beneficial for both IL and psil.
19.5 FIBERGLASS IN HVAC&R SYSTEMS
Problems
Fiberglass-made duct liners, duct boards, silencers, and insulation layers have been widely used in
HVAC&R systems for decades. On June 24, 1994, the U.S. Department of Health and Human Services
(DHHS) announced that fiberglass will be listed as a material “reasonably anticipated to be a
SOUND CONTROL 19.17
carcinogen.” According to ASHRAE, recently the use of fiberglass in several institutional, eduational,
and medical projects has been banned or severely limited because of the concerns that the
fibers may be carcinogenic and fiberglass may promote microbial growth.
The International Agency for Research on Cancer (IARC) has performed extensive research
concerning the carcinogenicity of fiberglass materials and found inadequate evidence to link insulation
of fiberglass with cancer in humans. IARC does indicate a possible link to cancer from glass
wool based on heavy dosage in animals. The Canadian Environment Protection Act indicated that
studies have failed to show any evidence that fiberglass is carcinogenic and classified it as “unlikely
to be carcinogenic to humans.”
Concerning the fiber erosion, more recent studies reported in the 6th International Conference
on Indoor Air Quality and Climate in Helsinki, Finland, 1993 that airborne fiber levels in buildings
with these products range from nondetectable up to only 0.006 fiber per cm3 (1 in.3 	 16.4 cm3).
Both liquid water and dirt are required for mold and microbial growth, and microbial growth can
occur on any surface in an HVAC&R system. Tinsley (1998) concludes that “. . . the presence of
high humidity air, by itself, is not sufficient to cause fungal growth. . . . fungal growth occurred
only where both standing water and nutrients (in the form of dirt, pollen, or other organics) could
be found.” Control of the wetted surface and dirt in the duct system is the most effective method to
reduce potential microbial growth.
Recommendations
1. For health care facilities, schools, semiconductor fabricating, food processing, pharmaceutical
manufacturing, clean rooms, and many more demanding projects, fiberglass duct liners and fiberglass
silencers should be replaced by any one or a combination of the following alternatives:
 Foil-coated products such as acrylic polymer-coated fiberglass duct liners or silencers
 Fibrous acoustic infill (fiberglass or mineral wool) encapsulated inside the polymer sheets
 Fiber-free sound absorber using sintered aluminum or acrylic polymer with microperforations
 Active silencers
2. For air systems that have wetted surfaces (standing water) from outdoor air intakes, cooling
coils, coil condensate droplet carryover, drain pans, and humidification processes, especially atomizing
devices, direct evaporative coolers, and air washers, all-metal fiber-free silencers or active silencers
should be used to attenuate duct-borne fan and airflow noise, if required. Interior surface of
ducts that contact with supply air must be galvanized metal sheet. Duct insulation should be
wrapped externally.
3. For air systems that employed draw-through fans and medium- or higher-efficiency air filters
and had no wetted surfaces (standing water) in contact with the air in the air system (except in the
coil section and condensate pan) having a face velocity of the cooling or DX coil of 500 fpm
(2.5 m/ s) or less, and properly designed, installed, operated, and maintained, fiberglass duct liners
and fiberglass silencers with perforated metal facing sheet can be used.
Periodic monitoring of the moisture and dirt conditions of the fiberglass duct liner and silencers
must be provided.
19.6 RADIATED NOISE AND TRANSMISSION LOSSES
Break-out and Break-in
When a duct section enters the ceiling plenum or directly passes over occupied spaces carrying
duct-borne fan noise that is not well attenuated, the noise radiates through the duct wall, raises the
19.18 CHAPTER NINETEEN
sound power level in the ceiling plenum or the occupied zone, and is transmitted to the receiver.
Noise that radiates through the duct wall and causes the duct wall to vibrate, caused either by internal
sound waves or by airflow turbulence, is called breakout. Noise can also be transmitted into a
duct and then travel along the duct-borne path, either discharging into a space through the duct
opening or breaking out into ambient air, where it is transmitted to the receiver. The transmission of
external noise into a duct section through the duct wall is called break-in.
Breakout fan noise from the discharge duct directly under a rooftop packaged unit into the ceiling
plenum may cause a serious acoustic problem in the underlying conditioned space.
Duct Rumble
Duct rumble is the result of air pressure fluctuations caused by variations in the speed of the fan,
fan motor, or fan belt or due to the airflow instabilities transmitted to the fan housing or nearby
ductwork. When the air pressure fluctuations are exerted on large, flat, and unreenforced duct surfaces
that have resonance frequencies near or equal to the fluctuating frequencies, the duct surfaces
vibrate. Poor fan outlet connections often generate duct rumbles. Duct rumble can produce sound
pressure levels of 65 to 95 dB at frequencies from 10 to 100 Hz.
Changing the fan speed as well as the frequencies of the air pressure fluctuations is one method
to reduce duct rumble. Another method is to increase the rigidity of the duct wall and thus change
the resonance frequencies. Low-frequency duct rumble noise is very difficult to attenuate by duct
liners.
Transmission Loss
Transmission loss (TL) is the reduction of sound power when sound is transmitted through a wall, a
partition, or a barrier. The relationship between transmission loss and mass of the material indicates
that the sound transmission loss of a homogeneous solid partition is a function of its mass and the
frequency of sound f transmitted through it. This relationship can be expressed in the following
form:
(19.7)
where ms 	 surface density of homogeneous partition, lb / ft2 (kg /m2). Concrete walls and galvanized
sheet ducts show that for each doubling of the mass partition, TL increases about 2 to 3 dB
for low-frequency noise (less than 800 Hz), and TL increases about 5 to 6 dB for high-frequency
sound (800 Hz and over).
The TL-mass relationship is valid only when sound is incident on the surface of the partition in
a normal direction and is an approximation. Actual measured data may show considerable deviation
from the predicted values. Factors that cause deviations are nonhomogeneity, cracks, stiffness, and
resonance. Cracks and gaps around doors, windows, duct and piping sleeves, or other openings
on the partition or wall may considerably reduce TL. The TL of a well-sealed partition of 100 ft2
(9.3 m2) surface area may drop from 40 to 20 dB because of a total of 1 ft2 (0.093 m2) of openings
on that partition.
Breakout and Break-in Sound Power Level
According to ASHRAE Handbook 1999, HVAC Applications, the breakout sound power level radiated
from the outer surface of duct walls Lw,out, in dB, can be calculated as
(19.8) Lw,out 	 Lw,in  10 log
S
Ai
 TLout
TL 	 F(log ms, log f )
SOUND CONTROL 19.19
where Lw,in 	 sound power level of sound inside duct, dB
S 	 surface area of outer sound-radiating duct surface, in.2 (mm2)
Ai 	 inner cross-sectional area of duct, in.2 (mm2)
TLout 	 duct breakout transmission loss, dB
Equation (19.8) assumes no sound attenuation along unlined interior duct surface. Such a breakout
transmission model is shown in Fig. 19.6.
19.20 CHAPTER NINETEEN
FIGURE 19.6 Breakout transmission loss TLout.
TABLE 19.8 Duct Breakout Transmission Loss TLout for Round Duct at Various Octave Band Center
Frequencies
TLout, dB
Diameter, Length, Octave band center frequency, Hz
in. ft Gauge 63 125 250 500 1000 2000 4000
Long seam ducts
8 15 26 45 (53) 55 52 44 35 34
14 15 24 50 60 54 36 34 31 25
22 15 22 47 53 37 33 33 27 25
32 15 22 (51) 46 26 26 24 22 38
Spiral-wound ducts
8 10 26 48 64 75 72 56 56 46
14 10 26 43 53 55 33 34 35 25
26 10 24 45 50 26 26 25 22 36
26 10 16 48 53 36 32 32 28 41
32 10 22 43 42 28 25 26 24 40
Note: In cases where background sound swamped the sound radiated from the duct walls, a lower limit on TLout is indicated
by a  sign. Parentheses indicate measurements in which background sound has produced a greater uncertainty than usual.
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted by permission.
TABLE 19.9 TLout for Rectangular Ducts at Various Octave Bands, dB
Duct size Octave band center frequency, Hz
in. Gauge 63 125 250 500 1000 2000 4000 8000
12 
 12 24 21 24 27 30 33 36 41 45
12 
 24 24 19 22 25 28 31 35 41 45
12 
 48 22 19 22 25 28 31 37 43 45
24 
 24 22 20 23 26 29 32 37 43 45
24 
 48 20 20 23 26 29 31 39 45 45
48 
 48 18 21 24 27 30 35 41 45 45
48 
 96 18 19 22 25 29 35 41 45 45
The data are tests on 20-ft-long ducts, but the TL values are for ducts of the cross section shown regardless of length.
Source: Adapted with permission from ASHRAE Handbook 1991, HVAC Applications.
As a result of duct sound breakout, the sound pressure level in an occupied space Lp, in dB, can
be calculated as
(19.9)
where r 	 distance between duct and receiver’s position where Lp is calculated, ft (m)
L 	 length of duct sound-radiating and breakout surface, ft (m)
The break-in sound power level transmitted into duct and then transmitted upstream or downstream
Lw,in, in dB, can be calculated as
(19.10)
where TLin 	 duct break-in transmission loss, dB.
Table 19.8 lists the duct breakout transmission losses TLout for round ducts at various octave
band center frequencies, Table 19.9 lists TLout for rectangular ducts, and Table 19.10 lists TLout for
flat oval ducts. Also Table 19.11 lists the duct break-in transmission losses TLin for round ducts at
various octave band center frequencies, Table 19.12 lists TLin for rectangular ducts, and Table 19.13
lists TLin for flat oval ducts.
In lower frequencies, round ducts have the highest TLout and TLin, and rectangular ducts have the
lowest TLout and TLin among these three kinds of ducts.
Lw,in 	 Lw,out  TLin  3
Lp 	 Lw,out  10 log rL  10
SOUND CONTROL 19.21
TABLE 19.10 TLout for Flat Oval Ducts at Various Octave Bands, dB
Duct size TLout, dB
(a 
 b) Octave band center frequency, Hz
in. Gauge 63 125 250 500 1000 2000 4000 8000
12 
 6 24 31 34 37 40 43 — — —
24 
 6 24 24 27 30 33 36 — — —
24 
 12 24 28 31 34 37 — — — —
48 
 12 22 23 26 29 32 — — — —
48 
 24 22 27 30 33 — — — — —
96 
 24 20 22 25 28 — — — — —
96 
 48 18 28 31 — — — — — —
The data are tests on 20-ft-long ducts, but the TL values are for ducts of the cross section shown regardless of length.
Source: Adapted with permission from ASHRAE Handbook 1991, HVAC Applications.
19.22 CHAPTER NINETEEN
TABLE 19.11 TLin for Round Ducts at Various Octave Bands, dB
TLin, dB
Diameter, Length, Octave band center frequency, Hz
in. ft Gauge 63 125 250 500 1000 2000 4000
Long seam ducts
8 15 26 17 (31) 39 42 41 32 31
14 15 24 27 43 43 31 31 28 22
22 15 22 28 40 30 30 30 24 22
32 15 22 (35) 36 23 23 21 19 35
Spiral-wound ducts
8 10 26 20 42 59 62 53 43 26
14 10 26 20 36 44 28 31 32 22
26 10 24 27 38 20 23 22 19 33
26 10 16 30 41 30 29 29 25 38
32 10 22 27 32 25 22 23 21 37
Note: In cases where background sound swamped the sound radiated from the duct walls, a lower limit on TLin is indicated
by a  sign. Parentheses indicate measurements in which background sound has produced a greater uncertainty than usual.
Source: ASHRAE Handbook 1995, HVAC Applications. Reprinted by permission.
The data are tests on 20-ft-long ducts, but the TL values are for ducts of the cross section shown regardless of length.
Source: Adapted with permission from ASHRAE Handbook 1991, HVAC Applications.
Transmission Loss for Selected Building Structures
The transmission losses for selected building structures—including walls, partitions, window glass,
ceiling and plenum, and acoustic equipment housing—are listed in Table 19.14. These are mainly
abridged from ASHRAE Handbook 1991, HVAC Applications. In Table 19.14, the combined
plenum/ ceiling cavity effect indicates the combined effect of the plenum sound absorption and the
transmission through the ceiling material. These values are based on data from several manufacturers’
laboratory and mock-up spaces.
19.7 RELATIONSHIP BETWEEN ROOM SOUND POWER LEVEL
AND ROOM SOUND PRESSURE LEVEL
The sound pressure level at a given location in a room corresponding with a particular point source
or an array of multiple sound sources is a function of the sound power level and transmission characteristic
of the sound source, the room’s acoustical properties, and the distance between the sound
source and the receiver.
Single or Multiple Sound Sources
Based on field measurements, Schultz (1985) recommended the following empirical formula to
estimate the room sound pressure level Lpr, in dB re 20 Pa, from the room sound power level of
SOUND CONTROL 19.23

For a single sound source in the room, Eq. (19.11) can be applied directly. If there are multiple
sound sources, estimate Lwr for each sound source and sum the contributions to obtain the total
sound pressure level at the receiver due to multiple sound sources.
The difference between the room sound pressure level and the room sound power level Lpr  Lwr
is called the room effect.
Array of Ceiling Diffusers
Offices and large rooms in commercial buildings often have several ceiling diffusers mounted in
a distributed array on the ceiling. Although the sound pressure levels decrease as the vertical
distance from the ceiling increases, the sound pressure level along any selected horizontal plane,
such as a reference plane of 5 ft (1.5 m) from the floor level, is nearly constant. According
to ASHRAE Handbook 1995, HVAC Applications, the calculation of sound pressure level for a distributed
array of ceiling diffusers at a reference plane of 5 ft (1.5 m) from the floor level can be
greatly simplified as
(19.12)
where Lw,s 	 sound power level of single outlet, dB re 1012 W. Values of D are listed in

19.8 NOISE CONTROL FOR A TYPICAL AIR SYSTEM
In many commercial buildings using a variable-air-volume air system employing VAV boxes, there
are two noise sources, as shown in Fig. 19.1:
 Fan noise due to the supply fan of the air-handling unit usually installed in a fan room
 Airflow noise due to the VAV box with a single-blade damper installed inside the ceiling plenum
Sound is transmitted from these two sources to the receiver in the conditioned space via the following
paths:
 Duct-borne path, supply side
 Duct-borne path, return side
 Radiated noise breakout from the supply duct in the ceiling plenum
 Radiated sound breakout from the VAV box casing in the ceiling plenum
 Airborne path through fan room walls to adjacent area
Combination of Supply Fan Noise and Terminal Noise
Variable-air-volume boxes are tested for radiated sound power level and discharge-side sound
power level in accordance with ADC/ ARI Industry Standard 880-83, and the results are listed in
the manufacturers’ catalogs. The attenuated duct-borne sound power level of the supply fan from
the air-handling unit at terminal unit LAHUi should be added to the discharge sound power level from
the VAV box Lter to form a resultant sound power level if their difference Lter  LAHUi  8 dB. If
their difference is greater than 8 dB, the smaller of these two sound power levels can be ignored.
Estimate Sound Pressure Level for Spaces Served by Terminal Units
Recently, the ADC and ARI jointly developed Industry Standard 885 to estimate the space sound
pressure levels when terminal units are installed inside the ceiling plenum. Consider a terminal unit
connected to two slot diffusers through flexible ducts to serve a conditioned space, as shown in
Fig. 19.7, where C represents the terminal casing, D the duct-borne or discharge side, D1 and D2 the
discharge outlets 1 and 2, and O1 the supply outlet 1. Another outlet, O2, is not shown in Fig. 19.7.
Along with the arrows, B represents the transmission loss of the breakout from duct and casing,
P the sound attenuation due to the plenum ceiling, and S the room effect.
There are altogether seven sound paths by which the receiver can hear fan, terminal, and slot
diffuser noises.
Path 1: Radiated sound power level of the terminal unit : plenum ceiling : room effect : receiver
Path 2: Discharge sound power level : breakout : plenum ceiling : room effect : receiver
Path 4: Duct-borne sound power level at the outlet of slot diffuser : room effect : receiver
Path 5: Sound power level of the slot diffuser : room effect : receiver
Sound path 3 transmits radiated sound breakout from another flexible duct after the terminal unit,
and it is similar to path 2. The duct-borne noise is transmited along sound path 6 at another slot diffuser
and is similar to sound path 4. The airflow noise of another slot diffuser transmits along sound
path 7 and is similar to sound path 5.
The radiated sound power level that breaks out at the inlet of the terminal unit has the same kind
of sound path as paths 2 and 3. The sound power level of duct-borne noise at Do1 equals the sound
power level at D1 or D2 minus the sound attenuation of the flexible duct and the slot diffuser.
SOUND CONTROL 19.25
Environmental Adjustment Factor
According to Industry Standard (IS) 880, sound power levels of the terminal unit as listed in the
manufacturers’ catalogs are based on free field (outdoor) calibration of the reference sound source.
At low frequencies, actual rooms are highly reverberant. The adjustment factor that takes into account
the difference in sound power levels of a commonly used reference sound source measured in
a free field and a reverberant field is called the environmental adjustment factor Ef, in dB. IS 885
recommends the following Ef values:
Plenum Ceiling Effect
Sound attenuation of the plenum ceiling combination not only includes the transmission loss of the
ceiling, but also considers the effect of the plenum. For a -in.- (16-mm-) thick mineral fiber
acoustic tile ceiling with a density of 35 lb / ft3 (560 kg /m3), IS 885 recommends the following values,
which are the same as those in Table 19.14:

Example 19.1. A variable-air-volume box (terminal unit) receives conditioned air from an airhandling
unit through main and branch ducts. The VAV box is installed inside a ceiling plenum and
is connected to two slot diffusers by 10-in.- (250-mm-) diameter flexible ducts, each of which is 5 ft
(1.5 m) long. Information is available from manufacturers’ catalogs on the duct-borne fan noise at
the inlet of the VAV box, the radiated sound power level Lw,rad, the discharge sound power level
Lw,dis of the VAV box, and the sound power level of the slot diffuser Lw,s. From measurements, the
transmission loss of the breakout noise through flexible duct wall and the insertion loss of the flexible
duct are known. These data are summarized as follows:
If the volume of the room is 5040 ft3 (143 m3), estimate the total sound pressure level in the occupied
zone with an average distance from the receiver of 8 ft (2.4 m). Ignore the fan noise transmitted
through the return duct system.
Solution
1. From Eq. (19.11), the room effect at the octave band center frequency of 125 Hz of this occupied
zone is
The room effect of other octave bands can be estimated similarly as follows:
2. For path 1, after considering the influence of the environmental adjustment factor Ef , one
finds that the radiated sound power levels of the VAV box Lrad,r 	 Lw,rad-Ef, in dB, are as follows:
Because the transmission loss of the casing of the VAV box is rather high, the breakout of ductborne
fan noise through the casing is far smaller than Lrad,r, and is ignored.
The sound pressure level in the occupied zone due to the radiated sound power level from the
VAV box can be calculated by subtracting the plenum ceiling effect from Lrad,r and then converting
to the sound pressure level through the room effect.
Octave band center frequency, Hz 125 250 500 1000 2000 4000
Reverberant field, dB 57.5 53 49.5 45.5 40.5 36.5
Octave band center frequency, Hz 125 250 500 1000 2000 4000
Room effect, dB 9 9.5 11 11.5 12.5 13.5

Octave band center frequency, Hz
125 250 500 1000 2000 4000
Fan noise at inlet, dB 65 63 61 52 46 45
Lw,rad, dB 60.5 55 50.5 46.5 41.5 37.5
Lw,dis, dB 63 60.5 47 47.5 54.5 52
Lw,s, dB 40 39 35 33 31 24
Flexible, TLout, dB 8 11 14 17 20 25
Flexible, IL, dB/ ft 0.50 0.85 1.48 2.2 2.04 1.64
SOUND CONTROL 19.27
3. For path 2, the discharge sound power level of the VAV box changes from the free field Lw,dis
to the reverberant field Ldis,r and the duct-borne fan noise at the terminal unit Lw,f. Their combination
is as follows:
The sound pressure level in the occupied zone due to the combination Ldis,r  Lw,f transmitted along
sound path 2 can be estimated by subtracting the transmission loss of the flexible duct wall, the
plenum ceiling effect, and the room effect from Ldis,r  Lw, f .
4. For path 4, there are two flexible ducts and slot diffusers connected to the terminal unit; therefore,
the branch power division is 3 dB. The sound pressure level in the occupied zone due to ductborne
noise transmitted along sound path 4 can be evaluated by subtracting the branch power
division Lw,b, the insertion loss of the flexible duct ILflex, and the room effect from the combination
Ldis,r  Lw,f .
5. For path 5, change the sound power level of the slot diffuser from free field Lw,s to reverberant
field Ls,r.
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Ldis,r  Lw,f, dB 66.5 64.5 61.5 53.5 54.5 52
Lw,b, dB 3 3 3 3 3 3
ILflex, dB 2.5 4 7 11 10 8
Room effect, dB 9 9.5 11 11.5 12.5 13.5
Sound pressure level, dB 52 48 40.5 28 29 27.5
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Ldis,r  Lw,f , dB 66.5 64.5 61.5 53.5 54.5 52
Flexible, TLout, dB 8 11 14 17 20 25
Ceiling-plenum, dB 5 9 10 12 14 15
Room effect, dB 9 9.5 11 11.5 12.5 13.5
Sound pressure level, dB 44.5 35 26.5 13 8 0
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Ldis,r, dB 60 58.5 46 46.5 53.5 51
Fan noise Lw,f, dB 65 63 61 52 46 45
Ldis,r  Lw,f, dB 66.5 64.5 61.5 53.5 54.5 52
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Lrad,r, dB 57.5 53 49.5 45.5 40.5 36.5
Ceiling plenum effect, dB 5 9 10 12 14 15
Room effect, dB 9 9.5 11 11.5 12.5 13.5
Total sound pressure level, dB 43.5 34.5 28.5 22 14 8
19.28 CHAPTER NINETEEN
Then the sound pressure level in the occupied zone due to the slot diffuser can be estimated by subtracting
the room effect from Ls,r.
6. The resultant estimated sound pressure level in the occupied space Lp,t, in dB re 20 Pa, is the
sum of the contributions of all sound paths and can be calculated as follows:
These resultant sound pressure levels at the center frequencies of various octave bands are all equal
to or below the NC 40 curve.
19.9 ROOFTOP PACKAGED UNITS
Basics
Rooftop packaged units are widely used in low-rise commercial buildings. Because of the use of
large rooftop packaged units, many rooftop packaged units are curb-mounted on lightweight roof
deck construction. The distances between the fan and ceiling plenum, and between the fan and conditioned
space, are sometimes shorter than the distance between the fan in the AHU and space.
They may cause unique noise control problems. Objectionable noise from a rooftop packaged unit
usually includes direct transmission of fan and compressor noise through the building roof and ceiling,
structure-borne vibration, duct-borne fan noise, duct breakout noise, and duct rumble.
Locating the rooftop packaged unit at least 25 ft (7.6 m) horizontally away from areas with critical
sound control requirements, such as conference rooms and executive offices whose NC curve
should be 35 dB or less, is the most effective and economical method of noise control. A rooftop
packaged unit can be located over buffer spaces, low-use storage areas, high-noise loading areas,
and acoustically nonsensitive spaces such as rest rooms and stairwell buffer spaces.
Using a number of small packaged units instead of one or two large units divides the impact
over a greater roof area and is sometimes beneficial. A rooftop packaged unit with a lower sound
power level is often one size larger than the smallest unit for that capacity. The fan in such a unit is
properly sized, with less restricted inlet and outlet conditions inside the unit.
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Path 1 43.5 34.5 28.5 22 14 8
Path 2 44.5 35 26.5 13 8 0
Path 3 44.5 35 26.5 13 8 0
Path 4 52 48 40.5 28 29 27.5
Path 5 52 48 40.5 28 29 27.5
Path 6 28 27.5 23 20.5 17.5 9.5
Path 7 28 27.5 23 20.5 17.5 9.5
Lp,t, dB 56 51.5 44 32.5 32.5 31
Octave band center frequency, Hz
125 250 500 1000 2000 4000
Ls,r, dB 37 37 34 32 30 23
Room effect, dB 9 9.5 11 11.5 12.5 13.5
Sound pressure level, dB 28 27.5 23 20.5 17.5 9.5
SOUND CONTROL 19.29
Sound Sources and Paths
In a typical rooftop packaged unit, the paths by which fan and compressor noise is transmitted from
the rooftop packaged units to the receiver in the occupied space are shown in Fig. 19.8a:
1. Radiated fan noise breaks out from the supply duct in the ceiling plenum : plenum ceiling
effect : room effect : receiver.
2. Duct-borne fan noise on the discharge side : attenuation of ducts, elbows, and silencer :
end reflection : room effect : receiver.
3. Radiated fan noise breaks out from the return air duct in the ceiling plenum : plenum ceiling
effect : room effect : receiver.
4. Duct-borne fan noise on the return side : attenuation of ducts, elbows, and silencer : end
reflection : room effect : receiver.
19.30 CHAPTER NINETEEN
FIGURE 19.8 Sound path and sound control for a typical rooftop packaged unit: (a) installation of rooftop unit with
rectangular supply duct; (b) rectangular to multiple drop: round mitered elbows with turning vanes; (c) rectangular to
multiple drop: round mitered elbows with turning vanes, inner-lined round duct after mitered elbow.
5. Structure-borne fan or compressor noises : roof construction : plenum ceiling effect : receiver.
Discharge Side Duct Breakout
Experience has shown that the low-frequency noise breakout from the supply duct inside the ceiling
plenum (path 1) and subsequent transmissions to the receiver through the ceiling are the critical
noise problem in most rooftop packaged units. A supply fan that has a down shot outlet and that
penetrates the roof directly produces many noise problems, such as very loud breakout sound inside
the ceiling plenum. A supply fan with horizontal outlet discharge to an inner-lined plenum inside
the rooftop packaged unit is better. Use 2-in. (50-mm) acoustic linnings for fan, compressor, and
condenser sections to reduce sound transmission directly through the roof. An even better sound attenuation
arrangement is to run the supply duct horizontally from the rooftop unit 10 to 20 ft (3 to
6 m), combined with silencers along the roof, and then penetrate into the building, directly above
acoustically sensitive spaces. Duct exposed to outside atmosphere must be inner-lined, externally
insulated, weatherproof enclosures that are covered with noise barriers.
Round ducts have a far greater TLout than rectangular ducts. Based on laboratory tests, Beatty
(1987) recommended using rectangular to multiple round duct drops with round mitered elbows
and turning vanes, as shown in Fig. 19.8b and c to contain low-frequency rumble noise. Compared
with rectangular ducts without turning vanes, these configurations have an additional noise reduction
of 17.5 dB in octave band 1, 11.5 dB in octave band 2, and 13 dB in octave band 3 when the
configuration is the same as in Fig. 19.8b. If the multiple drops of round duct have the configuration
shown in Fig. 19.8c, there will be an additional noise reduction of 18 dB in octave band 1, 13 dB in
octave band 2, and 16 dB in octave band 3 compared to rectangular ducts without turning vanes.
Use heavier-gauge sheet metal to fabricate the rectangular plenum before the round duct drops
to reduce breakout noise.
If the first run of duct from the fan outlet must be rectangular and passes a noise sensitive area,
Harold (1993) recommended the following:
 Use 16-gauge sheet metal for duct wall for the first 25-ft (7.5-m) run of duct.
 Use 1500 fpm (7.5 m / s) as the maximum duct velocity for the rectangular duct.
 Enclose the supply duct in a drywall enclosure of two layers of -in. (16-mm) gypsum board
with freestanding walls that have a 4-in. (100-mm) airspace around the duct and contain a 3-in.-
(75-mm-) thick fiberglass blanket that touches only the duct or drywall.
 Employ an active silencer.
Careful design of the fan outlet configurations will prevent sharp turning of airstreams. Inner lining
of the ductwork and sometimes an additional silencer are necessary at the downstream side for
more effective noise control.
The duct system should not be designed for an excessive velocity, i.e., normally not greater than
3000 fpm (15 m / s). Design the duct fittings at minimum dynamic losses.
VAV boxes and reheating VAV boxes in a VAV reheat system should not be operated at a completely
shut-off condition because a high fan total pressure at a low volume flow rate is generally a
noisy operating point for centrifugal fans.
Sound Path on Return Side
On the return side, there are three types of return air systems:
 A rooftop unit whose return air system connects directly the return ceiling plenum to the return
intake of the rooftop unit through a short return duct system. This is the type of return air system
58
SOUND CONTROL 19.31
widely used for rooftop units. A rectangular return duct may be used if its duct wall is made of
16-gauge (1.6-mm thickness) galvanized sheet metal and breakout noise is not critical. Adopt a
duct velocity of 1500 fpm (7.5 m / s) or less. Use 2-in. (50-mm) thick inner lining in the return
side (including ductwork) of the rooftop unit. Use tees with a minimum leg length of 3 to 5 equivalent
diameters to increase end reflection loss.
 For ducted return air systems, the sound power level at the supply or return fan intake should be
attenuated by the return duct liners, return side silencer, and room effect. For a rooftop unit with a
collective return grille, a sound-attenuating plenum, as shown in Fig. 19.3, is often cost-effective
if space is available.
 For an open fan room (Fig. 16.7), return air is often transferred to the open fan room from the return
ceiling plenum through a transfer duct or a transfer wall passage. Both require an inner lining
2 in. (50 mm) thick with a perforated metal sheet facing. Air velocity in the return transfer duct or
passage should be 1000 fpm (5 m / s) or less.
Structure-Borne Noise
Structure-borne noise can be controlled by using suitable vibration isolators. Special curb-mounting
bases should be used to provide better sound and vibration control. Sound structural support should
be provided for rooftop packaged units. Borzym (1991) says that the best place for the rooftop
packaged unit is directly over a column or straddling a major beam close to a column. Good locations
are those near the intersection of beam with a heavy end close to a column, straddling a beam,
or close to a beam. The rooftop packaged unit should be mounted on a spring-isolated roof curb and
should float on the springs. The roof below the rooftop packaged unit should not vibrate.
The roof directly under the rooftop packaged unit within the roof curb should contain a good
acoustic barrier to prevent sound transmission through the bottom panel of the packaged unit, the
roof, and the ceiling plenum. A typical sound barrier is made of semirigid fiberglass insulation, at
least 4 in. (100 mm) thick, covered by two layers of -in. (16-mm) gypsum board. The staggered
gypsum boards are used to seal the joints and seams airtight.
REFERENCES
ASHRAE, ASHRAE Handbook 1991, HVAC Applications, ASHRAE Inc., Atlanta, GA, 1991.
ASHRAE, ASHRAE Handbook 1995, HVAC Applications, Atlanta, 1995.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, 1999.
Beatty, J., Discharge Duct Configurations to Control Rooftop Sound, Heating / Piping / Air Conditioning, July
1987, pp. 53–58.
Blazier,W. E., Noise Rating of Variable-Air-Volume Terminal Devices, ASHRAE Transactions, 1981, Part I,
pp. 140–152.
Bodley, J. D., An Analysis of Acoustically Lined Duct and Fittings and the State of the Art of Their Use,
ASHRAE Transactions, 1981, Part I, pp. 658–671.
Borzym, J. X., Acoustical Design Guidelines for Location of Packaged Rooftop Air Conditioners, ASHRAE
Transactions, 1991, Part I, pp. 437–441.
Cummings, A., Acoustic Noise Transmission through Duct Walls, ASHRAE Transactions, 1985, Part II A,
pp. 48–61.
Ebbing, C., and Waeldner,W. J., Industry Standard 885: An Overview, Estimating Space Sound Levels for Air
Terminal Devices, ASHRAE Transactions, 1989, Part I, pp. 529–533.
Ebbing, C. E., Fragnito, D., and Inglis, S., Control of Low Frequency Duct-Generated Noise in Building Air
Distribution Systems, ASHRAE Transactions, 1978, Part II, pp. 191–203.
Gelin, L. J., Active Noise Control: A Tutorial for HVAC Designers, ASHRAE Journal, no. 8, 1997, pp. 43–49.
58
19.32 CHAPTER NINETEEN
Grimm, N. R., and Rosaler, R. C., HVAC Systems and Components Handbook, McGraw-Hill, New York, 1997.
Harold, R. G., Achieving a Quiet Rooftop Installation, ASHRAE Journal, no. 12, 1993, pp. 32–38.
Hirschorn, M., Acoustic and Aerodynamic Characteristics of Duct Silencers for Air Handling Systems,
ASHRAE Transactions, 1981, Part I, pp. 625–646.
Phelps, G. R., Letters to Editor—Fibrous Facts, Engineered Systems, no. 10, 1997, pp. 13–16.
Reese, J., The Case for Certified Ratings, Heating / Piping / Air Conditioning, August 1986, pp. 85–87.
Reynolds, D. D., and Bledsoe, J. M., Sound Transmission through Mechanical Equipment Room Walls, Floor,
or Ceiling, ASHRAE Transactions, 1989, Part I, pp. 83–89.
Schultz, T. J., Relationship between Sound Power Level and Sound Pressure Level in Dwellings and Offices,
ASHRAE Transactions, 1985, Part I A, pp. 124–153.
Sessler, S. M., and Angevine, E. N., HVAC System Noise Calculation Procedure—An Update, ASHRAE Transactions,
1983, Part II B, pp. 697–709.
Smith, M. C., Industry Standard 885, Acoustical Level Estimation Procedure Compared to Actual Acoustic
Levels in an Air Distribution Mock-Up, ASHRAE Transactions, 1989, Part I, pp. 543–548.
Smith, M. C., and Int-Hout, D., Using Manufacturers’ Catalog Data to Predict Ambient Sound Levels, ASHRAE
Transactions, 1984, Part II B, pp. 85–96.
Tinsley, R. W., Duct Liner: Problem or Solution? HPAC, no. 5, 1998, pp. 65–70.
The Trane Company, Acoustics in Air Conditioning, American Standard Inc., La Crosse, WI, 1986.
Ver, I. L., Noise Generation and Noise Attenuation of Duct Fittings—A Review: Part II, ASHRAE Transactions,
1984, Part II A, pp. 383–390.
SOUND CONTROL 19.33
CHAPTER 20
AIR SYSTEMS: BASICS AND
CONSTANT-VOLUME SYSTEMS
20.1
20.1 AIR SYSTEM BASICS 20.2
Air Systems 20.2
Air Distribution Systems 20.3
Ventilation Systems 20.3
Mechanical Ventilation Systems 20.3
Regenerative Systems 20.3
Smoke Control Systems 20.4
Terminals 20.4
Primary, Secondary, and
Transfer Air 20.4
20.2 BUILDING LEAKAGE AREA AND
BUILDING TIGHTNESS 20.5
Effective Leakage Area 20.5
Building Air Leakage 20.5
Air Change Per Hour at 50 Pa
(ACH50) 20.6
20.3 SPACE PRESSURIZATION 20.7
Space Pressure Characteristics 20.7
Stack Effect 20.7
Stack Effect for High-Rise Buildings 20.9
Wind Effect 20.10
Air Systems and Mechanical Ventilation
Systems 20.11
Airflow Balance and Space Pressurization
by Differential Flow 20.11
20.4 INFILTRATION AND
EXFILTRATION 20.14
Volume Flow Rate of Infiltration 20.14
20.5 FAN-DUCT SYSTEMS 20.14
System Operating Point 20.15
Buckingham Method and
Fan Laws 20.15
20.6 SYSTEM EFFECT 20.17
Mechanism of System Effect 20.17
Inlet System Effect 20.18
Inlet System Effect Loss 20.19
Inlet System Effect Loss Coefficient 20.19
Outlet System Effect 20.20
Outlet System Effect Loss
Coefficient 20.22
Selecting Fan Considering System
Effect Losses 20.23
20.7 COMBINATION OF FAN-DUCT
SYSTEMS 20.24
Two Fan-Duct Systems
Connected in Series 20.25
Fans Combined in Parallel and
Connected in Series with a Duct
System 20.26
Two Parallel Fan-Duct Systems
Connected with Another
Duct System 20.28
20.8 MODULATION OF THE FAN-DUCT
SYSTEM 20.31
Modulation Curve 20.31
Modulation of Fan-Duct Systems 20.33
20.9 CLASSIFICATION OF AIR
SYSTEMS 20.39
20.10 CONSTANT-VOLUME SYSTEMS 20.40
System Characteristics 20.40
Energy per Unit Volume Flow 20.41
20.11 AIR CONDITIONING
PROCESSES 20.41
Sensible Heat Ratio 20.41
20.12 SPACE CONDITIONING AND
SENSIBLE COOLING AND HEATING
PROCESSES 20.43
Space Conditioning Process 20.43
Sensible Heating and Cooling
Processes 20.44
20.13 HUMIDIFYING PROCESSES 20.45
Steam Injection and Heating
Element Humidifier 20.45
Air Washer 20.46
Oversaturation 20.46
20.14 COOLING AND DEHUMIDIFYING
PROCESS 20.46
20.15 ADIABATIC MIXING AND BYPASS
MIXING PROCESSES 20.50
Two-Stream Adiabatic Mixing
Process 20.50
Bypass Mixing Process 20.52
20.16 SINGLE-ZONE, CONSTANT-VOLUME
SYSTEMS—COOLING
MODE OPERATION 20.53
Air Conditioning Cycle 20.53
Cooling Mode Operation in
Summer 20.53
Cooling Mode Operation in Winter
without Space Humidity Control 20.56
Cooling Mode Operation in Winter
with Space Humidity Control 20.57
Part-Load Operation and Controls 20.58
Outdoor Ventilation Air and
Exhaust Fans 20.58
20.17 SUPPLY VOLUME FLOW RATE 20.59
Based on Space Cooling and Heating
Load 20.59
Based on Requirements Other Than
Cooling Load 20.60
Rated Volume Flow of Supply and
Return Fans 20.61
20.1 AIR SYSTEM BASICS
Air Systems
Air systems and their controls directly affect the indoor environment and indoor air quality (IAQ).
In a broad sense, air systems are a group of subsystems in an air conditioning system. Air systems
inclued supply and return air systems (space recirculating systems), mechanical ventilation systems,
air distribution systems, regenerative systems, smoke control systems, and terminals. The main
functions of air systems are as follows:
 Conditioning the supply air including heating or cooling, humidification or dehumidification,
cleaning and purifying, and attenuation of objectional noise produced by fans, compressors, and
pumps
 Distributing the conditioned supply air with adequate outdoor air to the conditioned space, extracting
space air for recirculating, and exhausting or relieving unwanted space air to the outdoors
 Providing space pressurization, toxic gas exhaust, and smoke control for occupants’ safety and
fire protection
 Controlling and maintaining required space temperature, humidity, cleanliness, air movement,
sound level, and pressure differential within predetermined limits at optimum energy consumption
and cost
 Using high-temperature airstream to reactivate the desiccant, if any
In actual practice or in a narrower sense, an air system or an air-handling system often denotes
a more specific system that conditions, transports, distributes the supply air, recirculates the space
air, and controls the indoor environment according to requirements. Such an air system often consists
of fans, heat exchangers (including coils, evaporative coolers, and dehumidifiers), a direct-fired
gas heater, filters, a mixing box, dampers, ductwork, terminals, space diffusion devices, and controls.
In a central hydronic air conditioning system, an air system consists of an air-handling unit (air
is cooled and dehumidified by a water cooling coil), ducts, terminals, space diffusion devices, and
controls. In a unitary packaged air conditioning system, the air system consists of an air handler (in
which air is cooled and dehumidified by a DX coil or rotary dehumidifier), ducts, terminals, space
diffusion devices, and controls. In a space air conditioning system, its air system is a combination
of an outdoor air ventilation system and a space air recirculating system, which consists of a fan, a
coil, a filter, ductwork, and controls. A separate outdoor ventilation air system provides conditioned
20.2 CHAPTER TWENTY
Dual-Thermostat Year-Round Zone
Temperature Control 20.73
20.20 CONSTANT-VOLUME MULTIZONE
SYSTEM WITH REHEAT 20.74
Reheating, Recooling, and Mixing 20.74
Constant-Volume Multizone System
with Reheat 20.75
Control Systems 20.75
Operating Parameters and
Calculations 20.77
System Characteristics 20.78
REFERENCES 20.79
20.18 DETERMINATION OF THE SUPPLY
AIR CONDITION 20.62
Air Conditioning Rules 20.63
Graphical Method 20.63
Influence of Sensible Heat Ratio 20.64
20.19 CONSTANT-VOLUME SINGLE-ZONE
SYSTEMS—HEATING MODE
OPERATION 20.69
Heating Mode without Space Humidity
Control 20.69
Heating Mode with Space Humidity
Control 20.71
Part-Load Operation 20.73
outdoor air for occupants. A space recirculating system conditions the space recirculating air to offset
the space cooling or heating load. In an individual air conditioning system, the air system, heating
system, and refrigeration systems are integrated in a single packaged air conditioner. It is not
necessary to distinguish them separately. The air system in a thermal storage, desiccant cooling, or
an evaporative-cooling air conditioning system is similar to that of a central hydronic system.
The air system is the foundation in an air conditioning system. It directly contacts the occupants,
manufacturing products, and conditioned space. Air systems play a key role in maintaining a
healthy, safe, and comfortable indoor environment with an acceptable indoor air quality at optimum
energy use.
It is more specific, convenient, and clear that the systems which condition, transport, distribute,
and control the supply air and recirculate the space air are called air systems; and all the other air
systems in a broad sense that do not cool and dehumidify the air are called mechanical ventilation
systems, exhaust systems, and smoke control systems.
Air Distribution Systems
An air distribution system is a subsystem of an air system which consists of ductwork, terminals,
space diffusion devices, and controls to transport and distribute the conditioned air to the occupied
zone as well as to control the indoor environment within required limits.
Ventilation Systems
As discussed in Sec. 4.10, ventilation air comprises outdoor air plus treated recirculated air with
acceptable indoor air quality. Ventilation air is also the air used for the dilution of the contaminant
concentrations in the conditioned space. Ventilation air may or may not be conditioned. Ventilation
indicates the process to supply or to remove ventilation air. Because of the importance of the ventilation
air supply, a ventilation system is usually considered as part of air systems that use fans to
supply and to remove ventilation air to and from any space. Natural ventilation (including infiltration)
is a process that provides outdoor ventilation air by stack or wind effect, or both.
Mechanical Ventilation Systems
A mechanical ventilation system uses mechanical means (fans) to transport the air without cooling
and dehumidifying it. The following are the types of mechanical ventilation systems currently
used:
 Mechanical ventilation systems for cooling and contaminant dilution, which supply outdoor air to
a room to cool equipment or to dilute air contamination to acceptable level, as in a transformer
room or a garage
 Exhaust systems, which discharge indoor air, or exhaust toxic gas and contaminated air from localized
sources such as exhaust systems from bathrooms or from chemical hoods in laboratories
 Makeup air systems as a part of mechanical ventilation systems in which outdoor air is often
heated and filtered to compensate for the exhausted air
Regenerative Systems
In a regenerative or drying system, outdoor air is heated to a higher temperature such as 150 to
200°F (65 to 93°C) and is then used to remove the absorbed moisture from the desiccant by drying.
After drying, the desiccant is reactivated and is ready to dehumidify the process air again.
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.3
Smoke Control Systems
Smoke control systems include stairwell pressurizing systems and zone smoke control systems.
Stairwell pressurizing systems pressurize stairwells as an escape route in high-rise buildings during
building fire. Zone smoke control systems exhaust the smoke in the fired floor and pressurize the
floors immediately above and below the fired floor to prevent smoke contamination. Smoke control
systems are discussed in Chap. 22.
Terminals
A terminal is a factory-made device installed inside a control zone either in a ceiling plenum above
the conditioned space or sometimes directly mounted adjacent to the external wall of the conditioned
space or in a floor plenum. A terminal usually has an individual control system to maintain a
selected indoor enviromental parameter within predetermined limits for a control zone.
Conditioned air including outdoor air is supplied to a terminal from the supply main duct
through branch takeoffs. It may be mixed with recirculating air and is supplied to the control zone
served by the terminal in a multizone air system.
A terminal may perform any of the following functions or combinations thereof:
 Heating / cooling or filtration
 Mixing of outdoor ventilation air or conditioned air with recirculating air, or mixing of cold supply
air with warm supply air
 Controlling and maintaining a predetermined zone temperature by modulating the volume flow
rate of supply air, the temperature of supply air, or the ratio of the volume flow of cold and warm
air supplies. VAV boxes, reheating VAV boxes, mixing VAV boxes, fan-powered VAV boxes, and
fan-coil units are all terminals, and will be discussed in Chap. 21.
Primary, Secondary, and Transfer Air
Primary air is conditioned makeup outdoor air or a conditioned mixture of outdoor and recirculating
air. Primary air is either conditioned and supplied from a separate makeup AHU (PU) or from
an AHU (PU) using a mixture of outdoor and recirculating air to the terminals. Sometimes the
primary air is supplied directly to the space.
Primary air is usually mixed with the recirculating air, which is often called secondary air. The
mixing most often takes place in the terminals, or another AHU or PU, or sometimes directly in
the conditioned space. The mixture of primary air and recirculating air (secondary air) may be
conditioned again in the terminals. It is then supplied to the conditioned space to offset the space
load. The ratio of the volume flow rate of primary air to the mixture (zone supply air) varies from
0.15 to 0.5.
Recirculating air is the air recirculated from the conditioned space, through an AHU, a PU, or a
terminal, and supplied again to the conditioned space. Recirculating air is different from the return
air which returns from the conditioned space to an AHU, or PU, and exhausts outdoors. The purpose
of mixing primary air with recirculating air in the terminal is to save energy including fan energy,
as well as to save investment by using smaller supply and return ducts, which occupy less
space.
Transfer air is indoor air that moves or supplies to a conditioned space from an adjacent
area. Transfer air is entirely composed of recirculating air (secondary air). Transfer air often
has less unused outdoor air than a mixture of outdoor and recirculating air. Unused outdoor air is
outdoor air whose capability of dilution of air contaminants has not been expended. Recirculating
air still may contain a certain amount of unused outdoor air. Restrooms are normally supplied by
transfer air.
20.4 CHAPTER TWENTY
20.2 BUILDING LEAKAGE AREA AND BUILDING TIGHTNESS
Currently, the building tightness or the building air leakage can be indicated by the following
indices: (1) an effective leakage area, (2) air leakage per unit area of building shell, and (3) air
change per hour (ACH) at a specific indoor-outdoor pressure differential.
Effective Leakage Area
The effective leakage area of a building Ae,l, in ft2 (m2), is the amount of open wall area that allows
the same volume flow rate of air flowing through it as flowing through the actual building when the
pressure difference across the open wall area and the shell of the actual building is the same. The
Ae,l value is usually determined by a fan pressurization method, in which a blower installed at the
door of a building is used to maintain a pressure difference across the building shell. The airflow
rate , in cfm (m3), can then be measured , and the Ae,l value is calculated as
(20.1)
where Cd  discharge coefficient; its value is 1 for effective leakage area
 air density, lb/ ft3 (kg /m3)
p  pressure differential between outdoor and indoor air at same level, in.
WC (Pa)
According to ASHRAE Handbook 1997, Fundamentals, effective air leakage areas for representative
residential building components at an outdoor-indoor pressure difference of 0.016 in. WC
(4 Pa) are as follows:
Building Air Leakage
When air flows through a crack or gap, the flow never becomes fully developed. The relationship
between overall air leakage rate leak, in cfm (L / s), and the indoor-outdoor pressure difference p,
in in. WC (Pa), can often be represented by a widely used air leakage equation as
(20.2) V?leak  Cflow Aw (p)n
V?
Best estimate, in.2 /ft2
External walls
Cast-in-place concrete 0.007
Clay brick cavity wall, finished 0.0098
Lightweight concrete block, painted 0.016
Window
Wood frame
Uncaulked 0.025
Caulked 0.004
Double-hung, not weatherstripped 0.12
Double-hung, weatherstripped 0.031
Door, average 0.015 in.2 / lftc (linear ft. crack)
Ceiling 0.026
V?
 Cd Ae,1?2 p

 Cd Ae,l 4005?p
V?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.5
where Cflow  flow coefficient, cfm/ ft2 (m3/hm2) at p  in. WC (Pa)
Aw  area of exterior wall or roof, ft2 (m2)
p  pressure difference across building envelope, in. WC (Pa)
Because the cracks have large flow resistance, in Eq. (20.2), the exponent n may have a value
closer to 1 than openings with less flow resistance. Shaw et al. (1993) recommended n  0.65
based on field-measured results. Persily (1999) found significant levels of air leakage in commercial
and institutional buildings. Among the data measured from 139 buildings around the world, flow
coefficients Cflow, in cfm / ft2 (m3/hm2), varies from 0.15 to 6.85 cfm / ft2 (2.7 to 124.5 m3/hm2) at
an indoor-outdoor pressure difference p of 0.3 in. WC (75 Pa) using the fan pressurization test
method with a mean Cflow  1.5 cfm / ft2 (27.1 m3/hm2) of exterior wall area at p  0.3 in. WC
(75 Pa). There is no correlation between air leakage and the age of the building tested. Building
height does show an impact on airtightness with high-rise buildings being tighter, especially onestory
buildings are leakier.
Based on Persily (1999) who analyzed building envelope airtightness data measured from 139
commercial and institutional buildings, flow coefficients Cflow, in cfm / ft2 (m3/hm2), per unit
exterior wall area at 0.3 in. WC (75 Pa) for tight, average, and leaky buildings can be estimated as
follows:
Air Change per Hour at 50 Pa (ACH50)
The air leakage rate of a building can also be expressed by the air changes per hour at a specific
oudoor-indoor pressure difference of which 0.2 in. (50 Pa) is most widely used. As in the effective
leakage area, a fan is installed at the door of a building and is used to maintain a specific pressure
difference across the building shell. The air leakage leak, in cfm (L / min), supplied by the fan is
squeezed outdoors through the cracks and gaps on the building shell and is then measured. The air
change per hour of air leakage, denoted by ACH, can be calculated as
(20.3)
where Vr  the space volume of room or enclosure, ft3 (m3). For residential buildings, in ASHRAE
Handbook 1997, Fundamentals, ACH50  3 is considered exceptionally tight, and for residential
buildings 5 	 ACH50  20 is considered conventional. There are 10 percent of the 204 lowincome
houses whose ACH50  47.5.
Withers et al. (1996) measured the ACH50 values for 18 small commercial buildings after repair
of duct leaks, the sealing of the building shell, and other repairs, four of them with ACH50  5.2,
seven of them with 10.7  ACH50  19.7, five of them with 21.2  ACH50  28.8, one with
ACH50  46, and the highest one with ACH50  52.8, a court office.
Because of the desire for energy conservation, the airtightness of buildings has been improved
during the past decades. Since the ratio of the exterior wall area to the floor area is quite different
for different types of buildings, it is difficult to compare the airtighness of different types of buildings,
especially using different air leakage indices.
ACH 
60 V?leak
Vr
V?
Cflow, cfm/ft2 (m3/hm2) of exterior
wall area at 0.3 in. WC (75 Pa)
Tight building 0.30
Average 1.00
Leaky building 2.00
20.6 CHAPTER TWENTY
20.3 SPACE PRESSURIZATION
Space Pressure Characteristics
Space pressurization or building pressurization is a process that maintains a static pressure difference
between space air and the air in the surrounding area, either outdoor air at atmospheric
pressure or air in adjacent rooms, for the purpose of providing a required indoor environmental control.
As mentioned in Chap. 4, normally, slightly positive space pressure should be maintained if the
required level of indoor air quality and environmental control of the conditioned space is higher
than that of the surrounding area. In laboratories with toxic gas exhaust systems, negative space
pressure should be maintained. As covered in Chap. 4, for comfort air conditioning systems for
low-rise buildings, the pressure differential between space air pr and outdoor air po is usually
between 	0.005 and 	0.05 in. WC (	1.25 and 	12.5 Pa), and 	0.01 to 	0.03 in. WC (5 to
7.5 Pa) is often adopted. For high-rise buildings, the pressure differences between pr and po are far
higher because of the stack effect on various floors. For clean rooms, the minimum pressure differential
between a clean room and any adjacent area of less strict cleanliness requirements should be
0.05 in. WC (12.5 Pa) with all entryways closed. A space positve pressure of 0.05 in. WC (12.5 Pa)
exerts a force of 4 lbf (18 N) on a door with an area of 16 ft2 (1.5 m2). An excessive pressure differential
may make opening and closing doors difficult.
A positive pressure means that pr  po or pr  psur. Here psur denotes static pressure in the surrounding
area, in in. WG (Pa). When a positive pressure is maintained in a conditioned space, in addition
to the air returned and exhausted from the space, a certain amount of air is squeezed out or
exfiltrated from the space through openings and cracks in the building shell. Unconditioned outdoor
air or contaminated air from the surrounding area is prevented from infiltrating the conditioned
space because of this positive space pressure. A negative pressure means that pr 	 po or pr 	 psur,
which induces infiltration.
Except in air jets, the velocity pressure in the conditioned space is small and is usually ignored
in space pressurization analysis.
The space pressure difference between the conditioned space and the surrounding area is affected
by the stack effect, wind effect, and operation of the air systems and exhaust systems as well
as the airtightness (amount of leakage area) of the building shell.
Stack Effect
The temperature difference between the cold outdoor and warm indoor air columns causes a density
difference between these air columns that, in turn, creates a pressure difference between the cold
and the warm air columns. This phenomenon is called the stack effect.
In any building under the stack effect alone, there is a level at which the indoor and outdoor
pressures are equal. This level is called the neutral pressure level (NPL), as shown in Fig. 20.1b.
During the heating season, outdoor cold air pressure is greater than indoor warm air pressure below
the NPL. This pressure difference causes the outdoor cold air to enter the building through the
lower inlets, and the indoor warm air to flow upward and discharge from the top openings. During
the cooling season, the reverse effect occurs, but the stack effect is smaller because of the smaller
temperature difference between outdoor and indoor air.
The pressure difference due to the stack effect between outdoor and indoor air pst, in in. WC
(Pa, 1 in. WC  248.6 Pa) at height H (in ft or m), from a reference level, is given as
(20.4)  0.19256 i
(H HNPL )(TRi TRo)
TRo
pst  pso psi  0.19256( o 
i)(H HNPL)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.7
where pso, psi  outdoor and indoor absolute static pressure, in. WG (Pa)
o, i  outdoor and indoor air density, lb/ ft3 (kg /m3)
HNPL  vertical distance between NPL and reference level, ft (m)
TRo,TRi  outdoor and indoor absolute temperature, °R
For a high-rise building with two openings on the external wall, NNPL, in ft (m), measured from
lower opening can be calculated as
(20.5)
where Ho  vertical distance between two openings, ft (m)
A1, A2  area of lower and higher openings, ft2 (m2)
TRi, TRo  absolute temperature of indoor and outdoor air, °R (K)
In Eq. (20.5), TRi  TRo. If the indoor air is cooler, that is, TRo  TRi, then TRi /TRo should be
changed to TRo /TRi. For conditioned spaces where TRi  TRo, the chimney and exhaust systems raise
the NPL level and the cold outdoor air supply system lowers the NPL level.
In Fig. 20.1b, it can be seen that the absolute static pressure of outdoor air pso is greater at the
low level than at the high level. This is because the air column is taller at the lower level.
In an ideal single-cell building with no inside partitions, with a vertical distance of 100 ft (30 m)
between the lower air inlet and upper air outlets and an indoor and outdoor air temperature difference
of 50°F (10°C), the static pressure difference pst is about 0.15 in. WC (38 Pa).
Most buildings have inside partitions; i.e., they are multicell buildings. They have doors at the
lower entering inlet, windows at the upper discharge outlets, and doors to the stairwells. Usually
HNPL 
Ho
1 	 (A1/A2)2(TRi / TRo)
20.8 CHAPTER TWENTY
(a)
(c)
(d)
(b)
Neutral pressure
level (NPL)
Elevator
shaft
Elevator
shaft
Elevator
shaft
Exhaust duct
for restrooms
Height
ft
Building
Building
Outside
Outside
Outside
pel po
p in WC
p between
space air
and air in
exhaust duct
(exhaust fan
is shut off )
Outside
door
Elevator
door
Building
FIGURE 20.1 Stack effect in a high-rise building (a) schematic diagram of building airflow; (b) pressure profiles; (c) pressure profiles
for tight building shell; (d ) pressure profiles for leaky building shell.
doors and windows are closed. When outdoor cold air flows through the cracks and gaps around
these doors, windows, and partitions, its velocity and flow rate are significantly reduced because of
these dynamic losses.
Stack effect depends on the magnitude of outdoor-indoor temperature difference and therefore
varies from maximum to zero as the outdoor climate changes.
Stack Effect for High-Rise Buildings
Lovatt and Wilson (1994) measured the pressure differences of the various elements of the building
shell and internal partitions of a 44-story high-rise office building of 577 ft (176 m) above grade
located in Montreal, Quebec, Canada. The air leakage rate for this building was found by Tamura
and Wilson (1967) to be 0.55 cfm / ft2 (2.8 L / s m2). The stairwells in this building are not directly
connected to a stairwell pressurization system. The results were as follows:
Condition Normal occupied
Outdoor temperature 20.3°F (6.5°C)
Lobby temperature 73.0°F (22.8°C)
Pressure difference p, in. WC (Pa)
Outside–lobby vestibule 0.29 (72)
Lobby vestibule–lobby 0.02 (5)
Low-rise elevators–lobby 0.10 (24)
Mid-rise elevators–lobby 0.12 (30)
High-rise elevators–lobby 0.16 (40)
M2 stairwell–lobby 0.008 (2)
Roof elevator–41st floor 0.23 (56)
Roof elevator–roof vestibule 0.32 (80)
Observation deck–roof vestibule 0.07 (18)
Roof–stairwell 0.78 (195)
43d Floor–stairwell 0.19 (47)
42d Floor vestibule–stairwell 0.07 (17)
Exhaust fan room–42d floor vestibule 0.06 (16)
42d Floor–42d floor vestibule 0.17(43)
High-rise elevator mech. room–43d floor 0.21 (53)
Mid-rise elevator mech. room–32d floor 0.02 (5)
Low-rise elevator mech. room–16th floor 0 (0)
The stack effect for high-rise buildings in cold climates may cause the following problems:
 Pressure differences across elevator doors pelev, d generally do not exceed the 0.1-in. WG (25-Pa)
limit. This is the limit recommended by the manufacturer for proper door operation. Sticky elevator
doors may occur when pelev, d  0.1 in. WC (25 Pa).
 Pressure difference across the stairwell doors psta,w is generally less than 0.04 in. WC (10 Pa).
These stairwell doors to occupied floors are easy to open and provide an escape route to outdoors
during a building fire.
 During cold weather, the pressure difference across the roof door proof may well be above the
NFPA limit (1988) between 0.21 and 0.45 in. WC (52.5 and 112.5 Pa).
 The lobby entrance doors show greater pressure difference across them.
 The stack effect may affect the airflow distribution of a vertical exhaust duct from restrooms.
For a tight high-rise building with less air leakage area on the building shell, the pressure profile
of the high-rise building is shown in Fig. 20.1c, where the pressure difference across the outside (entrance)
door is greater than the pressure difference across the elevator doors. For a leaky high-rise
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.9
building, the pressure profile is shown in Fig. 20.1d; the pressure difference across the outside door
is smaller than the pressure difference across the elevator doors.
Lovatt and Wilson (1994) recommended that if the pressure difference of the elevator doors of a
high-rise building during cold climate pelev,d  0.1 in. WC (25 Pa), then repair of the air leakage
area on the building shell is often the first remedy. Reduction of air leakage will reduce the infiltration
and the heat energy use as well.
Add airtight vestibules around all entrance doors, and around doors serving elevators and
stairwells. These setups divide the pressure difference of the entrance door, elevator doors, and
stairwell doors into two parts, and again they reduce infiltration. Add airtight separations in vertical
shafts whenever possible, to reduce the pressure difference across the roof door in the stairwell.
To pressurize or depressurize the entrance lobby or any other floor by a mechanical ventilation
system is often not an effective and economical means to reduce the pressure difference across
doors or other building elements, because one of the pressure differences is reduced at the expense
of increasing the other one.
Wind Effect
As wind flows over a building, it creates positive pressure on the windward side of the building and
negative pressure on the leeward side, as shown in Fig. 20.2. Wind pressure on the building surface
relative to the approaching wind speed pw, in in. WC (Pa), can be calculated as
(20.6)
where Cp  surface pressure coefficient
pvw  velocity pressure at approaching wind speed, in. WC (Pa)
vwa  approaching wind speed at wall height on windward side, fpm (m/ s)
If the wind direction is normal to the windward surface of a high-rise building with a width-depth
ratio of 4, the average Cp value is about 0.60 on the windward side, about 0.5 on the leeward side
and on the flat roof, and about 0.25 on the other two sides.
pw  CpPvw  Cp vwa
4005 2
20.10 CHAPTER TWENTY
FIGURE 20.2 Airflow around buildings.
Wind speed from a meteorological station vmet, in fpm (m / s), is usually measured at a height of
33 ft (10 m). The approaching wind speed at wall height vwa should be corrected for height and terrain
roughness as follows:
(20.7)
where Hw  height of wall where wind pressure is exerted, ft (m)
Hmet  height at which wind speed is measured at meteorological station, ft (m)
In Eq. (20.7), Ao is a correction factor for terrain roughness, and n is the velocity profile exponent.
For suburban areas, Ao  0.6 and n  0.28; for urban areas, Ao  0.35 and n  0.4; for airports,
Ao  1.0 and n  0.15.
The mean wind speeds measured at meteorological station vmet in most U.S. cities in winter are
below 1200 fpm (6 m / s). Wind pressure on a wall at a height of 33 ft (10 m) in suburban areas with
vmet  1200 fpm (6 m / s) is about 0.054 in. WC (13 Pa).
The influence of the wind effect on space and building pressure characteristics is as follows:
1. Rooms on the windward side of the building are usually at a positive pressure and on the
leeward side at a negative pressure relative to the corridor pressure. It is best to build clean
spaces, such as conference rooms, on the windward side of the prevailing wind, and laboratories
with toxic gas exhaust systems on the leeward side of the building.
2. Outdoor air intake should be located on the side with a positive surface pressure coefficient Cp
in the prevailing wind. Exhaust outlets should be located where Cp is negative, preferably on the
rooftop.
3. Sufficient total pressure must be provided by the supply fan to overcome the negative pressure at
the outdoor intake and by the exhaust fan to overcome the positive pressure at the exhaust outlet.
Alternatives should be provided to allow outdoor air intake and outlet when the wind direction is
different from that of the prevailing wind.
Wind speed may vary from zero to its maximum speed. Wind velocity is a vector quantity.
Air Systems and Mechanical Ventilation Systems
If a space is served by air systems including supply, return, and exhaust systems, the space pressure
characteristics are closely related to the characteristics of air systems during their operating time.
When a hurricane comes, the windward pressure may be as high as several inches WC (several hundred
Pa), and the leeward negative pressure has a similar negative pressure. Hurricanes and storms
last a comparatively short time. Hurricanes affect entrance doors, windows, as well as the performance
of outdoor intake and exhaust outlets in the air systems and HVAC&R equipment outdoors
that are not well shielded and protected.
For high-rise buildings in cold winter, the stack effect has a significant influence on the indooroutdoor
pressure difference on various floors. For low-rise buildings, air systems (including
mechanical ventilation systems) have a dominant influence on space pressure characteristics except
on stormy days.
Airflow Balance and Space Pressurization by Differential Flow
Consider a single-cell room model of an enclosed space surrounded by an area of uniform pressure.
Based on the principle of continuity of mass, for an incompressible steady flow process, the total
mass flow rate of air entering an enclosed space must equal the total mass flow rate of air leaving
vwa  Aovmet Hw
Hmet n
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.11
the space (see Fig. 20.3), or
(20.8)
where s  mass flow rate of supply air, lb/ min (kg / s)
rt  mass flow rate of return air, lb/ min (kg / s)
ex  mass flow rate of exhaust air, lb/ min (kg / s) m?
m
? m
?
m? s 	 m? inf  m? rt 	 m? ex
20.12 CHAPTER TWENTY
FIGURE 20.3 Mass flow balance of an enclosed space: (a) mass flow balance; (b) flow circuit.
In Eq. (20.8), inf indicates the mass flow rate of infiltration air, lb/ min (kg / s). The term inf
means the mass flow rate of exfiltration. Because  , for simplicity, if the difference in air
density is ignored, then
(20.9)
where s, inf, rt, ex  volume flow rate of supply, infiltration or exfiltration, return, and exhaust
air, cfm (L / s).
For convenience, if the stack effect and wind effect are zero, and the pressure of the surrounded
area or outside pressure po is equal to the atmospheric pressure and is assumed to be zero, that is,
po  pat  0, then the pressure of the enclosed space is represented by pi, in in. WG (Pa).
Let po,st  po,st pat be the increase of the outside pressure due to the stack effect, and let pow 
pow pat be the increase of the outside pressure due to the wind effect; both are in in. WC (Pa).
Space pressurization by using differential flow control is the most widely used method today.
When po,st and pow both are zero and the difference in air density is ignored, if an enclosed space
is pressurized or depressurized because of differential flow of air (supply and return/exhaust air-
flow) systems, the following relationships must hold:
(20.10 )
From Eq. (20.1)
and
(20.11)
From Eq. (20.10), if s  ( rt 	 ex), then the enclosed space pressure is pressurized from atmospheric
pressure po up to pim, in in. WG (Pa), and inf is negative, that is, inf, an exfiltration. If
s 	 rt 	 ex, pim 	 po, then inf is positive, an infiltration.
From Eq. (20.11), space pressurization (pim po) or space depressurization (pim po)
depends on the amount of differential flow s ( rt 	 ex)  inf that has been infiltrated or exfiltrated
through the building shell and the air leakage area Ae,l (airtightness of the building). The
greater the inf and the smaller the Ae,l, the higher will be the pressurization. Similarly, the greater
the inf and the smaller the Ae,l, the higher the depressurization.
Cummings et al. (1996) made field measurements of uncontrolled airflow and depressurization
of eight restaurants. From the measured results, the average building pressures (pim po) of five
restaurants was greater than or equal to 0.032 in. WC (0.006, 0.003, 0.024, 0.032, and
0.018 in. WC). These are the results of the following two restaurants:
Here OA represents outdoor air and MA the makeup air. The building pressure of the restaurant had
a smaller Ae,l  2.16 ft2 (0.242 m2), and a greater differential of airflow between exhaust flow and
OA and MA flow was depressurized further (0.172 in. WC or 43 Pa).
Coogan (1996) recommended a space pressurization by differential airflow control using a more
complicated room model surrounded by two zones, which is useful when the enclosed space is adjacent
to two spaces of different pressures.
Floor Building pressure,
area, ft ACH50 Ae,l, ft2 OA & MA, cfm Exhaust, cfm in. WC (Pa)
Chicken 1 3,161 14.81 3.90 8,410 10,606 0.032 (7.6)
Chicken 2 3,321 7.70 2.16 6,360 9,272 0.172 (43)
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
V?
( pim po)   V?inf
4005Cd Ae,l 
V?inf  4005Cd Ae,l ?( pim po)
V?inf  V?s (V?rt 	 V?ex)
V?
V?
V?
V?
V?
s 	 V?inf  V?rt 	 V?ex
V?
m?
m? m?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.13
20.4 INFILTRATION AND EXFILTRATION
Infiltration is the uncontrolled inward flow of outdoor air through cracks and openings in the building
shell when the pressure of the outdoor air is higher than the pressure of the indoor air at the
same level. Exfiltration is the uncontrolled outward flow of air through cracks and openings in the
building shell when the pressure of indoor air is higher than that of outdoor air. Exfiltration is negative
infiltration.
Infiltration and exfiltration can be induced by the stack effect, wind effect, air systems or mechanical
ventilating systems, or a combination thereof. If atmospheric pressure is pat zero, the outdoor
pressure po is the sum of the stack effect po,st and wind effect pow, both in in. WC (Pa), and
for the building shell at a normal direction to the approaching wind, outdoor pressure po, in in. WC
(Pa), can be calculated as
(20.12 )
If both po,st and pow are zero and the space pressure is not pressurized or depressurized, then the indoor
pressure p  pat  0. If the space is pressurized by the supply or exhaust system, pi  pim,
both in in. WC (Pa).
Volume Flow Rate of Infiltration
The volume flow rate of infiltration inf, in cfm (m3 /s), for a residential building can often be calculated
by Eq. (20.1) using the effective leakage area Ae,l. For low-rise commercial buildings in summer,
as discussed in Chap. 6, the stack effect is small; if windows are well sealed and the space is
pressurized with a positive pressure, the infiltration can be ignored. For low-rise buildings in winter,
according to field measurements, the following infiltration rates, in ACH, may be considered in
space heating load calculations for rooms with windows on one side of a commercial building that
are not well sealed:
For high-rise buildings in both summer and winter design conditions, the infiltration induced by
stack and wind effect should be considered. First, the outdoor pressure due to stack effect po,st and
wind effect pow for each floor can be calculated by Eqs. (20.4) and (20.6), and the outdoor air
pressure po can be calculated by Eq. (20.12). As the pressurized indoor pressure is pim, the pressure
differece across the building shell p  po pim. Second, from the building construction data, find
the area of the exterior wall, in ft2 (m2). Third, calculate the volume flow rate of the infiltration inf,
in cfm (m3 /s), for each floor by the air leakage equation, Eq. (20.2). According to the airtightness of
the building, select the flow coefficient Cflow as listed previously.
20.5 FAN-DUCT SYSTEMS
A fan-duct system is an idealized air system in which a fan is connected to ductwork or equipment
(a flow resistance), or several fans are connected to many sections of ductwork and equipment.
More precisely, a fan-duct system is a combination of fan and flow resistance. Fan-duct systems are
used to study and analyze the operating characteristics of air systems.
V?
Winter mean wind speed vmet, fpm (m/ s) Infiltration rate, ACH
1000 (5) 0.2
1200 (6) 0.3
V?
po  po,st 	 pow 	 pat
20.14 CHAPTER TWENTY
System Operating Point
A system operating point indicates the operating condition and characteristics of a fan-duct system.
Because the operating point of the fan must lie on the fan performance curve and the operating
point of the duct system must lie on the system curve, the system operating point of a fan-duct system
must be the intersection point P of the fan performance curve and the system curve, as shown
in Fig. 20.4. At the operating point of the fan-duct system, its volume flow rate is P, and its total
pressure loss is pP.
If the flow resistance of the duct system remains the same while fan performance changes, and
its characteristics are represented by a new fan performance curve, the operating point of the fanduct
system moves to point Q in Fig. 20.4. On the other hand, if the fan performance curve remains
the same while the flow resistance of the duct system is varied and represented by a new system
curve, the operating point of the fan-duct system moves from P to S.
The fan in a constant-volume air system should be selected with a and pt so that the system
operating point P is near the point of maximum fan total efficiency 
t. The actual operating point of
a constant-volume system is usually within a range of 10 percent of the wide-open volume from
the point of maximum 
t.
Buckingham Method and Fan Laws
The relationship between various interdependent variables that represent fan characteristics for dynamically
similar systems can be determined by using the Buckingham method. For two systems
V?
V?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.15
FIGURE 20.4 System operating point.
to be dynamically similar, they must be similar in shape and construction, i.e., geometrically similar.
Their velocity distribution or profile of fluid flow should also be similar. The Buckingham 
method gives the relationship as
(20.13)
where F  dependent variable such as , pt , or power
D  diameter of impeller, ft (m)
n  fan speed, rpm
 air density, lb/ ft3 (kg /m3)
In Eq. (20.13), a, b, c are indices to be determined so that the group as a whole is dimensionless.
Variables corresponding to fan-duct systems and their dimensions in terms of mass M, length L,
and time t are as follows:
For the dimensionless group 1 including dependent variable volume flow and independent variables
D, n, and , their relationship can be found by determining the indices as follows:
Setting up equations in M, L, and t, we have
That is,
Or, / (D3n)  constant, then
(20.14)
where Kv  volume flow constant. Similarly,
(20.15)
where Kp  pressure constant and KP  power constant.
Equations (20.14) and (20.15) can be used to determine the performance of a fan-duct system
under the following two conditions:
Pf  KPD5n3 
pt  KpD2n2
V?
 Kv D3n
V?
1  V?D3n1 
V?
D3n
t 0 1 b b  1
L 0  3 	 a 3c a  3
M 0  c
[0]  [L3t 1 ][L]a[t 1 ]b[ML3]c
1  V?Danb c
V?
V?

 FDanb c
20.16 CHAPTER TWENTY
Indices
Variable Unit M L t
Volume flow cfm (m3 /s) 0 3 1
Total pressure pt lbf /ft2 (N/m2) 1 1 2
Power Pf ft lbf /min (kJ / s) 1 2 3
Impeller diameter D ft (m) 0 1 0
Speed n rpm 0 0 1
Air density  lbm/ft3 (kg /m3) 1 3 0
V?
 Condition 1. For the same fan-duct system operating at different speeds D1  D2, if the
difference in air density is negligible, the volume flow rate ratio is proportional to the speed
ratio, or
(20.16)
In Eq. (20.16), subscripts 1 and 2 indicate the original and changed operating conditions, as
shown in Fig. 20.4. The total pressure rise ratio is equal to the square of the speed ratio
(20.17)
The fan power ratio is equal to the cube of the speed ratio
(20.18)
 Condition 2. For fan-duct systems with geometrically and dynamically similar fans and the same
duct system, i.e., the same type of fans having the same Kv, Kp, and KP, operating at high
Reynolds numbers and installed in the same duct system, the volume flow ratio, total pressure rise
ratio, and power ratio can be expressed as
(20.19)
Equations (20.16) through (20.19) relate the parameters of the fan-duct system at various operating
conditions. They are often called the fan laws.
20.6 SYSTEM EFFECT
Mechanism of System Effect
The system effect of a fan-duct system describes the loss of fan performance caused by uneven
and nonuniform velocity profiles at the inlet entering and after the fan outlet using actual operating
inlet and outlet connections, compared to the performance of that fan test unit during laboratory
ratings.
Individual fans are tested based on ANSI/ASHRAE Standard 51-1985 and ANSI/AMCA
Standard 210-85. The configuration of the test unit is shown in Fig. 20.5b. There are no inlet
connections. There is also a minimum length of 10 duct diameters of straight duct connected to the
fan outlet.
The system effect of a fan-duct system consists of inlet system effect and outlet system effect.
p2
p1

Kp D2 5n2 3 2
Kp D1 5n1 3 1

D2 5n2 3
D1 5n1 3
pt2
pt1

Kp D2 2n2 2 2
Kp D1 2n1 2 1

D2 2n2 2
D1 2n1 2
V?
2
V?
1

Kv D2 3 n2
Kv D1 3n1

D2 3 n2
D1 3	n1
P2
P1

KPD2 5n2 3 2
KPD1 5n1 3 1

n2 3
n1 3
pt 2
pt1

Kp D2 2n22 2
Kp D1 2 n1 2 1

n2 2
n1 2
V?
2
V?
1

Kv D2 3n2
Kv D1 3n1

n2
n1
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.17
Inlet System Effect
Consider a 90° elbow that is connected to the fan inlet with a very short connecting duct between
the elbow and the fan inlet. Eddies, large-scale turbulence, and a nonuniform velocity profile form
after the elbow and are not smoothed to a uniform velocity distribution because of the short connecting
duct. Air enters the fan with a nonuniform velocity profile that is quite different from the
20.18 CHAPTER TWENTY
FIGURE 20.5 Inlet system effect: (a) inlet system effect on the pt diagram; (b) fan test unit at laboratory;
(c) round elbow before fan inlet.
V?
fan inlet conditions in the standard laboratory test unit. This results in a fan performance curve with
lower fan total pressure and volume flow than the catalog fan performance curve, as shown in
Fig. 20.5a.
The difference between the fan total pressure in the fan-duct system with a nonuniform velocity
profile at fan inlet and the fan total pressure during standard rating at the laboratory is called the inlet
system effect. Fans must be selected at a fan total pressure that compensates for the inlet system
effect, so that the actual fan performance curve can meet the required fan total pressure at designed
volume flow at operating point Pd, as shown in Fig. 20.5a.
If a fan-duct system does not consider the pressure loss caused by the inlet system effect, it will
actually operate at point Pa with a lower volume flow and fan total pressure than the calculated values,
even though the calculated design operating point is at Pd.
Inlet System Effect Loss
Inlet system effect loss ps,i, in in. WC (Pa), is the difference in pressure loss between actual and
catalog performance curves caused by the inlet system effect, and it can be calculated as dynamic
loss, as defined in Eq. (17.52)
(20.20)
where Cs,i  inlet system effect loss coefficient. In Eq. (20.20), vfi is the air velocity at the fan inlet
based on the area of the inlet collar, fpm (m/ s).
Inlet system effect loss should be added to the system total pressure loss calculated at the designed
volume flow rate, as described in Eq. (17.23). The selected catalog fan performance curve
should meet the system total pressure loss, including the inlet system effect loss.
Inlet System Effect Loss Coefficient
According to the Air Movement and Control Association (AMCA) Fan and Systems, Publication 201
(1990), if an elbow is connected by a round duct to the inlet of a centrifugal fan and the length of the
connecting duct is equal to 2 duct diameters, the inlet system effect loss coefficients are as follows:
Here R is the radius of the elbow, and D is the duct diameter, both in ft (m). If the length of the
connecting duct is 5D, then Cs,i is approximately one-half the values listed above. If there is no
connecting duct, then Cs,i is about twice the values listed above.
According to AMCA (1973), for a square elbow (with or without turning vanes) with an inlet
transition and a connecting duct of length 2D, as shown in Fig. 20.6, the inlet system effect loss coefficients
are as follows:
Here H indicates the dimension of either side of the square duct, in ft (m).
R/H of square elbow 0.5 0.75 1.0 2.0
Cs,i without turning vanes 1.6 1.2 0.66 0.47
Cs,i with turning vanes 0.47 — 0.33 0.22
R/D of elbow 0.75 1.0 2.0
Cs,i 0.8 0.66 0.53
ps,i  Cs,i  vfi
4005 2
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.19
For a square elbow before a fan inlet with a connecting duct of 5D and with turning vanes and
an inlet transition, Cs,i is approximately 0.6 of the value listed above.
If a square elbow before a fan inlet has turning vanes and an inlet transition with no connecting
duct, for R/H  1, Cs,i  0.53; and for R/H  0.5, Cs,i  0.8.
For double-width double-inlet fan installed in a plenum or in a cabinet, the distance between the
fan inlet and the wall of the plenum or cabinet should not be less than 0.4 diameter of the fan inlet.
If it is only 0.2 diameter of the fan inlet, an inlet system effect loss of Cs,i  0.08 should be added
to the system total pressure loss at the design volume flow rate. For details, refer to AMCA Fan and
Systems, Publication 201 (1973).
The fans in most air-handling units and packaged units installed according to ASHRAE and ARI
standards are rated and tested as complete units. The condition of fan inlet of the actual operating
fan-duct system is usually the same as the rated condition. There is no inlet system effect loss in
such a unit.
Outlet System Effect
If a duct fitting is connected to a fan outlet with a short connecting duct, its dynamic loss pdy and
local loss coefficient Cdy increase because of the nonuniform velocity profile of the airstream discharged
from the fan outlet. The difference between the dynamic losses of a duct fitting connected
to a fan outlet with a nonuniform and uniform velocity profile at the duct fitting inlet is known as
the outlet system effect loss ps,o, in in. WC (Pa). The outlet system effect results in an increase of
flow resistance of the duct system, as shown in Fig. 20.7a.
As in Eq. (20.20), outlet system effect loss at design volume flow rate can be calculated as
(20.21)
where Cs,o  outlet system effect loss coefficient
vfo  fan outlet velocity based on fan outlet area, fpm (m/ s)
ps,o  Cs,o vfo
4005 2
20.20 CHAPTER TWENTY
FIGURE 20.6 Square elbow with turning vanes before fan inlet.
20.21
FIGURE 20.7 Outlet system effect: (a) loss on pt diagram; (b) effective duct length; (c) orientation
of elbow connected to duct.
V?
Outlet system effect loss at the design volume flow should be added to the calculated total system
pressure loss.
Outlet System Effect Loss Coefficient
Outlet system effect loss ps,o and outlet system effect loss coefficient Cs,o both depend on the effective
duct length ratio L/Leff, the air velocity at the fan outlet vfo, and the configuration of the duct
fitting with respect to the fan outlet. Here L indicates the length of duct section between fan outlet
and the duct fitting in duct diameters.
The effective duct length Leff is the length of duct between the fan outlet and a cross-sectional
plane at which the uniform velocity profile has been completely recovered, in duct diameters or
equivalent diameters D and L and Leff must be expressed in the same units. Effective duct length can
be evaluated as
(20.22)
For a single-width single-inlet (SWSI) centrifugal fan, the configurations of a 90° elbow with respect
to the fan outlet at the end of the connecting duct can be classified as 90° turn right or 90°
turn down, 90° turn left or 90° turn up, parallel with inlet, and opposite to inlet, as shown in
Fig. 20.7c.
According to AMCA Fan and Systems, Publication 201 (1973), the magnitudes of Cs,0.8 at various
L/Leff values and elbow orientations with a blast area to fan outlet area ratio Ablast /Aout  0.8 are
as follows:
Leff  2.5 	
vfo 2500
1000
20.22 CHAPTER TWENTY
FIGURE 20.7 (Continued)
For a SWSI centrifugal fan with Ablast /Aout  0.7, a multiplier Kblast  1.4 should be applied to the
above values. If Ablast /Aout  0.6, then Kblast  2; and if Ablast /Aout  0.9, then Kblast  0.75.
For double-width double-inlet (DWDI) centrifugal fans, an additional multiplier KDI should be
applied to the Cs,0.8 values. For a 90° elbow connected to the duct at horizontal turning orientation,
KDI  1.25. The outlet system effect coefficient Cs,o can then be determined as
(20.23)
For details, refer to AMCA, Fan and Systems (1973).
For fans installed in air-handling units and packaged units and rated as complete units, the ratings
are based on a straight duct length of 2 duct diameters from the unit outlet. The outlet system
effect loss is not included in the laboratory-rated fan performance curve.
Selecting Fans Considering System Effect Losses
If the calculated volume flow rate of the design operating point Pd (as shown in Fig. 20.8) is d, in
cfm (L / s), and the calculated system total pressure loss is psy, in in. WC (Pa), the selected fan that
V?
Cs,o  KDIKblastCs,0.8
L/Leff
0.12 0.25 0.5 1.0
90° turn right or down 0.64 0.44 0.20 0
90° turn left or up, parallel with inlet 1.16 0.76 0.36 0
90° turn left or up, opposite to inlet 0.80 0.52 0.24 0
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.23
FIGURE 20.8 Selecting fan with performance including system effect loss.
compensates for inlet and outlet system effect losses must operate on a performance curve that intersects
with point Ps, which has a volume flow rate of d and a total pressure rise of pss including
system effect losses, in in. WC (Pa). That is,
(20.24)
System effect directly affects the performance of the air system as well as the function of the air
conditioning system. Because the air velocity at the fan inlet and outlet is usually between 1500 and
3000 fpm (7.5 and 15 m / s), system effect losses may amount to 5 to 20 percent of the system pressure
loss of the air system.
Inlet and outlet system effect losses must be calculated correctly. Moreover, air systems must be
designed to reduce system effect losses to save energy.
20.7 COMBINATION OF FAN-DUCT SYSTEMS
The purposes of combining fan-duct systems are as follows:
 To increase the volume flow rate of the combined fan-duct system
 To provide the required pressure characteristics for the combined fan-duct system
 To balance the volume flow of the combined fan-duct system
 To maintain a required space pressure characteristic
pss  psy 	 ps,i 	 ps,o
V?
20.24 CHAPTER TWENTY
FIGURE 20.9 Two-fan fan-duct system connected in series.
Two Fan-Duct Systems Connected in Series
Figure 20.9 shows two fan-duct systems connected in series equipped with fans F1 and F2 and the
corresponding two duct systems represented by flow resistances R1 and R2.
Flow resistance R1 may be the combination of two flow resistances R1A and R1B connected in
series, and flow resistance R2 may be the combination of two flow resistances R2A and R2B. The
volume flow rate of the combined fan-duct system p, in cfm (L / s), must be the same as the
volume flow rate 1 or 2 of fan-duct system 1 or 2 and their components fan F1, fan F2, or
(20.25)
The total pressure loss across flow resistance R1 has a magnitude p1, and the relationship p1 
R1 1
2 holds. This relationship also determines the system curve S1. Here the total pressure loss in
the duct system includes system effect loss. Fan performance curve Ft1 can be obtained from the
data given in the manufacturer’s catalog.
Total pressure loss p1 and volume flow 1 also determine the operating point of fan-duct
system 1 at point Q, which is graphically the intersection of fan performance curve Ft1 and system
curve S1. For fan-duct system 2, p2, 2, R2, fan performance curve Ft2, system curve S2, and
system operating point T can be similarly determined.
V?
V?
V?
V?
p  V?1  V?2
V?
V?
V?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.25
FIGURE 20.10 Comparison of pressure characteristics: (a) combined fan-duct system; (b) single
fan-duct system.
The total pressure loss of the combined fan-duct system pt, in in. WC (Pa), is
(20.26)
Graphically, the combination of the curves of two fans connected in series on a pt- diagram can
be performed by drawing several constant-volume flow rate lines. For each constant-volume line,
the fan total pressure of the combined fan curve Ft1 	 Ft2 always equals the sum of the sections of
fan total pressure of fan curve 1, Ft1 and fan curve 2, Ft2, represented by p1 and p2, as shown in
Fig. 20.9. The purpose of connecting two fan-duct systems in series is to increase the fan total
pressure that can be provided by the combined system.
Figure 20.10 shows the system pressure characteristics of a combined fan-duct system with
fans F1 and F2 and a single fan-duct system that uses only one fan F. It can be seen that the static
pressure across the duct wall is considerably higher in a fan-duct system with a single fan than in
the combined fan-duct system with two fans. Of course, a single-fan system has a lower initial cost.
For two fan-duct systems connected in series, the volume flow rates of the two fans should be
similar. If a large fan is connected in series with a small fan, at large volume flow rates, the fan total
pressure of the combined system may be less than if the large fan were operated alone. This loss of
efficiency results from the influence of negative total pressure of the small fan when it is operated at
a volume flow rate greater than its wide-open volume flow.
Fans Combined in Parallel and Connected in Series with a Duct System
When two fans F1 and F2 combined in parallel are connected to a duct system represented by flow
resistance R, as shown in Fig. 20.11, the volume flow rate of the parallel combined fans P is the
sum of the volume flow rates of individual fans 1 and 2.
The purpose of such a combination is to increase the volume flow rate of the system. Usually,
the fans in parallel combination are identical fans, i.e., fans of the same model with the same rated
V?
V?
V?
V?
pt  p1 	 p2
20.26 CHAPTER TWENTY
FIGURE 20.11 Two fans combined in parallel and connected in series with a duct system.
volume flow and fan total pressure. As in fan-duct systems connected in series, a large fan should
not be combined in parallel with a small fan because of the possible negative effect of the small fan.
The procedure for determining the combined fan performance curve Ft1 	 Ft2 of the two identical
fans connected in parallel is as follows:
1. Plot the fan performance curves Ft1 and Ft2 according to the manufacturer’s data.
2. Draw the constant total pressure loss pt lines (horizontal lines). Points on the combined curve
along the constant pt lines have a volume flow of 2 1. Here 1 indicates the volume flow of
each identical fan along the constant pt lines.
3. Plot the combined curve Ft1 	 Ft2 by connecting points having  2 1 on the constant pt
lines.
The system curve can be plotted based on the relationship pt  R 2. The intersection of the combined
curve Ft1 	 Ft2 and the system curve is the system operating point P. The volume flow at the
system operating point P, in cfm (L / s), is the volume flow rate flowing through the duct system at
a flow resistance R. The combined system total pressure at the system operating point pP, in in.
WC (Pa), is the fan total pressure of the identical fans or the system total pressure loss of the duct
system.
Example 20.1. An air system uses two identical fans combined in parallel and connected in series
with a duct system, as shown in Fig. 20.11. The pressure-volume and power-volume characteristics
of each of the identical fans are listed below:
If the total pressure loss of this air system is known to be 1 in. WC (250 Pa) at an airflow of 10,000
cfm (4719 L / s), determine the following:
1. The system operating point, the corresponding air volume flow rate, and the system total pressure
when both fans are operating
2. The operating point, the corresponding airflow rate, and the fan total pressure when only one fan
is running, assuming that the air damper after the other fan is closed
3. The fan total efficiency when two fans or only one fan is running
Solution
1. From the given data, the flow resistance of this air system is
The system pressure loss at various airflow rates is as follows:
After the single-fan performance curve is plotted, the combined curve of the two fans in parallel
can be drawn. Plot the system curve from the data given above. The volume flow rate and the
R 
pt
V?
2 
1
(10,000)2  1  108 in. WC / (cfm)2
V?
V?
V?
V?
V?
V?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.27
, cfm 2500 5000 7500 10,000 12,500 15,000
pt, in. WC 5.00 4.88 4.54 4.00 3.13 1.75
Power Pf, hp 3.93 5.90 7.14 7.87 8.20 7.60
V?
, cfm 5000 10,000 15,000 20,000 25,000
pt, in. WC 0.25 1.0 2.25 4.0 6.25
V?
system pressure loss at operating point P, which is the intersection of the combined curve and
the system curve when two fans are running (as shown in Fig. 20.11), are found to be
2. When only one fan is operating and the damper after the other fan is closed, the intersection of
the single-fan curve and the system curve is at point Q (see Fig. 20.11). At point Q
Q  14,500 cfm (6840 L / s), pQ  2.06 in. WC (510 Pa)
3. When two fans are running, each fan has a volume flow of 20,000 /2  10,000 cfm and a fan total
pressure of 4 in. WC. From the power curve, the corresponding fan power input is 7.75 hp.
From Eq. (15.6), the fan total efficiency is then
When only one fan is running, the corresponding fan power input is also 7.75 hp. The fan total
efficiency is
Two Parallel Fan-Duct Systems Connected with Another Duct System
When two parallel fan-duct systems are connected in series to a third duct system represented by
flow resistance R3, as shown in Fig. 20.12, the volume flow rate 3 flowing through R3 is the sum of
volume flow rates flowing through the parallel combined fan-duct systems 1 and 2
Although the total pressure loss pAB between points A and B of parallel combined fan-duct systems
1 and 2 is the same, flow resistances R1 and R2 may be different from each other; therefore, at
their own operating points, the fan total pressure of fan F1, denoted by pt1, may also be different
from the fan total pressure of fan F2, pt2, even though fans F1 and F2 are identical.
Through the concept of residual pressure of a fan-duct system, the performance of two parallel
combined fan-duct systems can be determined. When fan F1 is connected in series with flow resistance
R1, the total pressure loss across R1 is pR1  R1
2. If the volume flow rate of air flowing
through R1 is of a given value 1, the residual pressure after flowing through the flow resistance R1,
pres1, in in. WC (Pa), is
(20.27)
where pF1, Ft1  fan total pressure of fan F1, in. WC (Pa)
pR1, S1  total pressure loss across flow resistance R1 or total pressure loss of duct system
1, in. WC (Pa)
For fan-duct system 2, residual pressure can be similarly calculated.
For two fan-duct systems connected in parallel, the residual pressure of fan-duct system 1 after
R1, pres1, must be equal to the residual pressure of fan-duct system 2 after R2, pres2, because there
is only one unique total pressure ptB at point B.
If two parallel combined fan-duct systems are connected in series with another duct system having
flow resistance R3, the residual pressure provided after fan-duct systems 1 and 2 can be used to
overcome the total pressure loss across flow resistance R3 is as follows:
(20.28) pres1  pres2  R3V?
3 2
pres 1  pF1 pR1  pF1 R1V?1
2  Ft1 S1
V?
V?
V?
BC  V?3  V?1 	 V?2
V?
V?
V?

one 
2.06  14,500
6356  7.75
 0.61, or 61 percent

two 
ptV?
6356Pf

4  10,000
6356  7.75
 0.81, or 81 percent
V?
V?
p  20,000 cfm (9440 L / s) pP  4.0 in. WC (995 Pa)
20.28 CHAPTER TWENTY
Because the pressure at points A and C is equal to atmospheric pressure (zero), the performance of
two parallel combined fan-duct systems connected in series with another duct system can be determined
as follows:
1. Plot two fan curves Ft1 and Ft2, the residual pressure curves Ft1 S1 and Ft2 S2, and the
combined residual pressure curve Ft1 S1 	 Ft2 S2. The residual pressure Ft S curve can be
plotted as follows:
 At a volume flow of , find a corresponding fan total pressure ptfA  FtA from the fan
curve given by the manufacturer.
 Calculate the pressure drop across the duct system at a volume flow
 Calculate residual pressure presA  FtA SA at a volume flow of .
 Plot point A on the residual pressure Ft S curve at a volume flow of and a residual pressure
of presA.
 Similarly, plot points B, C, . . . on the residual pressure curve.
 Residual pressure curve Ft S can be drawn by joining points A, B, C, . . . etc.
At each point on a residual curve, there is a corresponding residual pressure pres  (Ft RV 2)
that can be provided at a corresponding volume flow of to overcome the flow resistance R3 of a V?
V?
A
V?
A
V?
A : SA  ptRA  RV?A
2.
V?
A
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.29
FIGURE 20.12 Two parallel fan-duct systems connected with another duct system.
connected series. Combining of residual pressure curves is performed in the same way as combining
of fan curves.
2. Calculate R3 from the required volume flow rate and total pressure loss in that duct system
represented by R3. From the relationship , draw system curve S3.
3. Determine the intersection point P of the combined residual pressure curve Ft1 S1 	 Ft2 
S2 and the system curve S3, which denotes the operating condition at R3. The residual pressure at
point P is the pressure that can be provided by fan-duct systems 1 and 2 to overcome the total pressure
loss across R3 at that volume flow rate . In fan-duct system design, if the volume
flow rate through R3 thus found is greater or smaller than the required volume flow rate, select a
larger or smaller fan in duct systems 1 or 2 to make them approximately the same.
4. Draw a horizontal line PQ from point P, which intersects with residual pressure curves at
points T and Q, and draw vertical lines QQ and TT from points Q and T. Points Q and T on fan
curve F1 or F2 indicate the volume flow, fan total pressure of the fan-duct systems 1 and 2, and fan
power input to fan F1 and fan F2.
Example 20.2. An air system is equipped with two parallel combined fan-duct systems connected
in series with another duct system, as shown in Fig. 20.12. The fans in the fan-duct systems
are identical and have the same pressure-volume and power-volume characteristics as listed in Example
20.1.
Flow resistances R1, R2, and R3 have the following pressure-volume characteristics:
 R1 has a pt loss of 2.29 in. WC at a of 10,500 cfm.
 R2 has a pt loss of 2.82 in. WC at a of 9500 cfm.
 R3 has a pt loss of 1.50 in. WC at a of 20,000 cfm.
Determine
1. Air volume flow rate flowing through the duct system having flow resistance R3
2. Operating point of fan-duct systems 1 and 2
3. Fan total efficiency of fans 1 and 2
Solution
1. From the given data, the flow resistances can be calculated as
2. From Eq. (20.27), the residual pressure of fan-duct systems 1 and 2 after R1 and R2 at a volume
flow rate of 2500 cfm can be calculated as
pres2  Ft2 S2  5.0 3.12  108  (2500)2  4.80 in. WC
 4.87 in. WC
pres1  Ft1 S1  pF1 R1V?
1 2
 5.0 2.08  108  (2500)2
R3 
1.5
(20,000)2  3.75  109 in. WC/(cfm)2
R2 
2.82
(9500)2  3.12  108 in. WC/(cfm)2
R1 
pt
V?
2 
2.29
(10,500)2  2.08  108 in. WC / (cfm)2
V?
V?
V?
V?
3  V? 	 V?2
pt  R3V?
3 2
20.30 CHAPTER TWENTY
Similarly, the residual pressure of other volume flow rates can be calculated as listed below:
3. At various volume flow rates flowing through the duct system with flow resistance R3, the total
pressure loss pt  can be calculated as follows:
4. Plot the single-fan curve Ft, the residual pressure curves Ft1 S1 and Ft2 S2, the combined
residual pressure curve Ft1 S1 	 Ft2 S2, and the system curve S3 on pt diagram, as
shown in Fig. 20.12. The intersection point P of the combined residual pressure curve Ft1 
S1 	 Ft2 S2, and the system curve S3, gives the volume flow rate in the duct system with flow
resistance R3. From the diagram,  20,000 cfm (9438 L / s).
5. Draw horizontal line PQ from point P to intersect the Ft2 S2 curve at Q and Ft1 S1 curve
at point T. Again, draw vertical lines QQ and TT from points Qand T. The lines intersect fan
curve Ft at points Q and T. For fan-duct system 1, operating point T gives
For fan-duct system 2, the operating point Q gives
6. From the plotted power-volume curve in Fig. 20.12, for operating point T, the fan power input
Pf 1  7.8 hp, and for operating point Q, the fan power input Pf 2  7.6 hp. Fan total efficiency
of fan 1 is therefore
Fan total efficiency for fan 2 is
20.8 MODULATION OF THE FAN-DUCT SYSTEM
Modulation Curve
The modulation curve of a fan-duct system or, practically, an air system, is its operating curve when
its volume flow rate is modulated at part-load operation. A modulation curve is also the locus of the
system operating points at reduced system loads and volume flow rates.

2 
9500  4.15
6356  7.6
 0.82, or 82 percent

1 
V?
1 pR1
6356Pf1

10,500  3.8
6356  7.8
 0.80, or 80 percent
V?
2  9500 cfm(4480 L / s) and pR2  4.15 in. WC (1032 Pa)
V?
1  10,500 cfm(4955 L / s) and pR1  3.8 in. WC (945 Pa)
V?
p
V?
R3V?
3 2
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.31
, cfm
2500 5000 7500 10,000 12,500
Ft, in. WC 5.0 4.88 4.54 4.0 3.13
R1 , in.WC 0.13 0.52 1.17 2.08 3.25
R2 , in.WC 0.20 0.78 1.76 3.12 4.88
Ft1 S1, in. WC 4.87 4.36 3.37 1.92
Ft2 S2, in. WC 4.80 4.10 2.78 0.88
V?
2 2
V?
1 2
V?
, cfm 5000 10,000 15,000 20,000 25,000
, in. WC 0.09 0.38 0.84 1.50 2.34 S3  R3V?
3 2
V?
For a variable-air-volume system with duct static pressure control, the pressure loss of the air
system can be divided into fixed part pfix and variable part pvar when the volume flow rate varies.
This is described in detail in Chap. 21.
Figure 20.13 shows a typical modulation curve of a VAV system with duct static pressure control
using inlet vanes to modulate fan capacity. This VAV system has many fan curves Ft at various
volume flow rates and a system curve for the variable pressure loss Svar.
During reduction of the volume flow rate of a VAV system, the dampers in the VAV boxes close
partially. The variable part of the pressure loss of the air system pvar from the AHU or PU up to the
pressure sensor of duct static pressure control varies as the volume flow is reduced. Its pressure-volume
characteristics are indicated by the system curve Svar. However, the pressure loss of the duct
system, branch takeoffs, and VAV boxes beyond the duct static pressure sensor remains constant,
and is equal to the set point of duct static pressure control pfix. The modulation curve is actually
the system curve of the part of VAV air system with variable pressure loss Svar.
At the same time, the inlet vanes at the fan inlet also reduce their opening in response to the
sensed higher duct static pressure. Therefore, a new fan curve Ft2 is formed. The intersection of Ft2
and system curve Svar at the reduced volume flow rate is the system operating point B. Similarly, at
a further reduction of volume flow rate, the system operating points are C, D, E . . . .
As the volume flow rate of the air system approaches zero, theoretically, the system total pressure
will equal the fixed part of pressure loss pfix, which is equal to the set point of duct static
pressure control, as shown in Fig. 20.13.
The modulation curve of a fan-duct system should not enter the fan surge area or stall region, or
unstable operation will result. However, a forward-curved centrifugal fan may be operated in the
surge area if the system total pressure is below 1.5 in. WC. Refer to the manufacturer’s data for detailed
analysis.
20.32 CHAPTER TWENTY
FIGURE 20.13 Modulation curve of a VAV system installed with duct static pressure control.
Modulation of Fan-Duct Systems
The performance of a fan-duct system can be modulated either by changing the fan characteristics
or by varying the flow resistance of the duct system. This can be achieved by any of the following
methods.
Modulation Using Dampers. Use of a damper is the simplest method of modulation. A multiblade
damper is usually installed inside the main duct after the centrifugal supply fan. When the
damper closes, the system flow resistance increases and the system operating point moves from
point A to B on a different system curve. This method is known as “riding on the curve,” as shown
in Fig. 20.14a. If the fan Ft curve is flatter, then the increase in pt during reduction of the volume
flow rate of the fan-duct system is small.
Because of the considerable total pressure loss across the damper, power input to the fan is only
slightly decreased by this method. Modulation dampers have the lowest installation cost and only
small energy savings.
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.33
FIGURE 20.14 Modulation of fan-duct systems: (a) using dampers; (b) using inlet vanes; (c) using ac inverter
to vary fan speed.
Modulation Using Inlet Vanes. Modulation by varying the opening of inlet vanes at the centrifugal
fan inlet gives different fan performance curves and therefore different system operating points.
It is widely used in many VAV systems. The surge area of a fan with inlet vanes is smaller than that
of a fan without inlet vanes. For a VAV system with duct static pressure control, a modulation curve
and the corresponding fan power inputs are shown by points A, B, and C on the power-volume
curves, as shown in Fig. 20.14b.
Compared to modulation by damper, modulation using inlet vanes has a moderately higher installation
cost and considerable energy savings.
Modulation Using Inlet Cone. Moving the inlet cone of a backward-curved or airfoil centrifugal
fan also gives different fan performance curves, so that its modulation curve is similar to that of
20.34 CHAPTER TWENTY
FIGURE 20.14 (Continued)
inlet vanes. Compared to modulation by inlet vanes, modulation using inlet cone blocks far less fan
inlet area.
Modulation by inlet cone for backward-curved and airfoil centrifugal fans has a comparatively
lower installation cost and also considerable energy savings.
Modulation by Blade Pitch Variation of Axial Fan. Modulation of the volume flow of a fan-duct
system by blade pitch (blade angle) variation of an axial fan changes the fan characteristics and,
therefore, the system operating points. Modulation by automatic blade pitch variation significantly
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.35
FIGURE 20.14 (Continued)
lowers the fan energy consumption at reduced volume flow rate. Although automatic blade pitch
variation is rather expensive, it has become more popular then before.
Modulation by Varying Fan Speed. Modulation of the fan-duct system by varying the fan speed
renders new fan performance curves n2 and n3 and, therefore, new system operating points B and C
at reduced volume flow rates, as shown in Fig. 20.14c. The modulation curve of a VAV system with
duct static pressure control can be obtained in the same way as for inlet vanes.
Fan power input for fan speed modulation is lower than for damper or inlet vane modulations.
Because of the fan speed reduction at reduced , the fan sound power level is reduced accordingly.
Fan speed variation using adjustable-frequency variable-speed drives in a VAV system has become
more popular. An adjustable-frequency ac inverter consumes only 7 to 8 percent more energy
input to the fan motor. The high initial cost of adjustable-frequency variable-speed drives is often
cost-effective for large centrifugal fans. The initial cost may drop further in the future as more and
more variable-speed drives are installed.
Because of the differences in fan characteristics, inlet vane, inlet cone, and variable-speed drive
modulation is widely used in centrifugal fan-duct systems, and blade-pitch and variable-speed drive
modulation is used for axial fan-duct systems.
Example 20.3. A variable-air-volume system equipped with a backward-curved centrifugal fan
with airfoil blades operated at 1700 rpm has the following pressure-volume and power-volume
characteristics:
At design condition, this system has a volume flow rate of 20, 000 cfm (9438 L / s) and a system total
pressure loss of 4.65 in. WC (1156 Pa). When it is operated at 50 percent of its design flow rate and
its flow rate is modulated by inlet vanes, its performance curves have the following characteristics:
Determine the fan power input if the volume flow rate is reduced to 50 percent of its design flow
rate by the following methods:
1. Volume control damper
2. Inlet vanes, where system total pressure loss at 50 percent design flow is 4 in. WC
3. Fan speed variation by means of an adjustable-frequency variable-speed drive
Solution
1. From the given data, the flow resistance of this VAV system at design condition is
The system total pressure loss at various volume flow rates can then be calculated as
S  pt  RdV? 2  1.16  108 V? 2
Rd 
pt
V?
2 
4.65
(20,000)2  1.16  108 in. WC / (cfm)2
V?
20.36 CHAPTER TWENTY
, cfm 5000 10,000 15,000 20,000 25,000
pt, in. WC 6.0 6.0 6.0 4.65 2.70
Power, hp 10.0 15.2 18.7 21.0 20.0
V?
, cfm 5000 7500 10,000 12,500 15,000
pt, in. WC 4.88 4.55 4.0 3.12 1.87
Pf, hp 9.6 11.2 12.0 12.3 12.0
V?
Plot the fan performance curve Ft and the system curve S. The intersection of these curves is the
system operating point P at design condition, as shown in Fig. 20.15a.
2. When the volume flow rate is reduced by a damper to 50 percent of its design volume flow, that
is, 0.5  20,000  10,000 cfm (4719 L / s), its operating point must lie on the fan performance
curve with a  10,000 cfm (4719 L / s), point Q. The fan power input at point Q is 15.2 hp
(11.3 kW), as shown in Fig. 20.15a.
3. When the volume flow is reduced by modulation of inlet vanes to 10,000 cfm (4719 L / s), the
system total pressure loss is 4 in. WC (994 Pa), so the flow resistance of this VAV system at reduced
flow using inlet vanes is
From the given data, plot the fan performance curve using inlet vanes Fiv and the system curve
Siv  Riv , as shown in Fig. 20.15b. Also plot the power-volume curve. The intersection point
Piv of the Fiv and Siv curves is the system operating point at 10,000 cfm (4719 L / s) volume flow,
and its fan power input is 12 hp (8.95 kW), as shown in Fig. 20.15b.
V?
2
Riv 
4
(10,000)2  4  108 in. WC / (cfm)2
V?
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.37
FIGURE 20.15 Modulation of a VAV system using dampers, inlet vanes, and fan-speed variation: (a) using
dampers and fan-speed variation; (b) using inlet vanes.
4. For the same fan-duct system, from the fan law, the fan speed required at reduced volume flow
of 10,000 cfm (4719 L / s) is
Fan power input at a fan speed n2  850 rpm is
The difference in fan power inputs between using damper and variable-speed drive fan
modulation is the power consumption at the damper: 15.2 2.63  12.5 hp (9.32 kW).
At reduced fan speeds, fan total efficiency may be reduced. Because of the degradation of
motor efficiency and ac inverter loss at part-load operations actual power input at 50 percent
of part-load operation for variable-speed drive modulation is far higher than 2.63 hp
(1.96 kW).
P2  P1 n2
n1 3
 21  10,000
20,000 3
 2.63 hp (1.96 kW)
n2  n1 V?2
V?
1   1700  10,000
20,000   850 rpm
20.38 CHAPTER TWENTY
FIGURE 20.15 (Continued)
20.9 CLASSIFICATION OF AIR SYSTEMS
Air systems, in a narrower sense, can be classified into the following categories according to their
system characteristics:
 Air-handling systems, air handlers, and air distribution systems.
 Supply air systems, return air systems, recirulating air systems, and exhaust air systems. A supply
air system may supply conditioned air, outdoor ventilation air, makeup air (either conditioned air
or nonconditioned air), or recirculating air. A return air system often returns the conditioned space
air to the fanroom or machinery room. A recirculating air system transports the conditioned space
air to the AHU, PU, fan-coil unit, or water-source heat pump for conditioning or mixing with
outdoor air again. An exhaust system exhausts the contaminated air or space air to the outside
atmosphere.
 Single-zone or multizone systems. In a single-zone system, there is no terminal. Space temperature,
relative humidity, and volume flow rate are controlled directly by the coils, humidifiers, and
inlet vanes or ac inverter in the air-handling unit or packaged unit. In a multizone system, zone
temperature or zone supply volume flow rate is controlled by terminals.
 Fan combination systems. Three fan combination systems are often used: supply and exhaust fan
combination, supply and relief fan combination, and supply and return fan combination systems.
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.39
Optional
exhaust
fan
Exfiltration
Vent
Slot
diffuser
Return
slots
VAV box
Filter
Fan room
Fan coil
Space recirculating
air system
Ceiling
plenum
Outdoor
air
Return
air
Relief
fan
Return
fan
Return
duct
Recirculating
air
PU
AHU
O
O
I
Supply air
DXcoil
Primary
and
secondary
heat
exchanger
Condensing
unit
Conditioned
space
Conditioned
space
(a)
(c)
(b)
Outdoor
air
Dedicated outdoor
ventilation air system
o
Outdoor air AHU
Supply air
fc cf
m
T2
s
FIGURE 20.16 Various types of air systems: (a) constant-volume system; (b) variable-volume system; (c) dedicated ventilation and
space recirculating system.
 Single-duct or dual-duct systems. In a single-duct system, conditioned air is supplied to the conditioned
space by a single supply duct. In a dual-duct system, conditioned air is supplied to the
conditioned space in the perimeter zone through two supply ducts: a warm air duct and a cold air
duct. In the interior zone, only a cold air duct is needed.
 Constant-volume (CV) or variable-air-volume systems. In a constant-volume system, the temperature
of supply air is modulated to match the variation of space load during part-load operation,
as shown in Fig. 20.16a. In a VAV system, the supply volume flow rate is modulated to maintain a
predetermined space temperature as space load varies, as shown in Fig. 20.16b.
 Dedicated ventilation and space recirculating system. In a dedicated ventilation and space recirculating
system, a required amount of outdoor ventilation air is supplied by a separate ventilation air
system; at the same time, there is a parallel space recirculating system to condition the space recirculating
air to offset the space heating or cooling load, as shown in Fig. 20.16c. Outdoor ventilation
air is either mixed with the space recirculating air first or directly supplied to the conditioned
space. When the outdoor and recirculating air are mixed in the mixing box (mixed plenum)
first; the mixture is then conditioned and supplied to the conditioned space.
Because a constant-volume system is often a single-duct and outdoor recirculating air mixing
system, a variable-air volume system is usually also an outdoor recirculating air mixing system, and
a space conditioning system always employs a separate ventilation air and a parallel space recirculating
air system. Therefore, air systems can be mainly classified into following three primary categories:
 Constant-volume, single-duct outdoor recirculating air mixing systems, or simply constantvolume
systems
 Variable-air-volume outdoor recirculating air mixing systems, or simply variable-air-volume
systems
 Dedicated ventilation and space recirculating systems
Each category of air system has supply, return, recirculating, or exhaust systems and other
specific functions. Dedicated ventilation and space recirculating systems are discussed in Chap. 28.
20.10 CONSTANT-VOLUME SYSTEMS
System Characteristics
A single-zone, single-duct, constant-volume, and outdoor air and recirculating air mixing system is
shown in Fig. 20.16a. A constant-volume system means that there is a constant volume of supply
air throughout the operating period. Supply air temperature is varied when the space load reduces in
part-load operation. Single-zone states indicates that the system serves a conditioned space which
is controlled to maintain a unique indoor temperature, relative humidity, cleanliness, and pressure
differential.
Single-zone constant-volume systems have been widely used in residential buildings and small
retail stores in commercial buildings. They often employ a packaged unit with a DX coil instead of
water cooling coil in an air-handling unit. In the PU, there is often a single supply fan, and the conditioned
space is generally positively pressurized at 0.005 to 0.03 in. WC (1.25 to 7.5 Pa), so that the
amount of outdoor ventilation air intake at the packaged unit can be exfiltrated through the cracks
and gaps on the building shell; or sometimes a small exhaust fan is installed in the restroom. For a
shopping mall or an indoor stadium, an AHU with a return fan is often used. Many air systems that
serve clean rooms are single-zone constant-volume systems. These air systems are demanding, complicated,
and expensive, and are discussed in Chap. 30. In health care facilities and industrial applications,
a single-zone constant-volume system or sometimes a multizone system with reheat is used.
20.40 CHAPTER TWENTY
To save energy, most AHU and PU manufacturers also offer air economizers for single-zone
constant-volume systems when the supply volume flow rate of the air system is 2000 cfm (1000
L/ s) and greater. Most single-zone constant-volume systems employ zone temperature control,
night setback, diagnostics, emergency stop, and safety controls. Air economizer and air system controls
are discussed in the chapters that follow.
Energy per Unit Volume Flow
As specified in ASHRAE/IESNA Standard 90.1-1999 and discussed in Sec. 17.2, the fan power
input to the fan motors per unit volume flow in hp/cfm (W · s / L), for constant-volume air
systems as calculated by Eq. (17.24a):
(20.29)
For variable-air-volume (VAV) systems, the fan power input per unit volume flow can be calculated
by Eq. (17.24b) as:
20.11 AIR CONDITIONING PROCESSES
An air conditioning process determines the change in thermodynamic properties of moist air between
the initial and final states of conditioning and also the corresponding energy and mass transfer
between the moist air and a medium, such as water, refrigerant, or moist air itself during this
change. The energy balance and the conservation of mass in nonnuclear processes are the two principles
most often used in the analysis and calculation of the change of thermodynamic properties in
air conditioning processes.
In general, for a single air conditioning process, heat transfer or mass transfer is positive. However,
for calculations that involve several air conditioning processes, the heat supplied to the moist
air is taken to be positive, and the heat rejected from the moist air is taken to be negative.
Sensible Heat Ratio
The sensible heat ratio (SHR) of an air conditioning process is defined as the ratio of the absolute
value of sensible heat to the absolute value of total heat, or
(20.30)
In Eq. (20.30), total heat, in Btu / h (W), is given as
qtotal   qsen  	  ql 
SHR 
 qsen 
 qtotal 

 qsen 
 qsen  	  ql 
Psy
V?
sd
 0.0015 hp / cfm (0.0024 W s / L) when V?sd  20,000 cfm
Psy
V?
sd
 0.0017 hp / cfm (0.0027 W s / L) when V?sd 	 20,000 cfm
Psy
V?
sd
 0.0011 hp / cfm (0.00174 W s / L) when V?sd  20,000 cfm
Psy
V?
sd
 0.0012 hp / cfm (0.0019 W s / L) when V?sd 	 20,000 cfm (9440 L / s)
Psy / V?sd
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.41
Figure 20.17 shows an air conditioning process from an initial state s to final state r. Refer to Secs.
2.9, 2.10, and Appendix Table B.4 for the SI unit conversion of the parameters on the psychrometric
chart in Fig. 20.17 as well as other figures.
The sensible heat change qsen in this process, in Btu / h (W), can be calculated as
(20.31)
where  volume flow rate of supply air, cfm [m3/(60 s)]
 mass flow rate of supply air, lb/ min [kg / (60 s)]
 density of supply air, lb/ ft3(kg /m3)
cpa  specific heat of moist air, Btu / lb °F (J /kg · °C)
Ts,Tr  moist air temperature at initial and final states, °F (°C)
Latent heat change ql in this process (in Btu / h or W) is given by
ql  60V?s s(wr ws)hfg
s
m?a
V?
s
qsen  60V?s scpa(Tr Ts)  60m?acpa(Tr Ts)
20.42 CHAPTER TWENTY
FIGURE 20.17 Space conditioning lines.
where ws,wr  humidity ratio at the initial and final states, lb / lb (kg / kg). Theoretically, hfg, in
Btu/ lb (J / kg), is the latent heat of vaporization of water at the temperature where vaporization or
condensation occurs. But because the sensible heat change of is already
included in Eq. (20.31), it is convenient to use hfg,32 	 cps(T 32) to replace hfg. The error is usually
less than 0.03 percent. Then the latent heat change ql can be calculated as
(20.32)
In Fig. 20.17, the slope of the air conditioning process sr is given by
(20.33)
where  angle between air conditioning process and horizontal line on psychrometric
chart, deg
Lw, Lt  vertical and horizontal distance between state points r and s, ft (m)
Cw,Ct  scale factor for humidity ratio lines and temperature lines, lb / lb ft (kg /kg m) and
°F/ft (°C/m)
Substituting Eqs. (20.31) and (20.32) into Eq. (20.30) gives
(20.34)
In ASHRAE’s psychrometric chart, as shown in App. Fig. B.1, additional SHR lines are given by
joining the intersection of T  78°F,  50 percent, and the outer sensible heat ratio scale. An air
conditioning process that is parallel to any one of these SHR lines has the same SHR.
20.12 SPACE CONDITIONING AND SENSIBLE COOLING AND
HEATING PROCESSES
Space Conditioning Process
A space conditioning process is an air conditioning process in which either:
1. Heat and moisture are absorbed by the supply air and removed from the space, or
2. Heat, or sometimes heat and moisture, is supplied to the space to compensate for the transmission
and infiltration losses through the building shell, with moisture being given up by the supply
air to the space. The purpose of these processes is to maintain a desirable space temperature
and relative humidity.
Figure 20.17 shows two lines, sr and sr, to indicate these two space conditioning processes.
These lines are called space conditioning lines. The upper line, sr denotes the absorbtion of space
heat and moisture during summer, and the bottom line indicates the supply of heat and absorption
of moisture during winter.
In space conditioning processes, assuming that the kinetic energy difference between the supply
inlet s and return exit r is negligible, and that there is no work being done during these processes,

1
1 	 (hfg,32Cw tan ) / (cpaCt)
SHR 
m?acpa(Tr Ts)
V?
a acpa(Tr Ts ) 	 m?a(wr ws )hfg,32
tan 

Lw
Lt

Ct(wr ws)
Cw(Tr Ts)
 60m?a(wr ws) (1075)  64,500V?s s(wr ws)
ql  60V?s s(wr ws)hfg,32  60  1075V?s s(wr ws)
m?acpa(wr ws)(Tr Ts)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.43
the steady flow energy equation can be simplified to
(20.35)
where hr, hs  enthalpy of space air and supply air, Btu / lb (J / kg)
Qrc  heat to be removed from conditioned space, or space cooling load, Btu /h (W)
Rearranging the terms, then, we can calculate the space cooling load as
(20.36)
where  volume flow rate of supply air, cfm [m3/(60 s)]
s  density of supply air, lb/ ft3 (kg /m3)
Considering the mass balance during the space conditioning process
where wr, ws  humidity ratio at exit r and inlet s, lb/ lb (kg / kg)
 rate of moisture gain in conditioned space, lb/ min [kg / (60 s)]
Again, by rearranging the terms, the rate of space moisture gain can be evaluated as
(20.37)
Sensible Heating and Cooling Processes
A sensible heating process is a process in which heat is added to the moist air, resulting in an increase
in its temperature, while its humidity ratio remains constant, as discussed in Sec. 15.9. This process is
represented by line el in Fig. 15.29. The sensible heating process occurs when moist air flows through
a heating coil in which heat is transferred from the hot water inside the coil tubes to the moist air, or
through a heat exchanger where heat transfer takes place between two fluid streams. The rate of heat
transfer from the hot fluid to the cold fluid Qch, in Btu /h (W), is called the heating coil load or the
heating capacity of the heat exchanger. From Eq. (15.24) the heating coil load can be evaluated as
(20.38)
where he,hl  enthalpy of moist air entering and leaving coil (or heat exchanger), Btu/ lb (J / kg).
From Eq. (2.29),
and from Eq. (2.36)
Substituting Eq. (2.29) into Eq. (20.38), assuming that the differences between cpd1 and cpd2, and
cps1 and cps2, are negligible, and that for a sensible heating process the humidity ratio of moist air
entering is equal to leaving air we  wl,
(20.39)
A sensible cooling process removes heat from the moist air, resulting in a drop in its temperature
while maintaining a constant humidity ratio of the moist air, as discussed in Sec. 15.9. This process
is represented by line e l in Fig. 15.29. The sensible cooling process occurs when moist air flows
through a cooling coil and there is an indirect heat transfer from the moist air to the chilled water
inside the cooling coil; or in an air-to-air heat exchanger where two airstreams do not make contact.
In a sensible cooling process, the rate of heat transfer Qcs, in Btu / h (W), from the moist air to
the chilled water inside the cooling coil or from one airstream to another airstream in the heat
Qch  60m?acpa(Tl Te)  60V?s scpa(T1 T2)
cpa  cpd 	 wcps
h  cpdT 	 w (hg0 	 cpsT )
Qch  60m?a(hl he )  60V?s
s(hl he )
m?g  m?a(wr ws )  V?s s(wr ws )
m?g
m?aws 	 m?g  m?awr
V?
s
Qrc  60 m?a(hr hs)  60 V?s s(hr hs)
60 m?ahs 	 qrc  60 m?ahr
20.44 CHAPTER TWENTY
exchanger is called the sensible cooling coil load or sensible cooling capacity of the heat exchanger.
It can be calculated as
(20.40)
20.13 HUMIDIFYING PROCESSES
In a humidifying process, water vapor is added to the moist air, which increases the humidity ratio
if the initial moist air is unsaturated. As discussed in Chap. 15, air is humidified by the following
methods:
1. Steam injection or submerged heating element
2. Atomizing devices
3. Wetted media evaporation or spraying air washers
Steam Injection and Heating Element Humidifier
In a humidifying process using steam injection, steam is often supplied from the main line to a
grid-type humidifier and then injected directly into the air through small holes, as discussed in Sec.
15.19 and shown in Fig. 20.18. The steam injection humidifying process is indicated by line 12,
Qcs  60m?acpa(Te Tl)  60V?s scpa(Te Tl)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.45
FIGURE 20.18 Humidifying processes.
which is approximately parallel to the constant temperature lines on the psychrometric chart. The
slight inclination toward the right-hand side at the top of line 12 is due to the high temperature of
the injected steam. The increase in the moist air temperature because of the steam injection can be
calculated as follows:
When a mass flow rate of lb /h [kg / (60 s)] of dry saturated steam at low pressure is injected
into a moist airstream of a mass flow rate of , lb/ h [kg / (60 s)], according to the principle of heat
balance
(20.41)
where h1, h2  enthalpy of moist air entering and leaving steam injection humidifier, Btu / lb
(J / kg)
hs  enthalpy of injecting steam, Btu/ lb (J / kg)
The mass flow rate of injecting steam is defined as
(20.42)
where w1, w2  humidity ratio of moist air entering and leaving steam injection humidifier, lb/ lb
(kg / kg). Let ; then
(20.43)
Assuming that cpd and cps are constants, as wcps 		 cpd, let w12  (w1 	 w2) /2 and replace w1 and
w2 by w12. Then
(20.44)
where Ts  steam temperature, °F (°C)
T1, T2  temperature of moist air entering and leaving steam injection humidifier, °F (°C)
A heating element humidifier employs a steam coil or electric heating element to provide the latent
heat of vaporized water. The saturated water vapor added to the moist air is generally at a temperature
higher than that of the airstream.
Air Washer
An air washer is a device that sprays water into air to humidify, to cool and dehumidify, and to
clean the air, as discussed in Sec. 15.21. When moist air flows through an air washer sprayed with
recirculating water, as shown in Fig. 20.18b, or water is distributed over a wetted medium, the
moist air is humidified, tending to approach saturation. This actual adiabatic saturation process follows
the thermodynamic wet-bulb temperature line on the psychrometric chart represented by line
12 in Fig. 20.18. This process increases the humidity ratio of the airstream while resulting in a reduced
air temperature. The cooling effect of the adiabatic saturation process is often called direct
evaporative cooling and is covered in greater detail in Chap. 27.
A saturated end state is not usually achieved by such a process. The saturation efficiency sat of
the moist air leaving the air washer is defined by Eq. (15.54) and discussed in Sec. 15.21.
Oversaturation
In water atomization for air humidification, an atomizing device breaks water into fine particles that
are injected into the airstream. Whether the atomization is accomplished by the centrifugal force,
T2 T1 
wsmcpsTs
cpd 	 w12cps
h1 	 wsmhs  h2
wsm  m?s / m?a  w2 w1
m?s  m?a(w2 w1)
m?ah1 	 m?shs  m?ah2
m?a
m?s
20.46 CHAPTER TWENTY
compressed air, or ultrasonic force, the humidification process is an actual adiabatic saturation
process. With some atomizing devices, moist air leaving the air washer or pulverizing fan (as shown
in Fig. 20.18c) can contain unevaporated water particles ww great enough to exceed the humidity ratio
of saturated air at the thermodynamic wet-bulb temperature . The excess amount of water particles
present in the moist air is called oversaturation. Oversaturation is defined as
(20.45)
The quantity of unevaporated water particles at state point 2, in lb / lb (kg / kg) is
(20.46)
where wo  sum of humidity ratio w2 and minute water particles ww at state point 2, lb/ lb (kg / kg)
w2  humidity ratio at state point 2, lb/ lb (kg / kg)
When adiabatic heat transfer occurs between the airstream and the minute water particles, some
evaporation takes place, so the humidity ratio of the moist air will increase. Such a transformation
still follows the thermodynamic wet-bulb temperature line 10, as shown in Fig. 20.18.
As the moist air flows through an air washer or an atomizing device, there can be minute water
particles present in the moist air, even if the water eliminators are employed. The magnitude of ww
depends mainly on the construction of the humidifying device and the water eliminators, as well as
the air velocity flowing through them. When moist air flows through an air washer, ww may vary
from 0.0002 to 0.001 lb / lb (kg / kg). If a pulverizing fan is used, ww may be as high as 0.00135
lb / lb (kg / kg).
Oversaturation is beneficial to a space or to a process where humidification is required. However,
oversaturation causes wetted surfaces and dampness and often facilitates the growth of mold
and fungus, resulting in a serious IAQ problem. Necessary remedies must be taken to eliminate
hazardous oversaturation and dampness along the airflow passage.
20.14 COOLING AND DEHUMIDIFYING PROCESS
In a cooling and dehumidifying process, both the temperature and the humidity ratio of the moist
air will drop. This process is represented by curve m-cc on the psychrometric chart in Fig. 20.19,
where m is the entering mixture temperature of the outside and recirculating air.
There are three types of heat exchangers commonly used in the cooling and dehumidifying
process:
1. Direct-expansion DX coil in which refrigerant evaporates directly inside the coil’s tubes, as discussed
in Sec. 10.2
2. Water cooling coils with chilled water flowing inside the coil’s tubes, as discussed in Sec. 15.10
3. Air washer where chilled water rather than recirculated water is used for spraying, as discussed
in Sec. 15.21
The evaporating temperature of the refrigerant Tev in a DX coil and the outer surface temperature
of the coil’s tubes Ts,t are usually lower than the dew point of entering air T ae. If the temperature
of the entering chilled water in a water cooling coil Twe and the outer surface temperature of the
coil’s tubes Ts,t are lower than the dew point of entering air T  ae, condensation occurs on the outer
surface of the coil tubes on both DX and water cooling coils. Temperature, humidity ratio, and
enthalpy of moist air are changed during such a cooling and dehumidifying process.
Based on the principle of heat balance,
 	 	
Heat energy
of condensate
Cooling coil load (or cooling
capacity of washer)
total enthalpy
of leaving air
Total enthalpy
of entering air
ww  wo w2
wo w*s  w2 	 ww w*s
ws
*
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.47
That is,
where hae, hcc  enthalpy of moist air entering and leaving the cooling coil or air washer, Btu / lb
(J / kg)
 mass flow rate of condensate, lb/ min [kg / (60 s)]
hw  enthalpy of condensate, Btu / lb (J / kg )
The cooling coil load or the cooling capacity of the air washer Qcc, in Btu / h (W), can be calculated
as
(20.47)
The mass flow rate of the condensate can be evaluated as
where wae, wcc  humidity ratio of moist air entering and leaving coil or air washer, lb/ lb
(kg / kg). Assuming that the temperature of the condensate Tw  Tcc, then we see that Eq. (20.47)
becomes
(20.48a)
where cpw  specific heat of water, Btu / lb °F (J /kg °C). In most cases, is less than m?whw
 60 V?s s[hae hcc (wae wcc)cpwTcc]
Qcc  60m?a[hae hcc (wae wcc)cpwTcc]
m?w  m?a(wae wcc)
 60V?s s(hae hcc) 60m?whw
Qcc  60m?a(hae hcc) 60m?whw
m?w
60m?ahae  60m?ahcc 	 Qcc 	 60m?whw
20.48 CHAPTER TWENTY
FIGURE 20.19 Cooling and dehumidifying process.
0.02 Qcc. Because is small compared to , for practical calculations, is often
neglected, so
(20.48b)
A straight line has been used to represent a cooling and dehumidifying process, and the intersection
of this straight line and the saturation curve is considered the effective surface temperature of
the coil and is called the apparatus dew point. This is a misconception. First, the cooling and dehumidifying
process is actually a curve instead of a straight line. Second, the effective surface temperature
as a fictitious reference point is affected by many factors and is difficult to determine in heatand
mass-transfer calculations.
It is more accurate to define the apparatus dew point as the dew point of the moist air leaving the
conditioning apparatus, the coil, the air washer, or other heat exchangers. Apparatus dew point is
represented by point ad on the saturation curve of the psychrometric chart in Fig. 20.19.
However, the sensible heat ratio of the cooling and dehumidifying process SHRc can be
indicated by the slope of the line joining points m and cc, as shown in Fig. 20.19. SHRc can be calculated
as the ratio of sensible heat removed during cooling and dehumidifying process Qcs to the
Qcc  60m?a(hae hcc)  60V?s s(hae hcc)
m?whw m?a(hae hcc) m?whw
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.49
FIGURE 20.20 Cooling and dehumidifying process based on manufacturers’ data.
total heat removed Qcc, both in Btu / h (W)
(20.49)
According to the data published by U.S. manufacturers, a cooling and dehumidifying curve can
be plotted for the moist air flowing through a water cooling coil. Fig. 20.20 shows such a curve.
The operating conditions are as follows:
Entering air 80°F (26.7°C) dry-bulb and 67°F (19.4°C)
wet-bulb temperatures
Face air velocity 500 fpm (2.5 m / s)
Entering water temperature 45°F (7.2°C)
Water temperature rise 10°F (5.6°C)
The number of rows of depth, i.e., the number of coil tubes along the airflow, varies from 2 to 10
rows, and the fin spacing varies from 8 to 18 fins per inch (1.4- to 3.2-mm fin spacing). Both of
these factors affect the outer surface area of the coil.
Note that the relative humidity of the air leaving the coil approaches saturation as the outer
surface area of the coil increases. The cooling and dehumidifying process for a DX coil is similar
to that of the water cooling coil. Therefore, for coils with fin spacing of less than 10 fins per
inch and for entering air at 80°F (26.7°C) dry-bulb and 68°F (20°C) wet-bulb temperatures, the
relative humidity of the conditioned air leaving the cooling coil may be estimated approximately as
follows:
4-row coils 90–95 percent
6-row and 8-row coils 96–98 percent
Many factors affect the final state of the conditioned air leaving the coil. Use manufacturer’s data or
refer to Secs. 10.2 and 15.10 for a more detailed analysis.
20.15 ADIABATIC MIXING AND BYPASS MIXING PROCESSES
Two-Stream Adiabatic Mixing Process
In a two-stream adiabatic mixing process, two moist airstreams (say, 1 and 2) are mixed together
adiabatically, forming a uniform mixture m in a mixing chamber. Such a mixing process is illustrated
by line 1-m1-2 on the psychrometric chart in Fig. 20.21. In actual practice, for the mixing of
either recirculating and outdoor airstreams or the conditioned airstream and bypass airstream, the
rate of heat transfer between the mixing chamber and the ambient air is small. It is usually ignored.
Based on the principle of heat balance and the continuation of mass
(20.50)
In Eqs. (20.50), subscripts 1, 2, and m represent airstream 1, airstream 2, and the mixture, respectively.
Because and , substituting into Eq. (20.50) gives
m?1/m?m 
h2 hm
h2 h1

w2 wm
w2 w1
 K1
m?2  m?m m?1 m?1  m?m m?2
m?1 	 m?2  m?m
m?1w1 	 m?2w2  m?mwm
m?1h1 	 m?2h2  m?mhm
SHRc 
Qcs
Qcc
20.50 CHAPTER TWENTY
and (20.51)
where K1, K2  fraction of airstreams 1 and 2 in the mixture.
From the above-mentioned relationships
1. The mixing point m must lie on the line that connects state points 1 and 2.
2. The ratio of the mass flow rate of airstream 1 to the mass flow rate of the mixture is exactly
equal to the ratio of segment m1-2 to segment 1-2, as shown in Fig. 20.21, that is,
(20.52)
Similarly, the ratio of the mass flow rate of airstream 2 to the mass flow rate of the mixture is
exactly equal to the ratio of segment m1-1 to segment 1-2, that is,
(20.53)
As the mass flow rate , substituting into Eqs. (20.50) gives
(20.54)
V?
1 1 	 V?2 2  V?m m
V?
1 1w1 	 V?2 2w2  V?m mwm
V?
1 1h1 	 V?2 2h2  V?m mhm
m?  V?
m?2
m?m

m1-1
1-2
m?1
m?m

m1-2
1-2
m?2 / m?m 
hm h1
h2 h1

wm w1
w2 w1
 K2
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.51
FIGURE 20.21 Adiabatic mixing and bypass mixing processes.
Because the magnitude of m always lies between 1 and 2, if the differences between m and 1, m
and 2, and the differences in specific heat cpa are negligible, then
(20.55)
Bypass Mixing Process
In many air-handling units, one airstream is divided into an upper hot deck air-stream and a
lower cold deck airstream, as shown in Fig. 20.21. During the cooling season, the lower airstream is
cooled and dehumidified by the cooling coil, and the hot water supply to the heating coil in
upper hot deck is shut off. Therefore, the upper airstream becomes a bypass airstream, i.e., it bypasses
the cooling coil. After the coils, the conditioned airstream is mixed with the bypass airstream
and forms a mixture at state point m2. This bypass mixing process is shown by line cc-m2-sf in
Fig. 20.21.
During the heating season, the cooling coil is not energized, but the heating coil is. Under such
circumstances, the upper conditioned warm airstream and the lower bypass airstream combine to
form a mixture m2, as indicated by line ch-m2-sf in Fig. 20.21.
Similar to the two-stream adiabatic mixing process, the ratio of the mass flow rate of the conditioned
stream to the mixture is
and
(20.56)
where Kcc, Kch = fraction of conditioning stream flowing through the cooling coil or heating coil.
Also
(20.57)
Subscript cc represents the condition off the cooling coil, ch the heating coil, by the bypass
airstream, which is the same as the state point of air at the supply fan discharge sf. According to the
principle of energy balance
(20.58)
Based on the mass flow rate of the mixture, the cooling coil load can be calculated as
(20.59) Qcc  KccVs s(hsf hcc)
wm 
m?ccwcc 	 m?bywsf
m?m

m?chhch 	 m?byhsf
m?m
hm 
m?cchcc 	 m?byhsf
m?m
m?mwm  m?ccwcc 	 m?by  wsf
m?m  m?cc 	 m?by  m?ch 	 m?by
m?ch
m?m

sf-m2
sf-ch
 Kch
m?cc
m?m

sf-m2
sf-cc
 Kcc
m?m m?cc
V?
1 	 V?2  V?m
V?
1T1 	 V?2T2  V?mTm
V?
1w1 	 V?2w2  V?mwm
V?
1h1 	 V?2h2  V?mhm
20.52 CHAPTER TWENTY
and the heating coil load is
(20.60)
20.16 SINGLE-ZONE, CONSTANT-VOLUME SYSTEMS—
COOLING MODE OPERATION
Air Conditioning Cycle
An air conditioning cycle consists of several air conditioning processes connected in a serial order.
Strictly speaking, such processes determine the operating performance of various components of
the air system in an air conditioning system including the heat, mass, and other energy transfer and
the changes in property of the moist air.
Different air systems are characterized by different types of air conditioning cycles. The psychrometric
analysis of an air conditioning cycle is an important tool to determine the properties of
moist air at various state points, the volume flow rates, and the capacities of the major components
of the air system.
Air conditioning cycles can be divided into two categories: open and closed cycles. In an open
air conditioning cycle, the moist air at the end state will not resume its original state. An air conditioning
cycle of a 100 percent outdoor air makeup system is an open cycle. In a closed air conditioning
cycle, moist air at the end state will resume its original state. The cycle of conditioning the
mixture of recirculating and outdoor air, supplying it to the space, maintaining it at a required temperature
and relative humidity, and recirculating it again is a closed cycle.
Based on space load characteristics, the air conditioning cycles for constant-volume single-zone
systems operate mainly in two modes: cooling mode or heating mode.
1. Cooling mode. When a single-zone, constant-volume system is turned on and its space temperature
Tr is raised above a predetermined limit, such as 75°F (24.4°C), the control system calls
for cooling (cold air supply), and the air economizer and refrigeration will be turned on in
sequence to maintain Tr within required limits. The constant-volume system is said to be operated
at cooling mode. Its equipment is selected to satisfy the summer outdoor and indoor design
conditions.
2. Heating mode. When a single-zone, constant-volume system is turned on and its Tr drops below
a predetermined limit, such as 70°F (21.1°C), the control system calls for heating (warm air supply)
to maintain a required Tr within predetermined limits. The constant-volume system is said
to be operated at heating mode. The same equipment is also selected to satisfy winter outdoor
and indoor design conditions.
Cooling Mode Operation in Summer
In Fig. 20.22, the upper air conditioning cycle is the summer cooling mode cycle for a constantvolume
single-zone system in a factory. Return air from the conditioned space may pass through a
return fan. Recirculating air enters the air-handling unit or packaged unit at state point ru and is
mixed with the outdoor air at point o in the mixing box. The mixture m is then cooled and dehumidified
at the cooling coil and leaves it at point cc. After that, the conditioned air flows through the
supply fan sf, the supply duct, and then supplies to the conditioned space through the supply outlet
at state point s. When supply air absorbs the sensible and latent loads from the space, it becomes
space air at point r. Space air then flows through ceiling plenum rp, absorbs heat released from the
light fixtures and the roof, and flows through the return duct / return fan rf. The recirculated portion
enters the AHU or PU again at point ru.
Qch  KchV?s scpa(Tch Tsf)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.53
20.54
FIGURE 20.22 Summer cooling mode air conditioning cycle for a constant-volume single-zone system in a factory.
Cooling mode operation in summer consists of the following processes:
1. Sensible heating process r-ru from the return system heat gain qr,s, in Btu / h (W), when return
air flows through the ceiling plenum, the return duct, and the return fan. It includes
 Heat gain from the electric lights recessed on the ceiling plenum qrp
 Return fan power heat gain qrf, if there is a return fan
 Return duct heat gain qrd
That is;
(20.61)
It is often convenient to use overall temperature rise to denote the system heat gain, such as, Tr,s,
in °F (°C). Then the return system heat gain is
(20.62)
Subscript ret indicates return air and ru the recirculating air entering the AHU or PU. The volume
flow rate of return air is often higher than that of the recirculated air entering the mixing box if
there is leakage or exhaust air through the exhaust / relief damper.
2. Adiabatic mixing process of recirculating air at state point ru and the outdoor air at state point
o in the mixing box, which is represented by line ru-m-o on the psychrometric chart in Fig. 20.22.
qr,s  60V?ret retcpa(Tru Tr)  60V?ret retcpa Tr,s
qr,s  qrp 	 qrf 	 qrd
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.55
FIGURE 20.23 Air conditioing cycle—winter cooling mode operation for a constant-volume single-zone system.
After mixing, the condition of the mixture is m, which is also the state point of the moist air entering
the cooling coil.
3. Cooling and dehumidifying process, which is indicated by the curve m-cc. The cooling coil
load can be calculated by Eq. (20.48b).
4. Sensible heating process cc-s from the supply system heat gain qs,s when supply air flows
through the supply fan and supply duct. It includes
 Supply fan power heat gain qsf represented by line cc-sf
 Supply duct heat gain qsd indicated by line sf-s.
Supply system heat gain is then given as
qs,s  qsf 	 qsd (20.63)
The supply fan power heat gain and the supply duct heat gain are far greater than the heat gains of
the return fan and duct. These differences are caused by the higher fan total pressure in the supply
fan and the larger temperature difference between the air inside and outside the supply duct.
Like the return system heat gain the supply system heat gain can be represented by an overall
temperature rise Ts,s, in °F (°C).
5. Space conditioning process, which is represented by line sr. From Eqs. (20.31) and (20.36),
the sensible heat removed from the conditioned space or the space sensible load Qrs, in Btu / h (W),
is
(20.64)
From Eq. (20.32), the latent heat removed from the conditioned space or the space latent load Qrl, in
Btu/ h (W), is
(20.65)
Therefore, the space cooling load Qrc, in Btu / h (W), is calculated as
Qrc  Qrs 	 Qrl (20.66)
Cooling Mode Operation in Winter Without Space Humidity Control
In winter, many factories that employ constant-volume single-zone systems still have a space cooling
load, need a cold air supply, and operate at cooling mode operation. These systems can be
divided into two categories: systems without space humidity control, as shown by the lower air conditioning
cycle in Fig. 20.23, and systems with space humidity control, as shown by the upper air
conditioning cycle in Fig. 20.23.
The air conditioning cycle of a constant-volume single-zone system without space humidity
control operated at cooling mode winter design conditions consists of the following air conditioning
processes, and most are similar to those in summer design conditions:
 Sensible heating process r-ru from the return system heat gain qr,s
 Adiabatic mixing process of recirculating air ru and the outdoor air o in the mixing box
 Sensible heating process m-sf-s from the supply system heat gain qs,s
 Space conditioning process sr
Most of the constant-volume single-zone systems operated at cooling mode winter design
conditions do not need a cooling and dehumidifying process because the mixture of outdoor and
Qrl  60m?a(wr ws)hfg,32  60V?s s(wr ws)hfg,32
Qrs  60m?acpa(Tr Ts)  60V?s scpa(Tr Ts)
20.56 CHAPTER TWENTY
recirculating air m is usually lower than the supply temperature plus the temperature rise due
to supply system heat gain. This winter cooling mode free cooling cycle can be constructed as
follows:
1. From Eq. (20.65), calculate the humidity ratio difference wr ws from the known space latent
load and supply volume flow rate.
2. The volume flow rate of outdoor air should be greater than the occupant requirements, and
sufficient to offset the space cooling load during winter without refrigeration. Based on the same
supply volume flow rate as in summer, use Eq. (20.31) to calculate the supply air temperature
difference Tr Ts from the design winter sensible cooling load, and determine Ts.
3. Temperature of the recirculating air entering the AHU or PU Tru can be calculated as
Tru  Tr 	 temperature rise due to return system heat gain
Mixing temperature of outdoor and recirculating air can be calculated as
Tm  Ts temperature rise due to supply system heat gain
4. Because wr  wru and wm  ws, the ratio
(20.67)
From Eq. (20.67), wr wo and wr can be determined.
Cooling Mode Operation In Winter with Space Humidity Control
Constant-volume single-zone systems using cold air supply with space humidity control during the
winter season have been adopted in many industrial applications, such as spinning and weaving
rooms in textile mills. In these workshops, high internal sensible heat loads are released by the
spinning machines and weaving looms, requiring cold air supply and a specific controlled relative
humidity, even in winter.
The upper air conditioning cycle of Fig. 20.23 shows such a cooling mode cycle in winter with
space humidity control using a spray-type air washer for a constant-volume single-zone system.
The adiabatic saturation process of evaporative cooling humidification in an air washer saves energy
to evaporate the liquid water which is especially suitable for conditioned space that needs a
cold air supply in winter. Air washers also cool and dehumidify the air in summer by chilled water
and clean the air by water spraying.
The following are the air conditioning processes in this cycle and their characteristics:
1. Process m-sc is an adiabatic saturation process. It proceeds along the thermodynamic wet-bulb
temperature line on the psychrometric chart.
2. State point sc indicates the conditioned air leaving the air washer. Process lines sc-sf, sf-s, and s-r
all undergo heating and humidifying processes. Because the airstream leaving the air washer is
oversaturated with minute water particles, as the airstream picks heat from the fan, ductwall, and
space machine loads, more evaporation occurs. The increase in the humidity ratios in each of
these processes varies from 0.0001 to 0.0003 lb / lb (kg / kg). These processes can be determined
from both the increase of the enthalpy and the humidity ratios.
3. Because of the high machine load, the sensible heat ratio of the space conditioning line SHRs
often exceeds 0.90.
4. Outdoor air is used for free cooling as well as for occupants and the dilution of air contaminants.
wru wm
wru wo

wr ws
wr wo

Tru Tm
Tru To
V?
o
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.57
Part-Load Operation and Controls
In a constant-volume single-zone system operated at cooling mode, when the space cooling load reduces,
two types of control are used to maintain the required space temperature or other controlled
variable during part-load operation:
Water Flow Rate Modulation Control. For an air system employing a water cooling coil in an
air-handling unit, the reduction of space sensible cooling load causes a drop in space temperature
Tr. When a drop of Tr is sensed, the controller closes the two-way valve of the chilled water supply
to maintain the required space temperature. The reduction of chilled water flowing through the coil
raises the air temperature and humidity ratio off the coil. The result is a higher space relative humidity
rp at part-load operation than at design load when the space temperature is maintained at a
nearly constant value.
The part-load performance of a water cooling coil, as calculated in Example 15.2, is shown
below:
If the load ratio drops to 0.5, then space relative humidity at part load rp may increase to 65 percent.
Two-Position or Cycling Control. For small constant-volume single-zone systems, such as using
a packaged unit whose supply volume flow rate is 1500 cfm (708 L / s) or less, on/off control of the
refrigeration compressors or the packaged unit is often used to adjust the cooling capacity of the
system within an operating cycle of a few minutes to 20 min. Both the air temperature off the coil
and the space temperature fluctuate. Because of the heat storage capacity of the air system and the
building structure in the conditioned space, the space temperature may fluctuate by 1 to 2°F (0.56 to
1.1°C).
For large packaged units, the fan is continuously operating. Only the refrigeration compressors
are cycling in groups, so that the cooling capacity is reduced at part-load operation. At the same
time, the amplitude of the fluctuation of the space temperature becomes smaller accordingly.
Outdoor Ventilation Air and Exhaust Fans
Because the supply volume flow is constant year-round for a constant-volume single-zone system
only when the filter is loaded with dirt, the fan total pressure increases and the supply volume flow
rate decreases accordingly.
Manually Adjusted Dampers. Many small constant-volume single-zone systems for residential
and retail stores using a rooftop or split packaged unit with a supply volume flow rate of 1500 cfm
(708 L / s) or less employ manually adjusted outdoor air dampers and operate as follows:
 Outdoor damper is field-adjustable to allow preset amount of outdoor air for year-round ventilation
purposes.
Design load Part load
Load ratio 1.0 0.8
Entering air 80°F dry /67°F wet 79°F dry /68°F wet
Leaving air 57.5°F dry / 56.5°F wet 61.3°F dry / 59.5°F wet
Water velocity 4 fps 2.5 fps
Water temperature rise 10°F (5.6°C) 16°F (8.9°C)
Space temperature 75°F (23.9°C) 75°F (23.9°C)
Space relative humidity 55 percent 59 percent
20.58 CHAPTER TWENTY
 Outdoor damper activates with supply fan to preset position and closes when supply fan is shut
off.
 Outdoor damper can be preset to admit up to 50 percent of outdoor air.
Air Economizer. For constant-volume systems with a supply volume flow rate of 2000 cfm
(944 L/ s) or greater, many packaged unit manufacturers offer air economizers to use outdoor air
free cooling for energy savings as well as to improve indoor air quality. An air economizer is
always used as the first stage of cooling before refrigeration is energized. Air economizers are discussed
in Chap. 21.
Exhaust Systems. In most small constant-volume single-zone systems using a PU, part of space air
is exfiltrated through the cracks and gaps on the building shell and the opening of entrance doors because
of the positive space pressure to balance the amount of outdoor air intake at the PU. Sometimes
an additional small exhaust fan is installed on the external window or wall of the restroom.
For a constant-volume system operated in an air economizer cycle that extracts 100 percent outdoor
air for free cooling, a return or relief fan must be used to exhaust the space air outdoors.
20.17 SUPPLY VOLUME FLOW RATE
Supply air volume flow rate and the refrigeration capacity are two primary air system characteristics.
Based on Space Cooling and Heating Load
For most air systems including constant-volume systems that are designed to maintain a healthy
and comfortable space temperature Tr, the supply air volume flow rate is often calculated from the
space cooling load. From Eqs. (20.31) and (20.36), the mass flow rate of the supply air, in lb /min
(g /min), can be calculated as
(20.68)
The supply volume flow rate, in cfm [m3 /(60 s)], can be calculated as
(20.69)
In Eq. (20.69), Tr is the space temperature and Ts the supply temperature at the outlet, both in °F
(°C). For a constant-volume single-zone system operated at cooling mode summer design conditions,
Tr is often the required indoor design temperature in summer, and Ts is the representative
supply temperature of this single zone, as shown in Fig. 20.22. For a multizone air system for comfort
air conditioning, Tr is often the same summer indoor design temperature, and Ts is the supply
temperature of the control zone that has the greatest supply system heat gain. In Eq. (20.69), Qrs
indicates the design sensible cooling load, in Btu / h (W). For a constant volume system Qrs is the
peak cooling load of that specific area which the air system serves because is not reduced at partload
operation. For a VAV system, Qrs is the block cooling load.
If Qrs and Tr and Ts are properly selected, Eq. (20.68) can be applied to both single-zone and
multizone as well as constant-volume and variable-air-volume air systems. The supply density ,
in cfm / ft2 (L/sm2) is defined as
V?
s,d 
V?
s
Afl
V?
s,d
V?
s
V?
s 
Qrc
60 s(hr hs)

Qrs
60 scpa(Tr Ts)
m?a 
Qrc
60(hr hs)

Qrs
60cpa(Tr Ts)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.59
where Afl  floor area of the conditioned space, ft2 (m2). Supply air density is often used in air
volume flow estimates.
The following items should be considered during the calculation of the supply air flow rate from
Eqs. (20.68) and (20.69):
1. Volumetric versus mass flow rate. The supply volume flow rate is widely used in calculations
to determine the size of fans, grilles, outlets, air-handling units, and packaged units. However,
mass flow rate is sometimes simpler to apply in cooling coil load calculations, or it may be more
appropriate when the variation of air density affects the result.
2. Temperature difference versus enthalpy difference. Theoretically, in Eq. (20.69), it is more accurate
to use the enthalpy difference to calculate the supply volume flow rate. This is due to the
elimination of the term cpa, which is actually a variable depending on the temperature and humidity
ratio. In actual practice, it is more convenient to use temperature differences instead of enthalpy difference
because usually the space temperature is controlled and maintained at a preset value. The
variation of cpa is often not significant enough to affect the calculated results.
3. Summer cooling load versus winter heating load. Summer cooling load is usually used to
calculate the supply volume flow rate because it is generally greater than using the winter heating
load. But if the winter weather is very cold, this may require a greater volume flow rate. Based on
winter heating load, the supply volume flow rate can be calculated as
(20.70)
where Qrh  space heating load, Btu /h (W). Use the greater volume flow rate of the two to select
the fans, AHU or PU, and other components.
Based on Requirements Other Than Cooling Load
The supply volume flow rate of many air conditioning systems can also be governed by the following
requirements:
1. To dilute the concentration of the air contaminants. Based on the principle of conservation of
mass, the mass generated rate of the air contaminants in the conditioned space , is given by
And the supply volume flow rate , in cfm (m3 /s), can be calculated as
(20.71)
where Ci, Cs  concentrations of air contaminants of space air and supply air, usually in mg /m3;
1 mg  0.01543 gr
 rate of contaminants generated in space, mg/ s
2. To maintain a required relative humidity r as well as a humidity ratio wr at a specific temperature
for the conditioned space. Then the supply volume flow rate , in cfm [m3 /(60 s)], can be
calculated as
(20.72)
V?
s 
Qrl
60 s(wr ws)hfg,32
V?
s
m?par
V?
s 
2118 m?par
Ci Cs
V?
s
m?par 
V?
s(Ci Cs )
2118
m?par
V?
s 
Qrh
scpa(Ts Tr)
m?a
V?
s
20.60 CHAPTER TWENTY
3. To provide a desirable air velocity vr , in fpm (m / s), within the working area in a clean room.
The supply volume flow rate , in cfm (m3 /s), is then
(20.73)
where Ar  cross-sectional area perpendicular to airflow in working area, ft2 (m2).
4. To exceed the outdoor air requirement for acceptable air quality for occupants. The supply
volume flow rate can then be calculated as
(20.74)
where n  number of occupants
 outdoor air requirement per person, cfm (m3 /s)
5. To exceed the sum of the volume flow rates of the exhaust air and the space air relief
through exfiltration to maintain a positive pressure in the conditioned space, both in cfm (m3 /s).
Then the supply volume flow rate is
(20.75)
The supply volume flow rate should be the one that is the greatest from any of the above requirements,
the space cooling load, or heating load, or outdoor air requirements.
Rated Volume Flow of Supply and Return Fans
The supply volume flow rate is determined based on the mass flow rate of supply air required
to offset the space load at the specific supply condition, with a specific supply air density s. Because
the supply volume flow rate must be provided by the supply fan, or the fan’s rated volumetric
capacity ,r against a certain fan total pressure pt, it is important to determine the relationship
between and .
If the supply fan is located downstream from the cooling coil, the humidity ratios of the moist
air between state points cc and s are constant, as shown in Fig. 20.22. According to the principle of
conservation of mass, the mass flow rate of supply air , in lb / min (kg / min), remains constant,
but the volume flow changes as its state changes. The relationships are
(20.76)
Subscript cc represents the state point of conditioned air leaving the cooling coil, sf represents the
moist air at the outlet of the supply fan, and s represents the supply air at the outlets.
Fans are rated at standard air condition. Standard air means dry air at a temperature of 70°F
(21.1°C) and a barometric pressure of 29.92 in. (760 mm) Hg (14.697 psi) with a density of 0.075
lb / ft3 (1.20 kg /m3). From Eqs. (20.76) and (20.68), the volume flow rate at the supply fan outlet
, in cfm (m3 /s), is calculated as
(20.77)
For the same fan at a given speed, its volume flow rate remains unchanged against a fixed pt even
if the air density varies during rated conditions. Therefore,
(20.78)
where  rated volume flow rate, cfm (m3 /s). From Eq. (2.32), the moist volume of 1 lb of
moist air v is
v 
RaTR(1 	 1.6078w)
pat
V?
sf,r
V?
sf,r  V?sf
V?
sf 
V?
s s
sf

Qrs
60 sfcpa (Tr Ts)
V?
sf
m?a  V?cc cc  V?sf sf  V?s s
m?a
V?
sf,r V?s
V?
sf
m?a V?s
V?
s  V?ex 	 V?exf
V?
exf V?ex
V?
oc
V?
s  nV?oc
V?
s  Arvr
V?
s
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.61
For normal summer and winter comfort applications, the humidity ratio varies less than 0.006 lb / lb
(kg / kg). Therefore, v 	 1.01RaTR /pat, that is, the influence of variation of humidity ratio is smaller
than 1 percent. Therefore, it can be ignored. If there is no infiltration and exfiltration at the conditioned
space, Eq. (20.76) can be rewritten as
(20.79)
Subscript rf represents the outlet of the return fan.
Similarly, the rated volume flow rate of the return fan is given by
(20.80)
For air temperature leaving the cooling coil Tcc at 55°F (12.8°C) with a relative humidity of 92 percent
and Tsf at 57°F (13.9°C), the psychrometric chart shows the moist volume vsf = 13.20 ft3 / lb
(0.824 m3 / kg). From Eq. (20.77), the rated volume flow rate of supply fan is
(20.81)
For cold air distribution, which is covered in detail in Chap. 18, Tcc may equal 40°F (4.4°C) with a
relative humidity of 98 percent; and if Tsf  42°F (5.6°C), then, vsf  12.8 ft3 / lb (0.8 m3 / kg), and
the rated supply volume flow is
(20.82)
For an air system in which the supply fan is upstream from the cooling coil, and if Tsf  82°F
(27.8°C) and sf  43 percent, then, from the psychrometric chart, vsf  13.87 ft3 / lb (0.866
m3 /kg), and the supply volume flow rate is
(20.83)
Thus, an upstream fan requires 9 percent greater fan capacity than a downstream fan with cold air
distribution even though Qrs and Tr Ts are the same.
20.18 DETERMINATION OF THE SUPPLY AIR CONDITION
The following parametrs are usually known before the constant-volume system cooling mode air
conditioning cycle is developed:
1. Summer and winter outdoor design temperatures: dry-bulb To and wet-bulb To
2. Summer and winter indoor space air temperature Tr and relative humidity r
3. Space sensible and latent load Qrs and Qrl
4. Outdoor air requirement
5. Estimated supply system heat gain qs,s
6. Estimated return system heat gain qr,s
If the state point of the supply air can be determined from the psychrometric chart, then the constant-
volume system cooling mode air conditioning cycle can be developed. Eventually, the supply
volume flow rate and the cooling coil load Qcc can also be determined. V?s
V?
o
V?
sf,r  V?sf 
13.87 Qrs
60  0.243(Tr Ts)

Qrs
1.05(Tr Ts)
V?
sf,r  V?sf 
12.8 Qrs
60  0.243(Tr Ts)

Qrs
1.14(Tr Ts)
V?
sf 
13.20 Qrs
60  0.243(Tr Ts)

Qrs
1.105(Tr Ts)
V?
rf,r  V?rf 
Qrs
60 rfcpa(Tr Ts)
V?
rf,r
m?a  V?cc cc  V?sf sf  V?s s  V?r r  V?rf rf
20.62 CHAPTER TWENTY
Air Conditioning Rules
To determine the supply air condition, the following air conditioning rules are helpful:
1. For a given summer indoor design temperature Tr, design space sensible cooling load Qrs, and
specific supply system heat gain qs,s,
 A lower air off-coil temperature Tcc and supply air temperature Ts always result in a greater
supply air temperature differential Ts  Tr Ts and a lower space relative humidity s.
 On the other hand, a higher Tcc means a smaller Ts and a higher r.
2. A greater Ts reduces the volume flow rate and, therefore, fan size, fan energy use, and terminal
and duct sizes. These reduce both first cost and operating costs. At the same time, a greater Ts,
needs a lower Tcc to maintain an indoor design temperature Tr. Excluding the cold air distribution
in the ice storage system, a Ts between 15 and 20°F (8.3 and 11.1°C) is most widely used.
3. Temperature of air leaving the cooling coil Tcc is closely related to the temperature of the chilled
water entering and leaving the coil, Twe and Twl, or to the evaporating temperature inside the DX
coil Tev. A lower Tcc or Twe needs a correspondingly lower Tev and, therefore, higher power input
to the refrigeration compressors.
4. For a given Tcc, Qrs, and qs,s, a supply volume flow rate that is greater than the required value
results in a lower space temperature Tr, a higher space relative humidity r, and a smaller Ts. On
the other hand, a smaller than required results in a higher Tr, a lower r, and a greater Ts.
5. When an air system is used to serve a single-zone conditioned space, Ts and Tcc should be properly
arranged so that the indoor design temperature and relative humidity can be maintained for
energy-efficient operation.
6. For an air system serving multizones, with many space temperatures and relative humidity requirements,
Twe and Tcc or Tev should satisfy the lowest Tcc requirement. In actual practice, Twe
and Tcc, or Tev, are often determined by previous experience with similar applications.
Graphical Method
The following graphical method may be used to determine the condition of supply air s for an air
system serving a single zone (including constant-volume single-zone systems) whose Tr and r
must be maintained.
1. According to the data given in Sec. 20.14, the relative humidity of the air leaving a cooling coil
cc with 10 or more fins per inch can be estimated as
4-row coil: cc  93 percent
6-row coil: cc  96 percent
8-row coil: cc  98 percent
2. Draw the space conditioning line sr from the state point r, parallel to the line of known sensible
heat ratio SHRs, as shown in Fig. 20.24.
3. State point of supply air s must lie on line sr.
4. The horizontal line cc-s represents the magnitude of the supply system heat gain. As segment
cc-s is moved up and down, there exists a position where cc lies exactly on the selected cc
curve for the coil, s lies on line sr, and Ts Tcc exactly matches the temperature increase caused
by supply system heat gain.
5. In practice, this required Tcc must be reconciled with the selected coil’s performance for a given
coil configuration, Twe, and air and water velocities.
V?
s
V?
s
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.63
Influence of Sensible Heat Ratio
For a specific design Tr and r, the sensible heat ratio of the space conditioning line SHRs has a significant
effect on Ts and Tcc. Fig. 20.25 shows such an influence. As SHRs becomes smaller, for the
sake of maintaining the required Tr and r, the temperature of conditioned air leaving the cooling
coil Tcc must be lowered. This requires a lower Twe or a lower evaporating temperature Tev in a DX
coil, which eventually increases the energy input to the refrigeration compressors, but reduces the
air volume flow that must be handled.
For a specific space cooling load Qrc, a lower SHRs always indicates a smaller Qrs and a greater
Ts. A higher Ts means a lower supply volume flow rate , smaller equipment size, lower fan energy
use, and reduced construction and operating costs.
In Fig. 20.25, when SHRs  0.7, the dew point of space air of 75°F (23.9°C), and 50 percent
relative humidity is 55°F (12.8°C), which is higher than the 54°F (12.2°C) supply air temperature.
Under such circumstances, condensation may occur at the supply outlet when its cold metal frame
contacts the space air at dew point 55°F (12.8°C) or higher.
Condensation can be avoided by the following measures:
 Add an insulation layer with vapor barrier, on the surface of ducts and outlets with sufficient
thickness, typically 1 to 2 in. (25 to 50 mm) that are exposed to space air or ambient air whose
dew point is higher than the external temperature of the exposed surface.
 Adopt supply outlet that can induce sufficient space air to mix with low-temperature supply air in
order to raise Ts.
For an air system to serve a single-zone conditioned space (including constant-volume single-zone
systems) with specific indoor Tr and r, say, 75°F (23.9°C) and 50 percent, and a fixed SHRs, there
will be only one supply air condition with a corresponding Ts that can provide the required indoor
Tr and r (see Fig. 20.26). If Tr remains constant at 75°F (23.9°C), an increase of Ts due to the
V?
s
20.64 CHAPTER TWENTY
FIGURE 20.24 Determination of state point of supply air.
20.65
FIGURE 20.25 Effect of sensible heat ratio of space conditioning line on the determination of the condition
of supply air.
FIGURE 20.26 Influence of supply temperature difference Tr Ts on Tcc and r.
reduction of supply volume flow rate at a fixed space sensible load Qrs is possible only when the
temperature of air leaving the cooling coil Tcc and supply temperature Ts are lowered accordingly.
The space relative humidity r also will be lower. This again needs a lower chilled water temperature
entering the coil Twe or a lower evaporating temperature inside the DX coil Tev.
Example 20.4. One workshop in a factory employs a constant-volume single-zone air system.
The summer space sensible cooling load is 350,000 Btu / h (102,550 W), and the summer space latent
load is 62,000 Btu / h (18,166 W). Other design data are as follows:
Summer outdoor design temperatures:
Dry-bulb 95°F (35°C)
Wet-bulb 75°F (23.9°C)
Summer indoor space conditions:
Air temperature 75°F (23.9°C)
Relative humidity 50 percent
Temperature rise:
In supply fan 2°F (1.11°C)
In supply duct 3°F (1.67°C)
Because of heat released from light
fixtures and roof 3°F (1.67°C)
In return duct and return fan 2°F (1.11°C)
Relative humidity of air leaving cooling coil 93 percent
Outdoor air requirement 3200 cfm (1510 L / s)
Determine
1. The condition of supply air at summer design temperatures
2. The rated volume flow rate of supply fan
3. The cooling coil load
Solution
1. From Eq. (20.30) and the given data, the sensible heat ratio of the space conditioning line can
be calculated as
On the psychrometric chart in the upper summer cool mode air conditioning cycle in Fig. 20.22,
draw a space conditioning line sr from point r with SHRs  0.85. From the given data, the total
temperature rise of the supply system heat gain is 2 	 3  5°F. Point cc will lie on cc 93 percent
line, and point s must lie on line sr. Let segment cc-s  5°F. Move cc-s until point cc is on the
93 percent curve and point s is on line sr. The state point of supply air s and the condition of air
leaving the cooling coil, point cc, are then determined. From the psychrometric chart,
Ts  60°F (15.6°C) s  78 percent and ws  0.0086 lb / lb (kg / kg)
Tcc  55°F (12.8°C) cc  93 percent wcc  0.0086 lb / lb (kg / kg)
hcc  22.6 Btu / lb (52.6 kJ / kg)
 0.85
SHRs 
| Qrs |
| qrs | 	 | qrl |

350,000
350,000 	 62,000
V?
s
20.66 CHAPTER TWENTY
2. Because the air temperature at the supply fan outlet Tsf  55 	 2  57°F, and wsf  wcc 
0.0086 lb / lb, from the psychrometric chart, moist volume vsf  13.21 ft3 / lb. From Eqs. (20.77) and
(20.78) the rated supply volume flow rate of the supply fan is
3. State point m can be graphically determined or calculated. The total temperature rise due to
the return system heat gain is 3 	 2  5°F; therefore, Trf  Tru  75 	 5  80°F. Because line
r-ru is a horizontal line, point ru can be plotted on the psychrometric chart. From the given data, the
summer outdoor temperature To  95°F, and the summer design outdoor wet-bulb temperature
To  75°F, so the state point of outdoor air o can also be plotted. Connect line ru-o. Ignore the difference
in density between points ru, m and o, Then
From the psychrometric chart, the length of line section ru-o is 1.8 in.; then point m can be determined
as
Tm  82°F (27.8°C) m  44 percent and hm  30.9 Btu / lb (71.9 kJ / kg)
Mixing point m can also be determined by using the analytical method. From the psychrometric
chart, the enthalpy of outdoor air ho  38.4 Btu / lb and the enthalpy hru  29.8 Btu / lb. From
Eq. (20.55),
The differences between the enthalpies and temperatures determined from the graphical and analytical
methods are small, less than 1 percent. But the analytical method is more accurate because the
line segment lengths are small. Because , from Eq. (20.48b), the cooling coil load is
calculated as
Example 20.5. A factory uses a constant-volume single-zone system without space humidity
control. The operating parameters at winter design conditions are as follows:
 60 
21,140(31.1 22.6)
13.21
 816,154 Btu / h (239,133 W or 239 kW)
Qcc  60V?sf sf (hm hcc)
V?
s s  V?sf sf

3200  95 	 (21,140 3200)80
21,140
 82.3F(27.95C)
Tm 
V?
oTo 	 (V?s V?o)Tru
V?
m
 31.10 Btu/lb (72.3 kJ / kg)

3200  38.4 	 (21,140 3200)29.8
21,140
hm 
V?
oho 	 (V?s V?o)hru
V?
m
ru-m
ru-o

3200
21,140
 0.15

13.21  350,000
60  0.243(75-60)
 21,140 cfm (9976 L /s)
V?
sf,r  V?sf 
Qrs
60 sf cpa (Tr Ts)
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.67
Winter outdoor design conditions
Air temperature 32°F (0°C)
Relative humidity 30 percent
Winter indoor design conditions
Air temperature 72°F (22.2°C)
Space sensible cooling load 800,000 Btu /h (234,400 W)
Space latent load 600,000 Btu /h (175,800 W)
Temperature rise due to supply system heat gain 4°F (2.22°C)
Temperature rise due to return system heat gain 2°F (1.11°C)
Supply volume flow rate 100,000 cfm (47,190 L / s)
Minimum outdoor air requirement 26,250 cfm (12,387 L / s)
At winter design condition, a mixture of outdoor air and recirculating air forms the cold air supply
(cooling mode operation) to offset the space cooling load.
Determine the space relative humidity at winter design conditions and the actual amount of outdoor
air intake.
Solution. Assume that the moist volume of the supply air vs  13.3 ft3 / lb. From Eq. (20.65)
and the given data, the humidity ratio difference between the space and supply air can be calculated
as
From Eq. (20.31) and the given data, the supply air temperature difference Tr Ts can be calculated
as
From the given data
Tru  Tr 	 2  72 	 2  74°F
and Tm  Tr (Tr Ts) 4  72 7.3 4  60.7°F (15.9°C)
Because wru  wr and wm  ws, from Eq. (20.67)
Then
From the psychrometric chart, at a temperature of 32°F and a relative humidity of 30 percent, wo 
0.0012 lb / lb, so
wr  wo 	 0.00394  0.0012 	 0.00394  0.00514 lb / lb
wr wo 
wr ws
0.317
 0.00394 lb / lb (kg / kg)

74 60.7
74 32
 0.317
wr ws
wru wo

Tru Tm
Tru To

800,000  13.3
60  100,000  0.243
 7.30 F (4.1C)
Tr Ts 
Qrs
60V?s scpa

600,000  13.3
60  100,000  1075
 0.00125 lb / lb (kg / kg)
wr ws 
Qrl
60V?s shfg,32
20.68 CHAPTER TWENTY
At a temperature of 72°F (22.2°C) and wr  0.00514 lb / lb (kg / kg), the space relative humidity
r  31 percent, as shown in the lower part of the psychrometric chart in Fig. 20.23. The amount of
outdoor air intake can be calculated as
This amount is greater than the amount of outdoor air required 26,250 cfm (12,387 L / s). At Ts 
72 7.3  64.7°F and ws  0.00514 0.00125  0.00389 lb / lb, from the psychrometric chart,
vs  13.3 ft3/ lb (0.830 m3/ kg), which is equal to the assumed value.
20.19 CONSTANT-VOLUME SINGLE-ZONE SYSTEMS—HEATING
MODE OPERATION
Heating Mode without Space Humidity Control
In many residential and small commercial buildings, the sum of the heat loss due to transmission
through the windows, walls, and roofs and the heat loss due to infiltration in winter is greater than
the sum of the internal heat gains from occupants, electric lights, equipment, and appliances. Then a
warm air supply is required to maintain a desirable indoor temperature. A constant-volume singlezone
system operated in the heating mode without space humidity is often used.
The lower air conditioning cycle in Fig. 20.22 shows such a heating mode cycle in winter. Return
air at state point r flows through the return grilles, return duct, and return fan. After that, recirculating
air at point ru enters the air-handling unit and mixes with the outdoor air. The mixture, at
point m, passes through the cooling coil which is not energized, and is then heated at the heating
coil. Warm air leaves the heating coil at point ch. It leaves the supply fan at point sf after absorbing
the fan power heat gain. The warm supply air loses heat in the supply duct and supplies to the conditioned
space at point s. At the conditioned space, the supply air absorbs the latent heat from the
occupants; supplies heat to compensate for transmission loss through external windows, walls, and
roofs; and finally changes to the point r.
The air conditioning cycle in heating mode operation without space humidity control during
winter design condition (as shown in Fig. 20.22) consists of the following processes:
1. Sensible heating process r-ru due to the return system heat gain qr,s. In winter, the return system
heat gain consists of mainly the heat released from the electric lights qrp and the return fan
power heat gain qrf, if any. That is,
qr,s  qrp 	 qrf (20.84)
The temperature difference between the air inside and outside the return duct is generally small
in winter and can be ignored.
2. Adiabatic mixing process ru-m-o. The ratio of the minimum volume flow rate of outdoor air to
the supply volume flow rate is the same as in the cooling mode in summer design condition in
order to provide the required amount of outdoor air for occupants.
3. Sensible heating process m-ch at the heating coil. The heating coil load can be calculated according
to Eq. (20.39) as
(20.85)
where Tm, Tch  temperature of mixture entering coil and air off heating coil,°F (°C).
4. Sensible heating process ch-sf due to the supply fan power heat gain qsf.
Qch  60 V?s scpa(Tch Tm)
 100,000  0.317  31,700 cfm (14,959 L / s)
V?
o 
V?
s(Tru Trn)
Tru To
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.69
5. Sensible cooling process sf-s due to the heat loss from the warm air inside the supply duct
through the duct wall to the ambient air qsd.
6. Space conditioning process sr. During heating mode operation in winter, heat is supplied to the
space to compensate for the transmission loss, and moisture is absorbed by the supply air due to
the space latent load. Supply air at point s is then changed to r.
The heat supplied to the conditioned space to offset the space heating load Qrh, in Btu / h (W),
can be calculated as
(20.86)
And the latent heat absorbed by the supply air Qrl, in Btu / h (W), is
(20.87)
Example 20.6. For the same constant-volume single-zone system as that in Example 20.4, shown
in Fig. 20.22, the following winter design conditions are given:
Winter outdoor design conditions:
Dry-bulb temperature 15°F (9.4°C)
Relative humidity 30 percent
Winter indoor design temperature 72°F (22.2°C)
Outdoor air requirement 3200 cfm (1510 L / s)
Temperature rise:
Due to heat release from light fixtures 2°F (1.11°C)
Return fan power heat gain 1°F (0.56°C)
Supply fan 2°F (1.11°C)
Temperature drop of the supply duct 1°F (0.56°C)
Space heating load 225,000 Btu /h (65,925 W)
Space latent load 60,000 Btu /h (17,580 W)
Supply volume flow rate 21,140 cfm (9976 L / s)
Determine
1. The space relative humidity at winter design conditions
2. The temperature of supply air
3. The heating coil load
Solution
1. To develop a winter heating mode air conditioning cycle for a constant-volume single-zone
system, the space humidity ratio wr must be determined. Because the winter indoor temperature is
given, the space relative humidity can then be found from the psychrometric chart.
Assume that the moist volume of the warm supply air vs  13.75 ft3 /lb (0.86 m3 /kg). Because
s  1/vs, from Eq. (20.87)
 0.000605 lb / lb

60,000  13.75
60  21,140  1075
wr ws 
Qrl
60V?s s hfg,32
Qrl  60V?s s(wr ws)hfg,32
Qrh  60 V?s scpa(Ts Tr)
20.70 CHAPTER TWENTY
Because wm  ws,
Therefore,
From the psychrometric chart, at To  15°F and o  30 percent, wo  0.0005 lb / lb. Then
wr  wo 	 0.0041  0.0005 	 0.0041  0.0046 lb / lb (kg / kg)
Point r can thus be plotted on the psychrometric chart. The relative humidity of the space air at winter
design condition is 30 percent.
2. From the given data, Tru  Tr 	 2 	 1  72 	 2 	 1  75°F. Because wru  wr, points ru
and o can be plotted from the psychrometric chart, as shown in Fig. 20.22, upper figure. Connect
ru-o. For an adiabatic mixing process
Section ru-o measured from the psychrometric chart is 6.15 in. Then mixing point m can be determined
from the psychrometric chart:
Tm  64.3°F (17.9°C) and wm  0.004 lb / lb (kg / kg)
From Eq. (20.86), the supply air temperature
Because wm  wsf  0.004 lb / lb, point s can be plotted. From the psychrometric chart, vs  13.72
ft3/ lb (0.856 m3 /kg) which is approximately equal to the assumed value. If the assumed value is too
high or too low, the calculation will be repeated with a new value of vs.
3.Warm air temperature off the heating coil at winter heating mode can be calculated by
Tch  Ts 	 1 2  80.0 	 1 2  79.0°F
And from Eq. (20.85), if s  1/ 13.70, the heating coil load is
Heating Mode with Space Humidity Control
In health care buildings with special requirements, or in industrial buildings humidification may
have to be incorporated in the warm air supply in winter. Figure. 20.27 shows an air conditioning
cycle operated in winter heating mode with steam injection humidification and the corresponding
schematic diagram of the constant-volume single-zone system. This cycle consists of the following
processes:
 60  21,140 
1
13.70
(0.243)(79.0 64.3)  330,719 Btu / h (96,901 W)
Qch  60 V?s scpa(Tch Tm)
 70 	
225,000  13.75
60  21,140  0.243
 80.0F (26.7C)
Ts  Tr 	
Qrh
60 V?s scpa
ru-m
ru-o

3200
21,140
 0.15
wr wo 
0.000605
0.15
 0.0041 lb / lb
wr ws
wr wo

3200
21,140
 0.15
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.71
1. Sensible heating process r-ru from return system heat gain
2. Mixing process ru-m-o of recirculated air and outdoor air
3. Sensible heating process m-ch at the heating coil
4. Steam injection or heated element humidifying process ch-h
5. Sensible heating process h-sf from supply fan power heat gain
6. Sensible cooling process sf-s from duct heat loss
7. Space conditioning process sr
The procedure for constructing this cycle is as follows:
1. Plot the state points of the conditioned space r and outdoor air o.
2. Draw the space conditioning line sr from point r with known sensible heat ratio SHRs.
3. From Eq. (20.87), calculate wr ws. Since wr, Qrl, and the supply volume flow rate from summer
cooling mode calculations are known values, ws can be calculated. Point s must lie on line
sr, so the intersection of sr and ws line determines point s.
4. Draw a horizontal line r-ru from point r, representing the return system sensible heat gain. Plot
point ru from its known temperature rise.
5. Connect points ru and o. From the ratio of the volume flow rate of the outdoor air to the supply
air , point m can be found and located on line ru-o.
6. Draw horizontal line m-ch from point m. Draw also a horizontal line sf-s-h through point s. Temperature
difference Tsf Th represents the temperature rise of the supply fan heat gain, and
Tsf Ts is the supply duct heat loss. This defines points sf and h.
V?
o / V?s
20.72 CHAPTER TWENTY
FIGURE 20.27 Air conditioning cycle—winter heating mode operation with space humidity control for a constant-volume single-zone
system.
7. Draw a vertical line through ch to line h-sf extended from point h. The intersection of lines h-ch
and m-ch locates point ch. Because wm  wch, the mass flow rate of the injected steam is the
mass flow rate of the water vapor , in lb / min (kg / s), added to the moist air, and can be calculated
from Eq. (20.42)
(20.88)
where wh  humidity ratio of air leaving humidifier, lb/ lb (kg / kg)
wm  humidity ratio of mixture, lb / lb (kg / kg)
Part-Load Operation
For a constant-volume single-zone system, when the space heating load is reduced during part-load
operation, as in the the cooling mode operation, hot water flow rate modulation is used to reduce
the supply air temperature and the heating capacity if a hot water heating coil is employed in an airhandling
unit. For a packaged unit using a gas-fired furnace as a heating source, a DDC unit controller
will directly adjust the amount of gas supplied to the burner when the temperature senser
senses the rise of space temperature during part-load operation.
Dual-Thermostat Year-Round Zone Temperature Control
For most of the single-zone constant-volume systems employed in residential buildings and small
retail stores, two separate thermostats with one set point for cooling mode operation and another set
point for heating mode operation are often used, as shown in Fig. 20.28. During cooling mode operation,
when the zone temperature drops below the cooling set point, the refrigeration system is shut
down; or during heating mode operation, when the zone temperature rises above the heating set
point, the gas furnace is shut off. On /off control is most often used.
Between the cooling and heating mode operation, there is a dead band in which neither cooling
nor heating is provided. During the dead band mode, the constant-volume single-zone system is
m?s  V?s s(wh wm)
m?s
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.73
Heating
mode
Cooling
mode
Dead
band
70
0
76
Tr, F
Dead
band
Output 100%
FIGURE 20.28 Dual-thermostat year-round zone temperature control.
often shut down, and sometimes the fan remains continuously operating to provide necessary outdoor
ventilation air for interior spaces. The on /off control of the cooling or heating mode operation
for such a constant-volume single-zone system is performed manually or automatically. It is possible
to have a dead band width of 5°F (2.8°C) or even wider for such a dual dual-thermostat yearround
zone-temperature control system.
If the cooling capacity of a single-zone constant-volume system is greater than 5 tons
(17.6 kW), multiple single-zone constant-volume packaged units may be used.
20.20 CONSTANT-VOLUME MULTIZONE SYSTEM WITH REHEAT
Reheating, Recooling, and Mixing
In zone controls for multizone systems, reheating is a process in which air is reheated after it has
been cooled. Recooling is a process in which air is cooled after it has been previously heated by
using combustion fuels, and in the mixing process a cooler airstream is often mixed with warmer
recirculating air, or a warmer airstream mixed with cooler outdoor air. Reheating, recooling, and
mixing processes are all simultaneous cooling and heating processes in air conditioning cycles, and
are energy-inefficient air conditioning processes.
ASHRAE/IESNA Standard 90.1-1999 specifies that zone thermostatic controls (either in constant-
volume or VAV systems) shall be capable of operating in sequence the supply of heating and
cooling energy to the zone. Zone controls shall prevent (1) reheating, (2) recooling, (3) mixing or
simultaneously supplying air that has been previously mechanically heated with air that has been
previously cooled by refrigeration or by an economizer, and (4) other simultaneous operation of
heating and cooling systems to the same zone except the following (exceptions for VAV systems
will be listed and discussed in Chap. 21):
a. Zones in which the volume flow of reheated, recooled, or mixed air is no greater than the larger
of the following:
 The volume flow of outside air to meet the ventilation requirement of ASHRAE Standard
62-1999.
 0.4 cfm / ft2 (2L/sm2) of the zone conditioned floor area provided that primary air temperature
is controlled:
0 to 12°F (0 to 6.7°C) below the design space heating temperature Tr,h when outdoor air temperature
To  60°F (15.6°C) for reheat systems and the cold deck mixing systems.
0 to 12°F (0 to 6.7°C) above design space temperature Tr when To  60°F (15.6°C) for recooling
system and the hot deck mixing systems.
Each zone need not comply with this exception provided the average of all zones served by
this system that have both heating and cooling ability comply.
 300 cfm (142 L / s). This exception is for zones whose peak flow rate totals no more than 10
percent of the total fan system (air system) volume flow rate.
 Any higher rate that is satisfied by the authority having jurisdication to reduce overall air
system annual energy usage by offsetting reheat / recool energy losses through a reduction in
outdoor air intake in accordance with the multiple space requirements in ASHRAE Standard
62-1999.
b. Zones where special pressurization relationships, cross-contamination requirements, or
code-required minimum circulation rates are such that variable-air-volume systems are impractical.
c. Zones where at least 75 percent of the energy for reheating or for warm air mixing systems is
provided from a site recovered (including condenser heat) or site solar energy source.
20.74 CHAPTER TWENTY
Reheating is a simple and effective means of controlling space temperature and relative humidity at
part load. Reheating is used with compliance to ASHRAE/IESNA Standard 90.1-1999 for VAV
systems at minimum volume flow rate settings for comfort air conditioning, and in hospitals, laboratories,
computer rooms, museums, and precision manufacturing environments (such as semiconductor
manufacturing plants) in order to maintain a close tolerance in space temperature (1°F or
0.56°C) or space relative humidity for processing air conditioning.
Constant-Volume Multizone System with Reheat
A constant-volume, multizone system with reheat is shown in Fig. 20.29. During cooling mode
operation, as shown in Fig. 20.29, outdoor air at point o is mixed with recirculating air at point ru.
The mixture m flows through the filter. It is then cooled and dehumidified to point cc at the cooling
coil and discharged from the AHU or PU at the supply fan outlet sf. After absorbing the duct heat
gain, the conditioned air enters the reheating coils 1 and 2 at point s.
During summer design load operation, theoretically, reheating coil 1 is not energized. Air is
supplied to room 1 at condition s1. Conditioned air is reheated at reheating coil 2 to maintain the
required room temperature in room 2. After absorbing the space cooling load, the supplied air becomes
room air r1 and r2. Returned air from rooms 1 and 2 is mixed together and forms mixture rm
in the ceiling plenum. The return air is then returned to the AHU or PU, and the recirculating air
enters the AHU or PU at point ru. A portion of return air is exhausted through the relief fan to the
outdoors.
During winter operation, outdoor air is mixed with recirculating air in a ratio such that its
amount is always greater than the minimum outdoor air requirement. The mixture m flows through
the filter, the deenergized cooling coil, and the supply duct and is reheated at the reheating coils 1
and 2. The heated air is supplied to rooms 1 and 2 to offset the heating load there. Both hot water
and electric reheating coils can be used in multizone systems. When electric reheating coils are
used, the design and installation requirements described in Sec. 8.4 and the requirements of the
National Electric Code and local codes must be followed.
Control Systems
As described in Sec. 5.14, the controls for a constant-volume multizone system with reheat using an
AHU or a PU may include the following:
 Discharge air temperature control
 Outdoor ventilation air control
 Space pressurization control
 Air economizer control
 Zone smoke control
All these functional controls for an air system are discussed in Chap. 23 except zone smoke control,
discussed in Chap. 22.
If there are only a few reheating coils, both the AHU or PU and the reheating coils can be controlled
by a DDC unit controller. If there are many reheating coils, the AHU or PU and the reheating
coils should be controlled by a DDC system controller.
When room temperature sensors T2 and T3 sense a drop in either room temperature below a
limit, a signal is sent to the DDC controller. The reheating coil is then energized. For a hot water reheating
coil, the two-way valve is modulated so that room temperature is maintained within predetermined
limits. For an electric reheating coil, heating is added or decreased in steps. To prevent
fires, electric heating is energized only when airflow is sensed by a flow switch.
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.75
20.76
FIGURE 20.29 Constant-volume multizone system with reheat: (a) schematic diagram; (b) air conditioning cycle.
Operating Parameters and Calculations
1. The zone supply air condition should be determined so that the space temperature and relative
humidity in each zone or room can be maintained at design load operation by using the reheating
processes.
2. The zone n supply mass flow rate , in lb / min [kg / (60 s)], and volume flow rate , in cfm
[kg / (60 s)], at summer design condition can be calculated as
(20.89)
(20.90)
where Qrsn  sensible cooling load for zone n, Btu /h (W)
Trn  temperature of zone n,°F (°C)
Ts  supply temperature of zone n,°F (°C)
vs  1/ s, moist volume of supply air, ft3 / lb (m3 / kg)
3. The zone n supply temperature at minimum part-load operation Tsnp, in°F (°C), can be calculated
as
(20.91)
where Trnp  temperature of zone at minimum part-load, operation °F (°C).
4. The zone n reheating coil load Qhn, in Btu / h (W), at minimum part-load operation can be calculated
as
(20.92)
System supply temperature Ts varies along the supply main duct. For simplicity, Ts can be taken as
an average value.
5. The weighted mean value of zone temperature Trm, in °F (°C), and mean humidity ratio wrm,
in lb / lb (kg / kg), can be calculated as
(20.93)
and (20.94)
where  mass flow rate of supply air for zones 1, 2, . . . , n, lb/ min (kg / s)
Tr1, Tr2, . . ., Trn  air temperatures in zones 1, 2, . . . , n,°F (°C)
wr1, wr2, . . ., wrn  air humidity ratio in zones 1, 2, . . . , n, lb/ lb (kg / kg)
6. The system supply volume flow rate , in cfm (m3 /s), of the AHU or PU is calculated as
(20.95)
where supply volume flow rate for zones 1, 2, . . . , n, cfm (m3 /s).
7. The condition of recirculating air entering the AHU or PU, point ru, and the mixture of outdoor
and recirculating air m can be determined graphically on the psychrometric chart.The cooling
coil load Qcc, in Btu / h (W), can therefore be calculated as
(20.96) Qcc  60(m?a1 	 m?a2 	    	 m?an)(hm hcc)
V?
s1, V?s2, . . . , V?sn 
V?
s  V?s1 	 V?s2 	    	 V?sn
V?
s
m?a1, m?a2,   , m?an
wrm 
m?a1wr1 	 m?a2wr2 	    	 m?anwrn
m?a1 	 m?a2 	    	 m?an
Trm 
m?a1Tr1 	 m?a2Tr2 	    	 m?anTrn
m?a1 	 m?a2 	    	 m?an
Qhnp  60m?ancpa(Tsnp Ts)
Tsnp  Trnp 
Qrsn
60m?ancpa
V?
sn  vsm?an
m?an 
Qrsn
60cpa(Trn Ts)
V?
sn m?an
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.77
where hm, hcc  enthalpy of mixture and conditioned air off coil, Btu/ h (W)
 mass flow rate of mixture, cfm [kg / (60 s)]
8. At part-load operation, space cooling load reduces and more reheating is required. The condition
of zone supply air point snp can be determined because snp must lie on horizontal line ccsnp.
Zone supply air temperature at part load Tsnp can be calculated from Eq. (20.91). Draw a line from
snp with the known sensible heat ratio for zone n at part load SHRnp. This line intersects zone temperature
line Trnp at point rnp, which is the state point of zone air at part load. Zone relative humidity
rnp can then be determined.
The operating parameters during winter heating mode operation at design load can be determined
as follows:
 At winter design load, wrm  wru and ws  wm. Here wru indicates the humidity ratio of the recirculating
air, ws is the humidity ratio of system supply air before entering the reheating coil, and
wm is the humidity ratio of the mixture of the outdoor air and recirculating air, all in lb / lb
(kg / kg). Then
(20.97)
Also, (20.98)
where wo  humidity ratio, lb / lb (kg / kg);  volume flow rate of outdoor air, cfm [kg / (60 s)].
 Zone supply temperature Tsn, in °F (°C), can be calculated as
(20.99)
where Qrhn  space heating load for zone n, Btu /h (W).
 The reheating coil load for zone n during winter heating mode operation Qhn, in Btu / h (W), is
(20.100)
where Ts  supply air temperature before reheating coil, °F (°C). Usually, the reheating coil load
during winter heating mode operation is greater than in summer cooling mode minimum partload
operation, and should be taken as the design capacity of the reheating coil. However, it is
necessary to select the greater value of these two as the reheating coil design capacity.
 At part-load operation, space heating load reduces, and the reheating coil load decreases accordingly.
System Characteristics
If the humidity ratio of the supply temperature Ts remains the same during summer cooling mode or
winter heating mode operation, the space relative humidity of a constant-volume multizone system
with reheat at part load is always lower than that at design load.
There is also another constant-volume multizone system that has a hot deck and a cold deck. For
each control zone, a separate air duct is connected to the hot deck, and another separate air duct is
connected to the cold deck; each has a damper. Constant-volume multizone systems waste energy
and also result in a very complicated control system. They are now replaced by other more energyefficient
systems.
Qhn  60m?ancpa(Tsn Ts)
Tsn  Trn 	
Qrhn
60V?sn scpa
V?
o
wru wm
wru wo

V?
o
V?
s
wru  ws 	
Qrl
60(m?a1 	 m?a2 	    	 m?an)hfg,32
m?a
20.78 CHAPTER TWENTY
REFERENCES
AMCA, Fan and Systems, Application Guide 201–90, Air Movement and Control Association, Arlington
Heights, IL., 1990.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, 1997.
ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except New Low-Rise Residential Buildings,
ASHRAE Inc., Atlanta, 1999.
Avery, G., No More Reheat: Banking Center Retrofit, Heating / Piping / Air Conditioning, no. 3, 1984,
pp. 98–99.
Carrier Corporation, Handbook of Air Conditioning System Design, McGraw-Hill, New York, 1965.
Coad,W. J., The Air System in Perspective, Heating / Piping / Air Conditioning, no. 10, 1989, pp. 124–125.
Coogan, J. J., Effects of Surrounding Spaces on Rooms Pressurized by Differential Flow Control, ASHRAE
Transactions, 1996, Part I, pp. 18–25.
Coward, C. W., Jr., A Summary of Pressure Loss Values for Various Fan Inlet and Outlet Duct Fittings, ASHRAE
Transactions, 1983, Part I B, pp. 781–789.
Cummings, J. B.,Withers, C. R., Jr., Moyer, N. A., Fairey, P. W., and McKendry, B. B., Field Measurement of
Uncontrolled Airflow and Depressurization in Restaurants, ASHRAE Transactions, 1996, Part I, pp. 859–869.
Desmone, C. L., and Frank, P. L., Air Conditioning for Precision Manufacturing, Heating / Piping / Air Conditioning,
no. 2, 1992, pp. 35–44.
Driscoll, D. J., System Effect–The Balancer’s Dilemma? ASHRAE Transactions, 1983, Part I B, pp. 795–801.
Energy Information Administrations, Commercial Buildings Characteristics, EIA,Washington, 1994.
Holness, G. R., Pressurization Control: Facts and Fallacies, Heating / Piping / Air Conditioning, no. 2, 1989,
pp. 47–51.
Lovatt, J. E., and Wilson, A. G., Stack Effect in Tall Buildings, ASHRAE Transactions, 1994, Part II,
pp. 420–431.
Manley, D. L., Bowlen, K. L., and Cohen, B. M., Evaluation of Gas-Fired Desiccant-Based Space Conditioning
for Supermarkets, ASHRAE Transactions, 1985, Part I B, pp. 447–456.
Meckler, G., Efficient Integration of Desiccant Cooling in Commercial HAVC Systems, ASHRAE Transactions,
1988, Part II, pp. 2033–2042.
O’Connor, J. E., The System Effect and How It Changes Fan Performance, ASHRAE Transactions, 1983,
Part I B, pp. 771–775.
Persily, A. K., Myths about Building Envelopes, ASHRAE Journal, no. 3, 1999, pp. 39–47.
Persily, A. K., and Linteris, G. T., A Comparison of Measured and Predicted Infiltration Rates, ASHRAE Transactions,
1983, Part II B, pp. 183–200.
Shaw, C. Y., Reardon, J. T., and Cheung, M. S., Changes in Air Leakage Levels of Six Canadian Office Buildings,
ASHRAE Journal, no. 2, 1993, pp. 34–36.
Tamblyn, R. T., Coping with Air Pressure Problems in Tall Buildings, ASHRAE Transactions, 1991, Part I,
pp. 824–827.
Tamura, G. T., and Wilson, A. G., Pressure Differences Caused by Chimney Effect in Three Tall Buildings,
ASHRAE Transactions, vol. 73, no. 2, 1967.
Tamura, G. T., and Shaw, C. Y., Studies on Exterior Wall Air Tightness and Air Infiltration of Tall Buildings,
ASHRAE Transactions, 1976, Part I, pp. 122–134.
The Trane Company, Building Pressurization Control, The Trane Company, La Crosse, WI, 1982.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechnic, Hong Kong, 1987.
Withers, C. R., Cummings, J. B., Moyer, N. A., Fairey, P. W., and McKendry, B. B., Energy Savings from
Repair of Uncontrolled Airflow in 18 Small Commercial Buildings, ASHRAE Transactions, 1996, Part II,
pp. 549–561.
AIR SYSTEMS: BASICS AND CONSTANT-VOLUME SYSTEMS 20.79
CHAPTER 21
AIR SYSTEMS:
VARIABLE-AIR-VOLUME
SYSTEMS
21.1
Stability of Zone Control Using VAV
Boxes and Reheating VAV Boxes 21.26
Case Study: A VAV Reheat System 21.27
Conditioned Air Off-Coil and Supply
Temperature Differential 21.28
System Volume Flow Rate and Coil
Load 21.28
Summer Cooling Mode Part-Load
Operation 21.29
Winter Reheating in Perimeter Zone 21.29
Winter Cooling Mode Operation in
Interior Zone 21.32
21.4 DUAL-DUCT VAV SYSTEMS 21.33
System Description 21.33
Number of Supply Fans 21.36
Mixing VAV Box 21.36
Mixing Mode Operation 21.38
Zone Controls and Sequence of
Operations of Dual-Fan Dual-Duct
VAV System 21.38
Discharge Air Temperature
Control 21.40
Zone Supply Volume Flow Rate 21.41
Case Study: A Dual-Fan Dual-Duct VAV
System 21.42
Winter Heating and Winter Cooling Mode
Operation 21.43
Part-Load Operation 21.43
21.5 FAN-POWERED VAV SYSTEMS 21.44
System Description 21.44
Fan-Powered VAV Boxes 21.48
Fan Characteristics in Parallel Fan-
Powered Boxes 21.50
Zone Control and Sequence of Operations
of a Fan-Powered VAV System with
Parallel Fan-Powered Box 21.52
Supply Volume Flow Rate 21.53
Fan Energy Use 21.54
Design Considerations 21.55
21.6 COMPARISON BETWEEN VARIOUS
VAV SYSTEMS 21.56
REFERENCES 21.56
21.1 SYSTEM CHARACTERISTICS OF
VARIABLE-AIR-VOLUME SYSTEMS 21.2
Variable-Air-Volume Systems 21.2
Types of Vairable-Air-Volume
Systems 21.2
21.2 SINGLE-ZONE VAV SYSTEMS 21.3
System Description 21.3
Year-Round Operation of a Single-Zone
VAV System 21.5
Economizer Cycle and Economizers 21.8
Differential Enthalpy, Electronic
Enthalpy, and Fixed Enthalpy Economizer
Control 21.8
Fixed Dry-Bulb and Differential Dry-Bulb
Economizer Control 21.9
Comparison of Enthalpy-Based and
Temperature-Based Economizer
Control 21.10
Water Economizer and Control 21.12
Comparison of Air and Water
Economizers 21.14
ASHRAE/IESNA Standard 90.1-1999
Economizer Control
Specifications 21.14
Air Conditioning Cycle and System
Calculations 21.16
Zone Temperature Control—Sequence
of Operations 21.17
21.3 VAV COOLING, VAV REHEAT, AND
PERIMETER HEATING VAV
SYSTEMS 21.18
VAV Cooling Systems 21.19
VAV Reheat Systems 21.20
System Description 21.20
Perimeter Heating VAV System 21.20
VAV Box 21.21
Reheating VAV Box 21.23
Sound Power Level of a
VAV Box 21.23
VAV Reheat Zone Temperature
Control and Sequence of
Operations 21.25
21.1 SYSTEM CHARACTERISTICS OF VARIABLE-AIR-VOLUME
SYSTEMS
Variable-Air-Volume Systems
A variable-air-volume (VAV) system is an air system that varies its supply air volume flow rate to
match the reduction of space load during part-load operation to maintain a predetermined space
parameter, usually air temperature, and to conserve fan power at reduced volume flow. A constantvolume
system varies its supply air temperature to match the reduction of space load during partload
operation to maintain a predetermined space air temperature.
Compared with a constant-volume system, a VAV system has mainly the following advantages:
 Reduced fan energy use during part-load operation when the supply volume flow rate is reduced
 A slightly lower or nearly the same zone relative humidity when the supply volume flow rate is
reduced during summer cooling mode part-load operation
 More individual control zones
 Reduction of the construction cost because of taking into consideration of the supply air volume
flow diversity factor instead of the sum of zone peak loads
 Capability of self-balancing of zone supply volume flow rates
 Convenience during the relocation of the terminals and space diffusion devices during future
expansion or retrofit
Compared with a constant-volume system, the primary disadvantages of a VAV system are:
 Inadequate outdoor ventilation air when the supply volume flow rate is reduced
 A more complicated system structure and controls, which need more demanding design, installation,
operation, and maintenance
VAV systems are applicable to air systems whose space load varies significantly so that there are
fan energy savings. VAV systems became popular after the energy crisis in 1973. They are widely
used in large commercial buildings in the United States.
Types of Variable-Air-Volume Systems
Most medium-size and large buildings need multizone air systems. However, many indoor stadiums,
convention centers, factories, residential buildings, and small retail stores employ single-zone
air systems. Currently used variable-air-volume systems can be mainly classified into the following
types:
 Single-zone VAV systems
 VAV cooling systems
 VAV reheat systems
 Dual-duct VAV systems
 Fan-powered VAV systems
Recently, another VAV system has been developed called the variable diffuser VAV system. In
variable diffuser systems, the aperture of each diffuser can be varied so that the discharge velocity
is relatively constant while the supply volume flow is reduced, and the throw from the variable
diffuser may also be maintained above a certain limit. Variable diffuser VAV systems provide more
individual control zones as well as a more complicated diffuser construction and controls. More
field perfomance data and experience are needed to make an appropriate selection.
21.2 CHAPTER TWENTY-ONE
Conditioned spaces in buildings are primarily divided into two categories: (1) perimeter zone,
where there are external wall, roofs, and windows and space load varies depending on solar heat,
outdoor-indoor temperature difference and internal load; and (2) interior zone where space loads
are mainly internal loads. In an interior zone, the air system often operates at cooling mode in both
summer and winter except for morning warm-up. A perimeter zone can often be subdivided into
four subzones according to their orientations. Although all these subzones operate at cooling mode
in summer, in fall and winter, zones facing south in the northern hemisphere may need cooling,
whereas zones facing north may need heating. Under such circumstances, a VAV system that serves
a perimeter zone is preferable to having a modulating cooling device and a modulating heating
device to meet the variable cooling and heating requirements for each of its subzones facing different
orientations.
The system characteristics of different types of VAV systems vary mainly because of the zone
control of their different types of VAV boxes to serve perimeter zone, except a single-zone VAV system
which directly varies the fan speed or the inlet vanes and a dual-duct VAV system in which two
supply ducts (a cold air duct and a warm air duct) are used. A VAV cooling system with VAV boxes
usually serves an interior zone, and reheating VAV boxes or a VAV cooling system plus a perimeter
heating system (such as a hot water heating system) will serve a perimeter zone. A single-zone VAV
system can be used for either a perimeter zone or an interior zone. Excluding the terminals and
zone controls, the following system characteristics of different VAV systems are similar to each
other, and are discussed in later chapters.
 Fan combination
 System pressure analysis and space pressurization
 Smoke control
 Outdoor ventilation air
 Controls, including discharge temperature, duct static pressure, and safety controls
ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except Low-Rise Residential
Buildings, specifies that individual variable-air-volume (VAV) fans with 30-hp (22-kW) motors
and larger shall have other controls and devices (such as adjustable-frequency, variable-speed
drives) that will result in fan motor demand of no more than 30 percent of design wattage at 50 percent
of design volume flow rate when fan static pressure equals one-third of the total design static
pressure based on manufacturer’s certified data.
21.2 SINGLE-ZONE VAV SYSTEMS
A single-zone VAV system is a VAV system that varies the air volume flow rates supplied to and
returned from a single-zone conditioned space to maintain a predetermined space parameter at
reduced load and to conserve fan power. Single-zone systems are widely used in indoor stadiums,
assembly halls, shopping malls, and factories.
System Description
Figure 21.1 shows the schematic diagram and the air conditioning cycles for a typical single-zone
VAV system. During the summer cooling mode operation, the mixture of outdoor air and recirculating
air at point m is filtered at the filter and cooled and dehumidified at the cooling coil. The conditioned
air leaves the coil at point cc. It then flows through the supply fan of the air-handling unit
(AHU) or packaged unit (PU) and the supply duct before it is discharged to the single-zone conditioned
space at point s. After absorbing the space cooling load, the supply air becomes the space air
r and returns to the AHU or PU through the return air system.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.3
21.4
FIGURE 21.1 A single-zone VAV system: (a) schematic diagram; (b) air conditioning cycle.
When air economizer cycles are used for free cooling during spring and fall seasons, outdoor air intake
may vary from minimal to 100 percent. A relief fan should be installed in combination with the supply
fan, if the pressure drop of the return duct system is small. If the pressure drop of the return duct system
is greater than 0.6 in. WC (149 Pa), a return fan in series with a supply fan should be considered.
During summer cooling mode part-load operation, the supply and return volume flow rates are
reduced to match the lower space cooling load. The temperature rise from the fan power heat gain
is reduced accordingly. At the same time, duct heat gain remains approximately the same, which
results in a higher temperature rise for supply and return ducts during summer cooling mode partload
operation.
During winter, if the solar heat and internal load is greater than the space transmission and infiltration
loss, then a cooling mode operation of a cold air supply is still needed, such as in fully occupied
indoor stadium and arenas. Outdoor air is often mixed with the recirculating air to form a
mixture at a temperature around 55°F (12.8°C) during the air economizer cycle for free cooling.
Warm air may be needed only during the morning warm-up period to raise the space temperature to
an acceptable value before the space is to be occupied.
If solar heat and the internal load are less than the transmission and infiltration loss, the mixture of
outdoor and recirculating air is heated at the heating coil to point hc. It is extracted by the supply fan and
forced through the supply duct. Warm air is then supplied to the conditioned space at point s, and the recirculating
air is returned to the AHU or PU through the ceiling plenum and connecting return ducts.
For comfort air conditioning without humidification in winter, the space relative humidity
depends mainly on the space latent load and the volume flow ratio of outdoor air to supply air.
For many VAV systems in commercial buildings, the ceiling plenum is often used as the return
plenum. The primary sound transmission path of supply fan noise from the AHU or PU to the space is
often duct-borne noise through the return system. A sound trap or silencer should be added in the fan
room between the AHU or PU and the connecting duct before penetrating the fan room partition wall.
Year-Round Operation of a Single-Zone VAV System
In this single-zone VAV system, cooling is provided by a refrigerant plant and evaporative coolers,
and heating is provided by the hot water boiler through a hot water heating coil. For convenience in
analysis, consider an interior zone with zone air temperature maintained at 75°F (23.9°C) yearround.
During cooling mode operation, the required off-coil temperature is 55°F (12.8°C) with a
relative humidity of 95 percent. The recirculating air enters the mixing box at 78°F (25.5°C). It is
mixed with an amount of outdoor air approximately equal to 20 percent of the total design supply
volume flow rate and is cooled and dehumidified in a cooling coil. During winter heating mode
operation, cold air supply is still required to maintain a year-round space temperature of 75°F
(23.9°C); outdoor air and recirculating air are mixed together to provide a mixture of 55°F
(12.8°C). It is also assumed that the building is in a location where no humidification is required to
maintain a space relative humidity no lower than 30 percent during the winter season.
To achieve effective and energy-efficient operation when the outside weather changes from hot
summer to cold winter, the year-round operation of a single-zone VAV air system described above
can be divided into the following operating modes, corresponding to the various outside weather
regions on the psychrometric chart, as shown in Fig. 21.2:
Region I: Refrigeration and Evaporative Cooling. The lower left boundary line of region I is the
(dashed) enthalpy line that passes through the state point ru of recirculating air entering the mixing
box. When the outside weather at point o is inside region I, the enthalpy of the outside air is higher
than that of the recirculating air, that is, ho  hru. In these instances, it is more energy-efficient to
condition (cool and dehumidify) the recirculating air than the outside air. As such, the quantity
of outdoor air mixed with the recirculating air will be the minimum needed to meet the required
outdoor ventilation rate in the conditioned space.
In region I, refrigeration or evaporative cooling or both are required to cool and dehumidify or
to cool and humidify the mixture of outdoor and recirculating air to point cc in order to maintain
the predetermined space condition r.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.5
21.6 CHAPTER TWENTY-ONE
FIGURE 21.2 Enthalpy economizer control: (a) outside weather regions; (b) enthalpy economizer.
For region I, the operating characteristics of the VAV system include the following:
 The outdoor damper is opened to such a degree that only that minimum required amount of
outdoor air is extracted.
 The recirculating damper is fully opened.
 Supply air is cooled and dehumidified, or cooled and humidified to point cc.
 A refrigeration system, evaporative coolers, or both are operating.
Region II: Free Cooling and Refrigeration / Evaporative Cooling. Region II is enclosed by four
boundary lines: the enthalpy line through point ru at the upper right, part of the saturation curve, the
55°F (12.8°C) temperature line, and the cc-ru line.
In region II, the enthalpy of outdoor air is lower than that of recirculating air, that is, ho  hru.
Less energy will be used in conditioning if 100 percent of outside air is cooled and dehumidified
instead of a mixture outdoor air and recirculating air. Both free cooling by the outside air and refrigeration
/ indirect evaporative cooling (to be discussed in Chap. 27) are required to bring the 100
percent outside air down to 55°F (12.8°C) with a relative humidity of 95 percent.
The operating characteristics of this VAV system when outdoor air is inside region II include the
following:
 The outdoor damper and the exhaust damper or relief damper both are fully opened.
 The recirculating damper is closed.
 Outdoor air is cooled and dehumidified to point cc. Point cc is at a temperature of 55°F (12.8°C)
and a humidity ratio of 0.0088 lb / lb (kg / kg).
 Refrigeration system / and indirect evaporative cooler are operating.
Region III: Free Cooling and Evaporative Cooling / Refrigeration. Region III is bounded by the
line cc-ru, 55°F (12.8°C) temperature lines, and enthalpy lines passing through points ru and cc for
wcc  0.005 lb / lb (kg / kg), as shown in Fig. 21.2. For outdoor air inside this region, ho  hru, the
humidity ratio of outdoor air wo is approximately less than 0.009 lb / lb (kg / kg). Indirect and direct
evaporative cooling plus necessary refrigeration is required to maintain a space temperature of 75°F
(23.8°C) and a space relative humidity greater than or equal to 30 percent.
When outdoor temperature To  78°F (25.5°C), it is energy-efficient to extract 100 percent
outdoor air (to make use of free cooling) and to use direct and indirect evaporative cooling plus
necessary refrigeration to cool and humidify the outdoor air to a temperature of 55°F (12.8°C) with
a humidity ratio wcc  0.005 lb / lb (kg / kg). When To 	 78°F (25.5°C), it is energy-efficient to
extract minimum outdoor air and mix it with recirculating air. The mixture is then evaporatively
cooled indirectly-directly with neccessary refrigeration to Tcc  55°F (12.8°C) and wcc 
0.005 lb / lb (kg / kg).
The operating characteristics of the single-zone VAV system when outdoor air is inside region
III and when 100 percent outdoor air is extracted are as follows:
 The outdoor air damper and the exhaust or relief air damper both are fully opened.
 The recirculating damper is closed.
 Evaporative coolers are used for evaporative cooling and humidifying.
 Necessary refrigeration is required.
When only a minimum of outdoor air is mixed with recirculating air, then
 The outdoor air damper is opened to minimum.
 The recirculating damper is fully opened.
 Evaporative coolers are used for evaporative cooling and humidifying.
 Sometimes refrigeration may be required.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.7
Region IV: Free Cooling, Humidification, and Heating. Region IV is enclosed by part of the saturation
curve, the 55°F (12.8°C) temperature line, and the enthalpy line passing through point cc for
wcc  0.005 lb / lb (kg / kg), as shown in Fig. 21.2. For outdoor air state points in region IV, either To 
55°F (12.8°C) and ho 	 hcc or To  78°F (25.5°C) and ho  hcc, when wcc  0.005 lb / lb (kg / kg).
In modulation of outdoor and recirculating dampers, or in locations with very cold winter, heating
may be required to provide a mixed air temperature of 55°F (12.8°C) when the volume flow of
outdoor air is equal to the minimum requirement. Free cooling of outdoor air will be utilized. Optimum
humidifying methods should be adopted to maintain a space relative humidity of preferably
not less than 25 percent.
In locations where the outdoor temperature may drop below 32°F (0°C) in winter, remedies
should be provided to protect water from freezing.
The operating characteristics of this VAV system include the following:
 Outdoor, exhaust, and recirculating air dampers are modulated.
 Heating, humidification, or cooling and humidification may be required.
 The refrigeration system is not operating.
Year-round operation of a multizone VAV system having perimeter and interior zones, is similar to
that for an interior zone VAV system with the exception that heating may be required in regions II,
III, and IV. For a multizone VAV system, the space air temperature can be considered as the average
value of multiple zones.
Economizer Cycle and Economizers
An economizer cycle is an air conditioning cycle that utilizes the free cooling capacity of outdoor
air either directly or to cool the condenser water in a cooling tower (or an evaporative cooler), and
then to cool the air indirectly, instead of using refrigeration to provide cooling / dehumidification so
as to maintain a required space temperature. The component and devices used in the operation of an
economizer cycle are collectively called an economizer, and the type of control used to operate the
economizer cycle effectively and energy-efficiently is called economizer control.
There are two types of economizers:
1. Air or air-side economizers. These bring in up to 100 percent outdoor air to utilize its free
cooling capacity, rather than use refrigeration or evaporative cooling, to offset the space cooling
load when the enthalpy of the outdoor air is lower than the enthalpy of the recirculating air, or the
outdoor air temperature is lower than a preset air temperature or a set point which is a function of
outdoor climate.
An air economizer consists of outdoor, exhaust, relief, and recirculating ducts and dampers in the
AHU or PU, as well as a control system to operate them (see Fig. 21.2b). Air economizer control can be
subdivided into enthalpy-based differential enthalpy, fixed enthalpy, and electronic enthalpy economizer
controls and also temperature-based fixed dry-bulb and differential dry-bulb economizer controls.
2. Water or water-side economizers. These economizers use the outdoor air to cool the condenser
water or cooling water in the cooling tower or evaporative cooler first, and then to cool the
the mixture of outdoor and recirculating air through a precooling coil. A water economizer consists
mainly of a cooling tower (or an evaporative cooler), a water precooling coil in the AHU or PU, a
circulating pump to circulate cooling water, and the associated control system.
Differential Enthalpy, Electronic Enthalpy, and Fixed Enthalpy
Economizer Control
The operation of an enthalpy-based economizer control system is based on the comparison of
air enthalpies or the comparison of air enthalpies and air temperatures. In a differential enthalpy
21.8 CHAPTER TWENTY-ONE
economizer, the on-and-off air economizer is based on the comparison of the enthalpy of the outdoor
air ho, in Btu / lb (kJ / kg), with the enthalpy of the recirulating air hru, in Btu / lb (kJ / kg). In cooling
mode operation, if ho  hru, the differential enthalpy economizer is turned on. If ho 	 hru, the differential
enthalpy economizer is shut off. Enthalpies of outdoor air and recirculating air can be calculated
from their measured dry-bulb and wet-bulb temperatures, or from their measured dew points.
The sequence of operations of a typical differential enthapy economizer control system based on
measured dry- and wet-bulb temperature of the outdoor and recirculating air is as follows:
1. The outdoor air temperature To is sensed by sensor T1, and the relative humidity of outdoor
air o is sensed by humidity sensor H1. Both signals are sent to a DDC controller, where calculations
are performed to combine the signals into an outdoor air enthalpy ho signal. This signal is
compared with the recirculating air enthalpy hru, which is the combined signal from temperature T2
and humidity sensor H2, at the DDC controller.
2. If ho 	 hru, the DDC controller sends signals to actuate the damper actuators so that the outdoor
air and exhaust air dampers will be at their minimum opening positions, as determined by the
manual position switch. Meanwhile, the recirculating damper is fully opened.
3. If ho  hru, the DDC controller sends signals to actuate the damper actuators to fully open the
outdoor air and exhaust air dampers; at the same time, the recirculating damper will be closed.
4. Temperature sensor T3 senses the mixed air temperature of outdoor and recirculating air Tm
for airflow leaving the mixing box. This signal is compared with the set point Tm,s  55°F (12.8°C)
in the DDC controller. When the outdoor air temperature To  55°F (12.8°C), under the condition
that the opening of the outdoor damper is equal to or greater than its minimum opening position,
the DDC controller then actuates the damper actuator to modulate the opening of the outdoor,
exhaust, and recirculating air dampers to maintain a predetermined mixed air temperature of Tm 
55°F (12.8°C). When To  55°F (12.8°C), the refrigeration compressors or chillers that serve this
air system can be shut down.
In an electronic enthalpy economizer control, outdoor air is compared with a set point A which
can be represented on the psychrometric chart by a curve that goes through a point at approximately
75°F (23.9°C) and 40 percent relative humidity. This curve is nearly parallel to dry-bulb lines at low
humidity level and nearly parallel to enthalpy lines at high humidity levels. In a fixed enthalpy
economizer control, outdoor air is often compared with a fixed enthalpy hoA of 75°F (23.9°C) and
50 percent relative humidity with an enthalpy of 28 Btu / lb (65 kJ / kg) at sea level. In cooling mode
operation, if ho 	 28 Btu / lb (65 kJ / kg) at sea level, the fixed enthalpy economizer shall be shut
off. If h  28 Btu / lb (65 kJ/ kg), the fixed enthalpy economizer shall be turned on.
Fixed Dry-Bulb and Differential Dry-Bulb Economizer Control
There are two temperature-based economizer controls: fixed dry bulb and differential dry bulb. In a
fixed dry-bulb economizer control, outdoor air temperature To is compared with a fixed dry-bulb
temperature Tdb. ASHRAE/IESNA Standard 90.1-1999 recommends the following fixed dry bulbs:
From the psychrometric chart, the enthalpies of 75°F, 50 percent relative humidity, 70°F, 64
percent relative humidity, and 65°F, 85 percent relative humidity are all near to the value 28 Btu / lb
(65 kJ / kg). In a differential dry-bulb economizer control, outdoor air temperature To is compared
with the fixed recirculating temperature Tru. In cooling mode operation, if To  Tru, the differential
Climate Fixed dry bulb High-limit shutoff
Dry 75 to 78°F (23.9 to 25.6°C) To 	 75 to 78°F
Intermediate 70°F (21.1°C) To 	 70°F
Humid 65°F (18.3°C) To 	 65°F
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.9
dry-bulb economizer control is turned on. If To 	 Tru, the differential dry-bulb economizer control
is shut off. For a space temperature Tr  75°F (23.9°C) in summer, Tru is usually equal to 75 	
3  78°F (25.6°C) after absorbing the lighting sensible heat in the return ceiling plenum.
Theoretically speaking, it is more energy-efficient to compare the enthalpy of outdoor air with
that of recirculating air during air economizer control. However, comparing enthalpies necessitates
having both temperature and humidity sensors or temperature and dew point sensors. In actual
practice, humidity sensors may demonstrate considerable errors (sometimes up to 10 percent) and
have extensive maintenance requirements. Dew point sensors are delicate and expensive and cause
maintenance difficulties. Therefore, it is simpler and more convenient to use only temperature
sensors and to compare the outdoor air temperature To with the recirculating temperature Tru (or a
predetermined set point) instead of sensing and comparing enthalpies. This method of control is
called temperature economizer control. Outside weather regions using temperature-based economizer
control are shown in Fig. 21.3a and are similar to the outside weather regions for enthalpy
economizer control.
The sequence of operations of a typical differential dry-bulb economizer control system is as
follows:
1. The outdoor air temperature To is sensed by temperature sensor T1, as shown in Fig. 21.3b. Then
To is compared with the recirculating temperature Tru sensed by temperature sensor T2. If To 	
Tru, the DDC controller sends signals to damper actuators so that the outdoor and exhaust air
dampers will be at their minimum opening, as determined by the manual positioning switch. The
recirculating damper should be fully opened.
2. If To  Tru, the outdoor and exhaust air dampers should be fully opened and the recirculating
damper should be closed.
3. Temperature sensor T3 senses the mixed air temperature Tm after the mixing box. When the outdoor
air temperature To  55°F (12.8°C), under the condition that the volume flow of outdoor
air equal or be greater than the minimum requirement, the DDC controller modulates the outdoor,
exhaust, and recirculating air dampers to maintain a nearly constant mixed air temperature
of 55°F (12.8°C).
4. When To  55°F (12.8°C), the refrigeration system used to serve this single-zone VAV air system
is shut off.
In a fixed dry-bulb economizer control, a single dry-bulb temperature set point Tdb, may be used as
an indicator of when to change from using the minimum amount of outdoor air to using 100 percent
outdoor air. The changeover is determined from a comparison of To with Tdb. The proper magnitude
of Tdb depends on the outside climate and the characteristics of the air system as recommended by
ASHRAE/IESNA Standard 90.1-1999. According to Wacker (1989), an investigation is required to
determine the optimum Tdb.
Comparison of Enthalpy-Based and Temperature-
Based Economizer Controls
In addition to the instrumentation and maintenance problems of enthalpy economizer control, enthalpy-
based and temperature-based economizer controls are differentiated by the location of the
status points of outside weather in the dotted area A or shaded area B shown in Fig. 21.3a. When
the outside weather is inside the dotted area A, differential-enthalpy economizer control using
minimum outdoor ventilation air recirculation mode often consumes less energy to cool and dehumidify
the mixture of outdoor and recirculating air than with the 100 percent outdoor air mode.
When the outside weather falls in the shaded area B, a differential dry-bulb economizer control
using minimum outdoor ventilation air recirculating mode may sensibly cool the mixture at a
smaller enthalpy differential than a 100 percent outdoor air intake in a differential enthalpy economizer
control.
V?
o
21.10 CHAPTER TWENTY-ONE
21.11
FIGURE 21.3 Temperature economizer control: (a) outside weather regions; (b) temperature economizer.
Consider a single-zone VAV system that serves a building located in Phoenix, Arizona. Assume
that the outdoor temperature is 85°F (29.4°C), the humidity ratio is 0.006 lb / lb (kg / kg),
and the conditioned air leaves the coil at a temperature of 55°F (12.8°C), with a relative humidity
of 95 percent. The recirculating air enters the AHU or PU at a temperature of 78°F (25.5°C) and
a humidity ratio of 0.0094 lb / lb (kg / kg). If a differential enthalpy economizer control is used,
the state point of outdoor air oh falls in region III, and if a differential dry-bulb economizer control
is used, the state point of outdoor air oT falls in region I. The properties of moist air are as
follows:
A differential dry-bulb economizer control is required to cool the mixture of outdoor and recirculating
air down to an enthalpy difference of 6.2 Btu / lb (14.4 kJ / kg) instead of 7.2 Btu / lb (16.7
kJ / kg) using a differential enthalpy economizer control.
Therefore, consider the following:
1. For locations having hot and humid climates where many status points of outdoor air fall in the
dotted area A in Fig. 21.3a, a differential enthalpy economizer control is recommended.
2. For most locations having moderate outdoor humidity ratios, according to Spitler et al. (1987),
the energy savings offered by using a differential enthalpy economizer control instead of a temperature-
based economizer are small. The two types of economizer control may differ by only
about 5 to 10 percent. A fixed dry-bulb economizer control is recommended.
3. For locations having dry climates, a fixed dry-bulb of economizer control with a temperature setting
of 75 to 78°F (23.9 to 25.1°C) sometimes even saves more cooling energy than a differential
enthalpy economizer control. A fixed dry-bulb economizer control is strongly recommended.
4. For locations with humid climates, where the space requires a lower relative humidity, a fixed
dry-bulb economizer control using a set point Tdb a few degrees Farenheit lower than space temperature
sometimes is beneficial.
5. A survey including the simulation and comparison of various air economizer control alternatives
is often beneficial.
Water Economizer and Control
Consider a typical water economizer that uses a precooling coil in either an AHU or PU, as shown
in Fig. 21.4b, and has the following operating characteristics:
 The cooling tower approach is assumed to be 5°F (2.8°C).
 The mixing temperature of outdoor air and recirculating air Tm, is 78°F (25.5°C).
 The minimum temperature difference between Tm and the condenser water temperature Tcon,w is
4°F (2.2°C).
 When the outdoor wet-bulb temperature To  40°F (4.4°C), the temperature of the condenser
water used to replace the entire refrigeration through a heat exchanger is less than 45°F (7.2°C).
Differential enthalpy Differential dry bulb
Outdoor Off-coil Recirculating Outdoor Mixture Off-coil
Temperature, °F 85 55 78 85 79.5 55
Relative humidity, % 24 65 45 24 40 90
Humidity ratio, lb / lb 0.006 0.006 0.0094 0.006 0.0086 0.0086
Enthalpy, Btu/ lb 27.0 19.8 28.6 22.4
Enthalpy difference, Btu / lb 7.2 6.2
Amount of outdoor air, % 100 20
21.12 CHAPTER TWENTY-ONE
21.13
FIGURE 21.4 Water economizer control: (a) outside weather regions; (b) water economizer of an indoor PU.
 The outdoor wet-bulb temperature To,we below which the water economizer can be operated
effectively is
To,we  78 5 4  69°F (20.6°C)
During the operation of a water economizer, the outside weather can be grouped into three
regions based on the outdoor wet-bulb temperature, as shown in Fig. 21.4a:
Region I: The water economizer cannot be operated effectively.
Region II: The water economizer is operating effectively and replaces part of the refrigeration.
Region III: The water economizer replaces the refrigeration up to 100 percent.
The dividing line between regions I and II is the wet-bulb To,we line. The dividing line between
regions II and III is the wet-bulb Tr,of line, where Tr,of represents the outdoor wet-bulb temperature
below which the refrigeration plant serving the air system installed with a water economizer can be
shut down.
In locations where the outdoor climate is cold, and when the temperature discharged from the cooling
tower drops below 35°F (1.7° C), the water economizer should be shut down for freeze protection.
Comparison of Air and Water Economizers
The economizer cycle is a popular means of saving energy in the year-round operation of air conditioning
systems. The amount of energy savings, compared with air systems for which the economizer
cycle is not adopted, ranges from 15 to 40 percent of total energy use in various locations.
The advantages of air economizers over water economizers include the following:
1. Because of the intake of a large amount of outdoor air, the indoor air quality is improved.
2. The energy input to an air conditioning system that uses an air economizer is approximately
20 percent less than that for a water economizer.
3. The cooling tower requires less makeup water, fewer water treatments, and reduced maintenance.
However, the disadvantages of air economizers are as follows:
1. For an AHU or PU located in the interior core fan room or machinery room and used to serve
several floors in a high-rise building, the large vertical shafts used to transport the outdoor air
and exhaust air often occupy a lot of valuable rental space.
2. When outdoor weather suddenly changes, the space air may also produce corresponding fluctuations.
Building pressurization is not as stable as with a water economizer.
3. During the winter season, the intake of drier outdoor air always results in a lower space relative
humidity than is experienced using recirculating air. Humidification may be required.
4. For packaged units, the first cost of an air economizer is about 25 to 40 percent of the installation
cost of the packaged unit. For a water economizer, the first cost is about 10 percent of the
installation cost of the PU.
5. Coil freeze protection is more critical in locations where the outdoor temperature drops below
32°F (0°C).
ASHRAE/IESNA Standard 90.1-1999 Economizer Control Specifications
For the use of free cooling of outdoor air to reduce refrigeration, to improve indoor air quality, and to
save energy use in HVAC&R systems ASHRAE/IESNA Standard 90.1-1999 specified the
following:
21.14 CHAPTER TWENTY-ONE
Installation of Economizers. Each air system that has a fan and cooling capacity equal to or greater
than that listed below shall include an air or water economizer and meet the following requirements:
Here, N. R. indicates “not required,” Tdb the dry-bulb temperature, and Twb the wet-bulb temperature.
Exceptions include:
 Air system used for the purpose of outdoor ventilation air cleaning to meet ASHRAE Standard 62
requirement.
 Where more than 25 percent of the supply air is used for space humidification at above 35°F
(1.7°C) dew point to satisfy process needs.
 Air systems that include a condenser heat recovery system.
 Air systems that serve residential spaces, or serve spaces whose sensible cooling load is rather small.
 Air systems expected to operate less than 20 hours per week.
 Where the use of outdoor air will affect supermarket open refrigerated display cases.
 When the cooling efficiency of the packaged unit or heat pump meet the following increasing
requirements:
For other packaged systems, when mandatory minimum EERs varies from 10.3 to 9.0, the corresponding
minimum increased cooling efficiencies (EER) required are from 12.5 to 9.9. For details,
refer to ASHRAE/IESNA Standard 90.1-1999.
Design and controls of economizers shall be such that economizer operation does not increase
the building heating energy use during normal operation except allowable reheating, recooling, and
mixing by Standard 90.1-1999 (refer to Sec. 20.20).
Air Economizers. Standard 90.1-1999 specifies the following for air economizers:
 Capacity control. Air economizers shall be capable of modulating outdoor air and return air
dampers to provide up to 100 percent of outdoor air for cooling. Air economizers shall be capable
of automatically reducing outdoor air intake to the required minimum outdoor ventilation air speci-
fied by ASHRAE Standard 62 when outdoor air intake is no longer reducing cooling energy usage.
Operators of air economizers using various types of temperature-based or enthalpy-based economizer
controls shall be aware that for dry climates (shaded area B in Fig. 21.3a) minimum outdoor
ventilation air (required by ASHRAE Standard 62) recirculating mode often needs less cooling energy
than 100 percent outdoor air economizer operation. The 100 percent air economizer operation
shall be shut off. Similarly, for humid climates (dotted area A in Fig. 21.3a) in which minimum
System Size, Mandatory
Cooling Degree Days (CDD50)
kBtu/ h Qrc Minimum EER 3601–5400 5401–7200 7201–9000 9001–10,800
65  Qrc  135 10.1 12.1 11.6 11.1 10.7
135  Qrc  240 9.3 11.3 10.8 10.4 9.9
240  Qrc  760 9.0 10.9 10.5 10.0 9.6
No. of hours between
8 a.m and 4 p.m. with Twb  69°F, 69°F  Twb  73°F, Twb 	 73°F,
55°F  Tdb  69°F Btu /h Btu/h Btu /h
200 – 599 135,000 N.R. N.R.
600 – 799 65,000 135,000 N.R.
800 – 999 65,000 135,000 135,000
1000 – 1199 65,000 65,000 135,000
	1199 65,000 65,000 65,000
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.15
outdoor ventilation air recirculating mode often needs less cooling energy than 100 percent outdoor
air economizer control and the air economizer operation shall also be shut off.
 Sequence control. Economizer dampers shall be in sequence with the refrigeration and shall not
be controlled only by mixed air temperature except for single-zone air systems.
 Damper leakage. Both recirculating and outdoor air dampers shall have a maximum leakage rate
of 20 cfm / ft2 (100 L / s 
m2) at a pressure difference of 4 in. WC (1000 Pa).
 Relief of excess outdoor air. Means to relieve excess outdoor air during air economizer operation
shall be provided to prevent overpressurization of the building. The relief outlet shall be located
to avoid recirculation into the building.
Water Economizers. Water economizers shall be capable of cooling supply air through indirect
evaporation and providing up to 100 percent of the system cooling coil load (precooling coil) at an
outdoor air temperature of 50°F (10°C) dry bulb /45°F (7.2°C) wet bulb and below [except when
the dehumidification requirement cannot be met at these outdoor dry- and wet-bulb temperatures
and the 100 percent system cooling (precooling) coil load must be satisfied at outdoor air temperatures
45°F (7.2°C) dry bulb /40°F (4.4°C) wet bulb].
Precooling coils and water-to-water heat exchangers as components of the water economizer
shall either have a water-side pressure drop of less than 15 ft (44.7 kPa), or using a secondary water
loop so that the pressure drop of the coil or heat exchanger is not taken care of by the circulating
pumps when the system is in normal cooling (noneconomizer) mode.
Air systems with water cooling and humidification to maintain space humidity at greater than
35°F (1.7°C) dew point temperature shall use a water economizer if an economizer is required because
the air economizer needs to extract outdoor air at lower temperature and humidity ratios and
thus increases the humidifying load.
Integrated Economizers Control. Economizers shall be integrated with refrigeration and can provide
partial cooling and even additional refrigeration if required to meet the remaining cooling coil
load, except the following:
 Direct-expansion systems that control and reduce the required quantity of outdoor air to prevent
coil frosting at the lowest step of compressor unloading if this lowest step is no greater than 25
percent of the total system capacity.
 Individual packaged unit has a capacity less than 65,000 Btu / h (19 kW) and uses nonintegrated
economizer controls.
 System in locations with less than 800 average hours annually between 8 a.m. and 4 p.m. when the
ambient (outdoor) dry-bulb temperatures are between 55°F (12.8°C) and 69°F (20.6°C) inclusive.
Air Conditioning Cycle and System Calculations
The air conditioning cycle of a single-zone VAV system during cooling mode operation at summer
design conditions and heating mode operation at winter design conditions without space humidity
control can be plotted on the psychrometric chart as shown in Fig. 21.1b During summer cooling
mode part-load operation, the state point of outdoor air o varies, and the sensible ratio of the space
conditioning line SHRsp usually becomes smaller than at design load. If the zone temperature Tr is
maintained at the same predetermined limit as in design load, the zone relative humidity will be
slightly greater than that in design load if SHRsp becomes smaller. During winter heating mode partload
operation, if the sensible heat ratio of space conditioning line SHRsp becomes smaller because
of the reduced volume flow, the zone relative humidity will also be greater than that in design load.
From Eq. (20.69), the design supply volume flow rate , in cfm [m3 / (60 s)], can be calculated
as
(21.1) V?s 
Qrs
60 s cpa (Tr Ts)
V?
s
21.16 CHAPTER TWENTY-ONE
where Qrs  zone sensible cooling load, Btu /h (W), and Tr, Ts  zone and supply air temperature,
°F (°C). In Eq. (21.1), [m3 / (60 s)] means that in SI units, the unit should be m3 /s.
From Eq. (20.48b), the design cooling coil load Qcc, in Btu / h (W), can be calculated as
(21.2)
where hm  enthalpy of the mixture of outdoor and recirculating air, Btu / lb (J / kg), and hcc 
enthalpy of conditioned air leaving the cooling coil, Btu/ lb (J / kg). From Eq. (20.85), the design
heating coil load Qch, in Btu / h (W), which offsets the space heating load or sometimes maintains a
setback temperature during unoccupied periods to reduce warm-up load next morning and prevent
freezing, can be calculated as
(21.3)
where Thc, Ten  air temperature leaving and entering heating coil, °F (°C).
Zone Temperature Control—Sequence of Operations
Consider a single-zone VAV system using an AHU with a water cooling coil, a water heating coil,
and a temperature economizer (air economizer), as shown in Fig. 21.1. The sequence of operations
of this VAV system to maintain a preset zone temperature Tr is as follows:
1. When the AHU is in the off position, the outdoor and exhaust dampers are closed, the two-way
valves of the cooling and heating coils are also closed, and supply and return fan motors are off.
2. When the AHU is in the on position, zone temperature sensor T1 senses the zone temperature
Tr , sends a signal to the DDC controller, and compares it with the set point in summer of 75°F
(23.9°F) in the controller. If Tr > 75°F (23.9°C), the DDC controller calls for cooling—cooling
mode operation. At the same time, the outdoor temperature To is measured by the outdoor temperature
sensor T2. If To 	 75°F (23.9°C) (a temperature economizer set point), the DDC controller
opens the outdoor and exhaust dampers to a minimum opening position, and fully opens the recirculating
damper. The DDC controller also starts supply and return fan motors with speed control
via variable speed drives.
3. The discharge air temperature Tdis is sensed by temperature sensor T4. During cooling
mode operation, if the off-coil temperature Tcc is 55°F (12.8°C), the set point of Tdis  55 	 2
(fan power)  57°F (13.9°C). When Tdis 	 57°F (13.9°C) is sensed by T4, the DDC controller
opens and modulates the two-way valve of the water cooling coil so as to maintain Tdis  57°F
(13.9°C).
4. If To  75°F (23.9°C), using the free cooling of outdoor air is always beneficial in order to
save energy. The DDC controller fully opens the outdoor and exhaust dampers and, at the same
time, closes the recirculating damper. After the air economizer is energized, if Tdis is still higher
than 57°F (13.9°C), the DDC unit controller will open and modulate the two-way valve of the water
cooling coil to maintain Tdis  57°F (13.9°C).
5. Zone temperature Tr is sensed by a temperature sensor T1. This signal is compared with the
zone temperature set point Tr,s, such as 75°F (23.9°C) in the DDC controller. If Tr drops below or
rises above 75°F (23.9°C), an output from the DDC controller actuates and modulates the supply
fan speed, and thus the supply volume flow rate through an adjustable-frequency variable-speed
drive, to maintain Tr  75°F (23.9°C). For single-zone VAV system serving a large conditioned
space, several temperature sensors can be located in different areas. Either the mean value or the
highest or lowest temperature can be taken as sensed input to the DDC controller.
6. A pressure sensor P1 is used to sense the zone pressure pr in the conditioned space. If pr 	
0.03 in. WC (7.5 Pa), the DDC controller starts the return fan and modulates its fan speed to maintain
a zone pressure pr  0.03 in. WC (7.5 Pa).
Qch  60V?s scpa(Thc Ten)
Qcc  60V?s s(hm hcc)
160 V?s
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.17
7. When To  55°F (12.8°C) and the system is operated in cooling mode operation, the DDC
controller
 Modulates the outdoor and recirculating damper to maintain a mixing temperature of 55°F
(12.8°C) and Tdis  57°F (13.9°C) with a volume flow of outdoor air not less than the ventilation
requirement
 Modulates the fan speed to maintain a zone temperature of 75°F (23.9°C)
 Closes the two-way valve of the water cooling coil
8. If the zone temperature sensed by sensor T1 is 72°F  Tr  74°F (22.2°C  Tr  23.3°C),
the system is operated in dead-band operation. Dead-band operation is neither a cooling nor heating
mode operation. In dead-band mode, the DDC controller
 Keeps both of the two-way valves of the cooling and heating coils closed
 Opens the outdoor damper to its minimum opening position and fully opens the recirculating
damper to provide necessary outdoor ventilation air for the occupants
9. If the zone temperature Tr drops below 72°F (22.2°C), the DDC controller calls for heating,
(heating mode operation). Also if To  55°F (12.8°C), then
 The outdoor damper is opened to its minimum opening position to extract only the necessary
amount of outdoor ventilation air.
 The DDC controller opens and modulates the two-way valve of the heating coil to maintain a
discharge air temperature during heating mode not higher than 15°F (8.3°C) above the space temperature,
say 72 	 15  87°F (30.6°C).
For the zone temperature control in heating mode operation, a reverse-acting relay is energized so that
the control action is reversed when the cold air supply is changed over to warm air supply. The DDC
controller also modulates the supply fan speed to maintain a zone temperature Tr  72°F (22.2°C).
During the above zone controls for a single-zone VAV system, to determine whether the DDC
controller calls for cooling or heating, either a zone temperature set point Tr,s or a specific outdoor
temperature set point To,s can be used as the comparing criterion. Zone temperature set point Tr,s directly
relates to the zone load and the actual zone conditions and is recommended. Because of the
high internal load, a zone may still need a cold air supply at a rather low outdoor temperature.
When a cooling mode is automatically changed to heating mode or vice versa, the width of the
dead band greater than 3°F (1.7°C) is not appropriate for thermal comfort. When two separate set
points are used and heating and cooling mode operations are not automatically changed over, only
then can a dead-band width of 5°F (2.8°C) or sometimes greater be used.
To save energy as well as to increase the zone supply volume flow rate, the discharge air temperature
is often reset to a higher value during cooling mode part-load operation, such as from 57 to
60°F or to 63°F (13.9 to 15.6°C or to 17.2°C).
21.3 VAV COOLING, VAV REHEAT, AND PERIMETER HEATING
VAV SYSTEMS
VAV Cooling Systems
A VAV cooling system is a multizone air system with VAV boxes to modulate the cold supply air
volume flow rate to match the variation of zone load to maintain a preset zone temperature, as
shown by the air system that serves the interior zone in Fig. 21.5a. VAV cooling systems with VAV
boxes only are widely used to serve the interior zone in commercial and industrial buildings.
21.18 CHAPTER TWENTY-ONE
21.19
FIGURE 21.5 A VAV reheat system: (a) schematic diagram; (b) air conditioning cycle.
VAV Reheat Systems
A VAV reheat system is a multizone VAV system with either VAV boxes or reheating VAV boxes to
modulate its cold supply volume flow rate for each of the control zones, to match the variation of
the zone load to maintain a preset zone temperature. For the interior zone in a building, VAV boxes
are used (VAV cooling); and for the perimeter zone in a building, reheating VAV boxes are used
(VAV reheat) in which a reheating coil is added to a cooling VAV box, as shown in Fig. 21.5.
To minimize energy consumption, as covered in Sec. 20.20, ASHRAE/IESNA Standard 90.1-
1999 specifies the following criteria: VAV reheat systems shall be capable of operating in sequence
the supply of heating and cooling energy to the zone to minimize simultaneous cooling and heating
except as follows:
1. 300 cfm (142 L / s), or a zone peak flow rate total of no more than 10 percent of the total fan system
flow rate
2. The minimum required to meet ventilation requirements
3. 0.4 cfm / ft2 (2.03 L / s 
m2) of zone conditioned floor area
VAV reheating is a simple and effective means of maintaining preset space temperature with a
lower relative humidity at cooling mode part-load operation and winter heating mode. For details
refer to Standard 90.1-1999.
System Description
The mixture of outdoor and recirculating air is conditioned either in an AHU or in a PU. Both
contain a supply fan, a return or relief fan, cooling coils, and medium-efficiency filters and are
distributed to various control zones via supply ducts, runouts, reheating VAV boxes, VAV boxes,
and outlets.
During summer cooling mode operation, both the reheating VAV boxes in the perimeter zone
and the VAV boxes in the interior zone modulate the zone supply volume flow rate to maintain a
predetermined zone temperature Tm. When the space sensible load in the perimeter zone drops
during summer cooling mode part-load operation, the VAV box closes its opening until its minimum
setting is reached. If the zone temperature Tm drops further and falls below a certain limit,
the reheating coil is energized and reheats the supply air to maintain a predetermined zone temperature.
During winter heating mode operation, when one of the zone temperatures in the perimeter zone
falls below a predetermined limit, the reheating coil in the reheating VAV box in this control zone is
energized to heat the air to offset the space heating load to maintain a preset zone temperature.
For a control zone in the interior zone, the zone sensible cooling load seldom drops below 30
percent of its peak load while the space is occupied, except during morning warm-ups and weekends.
Therefore, reheating VAV boxes are not used in the interior zone. In the interior zone, zone
temperature Tm is also maintained by modulating the supply volume flow rate of the VAV box in
each control zone.
If an AHU or PU serves both the perimeter and the interior zone, space air from the perimeter
and interior zone is often returned to the same AHU or PU in the fan room through return slots,
ceiling plenum, and connecting return ducts.
Perimeter Heating VAV System
A perimeter heating VAV system consists of a hot water perimeter heating system to offset the winter
transmission and infiltration loss in the perimeter zone and a VAV cooling system with VAV
21.20 CHAPTER TWENTY-ONE
boxes to provide cooling and outdoor ventilation air for both perimeter and interior zones, as shown
in Fig. 21.6. A perimeter heating system consists of hydronic hot water finned-tube baseboard
heaters with a zone control system. The baseboard heaters must be carefully zoned and controlled
in coordination with the VAV boxes. Each side of the building should have at least a separate control
zone to match the variation of the solar load and the infiltration loss as the wind direction shifts.
Perimeter heating systems are suitable for commercial buildings in locations with long, cold
winters.
During the early development of VAV systems, air skin VAV systems were used in office buildings.
An air skin VAV system uses a constant-volume variable supply temperature air skin system to
offset the transmission loss or transmission gain in the perimeter zone; it also has a VAV cooling
system with VAV boxes to provide cooling and outdoor air for the perimeter and interior zones. The
supply temperature of the air skin system is reset according to the outdoor temperature. If several
air skin systems are not used for various zones of different sides of a building, field surveys show
that during a cold day in winter, rooms facing south are sufficiently warmed by the incoming solar
radiation. The additional warm air supply from the constant-volume air skin system raises the zone
temperature to an unacceptably high level. Simultaneous cooling from the VAV boxes must be provided
to offset the warm air supply, which results in a waste of energy.
VAV Box
A variable-air-volume box or a cooling VAV box is a terminal device in which the supply volume
flow rate is modulated by varying the opening of the air passage by means of a single-blade butter-
fly damper, a multiblade damper, or an air valve, as shown in Fig. 21.7. A VAV box may have a single
outlet or multiple round outlets.
A single-blade damper VAV box, as shown in Fig. 21.7a and b, has a simple construction and is
simple to operate. A typical damper closes at an angle 30° from vertical and rotates in counterclockwise
direction to an angle of 60° in the fully open position.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.21
FIGURE 21.6 A perimeter heating VAV system.
An air valve, as shown in Fig. 21.7c, is a piston damper moving horizontally inside a hollow
cylinder. The opening of the air passage can be adjusted. The main advantage of an air valve is its
almost linear relationship between the modulated air volume and the displacement of the piston
damper.
Control Mechanisms. The control mechanism of a VAV box can be classified as pneumatic or direct
digital control. In a pneumatic control VAV box, the zone temperature is sensed by a pneumatic
thermostat, and the rotation of the single-blade damper is actuated by pneumatic power. When the
thermostat senses an increase in zone temperature, the increase of compressed air pressure on a diaphragm
moves the actuator at a stroke of 1 to 1.5 in. (25 to 38 mm) and rotates the damper through
the linkages, as shown in Fig. 21.7b. Therefore, the air passage opens wider. The spring pushes the
actuator back when the air pressure decreases.
In a DDC VAV box, the temperature sensor sends a signal to a DDC terminal controller. It actuates
the motorized operator, moves the actuator to a certain displacement, and opens the singleblade
damper wider. The damper is closed by either spring force or the reverse rotation of the
motorized operator. Usually, a PI control mode is used by such a DDC terminal controller.
21.22 CHAPTER TWENTY-ONE
FIGURE 21.7 VAV boxes: (a) single-blade, pressure-dependent; (b) single-blade, pressure-independent; (c) air
valve.
The pneumatic or direct digital control of an air valve is similar to that of a single-blade damper.
A pneumatic control VAV box has a faster response actuator and a lower initial cost than those of a
DDC VAV box. However, pneumatic systems require more maintenance to provide clean, dry compressed
air. More importantly, DDC VAV boxes provide many sophisticated control functions and
can communicate with other DDC controllers and the central computer.
There is also a type of self-powered or system-powered VAV box control. It requires a higher
supply air pressure to actuate the single-blade damper or air valve according to the input signal
from the temperature sensor. A self-powered VAV box also has complicated control mechanisms, so
its applications are very limited.
Influence of Duct Static Pressure. VAV boxes can be classified as pressure-dependent or pressure-
independent. In a pressure-dependent VAV box, the variation of the duct static pressure at the
inlet of the VAV box caused by the opening and closing of the dampers connected to the same main
supply duct influences the modulation of its supply volume flow rate. When the pressure at the inlet
varies, the airflow often oscillates. Pressure-dependent VAV boxes are least expensive. They are
used when the duct static pressure is more stable and in places where there is no need for maximum
and minimum limit control.
A pressure-independent VAV box modulates its supply volume flow rate regardless of the variation
of duct static pressure at its inlet. A typical pressure-independent VAV box is shown in Fig. 21.7c. The
temperature sensor resets the set point of the sensed velocity, and the controller actuates the damper
to a wider or a narrower position based on the sensed signal from the velocity probe and the reset set
point. Even if the static pressure at the VAV box inlet varies from 0.5 to 3 in. WG (125 to 750 Pag),
the volume flow is maintained according to the required value called for by the temperature sensor
and the controller. Pressure-independent VAV boxes are widely used in VAV systems.
The sizes of the VAV boxes made by one manufacturer range from 04 to 20, with a corresponding
volume flow rate from 225 to about 4000 cfm (106 to 1887 L / s). The pressure drop of these
VAV boxes at nominal maximum volume flow rate when the damper is fully open usually varies
from 0.2 to 0.5 in. WC (50 to 125 Pa).
To provide the required amount of outdoor air to the conditioned space for occupants, a VAV
box usually reduces its volume flow to a minimum setting, such as 30 percent of its design flow,
while the space is occupied.
Reheating VAV Box
A reheating VAV box is a VAV box with a reheating coil, as shown in Fig. 21.8. The reheating coil
is usually a two- to three-row hot water heating coil. If an electric coil is installed, it should be a
duct-mounted coil at least 4 ft (1.22 m) downstream from the VAV box. Federal and local safety
codes must be followed if an electric coil is installed. To prevent excessive pressure drop at the reheating
coil, the face velocity of the reheating coil is usually between 200 and 400 fpm (1 and
2 m/ s). For each row of the reheating coil, the additional pressure drop is usually between 0.1 and
0.2 in. WC (25 and 50 Pa).
In the perimeter zone, a two-way slot diffuser should be located about 1 ft (0.3 m) from the window.
Air should be discharged downward from one of the slots in a direction toward the window, as
described in Sec. 18.6. Air is discharged horizontally away from the window from another slot.
Sound Power Level of a VAV Box
The sound power level of a VAV box depends mainly on the following:
 Volume flow of supply air for a specific size box, cfm (L / s)
 Difference in static pressure ps across the VAV box, in in. WC (Pa), to provide a specific volume
flow, in cfm (L / s)
 The configuration of the VAV box, flexible ducts, and diffusers
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.23
The greater ps, the larger the static pressure at the box inlet. The smaller the damper opening, the
higher the sound power level of the VAV box.
Noise generated by the VAV box can be transmitted to occupants in the conditioned space by a
duct-borne path through flexible ducts and diffusers, or a radiated path from its casing through the
ceiling plenum.
Evaluating noise levels from VAV boxes on an occupied zone has always been difficult because
of the many variables that affect the results. The ADC and ARI have jointly developed Industry
21.24 CHAPTER TWENTY-ONE
FIGURE 21.8 DDC for a reheating VAV box.
Standard 885, Acoustical Level Estimation Procedure, as described in Chap. 19, to provide a reliable
method to predict NC or RC in an occupied space.
In Smith (1989), the sound pressure levels of two VAV boxes in a 29-ft 20-ft (8.8-m 6.1 m)
room with an 8-ft 7-in. (2.6-m) ceiling were estimated and measured. One VAV box was operated at
500 cfm (236 L / s) at 4.0 in. WG (1000 Pa) inlet static pressure and supplied 250 cfm (118 L / s) to
each of two diffusers through flexible ducts. The actual sound pressure level measured in the occupied
space was NC-41. Another VAV box was operated at 600 cfm (283 L / s) with 3.0 in. WG
(750 Pa) inlet static pressure and supplied 300 cfm (142 L / s) to each of two diffusers in the same
room. The actual sound pressure level measured was about NC-41. The measured sound pressure
levels were about 2 to 3 dB higher or lower than the estimated duct-borne and radiated paths according
to Industry Standard 885. If the sound power level at mid-frequency 2k and 4k Hz can be
further attenuated by flexible duct and the slot diffuser, the sound pressure level in the occupied
space may drop to NC-35. The sound power level of a VAV box listed in a manufacturer’s catalog
must be tested in a certified laboratory.
VAV Reheat Zone Temperature Control and Sequence of Operations
The year-round control of zone temperature served by a reheating VAV VV1 box in the perimeter
zone of a building served by a VAV reheat system can be divided into cooling mode, dead-band
mode, and heating mode operations.
1. On-off control is in the off position. When an AHU or a PU that serves this perimeter zone is
shut off, the single-blade damper of the VAV box VV1 and the two-way valve V1 of the reheating
coil both are closed.
2. On-off control is in the on position. When the zone temperature Trx1 sensed by the zone sensor
exceeds a set point, such as 75°F (23.9°C), the DDC system controller calls for cooling; the AHU
or PU and the reheating VAV box are both in cooling mode operation. The DDC terminal controller
opens the single-blade damper VV1 of the reheating VAV box to maximum opening position. The
two-way valve of the reheating coil is still closed.
3. If the zone temperature Trx1p at cooling mode part-load operation drops below 75°F (23.9°C)
because of a reduction in the zone sensible cooling load Qrsx1p, the DDC terminal controller receives
a corresponding analog input from the zone sensor. This signal is compared with the set point 75°F
(23.9°C). If Trx1p  75°F, (23.9°C), the DDC terminal controller sends an output to the damper actuator
and starts to reduce the opening of the flow passage of the reheating VAV box. The total pressure
loss across the VAV box is then increased, and so the volume flow rate of the conditioned air
supply to this control zone is reduced accordingly. If the reduced is balanced with the
zone sensible cooling load, Trx1p will remain at 75°F (23.9°C). The DDC terminal controller then
modulates the opening of the single-blade damper VV1 with a PI control mode to maintain a constant
75°F zone set point, as shown in Fig. 21.8.
4. Further drop in Qrsx1p causes damper VV1 to close more and more until it reaches the minimum
setting which is necessary to supply the required amount of outdoor ventilation air for this zone, such as
30 percent of the zone peak flow. After the damper reaches the minimum setting, if Qrsx1p further drops
and causes Trx1p to drop below 75.0°F (23.3°C), the reheating VAV box will operate in dead-band mode.
5. When the zone temperature drops within the range 72.0°F  Trx1p  75.0°F (22.2°C 
Trx1p  23.9°C), the reheating box and zone control operate in dead-band mode. In dead-band
mode, the DDC terminal controller actuates as follows:
 The damper in the reheating VAV boxes closes to the minimum setting, and the minimum cold
primary air supply is at a value, such as 30 percent of the zone peak flow, to provide outdoor
ventilation air for the control zone.
 The reheating coil is deenergized.
 The space relative humidity will be slightly lower than that at the design condition.
V?
sx1,min
V?
sx1p V?sx1p
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.25
In dead-band mode, the reheating box will reenter the cooling mode operation only if T  75°F
(23.9°C).
6. When the zone temperature Trx1  72.0°F (22.2°C), the reheating box and the associated zone
control are in heating mode operation. During heating mode operation, the DDC terminal controller actuates
the reverse action relay so that the direct-acting mode will be changed to reverse-acting mode.
The volume flow rate of the cold primary air supplied from the reheating box in heating mode
is still maintained in a minimum setting of 30 percent of zone peak volume flow and will remain
constant to provide cold primary air for ventilation requirement.
The DDC terminal controller opens and modulates the opening of the two-way valve of the reheating
coil as well as the mass flow rate of hot water flowing through the coil to maintain a preset
zone temperature Trx1  72°F (22.2°C). The zone relative humidity may be at a value between 25
and 50 percent depending on the zone internal latent loads, outdoor air humidity ratio, and percentage
of outdoor air intake.
At winter design conditions, the warm supply air temperature from the reheating box (after the
cold primary air is heated in the reheating coil) should not exceed 87°F (30.6°C), in order to reduce
the buoyancy effect of the warm supply air.
The dead band in a year-round zone-temperature control for a perimeter zone served by a VAV
reheat system is different from the dead band in a dual-thermostat year-round zone temperature
control system served by a constant-volume single-zone system, as discussed in Sec. 20.19. In a
VAV reheat system,
 The cooling mode is automatically changed over to heating mode operation and vice versa.
 The occupants may suffer the zone temperature variation from cooling mode to dead-band mode,
and then to heating mode and vice versa, continuously within a time period of several hours.
The width of the dead-band mode in such a circumstance is preferably limited to 1 to 3°F (0.6 to
1.7°C).
Stability of Zone Control Using VAV Boxes and Reheating VAV Boxes
VAV boxes including reheating VAV boxes using DDC terminal controllers can provide the following
functional controls:
 Zone temperature control by modulating the damper or air valve
 Zone temperature reset based on outdoor temperature or zone temperature
 Maximum and minimum settings of airflow for each zone
 Morning warm-up and cool-down control
A normally closed DDC VAV box can shut off its supply volume flow when the served control zone is
not occupied. This is especially useful for partially occupied floors after normal working hours.
Air flowing through VAV boxes is self-balanced along the supply main duct. However, airflow
imbalance does affect the modulation range of VAV boxes. A well-designed supply duct system
with minimized imbalance is recommended.
One characteristic of VAV boxes that directly affects their performance is control stability. Dean
and Ratzenberger (1985) described the primary differences between an efficient and inefficient design
of a VAV box and its control. An excellent design minimizes the open-loop gain and improves
the overall system stability. Gain is defined as the real part of the logarithm of the transfer function
of the control system. The primary factors that affect the open-loop gain and thus the stability of a
VAV box are as follows:
 Actuator size. In an efficient design, the stroke of the actuator is small [typically 1.0 in. (25 mm)],
and in a pneumatic system, the active area of the actuator is typically only 2 in.2 (1290 mm2)
V?
s,cxn
21.26 CHAPTER TWENTY-ONE
instead of 8 in.2 (5160 mm2 ). A larger actuator needs more compressed air to drive a full-range
stroke, so the gain is greater.
 Damper size. Gain is directly proportional to the size of the damper as well as the size of the VAV
box.
 Duct pressure. A lower excess pressure pexc, in in. WC (Pa), during maximum and minimum air-
flow affects open-loop gain. Excess pressure is defined as the amount of pressure above that required
to drive a nominal flow through a fully open damper.
According to Dean and Ratzenberger (1985), the typical measured performance of 26 efficient
B-size VAV boxes and 27 inefficient C-size VAV boxes is shown below:
For a step function response due to a step change in internal loads, such as turning lights on or off,
the deviation from set point in a pneumatic control system with a large actuator lasts about 40 min.
In a DDC system, it lasts only about 15 min.
Based on their analysis, Dean and Ratzenberger (1985) recommended the following:
 A VAV system using pneumatic control VAV boxes should have relatively low main duct pressures
and relatively small VAV boxes and actuators.
 VAV boxes using DDC systems have excellent operating characteristics, although their actuators
are a little slower than pneumatic controls.
 Pressure-independent VAV boxes with velocity resetting improve the control stability of poorly
designed pneumatic controls, make an unstable system more stable, and improve the performance
of a poorly designed supply duct system.
Case Study: A VAV Reheat System
Consider a multizone VAV reheat system using an AHU to serve a typical floor of 20,000 ft2
(1858 m2) of conditioned area in a high-rise office building that is divided nearly equally into
perimeter and interior zones. Reheating VAV boxes are used for summer cooling and winter heating
in perimeter zones. VAV boxes are used for summer and winter cooling in the interior zone when
conditioned space is occupied. The perimeter and interior zones may each be divided into several to
more than 20 control zones. The summer and winter design conditions are shown below:
Perimeter Interior
Summer indoor temperature, °F 75 75
Space relative humidity, percent 43 45
Summer outdoor temperature, °F 95 95
Summer outdoor wet-bulb temperature, °F 78 78
Winter indoor temperature, °F 72 72
Winter outdoor temperature, °F 20 20
Winter indoor relative humidity, percent 30 30
B size C size
Maximum flow, cfm 442 448
Corresponding pressure ps,max, in. WG. 1.07 1.40
Minimum flow, cfm 64 66
Corresponding pressure, ps,min, in. WG 1.24 1.59
Excess pressure, in. WC 0.08 0.62
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.27
For the perimeter zone in a typical floor of this office building, the average space cooling load
density at summer design condition is 30 Btu /h
 ft2 (94.6 W /m2 or 8.8 W/ ft2). The sensible heat ratio
of the space conditioning line SHRs for the perimeter zone at design load is usually about 0.95.
For the interior zone of a typical floor of this office building, the summer design load density is
15.5 Btu /h
 ft2 (4.5 W/ ft2 or 48.4 W/m2) depending on the equipment load density. Occupant
density is assumed to be 150 ft2 / person (14 m2 / person). The sensible heat ratio of the space conditioning
line SHRs for the interior zone may vary from 0.80 to 0.90.
Conditioned Air Off-Coil and Supply Temperature Differential
For a supply temperature differential of 20°F (11.1°C), if the zone supply temperature Ts  55°F
(12.8°C), for a typical AHU with a supply fan power temperature rise of 2°F (1.1°C) and a
supply duct temperature rise of 3°F (1.7°C) at summer design conditions, the temperature of conditioned
air leaving the cooling coil Tcc should be 55 (2 	 3)  50°F (10°C). If the relative
humidity of conditioned air off the coil cc is 95 percent, and hcc  20 Btu / lb (46.5 kJ / kg), then,
from the psychrometric chart shown in Fig. 21.5b, the relative humidity of the supply air s is
81 percent.
Psychrometric analysis shows that to maintain a zone temperature of 75°F (23.9°C) with an offcoil
temperature of Tcc  50°F (10°C) and cc  95 percent, a supply temperature Ts  55°F
(12.8°C), and a sensible heat ratio of space conditioning line in the perimeter zone SHRsx of 0.95,
or of interior zone SHRsi of 0.84, and a supply temperature differential Ts  20°F (11.1°C), the
relative humidity of the perimeter zone must be rx  43 percent and the relative humidity of the
interior zone ri  45 percent, as shown in Fig. 21.5b.
System Volume Flow Rate and Coil Load
For any control zone in the perimeter and interior zones, the VAV box, reheating VAV box, flexible
ducts, and slot diffusers should be sized according to the zone peak supply volume flow rate. The
zone peak supply volume flow rate of any control zone served by a VAV box or a reheating VAV
box , in cfm [m3 / (60 s)], at summer design conditions can be calculated as
(21.4)
where Qrsn  zone sensible cooling load at summer design conditions, Btu/ h (W). In a multizone
VAV system, the total supply volume flow rate for the perimeter zone , in cfm [m3 / (60 s)], at
summer design load is the maximum possible coincident total supply volume flow rate of various
control zones at the same instant, or the block supply volume flow rate, which corresponds to the
block load of the perimeter zone Qrsx, in Btu / h (W). The block supply volume flow rate in the
V?
sx
V?
sn 
Qrsn
60 scpa(Tr Ts)
V?
sn
Perimeter Interior
Winter outdoor humidity ratio, lb / lb 0.0017 0.0017
Summer block cooling load, Btu /h 300,000 160,700
Summer sensible cooling load, Btu /h 285,500 135,000
Winter heating linear density, Btu /h
 ft 150 —
Winter interior zone latent load, Btu /h 13,300 13,300
Outdoor air required, cfm 1,920 1,920
Summer supply system heat gain, °F 5 5
Summer return system heat gain, °F 3 3
21.28 CHAPTER TWENTY-ONE
perimeter zone at summer design conditions can be calculated as
(21.5)
Similarly, the total block supply volume flow rate of the interior zone at summer design condition
, in cfm [m / (60 s)], is
(21.6)
Assume that Qrsi and remain constant during occupied hours. The total supply volume flow
rate for the AHU at summer design conditions is calculated as
(21.7)
Trunks or main ducts that serve multizones should accommodate the maximum possible airflow
through these ducts at a given time. A diversity factor can be multiplied by the sum of individual
peak volume flows to estimate the possible maximum flow.
From the psychrometric chart, the weighted mean space relative humidity for perimeter and
interior zones rm  43.5 percent, and the weighted mean humidity ratio wrm  0.0079 lb / lb
(kg / kg). Because the temperature of recirculating air entering the AHU Tru  75 	 3  78°F,
and wru  wrm  0.0079 lb / lb, point ru can be plotted on the psychrometric chart. Draw line
ru-o.
so point m can then be determined: Tm  80.5°F (26.9°C), and hm  30.2 Btu / lb (70.2 kJ / kg). The
cooling coil load Qcc in the AHU which is also a block load, in Btu / h (W), is
(21.8)
Summer Cooling Mode Part-Load Operation
During summer cooling mode part-load operation, if the sensible cooling load in the perimeter
zone is reduced to about 50 percent of its design load and the sensible cooling load of the interior
zone is reduced to 85 percent of its design load, the total volume flow rate of the supply fan may
decrease to about two-thirds of its design value. It is assumed that the supply fan power heat gain
and the volume flow are both reduced to 0.65 of the design values. The temperature rise at the
supply fan at part load is still about 2°F (1.1°C). If the chilled water temperature entering the coil
Twe is reset from 45°F (7.2°C) at design load to 48°F (8.9°C) at about two-thirds part-load operation,
the discharge air temperature increases to Tdis  50 	 2 	 48 45  55°F (12.8°C). The
temperature rise from duct heat gain increases because of the reduction of the supply volume
flow rate. Because the increase of this temperature rise is proportional to the reduction of the
supply volume flow rate, the supply air temperature for the perimeter zone at part load can be
 60 19,227 0.075(30.2 20)  882,519 Btu / h (258,578 W)
Qcc  60V?s s(hm hcc)
ru-m
ru-o

3840
19,227
 0.20
V?
s  V?sx 	 V?si  13,054 	 6173  19,227 cfm (9073 L / s)
V?
si

135,000
60 0.075 0.243 (75 55)
 6173 cfm (2913 L / s)
V?
si 
Qrsi
60 scpa(Tri Ts)
V?
si

285,500
60 0.075 0.243(75 55)
 13,054 cfm (6160 L / s)
V?
sx 
Qrsx
60 scpa(Trx Ts)
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.29
calculated as
And the supply temperature for the interior zone at part load is
The supply volume flow rate for the perimeter zone at part load is
The supply volume flow rate for the interior zone at part load is
At summer cooling mode part-load operation, the total supply volume flow rate is
Assume the following conditions:
 The sensible ratio of the perimeter zone at part-load SHRsxp drops to 0.91.
 The SHRs for the interior zone at part-load SHRsip remains the same.
 The temperature rise of the return system heat gain from the heat released by the light troffer in
the ceiling plenum Tretp, in °F (°C), is proportional to the air volume flow; that is, Tretp 
3/ (15,684 / 19,227)  3.7°F (2.1°C).
From the psychrometric chart, Trup  75 	 3.7  78.7°F (25.9°C).
If, at summer cooling mode part-load operation, the required amount of outdoor air remains the
same as at design load, the ratio of outdoor air to supply air at summer cooling mode part-load operation
is therefore
If the dry- and wet-bulb temperatures of outdoor air are the same as those at design conditions, then
from the psychrometric chart Tmp  83°F (28.3°C), and hmp  31.9 Btu / lb (74.2 kJ / kg). As Tcc is
raised to 53°F (11.7°C) because of the 3°F (1.7°C) reset in Twe, the enthalpy of conditioned air off
the coil at 53°F, 95 percent relative humidity is 21.6 Btu / lb (50.2 kJ / kg), and the coil load at summer
cooling mode part-load operation is
(21.9)
Winter Reheating in Perimeter Zone
If there is only one AHU or PU to serve both the perimeter and interior zones of a typical floor in a
VAV reheat system, the air discharged from the AHU or PU at winter design conditions using an air
economizer cycle must fulfill the cold air supply requirement of the interior zone, typically at 52°F
 60 15,684 0.075(31.9 21.6)  726,953 Btu / h (212,997 W)
Qccp  60V?sp s(hmp hcc)
rup-mp
rup-o

3840
15,684
 0.25
V?
sp  V?sxp 	 V?sip  9325 	 6359  15,684 cfm (7401 L / s)
V?
sip 
0.85 35,000
60 0.075 0.243(75 58.5)
 6359 cfm (3001 L / s)
V?
sxp 
0.5 285,500
60 0.075 0.243(75 61)
 9325 cfm (4400 L / s)
Tsip  50 	 2 	 3 
3
0.85
 58.5F (14.7C)
Tsxp  50 	 2 	 3 	
3
0.5
 61F (16.1C)
21.30 CHAPTER TWENTY-ONE
(11.1°C), as shown in Fig. 21.5b. If the temperature rise from supply duct heat gain is 2°F (1.1°C),
the temperature of supply air entering the reheating coil in a reheating VAV box Ten is 54°F
(12.2°C).
One of the primary considerations in the design of warm air supply VAV systems (including
VAV reheat system) at winter design conditions is the control of stratification when warm air is supplied
from overhead slot and ceiling diffusers in the perimeter zone. As discussed in Sec. 18.6, it is
recommended that the supply air temperature differential Tr Ts not exceed 15°F (8.3°C), to prevent
excessive buoyancy effects. A terminal velocity of 150 fpm (0.75 m / s) is required at the
5-ft (1.5-m) level to offset the cold draft from the external window. An even distribution of the volume
flow rate of supply air between the warm airstream blowing downward toward the window
glass and the warm airstream horizontally discharged toward the partition wall across from the window
is preferable.
At a warm air supply volume flow rate equal to 30 percent of the summer peak volume flow rate
of 1.5 cfm / ft2 (9.14 L / s 
m2), the space heating load per ft2 of floor area that can be offset by warm
air supply in the perimeter zone during winter design conditions is
If the depth of the perimeter zone is 15 ft (4.6 m), the heating load linear density qh,ft is
qh,ft  7.4 15  111 Btu /h
 ft (107 W/m)
The maximum reheating coil load for any control zone in the perimeter zone to offset the zone heating
load at winter design conditions Qchxn, in Btu / h (W), can be calculated as
(21.10)
where Qrhxn  space heating load for control zone n in perimeter zone, Btu /h (W)
Qvenxn  ventilation load due to a 30 percent of peak volume flow of cold supply
air, Btu /h (W)
s,cxn  volume flow rate of cold primary air discharged to control zone from AHU of VAV
reheating system, cfm [m3/ (60 s)]
Ten,n, Trxn  cold primary air temperature and zone temperature of control zone n, °F (°C)
For a summer peak volume flow rate of 1.5 cfm / ft2 (9.14 L / s 
m2) in the perimeter zone, at 30 percent
of peak volume flow that means 0.3 1.5  0.45 cfm / ft2 (2.3 L / s 
m2) of cold supply air.
If the zone heating load linear density qh,ft is 111 Btu /h
 ft (107 W/ m) and the zone heating load
density is 7.4 Btu /h
 ft2, the reheating coil load for 1 ft2 (0.0929 m2) of floor area at winter design
conditions can be calculated as
If the AHU that serves the perimeter zone in a VAV reheat system is separated from the interior
zone, and a minimum outdoor air intake of 15 percent of the peak supply volume flow rate can be
set up for the perimeter zone during the winter heating mode operation, as shown in the air conditioning
cycle in Fig. 21.1b, the reheating coil load is then
A saving of about 8.9 3.9  5.0 Btu /h
 ft2 (15.7 W /m2) is possible.
 11.3 Btu / h
 ft2 (35.6 W / m2)
Qchn  7.4 	 60 0.45 0.075 0.243(72 64)  7.4 	 3.9
 16.3 Btu / h
 ft2 (51.4 W / m2)
Qchn
A
 7.4 	 60 0.045 0.075 0.243(72 54)  7.4 	 8.9
V?
 Qrhxn 	 60V?s, cxn scpa (Trxn Ten, n)
Qchxn  Qrhxn 	 Qvenxn
qrh
A
 0.3 1.5 60 0.075 0.243 15  7.4 Btu / h
 ft2 (23.3 W / m2)
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.31
For the winter heating mode operation in the perimeter zone of a VAV reheat system, the following
is recommended:
 If there is only one AHU or PU to serve both the perimeter and interior zones in a VAV reheat system
and if the zone heating load linear density qh,ft 	 120 Btu /h
 ft (115 W/ m), then the warm air
supply volume flow rate from the reheating box in heating mode operation (at cold primary air
minimum setting) should be proportionally increased to prevent excessive buoyancy effects.
 It is more energy-efficient to use a perimeter heating VAV system; a dual-fan, dual-duct VAV
system; or a fan-powered VAV system; or else the AHU or PU that serves the perimeter zone is
separated from another AHU or PU serving the interior zone.
Winter Cooling Mode Operation in Interior Zone
At winter design conditions, the outdoor humidity ratio wo  0.0017 lb / lb (kg / kg), and the space
sensible cooling load in the interior zone may still equal that of summer design conditions. Therefore,
the supply temperature differential for the interior zone is Tr Ts  20°F, or Ts  72 
20  52°F (11.1°C). The supply volume flow rate for the interior zone is still  6173 cfm
(2913 L/ s).
If the same AHU or PU is used for both perimeter and interior zones, a minimum supply volume
flow rate of cold primary air that equals 0.3 of the peak supply volume flow rate (including the
required outdoor ventilation) is supplied to the perimeter zone during winter heating mode and can
be calculated as
The difference in humidity ratio between supply and zone air is
where wr, ws  humidity ratio of zone and supply air, lb/ lb (kg / kg). The temperature rise from the
return system heat gain at winter design conditions is then
and
If the temperature rise from the supply fan power is still about 2°F (1.1°C) and the temperature rise
from supply duct heat gain is 3°F (1.7°C) for the interior zone, then
Tm  Ts (2 	 3)  52 (2 	 3)  47°F (7.2°C)
and
From the psychrometric chart, at a temperature Tr  72°F (22.2°C) and r  30 percent,
wr  0.005 lb / lb (kg / kg). Draw line r-s from point r with SHRs  0.84, which intersects the

V?
o
V?
s
 0.55
Tru Tm
Tru To

79.7 47
79.7 20
Tru  Tr 	 Tret,w  72 	 5.7 	 2  79.7F (26.5C)
Tret,w 
3
(6173 	 3916) / 19,227
 5.7F (3.2C)

13,300 	 13,300
60 (6173 	 3916) 0.075 1061
 0.00055 lb / lb (kg / kg)
wr ws 
Qrl
60V?s shfg, o
V?
sx,c  0.3 13,054  3916 cfm (1848 L / s)
V?
sx,c
V?
si
21.32 CHAPTER TWENTY-ONE
72 20  52°F (10.0°C) constant-temperature line at point s, and ws  0.005 0.00055 
0.00445 lb / lb (kg / kg), as shown in Fig. 21.5b. Because Tru  79.7°F (26.5°C) and wru  wr 
0.005 lb / lb (0.005 kg / kg), draw a line o-ru. Because om/o-ru  0.55, wm  0.0032 lb / lb (kg / kg).
To maintain a space relative humidity of 30 percent at winter design conditions, a steam humidi-
fier is required with a humidifying capacity s, in lb /h (g /s), calculated as follows:
From the above analysis, when a VAV reheat system is used in an office building, these primary
problems may arise:
 Outdoor ventilation air supply becomes insufficient during part-load operation. This is discussed
in Chaps. 22 and 23.
 If an air economizer cycle is used for free cooling in the interior zone during winter cooling mode
operation, for locations where the outdoor humidity ratio wo drops below 0.003 lb / lb (kg / kg), the
zone relative humidity may fall below 25 percent, or even 20 percent. A steam humidifier may be
used, as shown in Fig. 21.5.
 If the same AHU or PU serves both perimeter and interior zones, free cooling is used for the interior
zone, and the mixed air in the perimeter zone contains the same percentage of outdoor air as
in the interior zone, then a humidifier may be required for the perimeter zone otherwise the space
relative humidity may even drop to 20 percent.
When a steam humidifier is installed in an AHU or PU, it is important that its humidifier capacity
be strictly controlled so that wet surfaces will not occur inside the AHU, PU, and supply ducts,
which may cause indoor air quality problems.
21.4 DUAL-DUCT VAV SYSTEMS
System Description
A dual-duct VAV system uses two supply air ducts—a warm air duct and a cold air duct—to supply
both warm air and cold air to various control zones in the perimeter zone. Warm air and cold air
are mixed in the mixing VAV box in order to maintain a predetermined zone temperature, as shown
in Fig. 21.9a. For control zones in the interior zone, only a cold air duct is used to supply cold air to
offset the year-round zone cooling load through VAV boxes.
A dual-duct VAV system can be a built-up system with a cold deck and a hot deck, or it can be a
combination of two air-handling units. Packaged units are seldom used for dual-duct VAV systems
because of their more complicated configuration and controls.
A dual-duct VAV system may condition and supply air separately to perimeter zones and interior
zones. It is more flexible because the supply air temperature and the variation of supply volume
flow rate can be matched better according to the zone load changes and outdoor ventilation air requirements.
It is also more energy-efficient than other VAV systems because of
 Transfer of internal heat gains from the interior zone to the perimeter zone through the recirculating
air
 Different outdoor air mixing ratio that can be used for interior and perimeter zones for free cooling
According to Warden (1996), a dual-duct VAV system is less expensive than a perimeter heating
VAV system and is also less expensive than a VAV hot-water reheat system depending upon the design
and configurations.
 60(6713 	 3916)(0.075)(0.00445 0.0032)  56.8 lb / h (0.0072 kg / s)
m?s  60V?s s(ws wm)
m?
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.33
21.34
21.35
FIGURE 21.9 A dual-fan dual-duct VAV system: (a) schematic diagram; (b) air conditioning cycle.
It is possible to use the warm air duct as a cold air duct in summer cooling mode operation so
that there is a warm/ cold duct and a cold duct. Such an approach may reduce the size of the cold
duct and the warm air leakages. However, changeover from cold air supply to warm air supply often
produces many problems if the warm/ cold duct serves control zones with different orientations in
the perimeter zone. This approach is not advisable.
Number of Supply Fans
If an air economizer cycle is used, a relief fan or a return fan must be installed to relieve the space air
when 100 percent outdoor air is used for free cooling. If the ceiling plenum is used as the return plenum,
and if each typical floor in a multistory building is served by one or two air-handling units, then the pressure
drop of the return system is often less than 0.6 in. WC (150 Pa) and a relief fan is a better choice
than a return fan. The comparison between a return fan and a relief fan is made in Chap. 22.
A dual-duct VAV system may be installed with a single supply fan that delivers to both warm
and cold ducts, or two supply fans, one for the warm duct and one for the cold duct. Although a single
supply fan is simpler and lower in initial cost, and the air-handling unit can be a factory-made
multizone unit rather than a field-assembled built-up AHU, a dual-fan design has the following
advantages:
 It allows the use of different ratios of outdoor air to supply air for the cold duct and the warm
duct.
 The warm air supply fan is operated only when it is required, such as during the warm-up period,
or in spring, fall, and winter when the transmission and infiltration loss is greater than the internal
and solar loads in any control zone in the perimeter zone.
 A dual-fan design simplifies control and saves energy. Different control schemes are possible for
cold and warm air temperature resets, as well as for cold and warm air duct static pressure
controls.
Dual-fan dual-duct VAV systems are widely used in current dual-duct VAV systems design.
Mixing VAV Box
Figure 21.10 shows a pressure-independent mixing VAV box using a terminal DDC controller. This
box consists of two separate, equal-size air passages arranged in parallel—one for warm air and the
other for cold air. Each has a single-blade inlet mixing damper. These two air passages are then
combined together, and the mixture of warm and cold air is discharged through a discharge volume
damper, and then to the diffusers through a flexible duct, single and multiple outlets, as shown in
Fig. 21.10a. The DDC terminal controller controls the inlet mixing dampers directly according to
the signal from the temperature sensor and controls the discharge volume damper from the signal
sent by the airflow sensor. The temperature sensor also resets the airflow set point in the DDC controller.
As in the VAV box, a pressure-independent mixing VAV box improves system stability and
performance and is widely used.
A currently available DDC mixing VAV box has the following controls: zone temperature and
reset, minimum and maximum limit, minimum outdoor air, and warm-up and cool-down controls.
Most DDC mixing VAV boxes use proportional plus integral (PI) control mode.
The year-round operation of a mixing VAV box can be divided into cooling mode, mixing mode,
and heating mode operations, as shown in Fig. 21.10b. The cooling mode has the greatest supply
volume flow rate, and the cold air supply volume flow rate sc, in cfm (L / s), reduces from 100 percent
peak supply volume flow rate at summer design conditions down to the minimum setting. In
the mixing mode operation, the reduction of the cold air supply volume flow rate exactly equals the
increase of the warm air supply volume flow rate; i.e., the total supply volume flow rate remains
constant and is equal to the minimum setting. In the heating mode operation, the maximum warm
V?
21.36 CHAPTER TWENTY-ONE
21.37
Mixing Cooling Heating
100
80
60
40
20
0
Tc,o Th,o
Cold
(a)
(b)
Peak supply volume flow rate, percentage
Warm
Inlet
mixing
dampers
Zone
temperature
sensor
Air flow
sensor
Discharge
volume
damper
1 3
F
T
2 2 1
FIGURE 21.10 A pressure-independent mixing VAV box: (a) schematic diagram; (b) cooling, mixing, and heating
mode.
air supply volume flow rate sw, in cfm (L / s), may be varied between 50 and 100 percent of the
peak cold air supply volume flow rate, depending on the heating load linear density qh,ft, in Btu /h
 ft
(W/ m), in the perimeter zone, so that the maximum zone supply air temperature differential during
heating mode operation does not exceed 15°F (8.3°C). The warm air supply volume flow rate reduces
as the zone heating load decreases until it is equal to zero in the mixing mode operation.
As in a VAV reheat system, the minimum setting of zone supply volume flow from the mixing
VAV box in a dual-duct VAV system should follow the guidelines for VAV systems specified in
ASHRAE/IESNA Standard 90.1-1999.
Mixing Mode Operation
In Fig. 21.10b, the mixing mode starts at an outdoor air temperature Th,o when the warm air supply
reduces to zero and ends at Tc,o when the cold air supply drops to zero, both in °F (°C). When outdoor
temperature To  Th,o, the cold air supply volume flow scx1, p, in cfm (L/S) is reduced to
a mixing-mode setting, such as, 30 percent of the cold peak volume flow rate. Similarly, when
To  Tc,o, the warm air supply is also reduced to mixing-mode setting. Consider a mixing VAV box
VXn serving a control zone rxn in the perimeter zone; its peak cold air supply volume flow rate is
13,054 / 10,000  1.3 cfm / ft2 (6.6 L / s 
m2), excluding the influence of warm air leakage. If the
maximum warm air supply volume flow rate is equal to 70 percent of the peak cold air supply volume
flow rate, also when the outdoor air temperature To  Th,o, if the cold deck discharge air temperature
Tc,dis has been reset to 58°F (14.4°C), for a duct heat gain equal to 1°F (0.6°C), the zone
sensible cooling load is then
Qrsxn  0.3 1.3 60 0.0075 0.243(75 59)  6.82 Btu /h
 ft2 (21.5 W /m2)
When outdoor air temperature To  Tc,o, the control zone rxn that the mixing VAV box serves has
the following characteristics:
 There is only warm air supply except cold air leakage.
 Zone temperature should be reset to a value for heating mode operation, such as 72°F (22.2°C).
 Heating from hot water, electric heater or gas heater is usually required.
When To  Tc,o, if the hot deck discharge air temperature Th,dis  80°F (26.7°C), the warm air supply
can offset a zone heating load
Qrhxn  0.3 1.3 60 0.075 0.243(80 72)  3.4 Btu /h
 ft2 (10.7 W /m2)
For a dual-duct VAV system with a mixing VAV box, there may or may not be a dead band between
the changeover from the cooling mode to heating mode through mixing mode operation, as shown
in Fig. 21.10b.
Zone Controls and Sequence of Operations of Dual-Fan Dual-Duct
VAV System
Consider a mixing VAV box VX1 and a VAV box VI1 in a dual-fan dual-duct VAV system, as shown in
Fig. 21.9, that serves the perimeter zone and interior zone of a typical floor in an office building with a
summer indoor temperature of 75°F (23.9°C) and a winter indoor temperature of 72°F (22.2°C). The
following are the zone controls and the sequence of operations of this dual-fan dual-duct VAV system:
1. When time-of-day clock signals the AHUs of the dual-fan dual-duct VAV system to “off,” the
mixing VAV box VX1 and the VAV box VI1 are closed.
2. When the clock signals the AHU in the on position, the zone temperature sensor T1 measures
the zone temperature Trxn in the perimeter zone. If Trxn exceeds a predetermined limit, say,
V?
V?
21.38 CHAPTER TWENTY-ONE
77°F (24.4°C), the AHUs and the mixing VAV box VX1 and VAV box VI1 are in cooling mode
operation. The DDC terminal controller signals VX1 and VI1 to peak airflow supply to the served
control zones, and the DDC system controller signals the AHU to reset discharge air temperature to
maximum cooling set point, say, 52°F (11.1°C).
3. If the outdoor air temperature To exceeds a temperature economizer set point, that is, To 	
75°F (23.9°C), the DDC system controller closes the outdoor and exhaust dampers to a minimum
opening to extract a minimum amount of outdoor ventilation air, fully opens the recirculating
damper in both hot and cold decks, and opens and modulates the two-way valve of the water cooling
coil in the cold deck to maintain a required discharge air temperature Tdis.
At summer design conditions, the refrigeration capacity needed to produce chilled water is 100
percent full capacity. At the same time, the zone temperature sensors sense the zone temperature
Trx1 and Tri1, and the airflow sensors measure the airflow at the discharge outlet of the mixing VAV
box in the perimeter zone and the VAV box in the interior zone. In the control zone X1 of the
perimeter zone, the DDC terminal controller modulates and controls the inlet mixing dampers according
to the sensed temperature signal and the reset signal from the airflow sensor to maintain a
required zone temperature of 75°F (23.9°C). The cold air supply volume flow rate is gradually reduced
as the zone sensible cooling load decreases.
In the control zone I1 of the interior zone, the DDC terminal controller modulates the damper
opening of the VAV box VI1 to maintain a zone temperature of 75°F (23.9°C). The cold primary air
supplied from the VAV box is gradually reduced as the zone sensible cooling load decreases at cooling
mode part-load operation. In the cold deck, the discharge air temperature Tc,dis  52°F
(11.1°C); and in the hot deck, the discharge air temperature Th,dis  81°F, which is the mixing temperature
of recirculating air (85 percent of airflow) at 78°F (25.6°C) and an outdoor air temperature
(15 percent of airflow) of 95°F (35°C) at summer design conditions. The heating coil in the hot
deck is not energized.
4. When 50°F  To  75°F (10.0°C  To  23.9°C), the DDC system controller fully opens
the outdoor and exhaust dampers and closes the recirulating damper to extract 100 percent outdoor
air for free cooling in the cold deck. Refrigeration is at a capacity significantly less than the fullload
design capacity and is still required to maintain a cold deck discharge air temperature. In the
hot deck, the outdoor damper remains in the minimum opening position, and the recirculaing
damper is fully opened to provide a warm air supply.
In the mixing VAV box, the terminal DDC controller modulates the inlet mixing dampers and
the discharge volume damper according to the sensed zone temperature and airflow, to maintain a
required zone temperature of 75°F (23.9°C). The volume flow rate of cold air supply further reduces
as the space cooling load decreases.
The cold-deck discharge air temperature remains at 52°F (11.1°C) in order to meet the requirement
of a 15 to 17°F (8.3 to 9.4°C) supply temperature differential in the interior zone. However,
the cold supply air temperature from the mixing box may be 3 to 5°F (1.7 to 2.8°C) higher because
of the duct and fan power heat gains at part load. The hot deck discharge air temperature may be
varied from 80 to 76°F (26.1 to 24.4°C) because of the variation of the outdoor air temperature. The
heating coil in the hot deck is not energized.
5. If To  Th,o  50°F (10.0°C), the DDC system controller positions the outdoor and recirulating
dampers in both the cold deck and hot deck, first, to provide an outdoor air intake at a value not less than
the minimum ventilation requirment and, second, to maintain a mixing temperature Tm  50°F (10°C) if
it is possible. The DDC system controller also closes the two-way valve of the water cooling coil in the
cold deck of the AHU, so that no refrigeration is required for cooling mode operation.
At To  Th,o, when the cold air supply in the mixing VAV box s,cx1 is reduced to mixing mode
setting 30 percent of the peak cold air supply volume flow rate scx1,d, the mixing VAV box is in
mixing mode operation. The mixing operation ends at the warm air supply shx1  0.3 shx1,d. In
mixing mode, the DDC terminal controller for control zone X1 in the perimeter zone
 Gradually resets the zone temperature from Trx1  75°F (23.9°C) at s,cx1  0.3 scx1,d (To  Th,o)
to Trx1  72°F (22.2°C) at shx1  0.3 scx1,d (To  Tc,o) V?
V?
V?
V?
V?
V?
V?
V?
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.39
 Modulates the discharge volume damper of the mixing VAV box so that the total supply volume
flow rate is maintained at mixing mode setting 30 percent of the peak cold air supply scx1,d during
the mixing mode operation
The cold deck discharge air temperature Tc,dis is still 52°F (11.1°C). If the cold air supply in the
perimeter zone drops nearly to its minimum setting (30 percent of the peak supply volume flow
rate), the duct and fan power heat gain may have a temperature rise of 10°F (5.6°C) which raises
the cold supply air temperature in the perimeter zone to about 60°F (15.6°C).
If the hot-deck discharge air temperature Th,dis cannot be maintained at a value 5°F (2.8°C)
higher than the zone temperature Trxn, the DDC system controller opens the two-way valve of the
hot water coil and maintains a Th,dis  Trxn 	 5. And Th,dis increases as To falls.
The relationship between the outdoor air temperature To and Th,o as well as To and Tc,o depends
on the characteristics of the zone load and the construction of the building shell.
6. When the outdoor air temperature To  Tc,o, the mixing VAV box is in heating mode operation.
In heating mode operation, the DDC system controller positions the outdoor and recirculating
dampers to provide required minimum ventilation as well as to maintain a mixture temperature of
50°F (10°C) if possible. The cold deck discharge air temperature remains at 52°F (11.1°C). Also the
two-way valve of the water cooling coil remains closed.
If the duct heat loss is about 2°F (1.1°C), the hot deck discharge air temperature is equal to
Th,dis  72 	 15 	 2  89°F (31.7°C). The DDC system controller modulates the two-way valve
of the hot water coil in the AHU to maintain Th,dis  89°F (31.7°C).
In the mixing box X1 in the perimeter zone, the DDC terminal controller modulates the inlet
mixing dampers and the discharge volume damper according to the sensed zone temperature and
airflow signal, to maintain a preset zone temperature of 72°F (22.2°C). As the zone heating load
increases, the warm air supply volume flow rate increases accordingly to a maximum value at winter
design conditions with a warm air supply temperature differential not exceeding 15°F (8.3°C).
Discharge Air Temperature Control
In general, the discharge air temperature from the cold deck must meet the year-round cooling requirements
for the interior zone. An air economizer cycle using free cooling should be applied to
the cold deck whenever possible, to meet the outdoor air requirements and to minimize refrigeration
in order to save energy. In the hot deck, internal heat gains carried by the recirculating air
should be fully utilized.
Cold deck discharge air temperature Tc,dis is often reset according to zone demands at part load
or outdoor air temperature by means of a DDC unit controller. Hot deck discharge air temperature
Th,dis is often reset by outdoor temperature during winter heating mode.
Figure 21.11 shows the cold and warm deck discharge air temperature from the AHU of a
typical dual-fan dual-duct VAV system at various outdoor temperatures. The cold deck discharge
air temperature Tc,dis is maintained at 52°F at all outdoor air temperatures, in order to
meet high zone cooling loads in the interior zone. However, when the cold air supply volume
flow rate in the perimeter zone reduces to about 30 percent of its peak volume flow during
part-load operation, the temperature rise because of the fan power and duct heat gain may increase
to 10°F (5.6°C), which results in a supply air temperature of cold air up to 60°F
(15.6°C).
When the outdoor air temperature To drops from 70 to 60°F (21.1 to 15.6°C), the mixing of minimum
outdoor air and recirculating air lowers the hot deck discharge air temperature Th,dis from 80
to 76°F (26.7 to 24.4°C). If To  50°F (10.0°C), heating may be required in the perimeter zone, so
Th,dis increases as To falls. If the duct heat loss is 2°F (1.1°C), the warm air supply temperature differential
Th,dis 2 Trxn should not exceed 15°F (8.3°C) even at winter design conditions, in order
to avoid stratification.
V?
21.40 CHAPTER TWENTY-ONE
Zone Supply Volume Flow Rate
For each control zone in the perimeter zone served by a mixing VAV box, an air leakage of 0.03 to
0.07 of its peak supply volume flow should be considered for the shut off damper. If the difference
in air density between warm and cold air is ignored, the additional supply volume flow rate lk required
to compensate for the air leakage when the warm air damper is shut, in cfm (L / s), can be
calculated as
(21.11)
where Ts,h, Ts,c  temperature of warm and cold air supply at mixing VAV box, °F (°C)
Trxn  zone temperature, °F (°C)
As in a VAV reheat system, for any control zone in either the perimeter or interior zones, the
mixing VAV box, VAV box, flexible ducts, and slot or ceiling diffusers should be sized according to
the peak supply volume flow rate. If there is no air leakage, the control zone peak supply volume
flow rate , in cfm [m3 / (60 s)], is
(21.12) V?rxn 
Qrsxn
60 scpa(Trxn Ts,c)
V?
rxn
V?
lk  0.05V?sxn
Ts,h Trxn
Trxn Ts,c
V?
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.41
Air economizer
Outdoor air mixing,
refrigeration turned off
Refrigeration
Cooling and mixing modes Heating mode
Discharge air temperature Tdis, F
Hot deck
discharge air temperature
Cold deck
discharge air temperature
40
0
Outdoor temperature To, F
60
80
100
120
20 40 60 80 100
FIGURE 21.11 Warm and cold deck discharge air temperature in a dual-fan dual-duct VAV system.
where Qrsxn  control zone maximum sensible cooling load, Btu /h (W). The control zone peak
supply volume flow rate for cold air in the perimeter zone at summer design conditions is
(21.13)
For any control zone in the perimeter zone, the peak supply volume flow rate delivered to the
conditioned space from the mixing VAV box, including 5 percent air leakage from the warm air
damper at summer design conditions, in cfm (L / s), is therefore
(21.14)
Because only cold air from the VAV box is supplied to the interior zone, for any control zone in the interior
zone, the peak supply volume flow rate at summer design conditions, in cfm (L / s), is then
(21.15)
Peak supply volume flow rate for control zone n, , can be calculated by Eq. (21.12).
Case Study: A Dual-Fan Dual-Duct VAV System
Consider a dual-fan dual-duct VAV system serving a typical 20,000-ft2 floor in an office building
with the same operating parameters as the VAV reheat system described in Sec. 21.3, except the
winter outdoor design temperature is 30°F (1.1°C) and the outdoor humidity ratio is 0.003 lb / lb
(kg / kg). As in the VAV reheat system, the maximum supply volume flow rate for the interior zone
at summer design conditions based on block load is
For a control zone in the perimeter zone, if the temperature of the mixture of outdoor and recirculating
air Tmx  80.5°F (26.9°C), humidity ratio wmx  0.010 lb / lb (kg / kg), Tsx  Ts,c  55°F
(12.8°C), and sx  s,c  80 percent, draw line sx-mx as shown in Fig. 21.9b.
For a warm air leakage of 0.05 from the inlet damper, the state point of the mixture of cold
air supply and warm air leakage sxn can be determined. From the psychrometric chart, Tsxn 
56.5°F (13.6°C), and wsxn  0.0083 lb / lb (kg / kg).
The volume flow rate of cold air supply to the perimeter zone at summer design conditions, based
on the block load and including 5 percent air leakage from the warm air inlet damper, is therefore
The total cold air supply volume flow rate to both the perimeter zone and the interior zone at summer
design conditions is
(21.16)
The volume flow rate of warm air supply in the perimeter zone is usually expressed as a percentage
of peak supply volume flow rate of cold air in the perimeter zone, usually between 0.5 and
1.0 . It depends mainly on the space heating load linear density qh,ft, in Btu /h
ft (W / m), and the V?scx
V?
scx
 14,113 	 6713  20,826 cfm (9828 L / s)
V?
s,c  V?scx 	 V?sci
V?
scx 
285,500
60 0.075 0.243(75 56.5)
 14,113 cfm (6660 L / s)
V?
sxn

135,000
60 0.075 0.243(75 55)
 6173 cfm (2913 L / s)
V?
si 
Qrs
60 scpa(Tr Ts,c)
V?
sin
V?
sin  V?rin
V?
sin
V?
sxn 
V?
cxn
0.95
V?
sxn
V?
cxn  V?rxn 	 V?lk
21.42 CHAPTER TWENTY-ONE
supply temperature differential Ts Tr). If the volume flow rate of the warm air supply in the
perimeter zone at winter design condition , in cfm (L / s), is 0.7 , then
Other parameters such as cooling and heating coil loads can be calculated as for VAV reheat systems.
Winter Heating and Winter Cooling Mode Operation
In a dual-fan dual-duct VAV system, the cold deck and the hot deck have their own outdoor and recirculating
air supply. In the hot deck during winter heating mode operation, it is possible to have
an outdoor air ratio of only about 22 percent and a mixing temperature of outdoor and recirculating
air Tmx  62°F (16.7°C) for the perimeter zone, so that only necessary heating energy is provided
for a hot deck discharge air temperature of 87°F (30.6°C), as shown in Fig. 21.9b. In the cold deck
winter cooling mode operation, nearly 60 percent outdoor air is required to form a mixture of outdoor
and recirculating air with a temperature Tmi of about 50°F (10.0°C) and a cold deck discharge
air temperature of 52°F (11.1°C), so that outdoor air free cooling can be fully utilized to offset the
zone cooling load in the interior zone in winter.
Part-Load Operation
Consider a typical control zone in the perimeter zone of this dual-fan, dual-duct VAV system whose
operating parameters at summer design condition and summer part-load operation are as follows:
If the supply air temperature discharged from the mixing VAV box Tsxn  Tm2  56.5°F (13.6°C),
the volume flow rate of conditioning air supplied to this control zone , in cfm (L / s), including
5 percent warm air leakage from the damper at summer design conditions, is therefore
During part-load operation, as the zone sensible cooling load is reduced to 30 percent of the design
load, Tsxnp  50 	 10  60°F (15.6°C). Assuming that the percentage of warm air leakage and the
temperature rise due to this warm air leakage are the same at design condition, the supply temperature
of the mixture of cold and warm air of control zone xn in the perimeter zone at part-load Tsxnp is
now equal to 50 	 10 	 1.5  61.5°F (16.4°C).
The supply volume flow rate from the mixing VAV box in the perimeter zone at part-load can be
calculated as
(21.17)

4260
60 0.075 0.243(75 61.5)
 289 cfm (136 L / s)
V?
sxnp 
Qrsxnp
60 cpa (Trxn Tsxnp)
V?
sxn 
14,200
60 0.075 0.243(75 56.5)
 702 cfm (331 L / s)
V?
sxn
Full load Part load
Zone sensible cooling load, Btu /h 14,200 4,260
Leaving cooling-coil temperature, °F 50 50
Relative humidity, percent 95 95
Supply system heat gain, °F (°C) 5 (2.8) 10 (5.6)
Warm air supply temperature, °F (°C) 87 (30.6) 80 (26.7)
V?
shx  0.7 14,113  9879 cfm (4661 L / s)
V?
scx V?shx
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.43
From Eq. (21.14), the volume flow rate of cold air supply at part load can be calculated as
Also the warm air supply is
Figure 21.12 shows the air conditioning cycle of a typical control zone in the perimeter zone
served by a dual-fan dual-duct VAV system at summer cooling mode part-load operation. From the
psychrometric chart, zone relative humidity at summer part-load operation is higher than at fullload
condition. This is mainly because of the increase of the cold deck discharge air temperature
due to the duct heat gain and the possible chilled water reset at part load.
At winter heating mode part-load operation, the state point of supply air will move along the
horizontal line mx-hc as shown in Fig. 21.9b, to maintain a predetermined zone temperature.
21.5 FAN-POWERED VAV SYSTEMS
System Description
A fan-powered VAV system, as shown in Fig. 21.13, often employs fan-powered boxes in each
control zone of the perimeter zone and VAV boxes in each control zone of the interior zone in a
V?
s,hn  289 275  14 cfm (6 L / s)
V?
s,cn  0.95 289  275 cfm (130 L / s)
21.44 CHAPTER TWENTY-ONE
FIGURE 21.12 Air conditioning cycle of a typical control zone at summer part-load operation served by a mixing
VAV box of a dual-fan dual-duct VAV system.
21.45
FIGURE 21.13 A parallel fan-powered VAV system: (a) schematic diagram; (b) air conditioning cycle; (c) details of winter heating mode air conditioning cycle; (d) cold primary air
and heating coil output diagram.
21.46
FIGURE 21.13 (Continued)
building. Both fan-powered boxes and VAV boxes can provide cooling for the perimeter and interior
zones. The function and advantages of a fan-powered box include the following:
 It extracts warm recirculating air from the ceiling plenum to maintain a preset zone temperature
when the volume flow of the cold primary air supplied from the fan-powered box has dropped to the
minimum setting or other preset value at summer cooling part-load operation. This is the first step.
 If the zone temperature drops further in any of the control zones in the perimeter zone, a reheating
coil in the fan-powered box is energized to maintain predetermined zone temperature during
heating mode operation. This is the second step.
 It can increase the warm air supply volume flow rate to meet the requirement that the warm air
supply temperature differential does not exceed 15°F (8.3°C), in order to prevent excessive buoyancy
effects.
 It mixes the low-temperature primary air with warm plenum air to prevent the dump of a cold air
jet and possible surface condensation in cold air distribution.
If the fan-powered box is used for mixing low-temperature primary air with warm plenum air in
cold air distribution, fan-powered boxes can also be installed in the control zones of the interior
zone. The disadvantages of using a fan-powered box instead of a VAV box include higher energy
use, higher noise level, and greater initial cost and maintenance.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.47
Cooling Heating Dead band
100%
Output 100%
Minimum
0
0
72
75 Tr, F
Tr, F
Heating coil
Cold primary air
(d)
Vs,
cfm
.
FIGURE 21.13 (Continued)
Fan-Powered VAV Boxes
A fan-powered VAV box, or simply a fan-powered box, consists of a small forward-curved centrifugal
fan, a VAV box, a heating coil (hot water or electric heater), a low-efficiency disposable filter,
dampers, an outer casing, and corresponding controls, as shown in Fig. 21.14. Fan-powered VAV
boxes can be classified as series and parallel boxes according to their construction and operating
characteristics.
Series Fan-Powered Box. In a series fan-powered box, a small centrifugal fan is connected in series
with the cold primary airstream flowing through the VAV box, as shown in Fig. 21.14a. The
volume flow of the cold primary air is varied at part-load operation, but after this air mixes with the
induced plenum recirculating air, the fan delivers a continuously operated, nearly constant-volume
flow rate of the mixture of cold primary air and plenum recirculating air. The mixture is then supplied
to the control zones, either heated at the heating coil or without heating depending upon the
zone load.
Because the airflow from the series fan-powered box is constant, it is easier to select ceiling diffusers
to achieve good space air diffusion. In a series fan-powered box, cold primary air is thoroughly
mixed with plenum recirculating air. The main disadvantage of a series fan-powered box is
its larger fan, which consumes more energy than a parallel fan-powered box.
Parallel Fan-Powered Box. In a parallel fan-powered box, the induced recirculating plenum
airstream discharged from the centrifugal fan is connected in parallel with the cold primary
airstream discharged from the VAV box in the fan-powered box. These two airstreams are mixed either
before or after the heating coil and then are supplied to the control zones.
A parallel fan-powered box with a heating coil located upstream from the fan is called a drawthrough
parallel fan-powered box, as shown in Fig. 21.14b. When the heating coil is located downstream
from the fan, the box is called a blow-through parallel fan-powered box, shown in
Fig. 21.14c. The air temperature entering the heating coil of a draw-through unit is the temperature
of recirculating plenum air, which is higher than that of the mixture of cold primary air and recirculating
plenum air in a blow-through box.
A parallel fan-powered box operates only when there is a need to mix the warm plenum air with
the cold primary air at summer cooling mode part-load operation, when the volume flow of the cold
primary air is at its minimum setting or other preset value, or when the mixture of cold primary air
and warm plenum air is heated at the heating coil to offset the zone winter heating load. When a
parallel fan-powered box is used to mix 40 to 50°F (4.4 to 10.0°C) low-temperature primary air
with warm plenum air in a cold air distribution system, it is also operated continuously during summer
cooling mode.
In a parallel fan-powered box when the cold primary airstream is reduced to its minimum setting
during cooling mode part-load operation, the volume flow rate of the extracted plenum recirculating
air plus the cold primary air is usually equal to 0.80 to 1.0 of the zone summer peak supply volume
flow rate . The fan in a parallel fan-powered box extracts only recirculating plenum air, and
therefore it has a smaller volume flow rate and a lower fan total pressure than that in a series fanpowered
box. Because of its smaller fan size, it consumes less fan power in cooling mode
operation.
Current parallel fan-powered box designs also provide a good mixture of cold primary air and
extracted plenum air. Some of the recent models use a variable-speed small parallel fan. It energizes
as soon as the supply volume flow of the cold primary air drops at cooling mode part-load operation.
The extracted recirculating plenum air gradually increases as the volume flow of cold primary
air decreases up to a maximum limit between 0.4 and 0.5 . Under this condition, the zone
supply volume flow rate is kept nearly constant. Parallel fan-powered boxes are often used in cold
air distribution because of their lower initial cost and lower fan energy consumption.
In a parallel fan-powered box, a backdraft damper should be installed upstream from the fan.
This damper is normally closed to prevent the backward flow of cold primary air in case the fan is
deenergized. When the fan is turned on, the backdraft damper is fully opened.
V?
sn,d V?sn,d
V?
sn,d
21.48 CHAPTER TWENTY-ONE
21.49
FIGURE 21.14 Fan-powered boxes: (a) series box; (b) parallel box, draw-through;
(c) parallel box, blow-through.
The drawbacks of a parallel fan-powered box include the need for a careful balance of flow and
total pressure of two airstreams, the difficulty in selection of the size of the box, and its more complicated
controls.
The size and capacity of a fan-powered box is determined by the capacity of the fan, the VAV
box, and the heating coil. The heating coil in a fan-powered box can be a hot water coil or an electric
duct heater located downstream from the fan-powered box. Fan-powered boxes are usually
built in sizes corresponding to the associated VAV box primary air volume flow rates from 300 to
3200 cfm (150 to 1600 L / s) and fan supply volume flow rates of 250 to 4000 cfm (125 to 2000
L/ s). Most fan-powered boxes have a fan total pressure between 0.5 and 0.8 in. WC (125 and 200
Pa) and an external pressure loss of about 0.25 in. WC (63 Pa) to offset the pressure drop of downstream
flexible ducts and diffusers. The pressure drop of the VAV box and the corresponding pressure
drop of flexible ducts and diffusers due to cold primary air should be offset by the supply fan
in the primary air AHU or PU.
The speed of the centrifugal fan in the fan-powered box can be adjusted to high, medium, or low
or even variable-speed drive. Both radiated and duct-borne sound paths should be checked to determine
whether they can meet the space air required NC criteria. Because the fan total pressure is
higher in a fan-powered box than in a fan-coil unit, a temperature rise due to fan power heat gain of
0.75°F (0.4°C) should be considered in psychrometric analysis.
The VAV box a in fan-powered box can be either pressure-dependent or pressure-independent.
Pressure-independent fan-powered boxes are widely used.
Fan Characteristics in Parallel Fan-Powered Boxes
In a fan-powered VAV system using a parallel fan-powered box, two fan-duct systems are connected
in parallel: the cold primary airstream discharged from the supply fan F1 in the AHU or
PU and the recirculating plenum airstream extracted by fan F2, as shown in Fig. 21.15. At the
combining point A, the pressures of these two airstreams must be equal, and the volume flow rate
is the sum of the volume flow rates of the cold primary air and the recirculated plenum air
.
For a control zone in the perimeter zone using a parallel fan-powered box at its zone peak supply
volume flow rate with conventional air distribution (a supply air temperature of 55°F, or
12.8°C), in cfm (L / s), the distribution is as follows:
Zone supply volume flow rate of control zones
at summer design conditions
Zone supply volume flow of recirculating
plenum air from parallel fan-powered box 0
According to Yu et al. (1993), the zone supply volume flow rates of control zones in the interior
zone using a parallel fan-powered box with cold air distribution at summer design conditions are as
follows:
Supply volume flow rate of 44°F (6.7°C) cold
primary air, cfm (L / s)
Supply volume flow rate from fan-powered
box (recirculating plenum air), cfm (L / s)
Zone supply volume flow rate for control zones in the perimeter zone at winter heating design conditions
must meet the following requirements:
Cold primary air for ventilation
Supply temperature differential Ts Tr  15°F (8.3°C)
0.3V?sn,d
0.4V?sn,d
0.6V?sn,d
1.0V?sn,d
V?
sn,d
V?
sn,r
V?
sn,c
21.50 CHAPTER TWENTY-ONE
In a fan-powered VAV system, the fan total pressure of F2 in Fig. 21.15a is about 0.5 in. WC
(125 Pa), and the pressure drop across the flexible duct and the slot diffuser may be around 0.25 in.
WC (63 Pa).
During summer cooling mode operation, when the zone sensible cooling load is reduced to
about one-half of the design load, the volume flow rate of the cold primary air may be reduced from
0.6 to 0.3 . As the supply volume flow through the flexible duct and slot diffuser is reduced
accordingly, the pressure drop between points A and o also decreases. The system pressure drop of
the recirculating plenum air passage becomes smaller, and a new system curve S2 is formed, as
shown in Fig. 21.15b. The operating point of fan-duct system F2 and S2 moves from point P to
point Q. The result is a greater volume flow rate of the recirculating plenum air than that at
summer design conditions, and a smaller zone supply volume flow rate flowing through the flexible
duct and slot diffusers.
V?
sxn,r
V?
sn,d V?sn,d
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.51
FIGURE 21.15 Fan operating characteristics of a parallel fan-powered VAV system: (a) pressure characteristics;
(b) fan-duct system characteristics.
Zone Control and Sequence of Operations of a Fan-Powered VAV System
with Parallel Fan-Powered Box
Consider a control zone rx1 in the perimeter zone served by a parallel fan-powered box FBX1 that
mixes low-temperature cold primary air with warm plenum air, and a control zone ri1 in the interior
zone in which the cold primary air is directly supplied from the VAV box, as shown in Fig. 21.13.
Because the temperature of cold primary air from the AHU that serves fan-powered box FBX1 and
VAV box VI1 Tc,p  44°F (6.7°C), the summer indoor space relative humidity drops to about
40 percent. The summer indoor design temperature is still maintained at 75°F (23.9°C). At winter
design conditions, the zone temperature is maintained at 72°F (22.2°C) with a relative humidity of
20 to 30 percent.
1. When the time-of-day clock signals the AHU to “off,” the AHU and the fan-powered box
FBX1 and VAV box VI1, as shown in Fig. 21.13a, are deenergized.
2. When the clock signals the AHU in the on position, if the zone temperature sensed by the
temperature sensor T11, Trxn 	 75°F (23.9°C), then the AHU and fan-powered box FBX1 and VAV
box VI1 call for cooling. The DDC system controller then opens the two-way valve of the water
cooling coil and resets the discharge air temperature from the AHU Tdis  44°F (6.7°C). The DDC
terminal controllers also open the single-blade volume dampers in the fan-powered boxes and VAV
boxes to the maximum cooling position.
If the outdoor temperature To 	 75°F (set point of the temperature economizer cycle), the DDC
system controller also positions the outdoor damper to a minimum opening position and fully opens
the recirculating damper.
3. In the perimeter zone, the control zone temperature Trx1 is sensed by sensor T11. The DDC
terminal controller modulates the volume damper and thus the volume flow of cold primary air supply
in the fan-powered box FBX1 to maintain a preset zone temperature of Trx1  75°F (23.9°C)
with a PI control mode. The fan in the fan-powered box FBX1 is not energized.
The control zone temperature in the interior zone Trin is sensed by sensor T21. Another DDC
terminal controller modulates the volume damper in the VAV box VI1 to maintain a preset zone
temperature Tri1  75°F (23.9°C).
4. When 52°F  To  75°F (11.1°C  To  23.9°C), the DDC system controller fully opens
the outdoor air damper and closes the recirculating damper to use 100 percent outdoor air free cooling
as the first-stage cooling in the AHU. Based on the measured discharge air temperature by sensor
T1, the DDC system controller modulates the two-way valve of the water cooling coil as the secondstage
cooling simultaneously to maintain a preset discharge air temperature Tdis  44°F (6.7°C).
When the outdoor temperature To drops from 65 to 52°F (18.3 to 11.1°C), Tdis is reset inversely
proportional by the DDC system controller from 44 to 52°F (6.7 to 11.1°C). At the same time, the
DDC terminal controllers modulate the volume damper as well as the volume flow of cold primary
air supply to maintain a preset zone temperature of 75°F (23.9°C) in the control zones rx1 and ri1.
The fan in the fan-powered box FBX1 is not energized.
5. When 52°F  To  75°F (11.1°C  To  23.9°C) and the supply volume flow rate of the
cold primary air from the fan-powered box FBX1 in the perimeter zone has dropped to 0.40 ,
the fan in the fan-powered box FBX1 is energized in low-speed operation and extracts recirculating
ceiling plenum air to the fan-powered box, and the supply volume flow rate of recirculating plenum
air to the control zone rx1 is roughly increased to about .
If the volume flow rate of the cold primary air supplied from the fan-powered box FBX1 has
dropped to 0.35 , the fan in the fan-powered box is energized in high-speed operation, and its
supply volume flow of recirculating plenum air is further increased to 0.45 .
6. When 52°F  To  75°F (11.1°C  To  23.9°C) and the supply volume flow rate of the
cold primary air from the fan-powered box FBX1 has dropped to the minimum setting, that is,
0.3 , if the control zone temperature Trx1 sensed by sensor T11 drops below 75°F (23.9°C), that
is, 72°F  Trx1  75°F (22.2°C  Trx1  23.9°C), then the fan-powered box FBX1 is in dead-band
operation, as shown in Fig. 21.13d. In dead-band operation
V?
sn,d
V?
sn,d V?sn,r
V?
sn,d
0.40V?sn,d V?sn,r
V?
sn,d
21.52 CHAPTER TWENTY-ONE
 The supply volume flow rate of the cold primary air is at the minimum setting.
 The fan in the fan-powered box FBX1 is still energized.
 The heating coil is not energized.
7. When To  75°F (23.9°C) and the supply volume flow rate of cold primary air from the fan-powered
box FBX1 has dropped to the minimum setting, if the control zone temperature Trx1 sensed by sensor
T11 drops below 72°F (22.2°C), the fan-powered box is in heating mode operation. In heating
mode operation, the extracted recirculating plenum air is heated at the heating coil, flows through the
centrifugal fan, mixes with the cold primary air, and is then supplied to the control zone, as shown in
Fig. 21.13b. The DDC terminal controller opens and modulates the mass flow rate of hot water entering
the heating coil to maintain a control zone temperature T  72°F (22.2°C) in reverse-acting mode.
The operation of the fan in a DDC fan-powered box can be arranged in different alternatives.
Supply Volume Flow Rate
According to Eq. (20.69), at summer design conditions, the volume flow rate of cold primary air supplied
to a control zone with conventional air distribution in cfm [m3 / (60 s)], can be calculated as
where Qrsn  zone peak sensible cooling load, Btu /h (W) and
At summer design conditions, the volume flow rate of cold primary air supplied to a control zone
with cold air distribution in cfm [m3 / (60 s)], can be calculated as
where Qrsn  zone peak sensible cooling load, Btu /h (W)
K  a factor
Tr Ts,c dis  32°F (17.8°C)
At winter design conditions, the volume flow rate of the mixture of cold primary air and recirculating
plenum air , in cfm [m3 / (60 s)], supplied to a control zone in the perimeter zone can be calculated
as
(21.18)
where Qrnh  zone heating load, Btu /h (W)
Ts,h  supply temperature at winter design conditions, °F (°C)
Warm air supply temperature differential Ts,h Tr should not exceed 15°F (8.3°C) in order to prevent
buoyancy effects.
Supply volume flow rate of a control zone in the perimeter zone served by a fan-powered VAV
system with cold air distribution is shown in Fig. 21.16. In Fig. 21.16: , indicate supply
volume flow rate of cold primary air and recirculated plenum air to the perimeter control zone, in
cfm (L / s), and Ts,n the supply air temperature to the perimeter control zone xn, in °F (°C).
As in the VAV reheat system, the cooling and heating coil load can be similarly calculated by
Eqs. (21.8), (21.9), and (21.10).
V?
sn,r V?sn,c
V?
sn,h 
Qrnh
60 scpa(Ts,h Tr)
V?
sn,h
V?
sn,c dis  KV?sn,d 
Qrsn
60 scpa(Tr Ts,c dis)
V?
sn,c dis
Tr Ts  20F (11.1C)
V?
sn,d 
Qrsn
60 scpa(Tr Ts)
V?
sn,d
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.53
Fan Energy Use
In a fan-powered VAV system, there are two kinds of fan energy use: the supply fan energy use in
an AHU or a PU and the energy use of the small centrifugal fans in the fan-powered boxes. Typical
supply fan efficiency is about 60 percent, and the motor efficiency is around 90 percent, which
gives a combined fan and motor efficiency of about 55 percent.
According to Elleson (1993), the efficiency of fans in fan-powered boxes from four manufacturers
ranged from 16 to 35 percent, and the motor efficiency ranged from 45 to 55 percent. Assuming
a fan efficiency of 30 percent and a motor efficiency of 50 percent, the combined fan and motor effi-
ciency for the fan in the fan-powered box is 15 percent.
Elleson (1993) has collected detailed data for the 1988 cooling season from the fourth floor of a
12-story building in southern California using a fan-powered VAV system (parallel fan-powered
boxes). The results are as follows:
Intermittent operation indicates that fans in parallel fan-powered boxes operated only when the cold
primary airflow fell below 50 percent of design flow.
55°F (12.8°C) Supply air 45°F (7.2°C) Supply air
Fan-powered Fan-powered
Supply fan boxes Total Supply fan boxes Total
Continuous operation, kWh 11,720 7,000 18,720 7,794 7,000 14,794
Intermittent operation, kWh 11,720 1,861 13,581 7,794 2,140 9,934
No fan-powered boxes, kWh 11,720 0 11,720 8,417 0 8,417
21.54 CHAPTER TWENTY-ONE
Percentage of zone peak (design) supply flow
Cooling
Supply temp Tsxn, F
Heating
Dead
band
0
20
40
60
80
80
60
40
Vsn, c
Vsn, c
Tsn
Vsn, r 	
. .
Vsn, r
.
.
FIGURE 21.16 Supply volume flow rate of a control zone in the perimeter zone served by a fan-powered VAV system.
If series fan-powered boxes and parallel fan-powered boxes are both operated continuously, the
energy use of the supply fan plus the fans in the parallel fan-powered boxes consume less than onehalf
of the kWh used by the supply fan plus the fans in the series fan-powered boxes in 45°F
(7.2°C) cold air distribution.
Design Considerations
 Because there are two mixings (the mixing of outdoor air and recirculating air in the AHU or PU
and the mixing of cold primary air and plenum air in the fan-powered box), adequate supply of
outdoor air is critical in a fan-powered VAV system. This is discussed in Chap. 23.
 In a VAV system using cold air distribution, fan-powered boxes can be installed in both the
perimeter and interior zones, or in the perimeter zone only, for lower initial cost. Cold primary air
is supplied directly to the interior zone by high-induction ceiling diffusers, as shown in
Fig. 18.10a and b.
 The mixing ratio of cold primary air to zone supply air and the reset of the cold primary air supply
temperature for the perimeter zone are different from those of the interior zone. Different
AHUs or PUs are preferable for the perimeter zone and interior zone separately. An analysis of
energy savings and investment costs is often necessary.
 In a series fan-powered box, the supply fan in the AHU or PU offsets the system pressure drop up
to the inlet of the VAV box. The fan in the fan-powered box offsets the pressure drop of the VAV
box, elbows (if any), downstream heating coil, flexible duct, and diffuser.
In a parallel fan-powered box, the supply fan in the AHU or PU offsets the pressure drop of
the VAV box, flexible duct, diffuser, and straight-through mixing loss of the cold primary
airstream in the fan-powered box.
 At summer and winter design conditions, the ratio of the volume flow rate of cold primary air
supplied to the fan-powered box to the zone supply volume flow rate and the ratio of
the volume flow rate of warm recirculating plenum air to the zone supply volume flow rate
depend on the following factors: (1) zone sensible cooling load and heating load,
(2) zone supply temperature differential at summer and winter design conditions Trn Tsn,d, or
Tsn,d Trn, and (3) outdoor air requirements.
As mentioned previously, for cold air distribution is about 0.6 for a fan-powered
box. The winter heating in the perimeter zone often needs greater supply volume flow to match
the requirement of the warm air supply temperature differential not exceeding 15°F (8.3°C);
for cold air distribution is about 0.7.
 The supply main ducts, branch ducts, cold air passage in the VAV box, flexible ducts, and diffusers
must be well insulated especially for cold air distribution. Cold primary air and recirculated
plenum air should be mixed in a perpendicular flow direction to prevent downstream temperature
stratification.
 A low zone relative humidity and zone dew point are desirable to prevent possible surface condensation
during cold air distribution. When cold primary air at 44°F (6.7°C) is supplied directly
to the conditioned space through high induction diffusers, the surface of the induction diffusers
can be assumed to be 3°F (1.7°C) higher than the supply air; or the dew point of zone air should
be lower than 44 	 3  47°F (8.3°C), that is, a zone temperature of 75°F (23.9°C) and a space
relative humidity r  37 percent.
 At the start of cool-down period in a hot and humid summer, direct supply of 40 to 50°F (4.4 to
10.0°C) cold primary air to the fan-powered boxes may cause condensation because of the higher
dew point of stagnant air in the ceiling plenum before the cool-down period. The AHU or PU should
supply 55°F (12.8°C) air to the fan-powered boxes first, to reduce moisture and lower the dew point.
 If humidification is required to maintain an acceptable zone relative humidity in winter, a humidi-
fier can be installed in the AHU or PU.
V?
sn,h /V?sn,d
V?
sn,c /V?sn,d
V?
s,rec /V?sn,d
V?
sn,c /V?sn,d
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.55
 During the unoccupied period in winter, the AHU or PU can be shut off while the fan-powered
boxes provide the necessary heating.
 Fan-powered boxes installed inside the ceiling plenum must meet local safety code requirements.
21.6 COMPARISON BETWEEN VARIOUS VAV SYSTEMS
VAV systems are used in applications where space load is varied in order to maintain a required indoor
environment and to save fan energy. Single-zone VAV systems are mainly used for single-zone
conditioned spaces such as arenas, indoor stadiums, assembly halls, and many factories.
VAV reheat systems using reheating VAV boxes are simple and effective for perimeter heating in
winter and with VAV boxes provide cooling only in the interior zone in a building. Simultaneous
cooling and heating may occur in reheat systems at low zone cooling loads. The VAV reheat system
becomes energy-inefficient for locations where the winter heating load linear density exceeds 120
Btu/h
 ft (115 W/ m).
Perimeter heating VAV systems are used for buildings that need heating for their perimeter
zones in locations with cold and long winters.
Dual-fan dual-duct VAV systems provide satisfactory indoor environmental control for multizone
high-rise buildings and health care facilities with flexible operating characteristics. If the shutoff
air leakage in the mixing VAV box can be minimized, it is often an appropriate choice for more
demanding applications.
Fan-powered VAV systems provide low cooling load compensation and winter heating from a
two-stage sequence operation: intake warm plenum air and winter heating. The small centrifugal
fan in the fan-powered box also provides the mixing of cold primary air with the plenum air. Fanpowered
VAV systems are widely used in buildings where cold air distribution is needed, such as
office buildings and schools.
REFERENCES
Alley, R. L., Selecting and Sizing Outside and Return Air Dampers for VAV Economizer Systems, ASHRAE
Transactions, 1988, Part I, pp. 1457–1466.
ANSI/ASHRAE Standard 62-1999, Ventilation for Acceptable Indoor Air Quality, ASHRAE Inc., Atlanta, GA,
1999.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except New Low-Rise Residential Buildings,
ASHRAE Inc., Atlanta, GA, 1999.
Avery, G., The Myth of Pressure-Independent VAV Terminals, ASHRAE Journal, no. 8, 1989, pp. 28–30.
Avery, G., The Instability of VAV Systems, Heating / Piping / Air Conditioning, February 1992, pp. 47–50.
Avery, G., Selecting and Controlling Economizer Dampers, HPAC, no. 8, 1996, pp. 73–78.
Brothers, P. W., and Warren, M. L., Fan Energy Use in Variable Air Volume Systems, ASHRAE Transactions,
1986, Part II B, pp. 19–29.
Cappellin, T. E., VAV Systems—What Makes Them Succeed? What Makes Them Fail? ASHRAE Transactions,
1997, Part II, pp. 814–822.
Chen, S., and Demster, S., Variable Air Volume Systems for Environmental Quality, McGraw-Hill, New York,
1996.
Coggan, D., Mixed Air Control with DDC, Heating / Piping / Air Conditioning, no. 5, 1986, pp. 113–115.
Cox, R. L., Retrofit for a College Campus, 1996 ASHRAE Technology Award Case Study, ASHRAE Journal,
no. 11 1996, pp. 59–64.
21.56 CHAPTER TWENTY-ONE
Dean, R. H., and Ratzenberger, J., Stability of VAV Terminal Unit Controls, Heating / Piping / Air Conditioning,
October 1985, pp. 79–90.
Dorgan, C. E., and Elleson, J. S., Cold Air Distribution, ASHRAE Transactions, 1988, Part I, pp. 2008–2025.
Elleson, J. S., Energy Use of Fan-Powered Mixing Boxes with Cold Air Distribution, ASHRAE Transactions,
1993, Part I, pp. 1349–1358.
Ellis, R., and McKew, H., Back to Basics: VAV Terminal Building Automation—Design and Commissioning,
Engineered Systems, no. 3, 1997, pp. 9–10.
Englander, S. L., and Norford, L. K., Saving Fan Energy in VAV Systems—Part 1: Analysis of a Variable-
Speed-Drive Retrofit, ASHRAE Transactions, 1992, Part I, pp. 3–18.
Grimm, N. R., and Rosaler, R. C., HVAC Systems and Components Handbook, 2d ed., McGraw-Hill, New York,
1998.
Haines, R. W., Supply Fan Volume Control in a VAV System, Heating / Piping / Air Conditioning, August 1983,
pp. 107–111.
Int-Hout, III, D., Analysis of Three Perimeter Heating Systems by Air Diffusion Methods, ASHRAE Transactions,
1983, Part I B, pp. 101–112.
Int-Hout, III, D, Stand-Alone Microprocessor Control of Dual-Duct Terminals, ASHRAE Transactions, 1987,
Part II, pp. 1722–1733.
Kettler, J. P., Efficient Design and Control of Dual-Duct Variable-Volume Systems, ASHRAE Transactions,
1987, Part II, pp. 1734–1741.
Krajnovich, L., and Hittle, D. C., Measured Performance of Variable Air Volume Boxes, ASHRAE Transactions,
1986, Part II A, pp. 203–214.
Linder, R., and Dorgan, C. B., VAV Systems Work Despite Some Design and Application Problems, ASHRAE
Transactions, 1997, Part II, pp. 807–813.
Linford, R. G., Dual-Duct Variable Air Volume—Design / Build Viewpoint, ASHRAE Transactions, 1987,
Part II, pp. 1742–1748.
Lizardos, E. J., Economizer Cycle and Damper Sizing, Engineered Systems, no. 1, 1997, pp. 126–131.
Lynn, M., Balancing DDC—Controlled Boxes, Heating / Piping / Air Conditioning, July 1989, pp. 79–84.
Mutammara, A. W., and Hittle, D. C., Energy Effects of Various Control Strategies for Variable-Air-Volume
Systems, ASHRAE Transactions, 1990, Part I, pp. 98–102.
Muxen, S. A., and Chapman,W. F., Versatile Application-Specific Controllers for Hotel Guest Rooms, ASHRAE
Transactions, 1988, Part I, pp. 1530–1538.
Schuler, M., Dual-Fan, Dual-Duct System Meets Air Quality, Energy-Efficiency Needs, 1996 ASHRAE Technology
Awards Case Study, ASHRAE Journal, no. 3, 1996, pp. 39–41.
Shavit, G., Retrofit of Double-Duct Fan System to a VAV System, ASHRAE Transactions, 1989, Part I,
pp. 635–641.
Smith, M. C., Industry Standard 885, Acoustic Level Estimation Procedure Compared to Actual Acoustic Levels
in an Air Distribution Mock Up, ASHRAE Transactions, 1989, Part I, pp. 543–548.
Spitler, J. D., Pedersen, C. O., Hittle, D. C., and Johnson, D. L., Fan Electricity Consumption for Variable Air
Volume, ASHRAE Transactions, 1986, Part II B, pp. 5–18.
Spitler, J. D., Hittle, D. C., Johnson, D. L., and Pederson, C. O., A Comparative Study of the Performance of
Temperature-Based and Enthalpy-Based Economy Cycles, ASHRAE Transactions, 1987, Part II, pp. 13–22.
Straub, H. E., and Cooper, J. G., Space Heating with Ceiling Diffusers, Heating / Piping / Air Conditioning, May
1991, pp. 49–55.
Tackett, R. K., Case Study: Office Building Use Ice Storage, Heat Recovery, and Cold Air Distribution,
ASHRAE Transactions, 1989, Part I, pp. 1113–1121.
Taylor, S. T., Series Fan-Powered Boxes: Their Impact on Indoor Air Quality and Comfort, ASHRAE Journal,
no. 7, 1996, pp. 44–50.
The Trane Company, Variable Air Volume Systems Manual, American Standard Inc., La Crosse, WI, 1988.
Wacker, P. C., Economizer Saving Study, ASHRAE Transactions, 1989, Part I, pp. 47–51.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechnic, Hong Kong, 1987.
Warden, D., Dual Fan, Dual Duct Systems, Better Performance at a Low Cost, ASHRAE Journal, no. 1, 1996,
pp. 36–41.
AIR SYSTEMS: VARIABLE-AIR-VOLUME SYSTEMS 21.57
Wendes, H. C., Estimating VAV Retrofit Costs, Heating / Piping / Air Conditioning, August 1983, pp. 93–103.
Wheeler, A. E., Energy Conservation and Acceptable Indoor Quality in the Classroom, ASHRAE Journal, April
1992, pp. 26–32.
Williams, V. A., VAV System Interactive Controls, ASHRAE Transactions, 1988, Part I, pp. 1493–1499.
Yu, H. C., Harmon, J. J., and Galway, J. M., Renovation of the Watergate 600 Building, ASHRAE Journal,
no. 10, 1993, pp. 30–36.
21.58 CHAPTER TWENTY-ONE
CHAPTER 22
AIR SYSTEMS: VAV
SYSTEMS—FAN
COMBINATION, SYSTEM
PRESSURE, AND SMOKE
CONTROL
22.1
22.1 RETURN AND EXHAUST
SYSTEMS 22.2
Types of Return and Exhaust
Systems 22.2
Return Ceiling Plenum 22.2
Low-Level Return Systems and Enclosed
Parking Garage Ventilation 22.2
Exhaust Hoods 22.3
22.2 FAN COMBINATIONS AND
OPERATING MODES 22.4
Fan Combinations 22.4
Operating Modes 22.4
22.3 SYSTEM PRESSURE DIAGRAM 22.5
System Pressure and Duct Static
Pressure Control 22.5
Fan Characteristics 22.7
Mixing-Exhaust Section and Conditioned
Space 22.8
22.4 SUPPLY FAN AND EXHAUST FAN
COMBINATION 22.8
System Characteristics 22.8
Operating Characteristics 22.9
Recirculation Mode and Design Volume
Flow 22.9
Recirculation Mode, 50 Percent Design
Flow Rate 22.12
Air Economizer Cycle 22.13
Warm-Up and Cool-Down Mode 22.13
Pressure Variation at the Mixing Box 22.13
22.5 SUPPLY FAN AND RELIEF FAN
COMBINATION 22.14
Recirculation Mode 22.14
Air Economizer Mode and Design Volume
Flow Rate 22.14
Air Economizer Mode, 50 Percent Design
Flow 22.17
Warm-Up and Cool-Down Mode 22.17
Design Considerations and Controls 22.17
22.6 SUPPLY FAN AND RETURN FAN
COMBINATION 22.18
Recirculation Mode 22.18
Air Economizer Mode 22.20
Controls 22.21
22.7 COMPARISON OF THREE FAN
COMBINATION SYSTEMS 22.21
22.8 PRESSURE FLOW
CHARACTERISTICS 22.22
Pressure Flow Characteristics for a
Supply and Return Fan Combination
System 22.22
Variation of Pressure in the Mixing
Box 22.23
Field Survey of System Pressure
Characteristics of a VAV System
Using a Supply and Relief Fan
Combination 22.23
22.9 SMOKE CONTROL AND FIRE
SAFETY 22.24
Fire Safety in Buildings 22.24
Smoke Movements in Buildings 22.25
Effective Area and Flow Rate 22.27
22.10 EFFECT OF AUTOMATIC SPRINKLER
ON FIRE PROTECTION 22.27
22.11 SMOKE CONTROL IN ATRIA 22.28
ANSI/NFPA 92A and 92B 22.28
Smoke Management in Atria, Malls, and
Large Areas 22.28
22.12 STAIRWELL PRESSURIZATION AND
ZONE SMOKE CONTROL 22.29
Stairwell Pressurization 22.29
Characteristics of Stairwell
Pressurization 22.29
Overpressure Relief and Feedback
Control 22.30
Stair and Shaft Vents 22.31
Zone Smoke Control 22.31
Design Considerations 22.32
Volume Flow Rate 22.32
System Pressure Loss for Stairwell
Pressurization System 22.33
REFERENCES 22.38
22.1 RETURN AND EXHAUST SYSTEMS
Types of Return and Exhaust Systems
As described in Sec. 20.3, for any enclosed conditioned space in a building, if the difference of air
densities is ignored, the total volume flow rate of air entering the space must be equal to the total
volume flow rate of air leaving the space.
Among the three main categories of air systems described in Sec. 20.9 (Fig. 20.16):
 For most of the constant-volume systems, return air is often returned to the packaged unit
through door undercuts and a collective return grille in the closet, under the attic, or at the high
level in the basement. An exhaust fan is optional and is usually installed in the bathroom or
toilet.
 In many dedicated outdoor air and space recirculating systems, an exhaust system with a volume
flow rate less than the supply volume flow rate of the outdoor ventilation air is often employed in
restrooms and service rooms. This is discussed in Chap. 28.
Return Ceiling Plenum
Among the variable-air-volume and dedicated minimum outdoor fan systems, use of the ceiling
plenum as a return air plenum, as shown in Fig. 20.16b and c, is the most common return air
system. In a return ceiling air plenum system, return air is often returned to the ceiling plenum
through the return slots or grilles first. Usually a connected return duct connects the ceiling plenum
and the fan room. Part of the return air becomes recirculating air and enters the AHU or PU, and the
remaining air is exhausted outside.
For a return ceiling plenum, the following must be considered during design:
 Fire protection. Because the return air flows through the ceiling plenum, refer to federal and local
fire codes for the requirements, materials, and equipment installed inside the ceiling plenum, as
well as the fire wall and partitions that separate the ceiling plenum and adjacent fire compartments.
 Noise control. Fan noise that has been transmitted into the ceiling plenum is very difficult to
attenuate. Do not install silencers inside the ceiling plenum. Install silencers in the fan room.
Check both duct-borne noise and radiated noise; also check noise transmitted from both supply
and return air sides.
 Air leakage. Because most of the ceiling plenums use acoustical ceiling tile and T-bar construction,
the airtightness between the conditioned space and the ceiling plenum is hard to
maintain. Field measurement showed that the pressure differential between the conditioned
space and the ceiling plenum in a VAV system serving a small public library pr  pl  0.004
in. WC (1 Pa). The critical air leakage from a ceiling plenum is the air leakage to outdoors, or
vice versa.
Low-Level Return Systems and Enclosed Parking Garage Ventilation
Low-level return systems have been widely used in the following applications:
 In indoor stadiums and assembly halls, where there are many spectator seats. It is better to extract
the return air from the return grilles located at low-level seats to induce part of the diffused
22.2 CHAPTER TWENTY-TWO
airstreams flowing from the upper level to the low level, as shown in Fig. 18.20. The return air
then flows through the return air duct, pulled by a return fan to the fan room, and is either recirculated
or exhausted to outdoors.
 In clean rooms or clean spaces, in order to form a unidirectional airflow from the ceiling to the
floor, so that dust will not ascend from the low level to contaminate the working area, as shown in
Figs. 18.23 and 18.24.
Fully enclosed underground parking garages require mechanical ventilation systems including
exhaust systems, makeup air supply systems, or both to maintain a CO concentration
not exceeding 35 ppm for up to 1 hour exposure. ASHRAE sponsored a study that found
a ventilation requirement of 1.5 cfm / ft2 (7.5 L / s m2) or 6 air changes per hour (ach) is satisfactory.
ASHRAE/IESNA Standard 90.1– 1999 mandatorily specifies that garage ventilation air systems
with a total volume flow rate greater than 30,000 cfm (14,200 L / s) shall be permited to (a) have at
least an automatic control to stage fans or modulating fan volume to maintain CO concentration below
levels specified in ASHRAE Standard 62, 35 ppm up to 1 hour exposure, or (b) be capable of
shutting off fans or reducing fan volume when the garage is not in use.
Exhaust Hoods
ASHRAE/IESNA Standard 90.1– 1999 specifies that individual kitchen exhaust hoods larger than
5000 cfm (2360 L / s) shall be provided with a makeup air system with a capacity equal to at least
50 percent of exhaust volume flow rate that is (a) unheated or heated to no higher than 60°F
(10.6°C) and (b) uncooled or cooled without the use of mechanical cooling (refrigeration). Exceptions
include
 Exhaust hoods used to exhaust ventilation air which otherwise would exfiltrate or be exhausted by
other air systems
 Certified grease extractor hoods that require a face velocity no greater than 60 fpm (0.3 m / s)
Standard 90.1– 1999 also specifies that buildings with fume hood exhaust systems having a total
exhaust volume flow rate greater than 15,000 cfm (7080 L / s) shall include at least one of the following
features:
 Variable-air-volume hood exhaust and room supply systems capable of reducing exhaust and
makeup volume flow to 50 percent or less of design flows.
 Direct makeup air supply equal to no less than 75 percent of the exhaust rate, heated to no
warmer than 2°F (1.1°C) below room set point, cooled to no cooler than 3°F (1.7°C) above
room set point, no humidification allowed, and no simultaneous heating and cooling for dehumidification.
 Heat recovery system for makeup air from fume hood exhaust with same exceptions as in Sec.
12.5. Refer to Standard for details.
Standard 90.1– 1999 also mandatorily specifies that all outdoor air supply and exhaust hoods,
vents, and ventilators shall be installed with motorized dampers that will automatically shut when
the spaces served are not in use. Gravity (nonmotorized) dampers are acceptable in buildings less
than three stories in height above grade and for buildings of any height located in climates with less
than 2700 HDD65. Exceptions are for ventilation systems serving unconditioned spaces as well as
during preoccupancy building warmup, colddown, and setback, except when ventilation reduces energy
costs (such as, night purge). Both outdoor air supply and exhaust air dampers shall have a
maxium leakage rate of 3 cfm/ft2 at 1.0 in. WC (15 L/ s m2 at 250 Pa).
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.3
22.2 FAN COMBINATIONS AND OPERATING MODES
Fan Combinations
In commercial buildings, constant-volume systems with optional exhaust are widely used in small
air systems, and the VAV systems with return ceiling plenum are the most widely used in mediumsize
and large air systems. As discussed in Sec. 20.9, among these air systems, the following fan
combinations are often used:
 Supply and exhaust fan combination
 Supply and relief fan combination
 Supply and return (including unhoused plug / plenum) fan combination
Supply, exhaust, relief, and return fans can either be a single fan or multiple fans. Recently, some
manufacturers use unhoused plug / plenum fan as the return fan to form a supply and return fan
combination. As discussed in Sec. 15.2 an unhoused plug / plenum fan is simpler in construction
and has the flexibility of location of the discharge outlet. However, it requires more power than a
centrifugal fan of the same capacity and fan total pressure.
Air systems of various fan combinations are installed to meet the following requirements:
 To provide required outdoor ventilation air for occupants and conditioned space according to
ASHRAE Standard 62–1999
 To supply conditioned air to the space to offset space load
 To recirculate space air for energy conservation
 To operate in an air economizer cycle for energy saving
 To exhaust objectionable contaminated air from the space
 To maintain a proper indoor environment
Operating Modes
Constant-volume systems with optional exhaust and VAV systems with return ceiling plenum may
be operated in the following modes:
 Cooling mode is one of the basic operating modes in air systems. In cooling mode operation, outdoor
air or recirculating air, or a mixture of the two is cooled or is cooled and dehumidified in the
water cooling coil or DX coil, and the conditioned air is supplied to the conditioned space to offset
the cooling load. Cooling modes can again be subdivided into summer cooling mode and winter
cooling mode. Many interior zones that have heavy internal cooling loads still need cold air
supply in the winter cooling mode.
 Heating mode is another basic operating mode in air systems. In heating mode operation, outdoor
air, recirculating air, or their mixture is heated in the direct-fired furnace, electric heater or water
heating coil. Warm air is supplied to the conditioned space to offset the heating load. In heating
mode, humidification is optional.
 In part-load operation, the space cooling or heating load is less than the summer or winter design
load. This may be due to the variation of the space load, or change of the outdoor climate, or both.
Air systems operate at part load most of the time. During part-load operation, the supply volume
flow rate as well as the cooling and heating capacities of the air system reduces accordingly. It is
assumed that the supply volume flow rate may be reduced to 50 percent, and even dropped to 40
percent of design flows.
22.4 CHAPTER TWENTY-TWO
 In minimum outdoor air recirculating mode or simply recirculating mode, only minimum outdoor
air contained in the mixture of outdoor and recirculating air is supplied to the conditioned space
for the occupant and for dilution of contaminated air. Minimum outdoor air circulation mode can
be a subsidiary operating mode in a cooling or heating mode operation.
 In an air economizer mode, only the outdoor air or a mixture of outdoor and recirculating air is
supplied to the conditioned space to replace part of or all the mechanical cooling (cooling from
refrigeration). Air economizer mode is an outdoor air free-cooling mode and is often used as the
first stage of cooling to save energy.
 In purge mode, 100 percent outdoor air is extracted during the nighttime or in the morn-ing before
the space is occupied, to cool down the space with outdoor air or to flush the conditioned
space with cleaner outdoor air while at the same time squeezing out the indoor contaminated
air.
 In warm-up and cool-down mode, the space air temperature is warmed up or cooled down to predetermined
limits after the air system is started in the morning while the space is not occupied.
During warm-up or cool-down mode, the recirculating damper is fully open and the outdoor and
exhaust dampers are closed, except when the warm-up or cool-down mode is combined with the
purge mode.
Warm-up and cool-down mode operation should not take place prior to the purge mode.
 In nighttime setback mode, the space temperature is set back when the conditioned space is unoccupied
in winter, typically from 70 to 55°F (21.1 to 12.8°C), to prevent freezing indoors and to reduce
the warm-up temperature differential the next morning before the space is occupied.
 Occupied mode is for the operating hours when the conditioned space is occupied by the occupants,
and the unoccupied mode is for the operating hours when the conditioned space is not occupied
by the occupants.
22.3 SYSTEM PRESSURE DIAGRAM
The system pressure diagram of a VAV system during minimum outdoor air recirculating mode is
shown in Fig. 22.1. The system characteristics of a VAV system not only affect its fan energy use,
but also are closely related to the variation in its supply volume flow rate and operating characteristics.
System pressure is usually expressed in total pressure, in in. WG (Pa), in the form of a system
pressure diagram. In conditioned space and in places where velocity pressure is so small, it is often
ignored. The total pressure is then equal to static pressure, that is, pt  ps.
System Pressure and Duct Static Pressure Control
In a VAV system with duct static pressure control, the total pressure loss of the VAV system
can be divided into two parts: variable part pvar and fixed part pfix, both in in. WC (Pa). The
total pressure loss between the centerline of the recirculating damper in the PU and the static
pressure sensor of the duct static pressure control is expressed as psf,var, in in. WC (Pa). As
the supply volume flow rate in a VAV system varies, pvar varies accordingly. The fixed part
pfix is independent of the variation in the supply flow rate and remains constant as the volume
damper in the VAV box increases its pressure loss at reduced volume flow rate (Figs. 20.13 and
22.1).
For a VAV system installed with a supply fan and a return fan connected in series, as shown in
Figs. 22.1 and 22.4, its system pressure loss psys, in in. WC (Pa), is also divided into variable part
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.5
pvar and fixed part pfix, in in. WC (Pa), and it can be calculated as
(22.1)
where pr,s  pressure loss of return air system, in. WC (Pa)
pPU  pressure drop across PU, in. WC (Pa)
psd, var  pressure loss of supply duct system before duct static pressure sensor, in. WC (Pa)
psf, var  pressure loss of variable part of supply fan, in. WC (Pa)
Rvar  flow resistance representing variable part of VAV system, in. WC/ (cfm)2 (Pa  s2/m6)
 supply volume flow rate of VAV system, cfm (m3 / s)
In a VAV system with a supply and return fan combination connected in series, the supply
volume flow rate flowing through the supply fan is different from the return volume flow rate
flowing through the return air fan. This makes the pressure-volume operating characterirstics more
complicated.
If the pressure sensor is installed closer to the supply fan, the fan power savings at reduced volume
flow decreases because of the increase in pfix. The set point of the static pressure sensor may
vary from 0.7 to 1.5 in. WC (125 to 375 Pa).
In a VAV system serving a single floor in a multistory building, if the maximum design air velocity
in the main duct vmax is 3000 fpm (15 m / s), and if the length of main duct is 150 ft (45.7 m),
V?
var
 (pr, s  psf,var)  pfix
 (pr, s  pPU  psd,var)  pfix
psys  pvar  pfix  RvarV? var
2  pfix
22.6 CHAPTER TWENTY-TWO
Exhaust
damper
Recirculating
damper
Recirculating/
exhaust chamber
Mixing
exhaust
section
Return
fan
Return duct AFD AFD
O
O
Economizer
air damper
Minimum outdoor air damper
Mixed plenum
Supply fan
Supply duct d
Supply air
Outdoor
air Exhaust
air
Return
air
Space
VAV
box
ru
rec
m cc sf
rt
sfo
rfo
rfi
sfi
pr,s psf,var psf,fix
psf,fix
pt pt, in WG
po po
pro  0.03 in WG
pr
s
psf prf
psf
pom
pro
pret
P2
P1
4.0
3.0
2.0
1.0
1.0
2.0
0
FIGURE 22.1 System pressure diagram for a supply-return fan combination air system (connected in series) (rfi 
return fan inlet; rfo  return fan outlet; sfi  supply fan inlet; sfo  supply fan outlet; prf  return fan total pressure;
psf  supply fan total pressure; AFD  adjustable-frequency, variable-speed drive).
then the design total pressure loss of this supply duct system psd, var may vary from 0.75 to 1.5 in.
WC (187 to 375 Pa), typically 1 in. WC (250 Pa), depending on the duct fittings in the main duct.
Fan Characteristics
A VAV system with a supply and return fan or supply and relief fan combination is often connected
in series in a fan-duct system. The reasons to use a supply fan and a return fan instead of a single
supply fan are as follows:
 To meet the requirement of maintaining a slightly positive pressure in the conditioned space as
well as at the inlet of the AHU or PU point ru, and a negative pressure in the mixing box (mixed
plenum) to extract outdoor air
 To exhaust 100 percent free-cooled outdoor air during an air economizer cycle
 To set up different volume flow rates of supply air and return air in order to provide a required
amount of outdoor ventilation air as well as to maintain a desirable space pressure
During the design of a series-connected supply and return fan combination system, as shown in
Fig. 22.1, it is essential that the following parameters be carefully calculated and analyzed:
 Clearly divide the total pressure loss that should be undertaken by the supply fan, or by the
return fan during minimum outdoor air recirculating mode at summer design conditions. The
return system pressure loss undertaken by the return fan start from the inlet of return slot rt
includes the return slot, airflow inside the ceiling plenum which is often ignored because of
its lower air velocity, the return grille if any, the return duct and duct fittings, up to part of the
recirculating damper, as shown in Figs. 22.1 and 22.4. For the PU with a configuration shown in
Fig. 22.1, in minimum outdoor air recirculating mode, a 90° turn plus one-half of the total pressure
loss of the recirculating damper at full opening design flow is always greater than a 90° turn
plus the total pressure loss of the exhaust damper and louvers at minimum outdoor air recirculating
mode.
In a VAV system, the fan total pressure of the return fan pt,rf must be carefully calculated,
neither overestimated nor underestimated.
 Based on the principle of conservation of mass, from Eq. (20.8), the relationship between the
mass flow rates of supply, return, and exhaust flow rate can be expressed as
(22.2)
where , ,  mass flow rate of supply air, return air, and exhaust air, lb/ min (kg / min)
 mass flow rate of exfiltrated air, lb/ min (kg / s)
If the difference in air densities is ignored and a positive pressure is maintained in the conditioned
space, from Eq. (20.9), the relationship between the supply, return, and exhaust volume flow rates
can be expressed as
(22.3)
where  volume flow rates of supply air, return air, and exhaust air, cfm (L / s)
 volume flow rate of exfiltrated air, cfm (L / s)
If the supply air is at 55°F (12.8°C) dry-bulb and 54°F (12.2°C) wet-bulb temperature and the return
air is at 78°F (25.6°C) dry-bulb and 65°F (18.3°C) wet-bulb temperature, when the mass flow
rates of the supply and return air are the same, the volume flow rate of return air is about 4 percent
greater than that of the supply air. During cold air distribution, if the supply air is at 44°F (6.7°C)
dry-bulb and 43°F (6.1°C) wet-bulb temperature, and the return air remains the same, the volume
V?
exf
V?
s, V?rt, V?ex
V?
s  V?rt  V?ex  V?exf
m?exf
m?ex m?rt m?s
m?s  m?rt  m?ex  m?exf
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.7
flow rate of the return air is 7 percent greater. For accurate analysis, these differences should be
taken into consideration.
Mixing-Exhaust Section and Conditioned Space
In an air system, there are two places where the air is connected to the outside atmosphere: (1) the
mixing box–exhaust chamber (simply the mixing-exhaust section), where it is connected with the
outside atmosphere through the outdoor air intake and the exhaust passage, and (2) the conditioned
space, where it connects with the outside atmosphere through the openings in the building shell, as
shown in Fig. 22.1.
In a mixing box of a VAV system, the variation of the negative pressure affects the amount of
outdoor air intake and, therefore, the indoor air quality of the conditioned space. Within the conditioned
space, the variation of the supply air, return air, exhaust air, infiltration, and exfiltration
makes the space airflow balance and associated space pressure complicated.
22.4 SUPPLY FAN AND EXHAUST FAN COMBINATION
System Characteristics
Air systems equipped with a single supply fan and an exhaust fan are often used in buildings in
which the total pressure loss of the return duct system is low. A certain volume flow of air is usually
exhausted from rooms such as restrooms. A barometric relief damper is often installed in the conditioned
space or in the return plenum to avoid excessively high space pressure. When the space positive
pressure on the damper is greater than the weight of the damper, the damper opens and the
space pressure is relieved and is thus maintained below a predetermined value. Such an air system
may be equipped with a water economizer.
To analyze the pressure characteristics of the airstream at two cross-sectional planes along the
airflow in an air system, the total pressure is determined from Eq. (17.12): pt1  pt2  pf .
To determine whether air is flowing from one enclosed space to another, or from an enclosed
space to the outdoors, the static pressure difference between these two places ps  ps1  ps2
should be calculated. However, the static pressure inside an enclosure of uniform pressure, or the
mean static pressure on a cross-sectional plane along airflow, must be determined from Eq. (17.11):
ps  pt  pv. Here pv represents velocity pressure.
As the atmospheric pressure is assumed po  0 for the convenience of analysis, static pressure
at the mixing box pm must be negative in order to extract outdoor air. The total pressure difference
between the outdoor air and the air in the mixed plenum (mixing box) pom  po  pm consists
mainly of the pressure loss of the outdoor dampers and louver. If a space is maintained at a positive
pressure, the volume flow rate of exfiltration from the space , in cfm (L / s), can be calculated
from Eq. (20.11). If space is maintained at a negative pressure, the volume flow rate of infiltration
into the space , in cfm (L / s), can be similarly calculated. From Eq. (20.11), the pressure loss
across the building shell pro, in in. WC (Pa), can be calculated as
(22.4)
The flow resistance against exfiltrated air across building shell Rexf, in in. WC/ (cfm)2 (Pa s2/m6), is
Rexf   1
4005 Ae,l 2
pro  pr  po   1
4005 Ae,l 2
V?
exf
2  RexfV? exf
2
V?
inf
V?
exf
22.8 CHAPTER TWENTY-TWO
Because the volume flow rates of the supply air and recirculating air are different, the flow resistance
of the supply system and return system as well as the flow resistance of the fixed part and
variable part of the air system pressure loss should be calculated separately.
Operating Characteristics
Consider a VAV rooftop packaged system with a supply fan and an exhaust fan serving a typical
floor in a commercial building, as shown in Fig. 22.2. This air system has the following operating
characteristics at design conditions:
Supply volume flow rate 20,000 cfm (9438 L / s)
Total pressure losses:
Across recirculating damper at design flow 0.4 in. WC (100 Pa)
Filters and coils 2.5 in. WC (625 Pa)
Supply main duct, pd, var 0.85 in. WC (212 Pa)
Recirculating system 0.15 in. WC (37 Pa)
VAV box, branch duct, and diffuser 0.75 in. WC (187 Pa)
Effective leakage area on building shell 0.5 ft2 (0.46 m2)
Volume flow of exhaust fan 3000 cfm (1416 L / s)
Minimum outdoor ventilation air required 3350 cfm (1581 L / s)
Space pressure  0.03 in. WG (7.5 Pag)
For simplicity, the difference in air densities between supply, return, and exhaust air is ignored.
In minimum outdoor air recirculating mode, the pressure losses of the outdoor louver can be neglected.
A barometric relief damper is mounted in the conditioned space and is opened when the
space positive pressure is greater than 0.2 in. WG (50 Pag). The supply fan has the same
characteristics as in Example 20.3.
The following control systems are installed to maintain the required operating parameters:
 Zone temperature control
 Discharge temperature control
 Duct static pressure control
 Minimum outdoor ventilation air control
 High pressure limit control
All these control systems are described in Chap. 23.
In summer, this air system can be operated in cooling mode, recirculating mode, air economizer
mode, or warm-up or cool-down mode.
Recirculation Mode and Design Volume Flow
Minimum outdoor air is extracted through the outdoor intake and is mixed with the recirculating air
in the mixing box in a rooftop packaged unit. The mixture is conditioned in the packaged unit. After
conditioning, the supply air flows through the supply fan and the supply duct and is discharged to
the conditioned space. In the conditioned space, a certain volume flow rate is exhausted through an
exhaust fan, and another small portion is exfiltrated through the gaps and openings on the building
shell if the space is maintained at a positive pressure between  0.005 and  0.03 in. WG (1.25
and 7.5 Pag). The major portion of the space air is returned to the rooftop unit, where it is mixed
with the outdoor air for recirculation.
ptV?
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.9
FIGURE 22.2 Air system of supply and exhaust fan combination: (a) schematic diagram; (b) system characteristics
on diagram. pt  V?
22.10 CHAPTER TWENTY-TWO
When the air system of a single supply fan and exhaust fan combination is operated in recirculating
mode with design volume flow, the inlet vanes at the supply fan inlet are fully opened and the
outdoor air damper is at minimum opening position. The minimum outdoor ventilation air required
is about 3350 cfm (1581 L / s). The fixed part of the pressure loss of such an air system is pfix 
0.75 in. WC (187 Pa). The variable part of pressure loss of the air system consists of the following
two sections:
 Mixed plenum point m to the static pressure sensor, point d, at a supply volume flow rate of
20,000 cfm (9438 L / s):
pmd  2.5  0.85  3.35 in. WC (835 Pa)
Its corresponding flow resistance can be calculated as
 Recirculating system from point r to m:
The variable part of pressure loss of the air system is
(22.5)
Because of a positive space pressure of  0.03 in. WG ( 7.5 Pag) and an effective leakage area
Ae, l of 0.5 ft2, the volume flow rate of exfiltration from the conditioned space is calculated as
And the flow resistance of the opening in the building shell is
The recirculating volume flow rate through the recirculating system is then equal to
The flow resistance of the recirculating system based on pressure loss between points r and m is
And the flow resistance of the variable part of the pressure loss of the air system is
The supply fan total pressure at design volume flow rate is
psf  0.75  2.5  0.85  0.15  0.4  4.65 in. WC (1156 Pa)
 9.75 	 109 in. WC / (cfm)2 (5.14 	 103 Pa s2
/ m6)
 8.38 	 109  1.98 	 109
 16,650
20,000 2
Rmd,rm  Rmd  RrmKrm
2
Rrm 
prm
V?
rm
2 
0.55
(16,650)2  1.98 	 109 in. WC / (cfm)2 (1.04 	 103 Pa s2
/ m6)
V?
ru  V?s  (V?exf  V?ex)  20,000  (350  3000)  16,650 cfm (7857 L / s)
V?
ru
Rexf 
pro
V?
exf
2 
0.03
(350)2  2.45 	 107 in. WC / (cfm)2 (0.129 Pa s2
/ m6)
V?
exf  4005 Ae, l?pro  4005 	 0.5?0.03  350 cfm (165 L / s)
V?
exf
 Rmd, rmV? s
2
 RmdV? s
2  RrmK rm
2 V? s
2
pmd,rm  pmd  prm
prm  RrmV? ru
2  0.15  0.4  0.55 in. WC (137 Pa)
Rmd 
pmr
V?
s
2 
3.35
(20,000)2  8.38 	 109 in. WC / (cfm)2 (4.42 	 103 Pa s2
/ m6)
V?
o
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.11
The system operating point of this supply and exhaust fan combination P is plotted in Fig. 22.2b.
The pressure-volume characteristics for the key points in this system are shown below:
Recirculation Mode, 50 Percent Design Flow Rate
When the space sensible cooling load is reduced at part-load operation, the supply volume flow rate
may be reduced to 50 percent of the design volume flow rate, that is, 0.5 	 20,000  10,000 cfm
(4719 L / s).
As the dampers in the VAV boxes close to smaller openings than at design volume flow, the static
pressure in the supply main duct rises. The sensor senses the rise, and the DDC unit instructs the inlet
vanes to close to a smaller opening, until the duct static pressure is maintained approximately at its set
point at the location where the pressure sensor is mounted. The static pressure is still equal to the pressure
loss of the VAV box, flexible ducts, diffuser in the branch takeoff, and last section of the main duct
between the pressure sensor and end of the main duct plus the space static pressure: 0.75  0.03  0.78
in. WC (194 Pa). The fan performance curve is now a new fan curve Ft50, as shown in Fig. 22.2b.
The variable-part pressure loss between the mixing box and point d is now decreased to
Assume that the space pressure at 50 percent design flow is  0.03 in. WG. The volume flow rate
of infiltrated air into the space is
because the volume flow rate of an exhaust fan is closely related to its pressure. According to the
catalog of a fan manufacturer, within the operating range of a 24-in. (600-mm) diameter exhaust
fan, for an increase of 0.01 in. WC (2.5 Pa) in fan total pressure, there is a corresponding drop in
supply volume flow rate of 135 cfm (64 L / s). Compared with the design conditions, there is a need
for an increase of  0.03  (0.03)  0.06 in. WC (15 Pa) of fan total pressure of the exhaust
fan, i.e., a decrease of 0.06 / 0.01 	 135  810 cfm of volume flow rate. The volume flow rate of
recirculating air is then
The pressure loss between the space (point r) and the mixed plenum (point m) becomes
Therefore, the total pressure in the mixing box is
pm  0.03  0.132  0.162 in. WG (40 Pag)
The flow resistance of the outdoor passage is
Because of a pressure difference pom  0.162 in. WC, outdoor intake is therefore
V?
o  ?pom
Rom
 ? 0.162
4.63 	 108  1870 cfm (883 L / s)
Rom 
pom
V?
o 2
  0.52
3350 2
 4.63 	 108 in. WC / (cfm)2 (0.0244 Pa s2
/ m6)
prm  RrmV? rm50
2  1.98 	 109(8157)2  0.132 in. WC (33 Pa)
V?
ru50  V?s50  V?ex  V?inf50  10,000  (3000  810)  347  8157 cfm (3849 L / s)
V?
inf  4005 Ae, l?p  4005 	 0.5?0.03  347 cfm (164 L / s)
pmd50  RmdV? s50
2  8.38 	 109(10,000)2  0.84 in. WC (209 Pa)
Point o Point m Point r Point ru
p, in. WG 0 0.52 0.03 0.12
V ·
, cfm 3350 20,000 20,000 16,650
22.12 CHAPTER TWENTY-TWO
If the difference in air density is ignored, the volume flow balance between the incoming infiltrated
air and outward exhaust air in this air system becomes
The volume flow of incoming air and exhaust air entering and leaving the system is nearly balanced.
From the above analysis, the system operating point at 50 percent design volume flow rate, point
Q, has a supply volume flow of 10,000 cfm (4719 L / s) and a fan total pressure
pt50  0.75  0.84  0.13  1.72 in. WC (428 Pa)
During minimum outdoor air recirculating mode at 50 percent design volume flow, the volume flow
rate of outdoor air extracted through the outdoor intake passage decreases to 1870 cfm (883 L / s).
Space is maintained at a negative pressure of  0.03 in. WG (7.5 Pag). An infiltration rate of 347
cfm (164 L / s) enters the space through openings and cracks in the building shell.
The characteristics at key points during 50 percent design volume flow are as follows:
Air Economizer Cycle
When an air economizer cycle is operated at a design flow rate of 100 percent of outdoor air in this
air system with a single supply fan and an exhaust fan combination, the recirculating damper is
closed, and both the minimum outdoor damper and the other, large outdoor damper for the air economizer
cycle are fully opened. Outdoor air is extracted to the mixed plenum, through the filters and
coils, and is supplied to the conditioned space. Because only a small portion of supply air is exhausted
and exfiltrated, the space positive pressure increases to about 0.2 in. WG (50 Pag). The
barometric relief damper then opens. Most of the supply air is discharged through the relief damper
outdoors.
When the supply and exhaust fan combination is operated at 50 percent design volume flow in
the air economizer cycle, space pressure is still limited to 0.2 in. WG (50 Pa), and most of the supply
air is discharged through the relief damper outdoors, as in the design volume flow.
Warm-Up and Cool-Down Mode
During the warm-up and cool-down mode, the outdoor damper is closed and the recirculating
damper is fully open, while the exhaust fan is turned off. Recirculated air from the space is conditioned
in the packaged unit and is then supplied to the space. Because of the negative pressure in
the mixing box, outdoor air is leaked into the box through the closed outdoor dampers and is exfiltrated
from the space because of the higher positive space pressure.
Pressure Variation at the Mixing Box
During the recirculating mode, when the supply volume flow rate is reduced from design flow to 50
percent of design flow, the decrease in the recirculating volume flow rate causes a corresponding
drop in the pressure loss of the recirculation system. Therefore, the pressure at the mixed plenum,
pm increases from 0.52 to 0.16 in. WG (129 to 40 Pag). The greater the pressure loss of
the return system at the design flow rate, the higher the fluctuation of pom. Because of the reduced
volume flow rate at a higher fan total pressure of the exhaust fan at 50 percent design flow, in order
Point o Point m Point r Point ru
p, in. WG 0 0.162 0.03 0.065
V ·
, cfm 1870 10,000 10,000 8157
ptV?
V?
o  V?inf  V?ex  1870  347  (3000  810)  27 cfm (13 L / s)
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.13
to balance the airflow, the space pressure is changed from  0.03 to  0.03 in. WG (7.5 to 7.5
Pag), and the exfiltration changes to infiltration. The pressure loss at the outdoor air intake passage
(excluding the damper), like the intake louver and duct friction loss plou, should be taken into
account only in the outdoor air economizer cycle. During minimum outdoor air recirculating mode,
plou should be negligible.
Pressure fluctuation pom at design flow and at 50 percent design flow causes insufficient outdoor
air intake. The effect of this pressure fluctuation can be reduced by using minimum outdoor
ventilation air control, as discussed in Chap. 23.
22.5 SUPPLY FAN AND RELIEF FAN COMBINATION
Consider an air system with a supply fan, relief fan, and exhaust fan combination, as shown in
Fig. 22.3a, with the same operating characteristics as mentioned in Sec. 22.4. A relief fan is a fan
that is installed in the relief flow passage (exhaust flow passage), adjacent to the recirculating flow
passage leading to the mixing box, to relieve the undesirably high positive space pressure. A relief
fan is different from an exhaust fan, which is often used to exhaust fumes and contaminated or toxic
gases and to maintain a negative pressure in the space or in a localized enclosure. An axial fan is often
used as a relief fan because of its large volume flow and smaller fan total pressure. The control
systems used in such a combination are similar to those of a supply fan and exhaust fan combination,
except that a relief fan control is added. The relief fan has the following pressure-volume characteristics:
Recirculation Mode
During minimum outdoor air recirculating mode operation, the relief fan is not operating and the
relief damper is closed. Outdoor air drawn into the air system is balanced by the exhaust air and the
exfiltration through the openings in the building shell. In this case, the system becomes a supply fan
and exhaust fan combination, and its operating characteristics are the same as those described and
analyzed in Sec. 22.4.
Air Economizer Mode and Design Volume Flow Rate
In a VAV system, there are often two outdoor air dampers: a smaller minimum outdoor air damper
and a larger economizer damper. Both dampers have a total pressure loss of 0.4 in. WC (100 Pa) at
100 percent outdoor air intake. During air economizer mode with 100 percent outdoor air at the design
volume flow rate, the outdoor dampers are fully opened and the recirculating damper is closed.
The relief fan relieves the space air pressure and maintains it within predetermined limits. Outdoor
air is extracted through the intake louver and both outdoor dampers, and flows through the filter and
coils. The conditioned air is then supplied to the space. A small part of the space air is exhausted,
and another part is exfiltrated. The majority of space air is relieved by the relief fan through the
ceiling plenum and relief passage.
At design volume flow, the duct friction loss of the outdoor intake passage is negligible, and the
pressure loss of the intake louver is 0.2 in. WC (50 Pa). The space pressure depends mainly on the air-
flow balance in the space. The variable part of the pressure loss of the supply air system is then given as
pvar  0.2  0.4  2.5  0.85  0.03  3.98 in. WC (990 Pa)
V ·
, cfm 10,000 15,000 16,640 17,250
pt, in. WC 4.0 2.10 1.08 0.43
22.14 CHAPTER TWENTY-TWO
22.15
FIGURE 22.3 Supply fan and relief fan combination: (a) schematic diagram; (b) system characteristics on
diagram.
pt  V?
The fixed part of pressure loss of the air system is still 0.75 in. WC (186 Pa). The fan total pressure
of the supply fan is then equal to
0.75  3.98  4.73 in. WC (1176 Pa)
The flow resistance of the variable part of the supply air system can be calculated as
The fan total pressure of the relief fan is used to overcome the pressure loss of the return
system, relief damper, and louver in the relief flow passage. Usually, the duct friction loss of
the relief passage is ignored. At design volume flow rate, for a space pressure of  0.03 in. WG
(7.5 Pag), an exfiltration of about 350 cfm (165 L / s), and an exhaust volume flow rate of
3000 cfm (1416 L / s), the volume flow rate of relief fan is 20,000  3000  350  16,650 cfm
(7857 L / s).
As in the supply fan and exhaust fan combination, the pressure loss between points r and ru is
and flow resistance is
If the pressure loss of the relief louver at a volume flow of 16,650 cfm (7857 L/ s) is 0.2 in. WC
(50 Pa), the pressure loss of the relief damper is 0.4 in. WC (100 Pa), and the velocity pressure of
the axial fan is 0.38 in. WC (95 Pa), the flow resistance of the relief system is
The sum of flow resistance of the return and relief systems is
Because a total pressure of 0.03 in. WG (7.5 Pag) is provided by the space positive pressure
and is undertaken by the supply fan, therefore, the fan total pressure of the relief fan is
calculated as
prel  0.2  0.4  0.38  0.15  0.03  1.10 in. WC (274 Pa)
The supply fan curve Fs, the relief fan curve Frel, the system curve of the supply system
, and the system curve of the return and relief system can then be plotted.
At design flow of an air economizer cycle, the system operating point for supply system P has
cfm (9320 L / s) and pt  4.73 in. WC (1176 Pa). The system operating point of the
relief system has cfm (7857 L / s) and pt  1.10 in. WC (273 Pa).
The pressure-volume characteristics at key points for air economizer mode at design volume
flow are as follows:
Point o Point m Point r Point ru
p, in. WG 0 0.6 0.03 0.12
V ·
, cfm 19,750 19,750 19,750 16,650
V?
 16,650
V?
 19,750
pt  RroV?2 pt  RorV?2
 4.08 	 109 in. WC / (cfm)2 (2.15 	 103 Pa s2
/ m6)
Rro  Rr,ru  Rrel  5.41 	 1010  3.54 	 109
Rrel 
0.2  0.4  0.38
(16,650)2  3.54 	 109 in. WC / (cfm)2 (1.86 	 103 Pa s2
/ m6)
Rr,ru 
pr,ru
V?
ru
2  5.41 	 1010 in. WC / (cfm)2 (2.85 	 104 Pa s2
/ m6)
pr,ru  Rr, ruV? ru
2  0.15 in. WC (37 Pa)
Ror 
por
V?
s
2 
3.98
(20,000)2  9.95 	 109 in WC / (cfm)2 (5.24 	 103 Pa s2
/ m6)
22.16 CHAPTER TWENTY-TWO
Air Economizer Mode, 50 Percent Design Flow
At 50 percent design volume flow rate in the air economizer mode, the variable part of the pressure
loss of the supply air system when inlet vanes are partly closed is
pvar50  9.95 	 10 9(10,000)2  1.0 in. WC (249 Pa)
And the fan total pressure of the supply fan at 50 percent design flow is
psf50  1.0  0.75  0.03  1.78 in. WC (443 Pa)
Total pressure loss of the return and relief systems is then
pr, ru50  4.08 	 10 9(6650)2  0.18 in. WC (45 Pa)
At 50 percent design flow, the velocity pressure of the relief fan is
The fan total pressure of the relief fan at 50 percent design flow is therefore
pref50  0.18  0.03  0.06  0.21 in. WC (52 Pa)
Pressure-volume characteristics at key points in air economizer mode at 50 percent design volume
flow are as follows:
Warm-Up and Cool-Down Mode
During warm-up and cool-down mode, the relief fan is turned off and the relief damper is closed.
As in the supply fan and exhaust fan combination, the exhaust fan is also turned off, the outdoor
damper is closed, and the recirculating damper is fully opened. Air is recirculated from the space to
the AHU or PU and to the space again.
Design Considerations and Controls
 As in the supply fan and exhaust fan combination system, control systems such as zone temperature,
discharge temperature, duct static pressure, minimum outdoor air, and high-pressure limit
control systems are discussed in Chap. 23.
 The operation of recirculating mode or air economizer mode is actuated by the air economizer
control which is either temperature-based or enthalpy-based, as described in Sec. 21.2. Warm-up
and cool-down mode is activated when the sensed zone temperature deviates from the set point
after the AHU or PU is turned on, as described in Chap. 23.
 At minimum outdoor air recirculating mode, when the relief fan is not operating, the space positive
pressure is maintained because of the airflow balance in the space. Suppose that 20,000 cfm
(9438 L / s) of air is supplied to the space. If the space pressure is zero, as 3000 cfm (1416 L / s) of
air is exhausted from the space by the exhaust fan, only 16,650 cfm (7857 L / s) of recirculating
Point o Point m Point r Point ru
p, in. WG 0 0.15 0.03 0.02
V ·
, cfm 10,000 10,000 10,000 6650
pv,rf   6650
16,650 2
(0.38)  0.06 in. WC (15 Pa)
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.17
air is extracted by the supply fan to mix with the 3350 cfm (1581 L / s) of outdoor air in the mixed
plenum at design volume flow. The excessive amount (350 cfm or 165 L / s) of supply air raises
the space pressure until the 20,000 cfm (9438 L / s) supply air is balanced with the sum of 16,650
cfm (7857 L / s) recirculating air, 3000 cfm (1416 L / s) exhaust air, and 350 cfm (165 L / s) exfiltrated
air at  0.03 in. WG (7.5 Pag) space positive pressure.
 The relief fan is often controlled by a DDC unit controller according to the input from the space
pressure sensor to maintain a space positive pressure within predetermined limits.
 The main drawback of the supply and relief fan combination is the increase in pressure at the
mixed plenum pm when the supply volume flow rate of a VAV system is reduced at part-load operation
during minimum outdoor air recirculating mode. The higher the pressure loss of the return
system, the greater the increase in pm. Higher pressure in the mixed box and, therefore, a lower
pom ( po  pm) cause a deficiency in outdoor air intake. Various outdoor ventilation controls have
been developed recently and are discussed in Chap. 23. The operation of a relief fan reduces both
mixing box pressure and space pressure.
22.6 SUPPLY FAN AND RETURN FAN COMBINATION
Supply fan and return fan combinations are widely used, especially in air systems that serve large
conditioned areas. A return fan is located upstream from the junction of the recirculating flow passage
and the exhaust air passage, point ru, as shown in Fig. 22.4a. Consider an AHU or PU with a
supply fan and return fan combination and an exhaust system. Its operating characteristics are the
same as those in Sec. 22.4 except for the following:
 A return fan is installed with the following pressure-volume characteristics:
 The pressure loss of the exhaust damper after point ru in the exhaust passage is 0.4 in. WC (100
Pa) and is 0.2 in. WC (50 Pa) for the air louver at the maximum exhaust volume flow rate. At
minimum outdoor airflow, louver pressure losses can be ignored. As in the supply and relief fan
combination, a minimum outdoor air damper and an economizer damper are installed, and both
have a total pressure loss of 0.4 in. WC (100 Pa) at design airflow.
Recirculation Mode
During minimum outdoor air recirculation mode, outdoor air is drawn through the outdoor air louver,
duct, and damper and is mixed with recirculating air. The mixture then flows through the filters
and coils and is supplied to the conditioned space. At design volume flow, a small part is exhausted,
another part is exfiltrated, and the remaining portion is extracted by the return fan flowing through
the return system. At point ru, almost all the return air is recirculated through the recirculating
damper.
During minimum outdoor air recirculating mode, at design volume flow rate:
 If the return fan is not operating and the exhaust damper is closed, then the fan total pressure of
the supply fan is higher than 4.65 in. WC (1156 Pa) because of the air flowing through the turnedoff
return fan.
 If the return fan is operating, 3000 cfm (1416 L / s) of air is exhausted through the exhaust passage
and exhaust damper instead of from the separate exhaust fan. The total pressure loss of the return
V ·
, cfm 10,000 15,000 20,000
pt, in. WC 1.47 1.23 1.10
22.18 CHAPTER TWENTY-TWO
22.19
FIGURE 22.4 Supply fan and return fan combination: (a) schematic diagram; (b) system characteristics on
diagram.
pt  V?
system pret increased from 0.15 to 0.7 in. WC (37 to 74 Pa). The fan total pressure of the supply
fan is then
0.2  0.4  2.5  0.85  0.75  0.03  4.73 in. WC (1176 Pa)
During the minimum outdoor air recirculating mode, the recirculating air turns 90° and flows
through the recirculating damper and has a greater total pressure loss than the 3000 cfm (1416 L / s)
exhaust air exhausted through exhaust damper and louver. Assume that the dynamic loss of the 90°
turn of the recirculating air is 0.2 in. WC (50 Pa). Then the total pressure loss of the recirculating
damper is 0.4 in. WC (100 Pa) and the fan total pressure of the return fan is
0.7  0.2  0.4  0.03  1.27 in. WC (316 Pa)
The pressure-volume characteristics at the key points are shown below:
At 50 percent of the design flow rate, if the return fan is turned off and the exhaust damper after
point ru is closed, then the fan total pressure of the supply fan will be higher than 1.72 in. WC (428
Pa), as in the supply fan and exhaust fan combination. The space pressure pr drops to  0.03 in.
WG (7.5 Pag). Total pressure at point ru may be lower than 0.07 in. WG (17 Pag) and at
point m lower than 0.26 in. WG (65 Pa) because of the pressure loss of air flowing through the
turned-off return fan.
If the total pressure loss of the return system is at 0.7 in. WC (174 Pa) and 3000 cfm (1416 L / s)
of air is exhausted through the return fan instead of the exhaust fan, and also if the return fan is operating
at 50 percent of design flow rate, then the fan total pressure of the return fan is about
(10,350 / 20,000)2(0.67  0.2  0.4)  0.34 in. WC (85 Pa) and the total pressure at ru is about
0.16 in. WG (40 Pag).
Air Economizer Mode
During air economizer mode operation at the design volume flow rate, the supply and exhaust volume
flow rate and the pressure losses of various sections of the supply air system are the same as in
minimum outdoor air recirculating mode. The fan total pressure of the supply fan is 4.73 in. WC
(1176 Pa).
Because the total pressure at point ru is raised to 1.0 in. WG (249 Pag) to overcome the pressure
loss of the exhaust flow passage and the discharge velocity pressure at the outlet of 0.40 in. WC
(100 Pa), the return fan total pressure is then
pret.f  0.15  0.03  0.60  0.40  1.12 in. WC (278 Pa)
If the total pressure loss of the return system is increased to 0.70 in. WC (174 Pa) and the return air
is exhausted through the return fan and exhaust damper, during air economizer mode operation at
design flow rate, then the return fan total pressure is the same as in the recirculating mode
pret.f  (0.70  0.03)  0.60  0.40  1.67 in. WC (415 Pa)
Pressure-volume characteristics at the key points in air economizer mode at design volume flow,
when the total pressure loss of the return system is 0.15 in. WC (37 Pa) and space air is exhausted
through the exhaust fan, are as follows:
pret, in. WC Point o Point m Point r Point ru
0.15 p, in. WC 0  0.52 0.03  0.12
0.15 V ·, cfm 3,350 20,000 20,000 16,650
0.70 p, in. WG 0  0.6 0.03 1.0
0.70 V ·, cfm 3,350 20,000 20,000 19,650
22.20 CHAPTER TWENTY-TWO
During air economizer mode, if the total pressure loss of the return system is increased to 0.70
in. WC (174 Pa) and the return air is exhausted through the return fan and exhaust damper, at 50
percent of design volume flow, then the fan total pressure of the return fan is
and the pressure-volume characteristics at the key points are as follows:
Controls
As in the supply and relief fan combinations, the controls for supply and return fan combination
systems operated at minimum outdoor air recirculating mode and air economizer mode operations
are discussed in Chap. 23.
22.7 COMPARISON OF THREE FAN COMBINATION SYSTEMS
An air system with supply and exhaust fan combination is simpler and less expensive. Such a system
is not suitable for operation in air economizer mode for a high space pressure. In a supply and
exhaust fan combination, a comparatively lower pressure drop in the return system is necessary to
prevent an unacceptable low pressure in the mixed plenum, such as pm   1 in. WG (250 Pag).
Such low mixed plenum pressure induces air leakage and may impede minimum outdoor air
control.
The operating characteristics of the supply and relief fan combination and the supply and return
fan combination can be compared as follows:
 In a supply and relief fan combination in recirculating mode, total pressure at point ru may be
negative; in a supply and return fan combination, total pressure at ru must be positive, in order to
overcome the pressure loss of the exhaust passage.
 With a relief fan, the mixed plenum pressure pm is always negative. With a return fan, pm may be
positive if the return fan is too large. A positive pm means that there will be no outdoor air intake
into the mixing box. Such a condition should never occur.
 When a negative space pressure must be maintained, a relief fan causes the pressure in the mixed
plenum to be smaller (more negative) than a return fan during minimum outdoor air recirculating
mode.
 When there is a considerable pressure drop in the exhaust or relief flow passage, a return fan requires
high positive pressure at point ru, so energy is wasted across the recirculating damper.
 A supply and relief fan combination with a lower total pressure loss return system is often more
energy-efficient than a supply and return fan combination during recirculating mode operation,
for the following reasons: (1) The pressure drop across the recirculating damper is smaller.
Point o Point m Point r Point ru
p, in. WG 0 0.15 0.03 0.23
V ·
, cfm 10,000 10,000 10,000 9650
pret.f50   9650
20,000 2
(0.70  1.0)  0.03  0.37 in. WC (91 Pa)
Point o Point m Paint r Point ru
p, in. WG 0 0.6 0.03 1.0
V ·
, cfm 20,000 20,000 20,000 16,650
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.21
(2) The relief fan heat gain does not increase coil load. (3) If the return fan is not operating and
there is no bypass passage, there is an additional pressure drop across the idle return fan.
 An axial relief fan is sometimes simpler in layout and installation, and is therefore less expensive
to install, than a centrifugal return fan. If an axial fan is used, noise attenuation must be considered.
Therefore, for those air systems with a smaller total pressure loss in return duct system, usually less
than 0.5 in. WC (125 Pa), especially for those systems also with a considerable pressure drop in the
exhaust or relief flow passage, a supply and relief fan combination is recommended.
For air systems with a greater total pressure loss in the return system, or those that require a negative
space pressure, a supply and return fan combination may be more appropriate.
22.8 PRESSURE FLOW CHARACTERISTICS
Pressure Flow Characterisitcs for a Supply
and Return Fan Combination System
During a supply and return fan combination VAV system design, it is critical to have low enough
negative static pressure in the mixing box to extract the required amount of outdoor ventilation air
and a positive pressure in the return chamber (junction of the return system and exhaust system) to
exhaust the proper amount of return air to outdoors to maintain a predetermined space pressure.
According to the calculations and analyses in the previous sections, to achieve these:
 Correctly divide the system total pressure loss into two parts: (1) that part undertaken by the fan
total pressure of the return fan, which starts from the return slot inlet rt and includes obstructions
in the ceiling plenum, if any, return ducts and fittings, 90° elbow in the return chamber, the
positve pressure required in the return chamber to exhaust return air through the exhaust damper
and louver or the recirculating damper, as shown in Fig. 22.1; and (2) that part undertaken by the
fan total pressure of the supply fan which starts from the negative pressure in the mixing chamber,
total pressure loss of the filter, coils, direct-fired furnace, 90° elbow, supply duct and duct
fittings, flexible ducts, VAV box, diffuser, and the positive pressure maintained in the conditioned
space.
 For a rooftop unit with system components as shown in Fig. 22.1, the total pressure in the return
chamber point ru must be high enough to exhaust part of the return air from the exhaust damper
and louver, or to overcome the total pressure loss of a 90° elbow of airflow and the recirculating
damper during the recirculating mode at summer design conditions. The total pressure loss of the
90° turn and the recirculating damper is often greater.
 In VAV systems, they are often installed with two outdoor dampers: minimum outdoor damper
and 100 percent outdoor damper. During the minimum outdoor air recirculating mode, only the
minimum outdoor air damper is activated. The total pressure loss of the minimum outdoor 0.4 
0.2 = 0.6 damper and louvers is still assumed to be 0.4  0.2 = 0.6 in. WC (150 Pa) or less for
effective control and energy saving. Then a negative pressure of 0.6 in. WC (150 Pa) or less is
needed during minimum outdoor air recirculating mode.
During the air economizer mode, 100 percent outdoor air is required at the design flow rate.
Both outdoor air dampers are opened. A negative pressure of 0.6 in. WC (150 Pa) or less is required
to extract the required amount of outdoor air.
 For a supply and return fan combination VAV system, accurately calculate the total pressure loss
of the system components, supply and return ducts and duct fittings, flexible ducts, VAV box, and
diffusers. It is necessary to draw a system pressure diagram for a supply and return fan combination
system and ensure that the negative pressure, at a low enough magnitude, is in the mixing
box and that there is adequate positive pressure at the junction of the recirculating passage and the
exhaust passage, point ru (return chamber, see Fig. 22.1).
22.22 CHAPTER TWENTY-TWO
System pressure diagram should be drawn by system total pressure. During the construction of
a system pressure diagram of a supply fan, the static regain in the supply duct is obtained at the
expense of the velocity pressure of the supply fan. Only during the minimum outdoor air recirculating
mode is the velocity pressure of the return fan utilized.
 It is often cost-effective to install adjustable-frequency variable-speed drive for both supply and
return fans to adjust the fan speed for large fans during operation. Set the fan total pressure of the
supply and return fans as well as the damper position during commissioning to achieve a required
system pressure distribution and an optimum energy use.
Variation of Pressure in the Mixing Box
The negative total pressure in the mixing box  pm is related to the upstream outside atmospheric
pressure po, all in in. WG (Pa), as follows:
 pm  po  pom (22.6)
where pom  pressure differential between the outside atmosphere and the air pressure in the
mixing box, in. WC (Pa). The negative total pressure in the mixing box  pm is also related to the
outside atmospheric pressure po through the space pressure  pr, in in. WG (Pag), by the following
relationship:
(22.7)
where por  pressure differential between outside atmosphere and space, in. WC (Pa)
pr,s  total pressure loss of return system, in. WC (Pa)
prec,d  total pressure loss of recirculating damper, in. WC (Pa)
During recirculating mode, part-load operation, the volume flow rates of the supply and return air
are reduced, and the total pressure loss of the return system pr,s reduces accordingly. At the same
time, the variation of the mass flow rates of the supply, return, exhaust, and exfiltrated air forms
new airflow balances and new space pressures. Because of the relationship between  pm, pr, and
po, as shown in Eq. (22.7), the result is often a higher pm (less negative pm) and less extracted outdoor
ventilation air through the outdoor dampers at recirculating mode during part-load operation,
as analyzed in Secs. 22.4, 22.5, and 22.6. The opening of the outdoor and recirculating dampers
should be adjusted in order to extract the required amount of outdoor air.
Field Survey of System Pressure Characteristics of a VAV System
Using a Supply and Relief Fan Combination
Figure 22.5 shows the pressure characteristics of a VAV system using a supply and relief fan combination.
This VAV cooling system serves the fifth floor of a high-rise public library with a conditioned
floor area of about 16,500 ft2 (1534 m2). There are two AHUs and two fan rooms, each with
a volume flow rate of about 12,650 cfm (5970 L / s) and a fan total pressure of 4.25 in. WC (1057
Pa) to serve this floor. Each fan room has three relief fans and three relief dampers. The volume
flow rate of relief air is modulated by relief dampers, which are controlled by an electronic controller
actuated by a signal from a pressure sensor in the ceiling plenum. In the perimeter zone, winter
heating is provided by the electric heating coil located downstream from the VAV box.
On September 12, 1983, the system pressure characteristics of the VAV cooling system were
measured. Control of relief dampers was deliberately deenergized. The supply volume flow rate of
AHU 2 measured was about 12,000 cfm (5663 L / s). The pressure differences between various
points in fan room 2 in the VAV cooling system during minimum outdoor air recirculating mode
and 100 percent air economizer cycle were as follows:
pm  pr  pr,s  prec,d
pm  po  por  pr,s  prec,d
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.23
During the air economizer cycle, when the outdoor damper is fully open and the recirculating
damper was closed, the pressure difference between the ceiling plenum and outdoors ppl,o was
0.038 in. WC (9.5 Pa) when only two relief fans were operating, and ppl,o was 0.04 in. WC
(10 Pa) when only one relief fan was operating.
The pressure difference between the conditioned space and the air in the ceiling plenum pr,pl
was so small because of gaps between the ceiling tiles. During recirculation mode, the space pressure
is about 0.064 in. WG (16 Pag). Space pressure was lower when more relief fans were operating
during the air economizer cycle.
22.9 SMOKE CONTROL AND FIRE SAFETY
Fire Safety in Buildings
Fire safety is a critical design factor in high-rise buildings. On November 21, 1980, in Las Vegas,
the MGM Grand Hotel fire took 85 lives. Smoke inhalation is the primary killer in building fires.
According to the annual survey taken by the National Fire Protection Association (NFPA), there
was a slight increase in civilian fire deaths to a total of 4585 people who lost their lives in 1995,
Recirculating mode Air economizer
Outdoor damper 10% open 100% open
Outdoor air, cfm 3330
Recirculating damper 100% open 100% closed
Relief fans operated 1 fan All 3 fans
Relief damper 15% open 100% open
ppl,o in. WC (Pa) 0.06 (15) 0.018 (4.5)
ppl,m in. WC (Pa) 0.20 (50)
po,m in. WC (Pa) 0.14 (35)
pr,pl in. WC (Pa) 0.004 (1) 0.004 (1)
22.24 CHAPTER TWENTY-TWO
FIGURE 22.5 Pressure characteristics of a VAV cooling system using supply-relief fan combination.
which is a 7.3 percent increase over a record low in 1994. In 1995, in the United States, public fire
departments responded to approximately 2 million fires, and property damage due to fire was nearly
$9 billion.
McGreal (1997) noted that building fires produce both smoke and heat. Smoke is the real killer.
Smoke consists of the airborne particulates, in the form of either solid or liquid, and gases which
evolve when a material undergoes pyrolytic combustion.
Studies around 1980 found that more than 80 percent of fire deaths were due to the inhalation of
fire combustion products. About one-half of the deaths are due to the inhalation of carbon monoxide
(CO). Another 16 percent of the deaths were caused by the combined effect of CO and hydrogen
cyanide and heart disease.
Fire protection and fire safety in buildings include the following measures:
 Fire compartmentalization
 Fire-resistant construction
 Fire alarm system
 Automatic sprinkler system
 Smoke control system including stairwell pressurization and zone smoke control system
 Fire protection management and coordination
Smoke control systems, which include stairwell pressurization and zone smoke control, are features
of HVAC&R system design and are discussed here.
National codes and local codes must be followed during the design of the smoke control system
in buildings.
Smoke Movements in Buildings
Figure 22.6 shows smoke movements in a 10-story experimental building. According to ASHRAE
Handbook 1999, HVAC Applications, the typical leakage area ratio Aleak /Afl, which is the ratio of
leakage area Aleak to floor area Afl, both in ft2 (m2) for commercial buildings of average tightness, is as
follows:
External walls, including cracks around
windows and doors 0.00021
Stairwell walls, construction cracks only 0.00011
Elevator shaft walls, construction cracks only 0.00084
Floors, construction cracks, and area around
penetrations 0.000052
The cracks around the stairwell door can be typically taken as 0.25 ft2 (0.023 m2), and the cracks
around each elevator on each floor can also be taken as 0.25 ft2 (0.023 m2).
A building fire can be simulated by burning wooden sticks in a second-floor corridor with a peak
energy release of 900 Btu / s (950 kJ / s) for a certain period while the building is under the following
operating conditions:
 The stair pressurization system and smoke control system are not operated.
 All stairwell doors are closed except that on the second floor, which is opened in. (13 mm) by
the high temperature of the hot gas.
 All windows and exit doors are closed.
 Outdoor temperature is 30°F (1.1°C), and there is a north wind at 5 mph (2.23 m / s).
12
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.25
When the fire is ignited, smoke movements are mainly caused by the expansion and buoyancy
forces of hot gas at a maximum temperature between 850 and 1000°F (454 and 538°C), the stack
effect from outdoor-indoor temperature differences, and the wind effect.
Smoke moves from fire on the second floor to the upper floors through stairwells, elevator
shafts, service shafts, vertical risers, and floor cracks; and smoke discharges to the outdoor atmosphere
through window cracks, elevator machine-room openings, and other openings in the upper
floors. Outdoor air enters the building below the neutral plane and discharges to the outdoors above
the neutral plane because of the stack effect. Oxygen supply to the fire enters through second-floor
window cracks, openings in the building envelope, and vertical air passages from the first and third
floors.
According to Klote (1990), the CO2 and CO levels on the highest floor of the experimental
building during tests without stairwell pressurization and smoke control are 0.15 percent CO2 and
0.015 percent CO. In tests with stairwell pressurization and smoke control, the levels are 0.002 percent
CO2 and 0.001 percent CO.
In winter, the stack effect assists the stairwell pressurization in preventing the smoke from contaminating
the stairwell. This result is verified by experiments in Tamura (1990b).
22.26 CHAPTER TWENTY-TWO
FIGURE 22.6 Smoke movements in a 10-story experimental building.
Effective Area and Flow Rates
In several airflow passages connected in parallel, the effective area Ae, in ft2 (m2), can be calculated as
Ae  A1  A2     An (22.8)
where A1, A2,    , An  airflow areas for path 1, 2,    , n, ft2 (m2)
In airflow paths connected in series, each with a flow area A1, A2,    , An, in ft2 (m2), the effective
area for these airflow paths connected in series Ae can be calculated as
(22.9)
If the flow coefficient is taken as 0.65 and air density 	a  0.075 lb / ft3 (1.2 kg /m3), the air volume
flow rate , in cfm, flowing through a crack, gap, or opening can be calculated as
(22.10)
where A  flow area or effective area, ft2 (m2)
p  pressure difference across flow path or opening, in. WC (Pa)
22.10 EFFECT OF AUTOMATIC SPRINKLER
ON FIRE PROTECTION
Automatic sprinkler systems are effective and reliable fire protection systems and should be
installed in buildings to provide fire protection for the occupants during a building fire. Based on
experimental results, Mawhinney and Tamura (1994) summarized the effect of automatic sprinklers
on fire protection as follows:
 For unshielded fires, automatic fire sprinklers reduced both smoke and fire hazard to negligible
levels. Unshielded fire is a fire exposed to the water sprays from activated sprinklers.
 For shielded fires with activated water sprays from the sprinklers, fire continues to burn at a reduced
rate and to produce smoke until the fuel inside the shielded area has all burned. Concentrations
of CO2 and CO in the smoke were dangerously high. Shield fires represent fires that are
shielded from the water sprays delivered by the sprinklers above because of a desk, table, or other
furniture.
According to experiments, shielded crib fires produced CO concentrations as high as 8 and 9
percent and CO concentrations between 1.0 and 1.5 percent (10,000 and 15,000 ppm) in a multistory
building. Peacock et al. (1989) indicated that a concentration of 15,000 ppm CO would
present a severe life-safety hazard to any one trapped on the fire floor.
 Sprinklers reduced the temperature and radiant heat from shielded fires to nonthreatening levels
within the boundary of the area covered by the sprinkler system. For a sprinkler system design
that meets NFPA Standard 13 requirements, a moderate increase in ventilation to the fire floor
during a building fire should not significantly increase the fire temperatures.
 Buoyancy forces for shielded, sprinklered fire were almost negligible. The recommendation for
the pressure difference between different smoke control zones (fire zone and adjacent zones)
p  0.05 in. WC (12.5 Pa) in NFPA (1988) for zone smoke control design in sprinklered buildings
(ceiling height  9 ft or 2.7 m) is more than sufficient to prevent smoke movement, provided
that the door to the fire floor remains closed.
 Recommended practice for zone smoke control design should allow for some airflow into the
stairwell to prevent the spread of smoke into the stairwell when the door to the fire floor is
opened.
V?
 2610A(p)1 / 2
V?
Ae   1
A1 2 
1
A2 2     
1
An 2 1 / 2
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.27
 The assumption that smoke will never become a threat to life safety in a fully sprinklered building
needs to be reexamined.
22.11 SMOKE CONTROL IN ATRIA
ANSI/NFPA 92A and 92B
The NFPA first published ANSI/NFPA 92A, Recommended Practice for Smoke Control Systems, in
1987. It covered smoke control barriers, airflows, and the pressure differential between different
control zones. It also contained the guidelines to provide a smokeproof enclosure using stairwell
pressurization, as well as smoke control for elevator shafts by using fire floor exhaust, and elevator
lobby pressurization. The latest NFPA 92A edition, in 1996, includes smoke control system supervision
and instrumentation.
Because the smoke control in large zones is quite complicated, another document, ANSI/NFPA
92B, Guide for Smoke Management Systems in Malls, Atria, and Large Areas, was first published in
1991. The latest edition is the 1995 edition.
Smoke Management in Atria, Malls, and Large Areas
Klote (1997) and ANSI/NFPA 92B recommended the following smoke management in atria:
 Exhaust the smoke from the top of the atrium in order to achieve a steady, clear height for a
steady fire. Steady fire is an idealization of actual unsteady fire. Steady fire has a constant heat
release rate. Consider that the only flow into the smoke layer formed at the upper layer of a
atrium is from the plume, and the only flow from the smoke layer is the smoke exhaust. Plume
comprises the combustion products above the fire. The exhaust flow must equal the flow from
the plume.
This method and approach have many simplified assumptions, such as that the simple plume
mass flow equation is valid, there is a constant heat release rate, clear height is greater than the
mean flame height, smoke layer is adiabatic, and plume flow and exhaust are the only significant
mass flows into or out from the smoke layer.
 Upper-layer exhaust is not necessary for an atrium with sufficient smoke-filling capacity. Occupants
have time for both decision making and evacuation before the smoke fills the place where
they are located. Various approaches to smoke filling can be used to conservatively estimate the
filling time.
 Air can be supplied to the communicating space to provide a specific average velocity at the
opening to the atrium to prevent smoke flowing from the atrium to the communicating space.
 When the above basic methods are not applicable, physical modeling and computing fluid dynamics
(CFD) analysis can be used.
 Because a layer of hot air is formed under the ceiling of an atrium, and this layer of hot air
prevents the smoke from reaching the ceiling, ceiling-mounted smoke detectors are usually not
recommended for atrium applications. Beam smoke detectors mounted on balconies oriented
horizontally are recommended to detect smoke in the plume.
 The exhaust system can pull some air below the smoke layer into the exhaust inlet if the smoke
layers are relatively thin. This plugholing of outdoor air can lower the smoke layer and expose occupants
to smoke. Klote (1999) noted the maximum flow Qsmoke of smoke without plugholing depends
on the depth of the smoke layer and the temperature of the smoke. If the total smoke exhaust
needed is greater than Qsmoke, a number of inlets to compensate exhaust are needed. These
inlets should be placed far enough from each other to avoid influence on their flow.
22.28 CHAPTER TWENTY-TWO
22.12 STAIRWELL PRESSURIZATION AND
ZONE SMOKE CONTROL
Stairwell Pressurization
A stairwell pressurization system uses fans to pressurize the stairwells to provide a smokefree escape
route for the occupants in case of a building fire. A stairwell pressurization system is a kind of
smoke control system. National and local codes require stairwell pressurization systems in high-rise
buildings. In a stairwell pressurization system, all interior stairwells are pressurized to a minimum
of 0.15 in. WC (37 Pa) and a maximum of 0.35 in. WC (87 Pa) when all stairwell doors are closed.
City of New York Local Code 1979 requires a minimum air supply flow rate of 24,000 cfm
(11,326 L / s) plus 200 cfm (94 L / s) per floor for the stairwell pressurization system. The maximum
allowable pressure difference between the stairwell and the floor space is 0.40 in. WC (100 Pa)
whether stairwell doors are opened or closed. The minimum allowable pressure difference is 0.10
in. WC (25 Pa) when all stairwell doors are closed, or 0.05 in. WC (13 Pa) when any three stairwell
doors are open. An alternative is to maintain at least 0.05 in. WC (13 Pa) or a minimum average air
velocity of 400 fpm (2 m / s) at the stairwell door when any three stairwell doors are opened. The
force required to open a stairwell door must not exceed 25 lbf (111 N) at the doorknob.
A stairwell pressurization system consists of centrifugal or vane-axial fans, a stairwell pressurization
supply duct with several supply air inlets, relief vents, and a control system, as shown in
Fig. 22.7. Outdoor air is extracted directly by the centrifugal fan. It is forced into the supply duct
and then supplied to the stairwell through supply inlets. When the stairwell is pressurized to a pressure
typically 0.10 to 0.40 in. WC (25 to 100 Pa) higher than that of the air outside the stairwell on
various floors across the stairwell doors, the smoke will not enter the stairwell, even an open stairwell
door. Air supplied into the stairwell is discharged through the open stairwell doors, leakage
area around closed stairwell doors, relief vents, or other openings to the rest of the building, and
then is discharged to the outdoors.
If the stairwell is an airtight enclosure or its doors have very small leakage areas, it will be overpressurized
when all stairwell doors are closed. The pressure difference across the stairwell doors
may be greater than 0.3 in. WC (75 Pa). Often, a large force is required to turn the stairwell doorknob
to open it. The total force required to open a stairwell door should not exceed 25 to 30 lbf (111
to 134 N), or it will be too difficult to open stairwell doors during evacuation. Methods of overpressure
relief are discussed later.
The air velocity at the opened stairwell door on the fire floor required to prevent the backflow of
smoke from the fire to the stairwell is called the critical velocity vcrit, in fpm (m / s). The outward
flow air velocity from the stairwell through the open door on the fire floor should be greater
than vcrit.
Characteristics of Stairwell Pressurization
Centrifugal or vane-axial fans can be used for stairwell pressurization. Fans can be installed either
at the bottom level of the building (bottom injection) or on the rooftop (top injection). Bottom injection
is preferable because it minimizes the possibility of smoke inhalation and optimizes the
stack effect to assist stairwell pressurization during winter. In any case, the fan intake must be remote
from the smoke exhaust during a building fire. If the fan room for stairwell pressurization is
located on the rooftop, facilities must be provided to minimize the influence of wind pressure on
fan performance.
Multiple injections, in which air is supplied from multiple inlets into the stairwell, provide a
more even pressure distribution along the stairwell than a single injection from the top or bottom.
Typically, each supply inlet serves two or three floors.
Open-tread stairs provide less flow resistance than closed-tread stairs. This difference becomes
more prominent when occupants are walking on the stairs during evacuation. Compartmentalization
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.29
of the stairwell into many sections, such as serving four to five floors, may not provide the airflow rate
necessary for stairwell pressurization when two or three stairwell doors are open at the same time.
Overpressure Relief and Feedback Control
When a stairwell pressurization system is operating and all the stairwell doors are closed, the pressure
difference across the stairwell doors eventually exceeds the maximum permissible limit. Two
methods are currently used to relieve the stairwell pressure. Overpressure relief is achieved by
opening first-floor exit doors (as shown in Fig. 22.7a) or overpressure relief vents (barometric
dampers, as shown in Fig. 22.7b). Variable volume of supply air is achieved by means of feedback
control (as shown in Fig. 22.7a).
Exit Door Relief. When a stairwell pressurization system is turned on, the interlocked control system
automatically opens the first-floor exit door to relieve stairwell pressure. Overpressurized air in the
stairwell is discharged to the outdoors. Because it is necessary to open the first-floor exit door to evacuate
the occupants during a building fire, this is a simple and effective means of overpressure relief.
Test results in Tamura (1990b) showed that when stairwell pressurization was activated and the
first-floor exit door was used as overpressure relief, the maximum pressure difference across the
22.30 CHAPTER TWENTY-TWO
FIGURE 22.7 Stairwell pressurization and zone smoke control systems: (a) fan bypass overpressure relief; (b) overpressure
relief vents (barometric damper).
stairwell doors was lower than 0.25 in. WC (65 Pa) in summer and winter. During nonfire conditions,
pressure differences were between 0.05 and 0.1 in. WC (13 and 25 Pa) across the stairwell
doors if there was a vent in the external wall on the second floor. When a building fire was simulated
on the second floor, there was no smoke backflow from the fire floor through the stairwell
door when the stairwell doors on the first and second floors were opened and the vent on the second
floor external wall was also opened. Smoke backflow occurred when stairwell doors on the first,
second, and third floors were open.
Overpressure Relief Vents. A relief vent is typically an assembly in which a fire damper is connected
in series with a barometric damper. The barometric damper is normally closed. If the stairwell
is pressurized above a predetermined limit, the vents open and relieve air to various building floors. A
counterweight in the barometric damper sets the maximum pressure limit. Fire dampers are normally
closed, and open when the stairwell pressurization system is turned on in case of a building fire.
According to the experimental results of Tamura (1990b), the performance of the overpressure
relief vents at a supply flow rate of 28,000 cfm (13,210 L / s) was slightly better than the performance
of a ground-floor exit door at a supply flow rate of 17,800 cfm (8400 L / s). If supply flow
rates are the same, the difference is further reduced.
Fan Bypass. When the stairwell is overpressurized, a pressure sensor located inside the stairwell
signals a feedback control to open a fan bypass damper so that part of the supply air returns to the
centrifugal fan inlet, as shown in Fig. 22.7a. The air volume flow supplied to the stairwell is reduced
until the pressure at the pressure sensor drops below a preset value. If the pressure sensor is
located on the rooftop or outdoors, it should be shielded from the influence of wind.
Variable-Speed or Controllable-Pitch Fan. When excessive pressure is detected in the stairwell,
a controller actuates an adjustable-frequency variable-speed drive to reduce the speed of a centrifugal
fan or to vary the blade pitch of a vane-axial fan to maintain the required pressure inside the
stairwell at the pressure sensor. The test results in Tamura (1990b) showed that both fan bypass and
variable-speed controls require a response time of more than 5 min. Variable-speed control is
slightly faster. Because reliability is the primary factor in stairwell pressurization control, and because
of its very short operating period, it may not be worthwhile to install an expensive adjustablefrequency
ac inverter to provide variable-speed control.
Stair and Shaft Vents
ASHRAE/IESNA Standard 90.1-1999 mandatorily specifies that stair and elevator shaft vents shall
be equipped with motorized dampers which are automatically closed during normal building operation
and are interlocked to open by fire and smoke control detection systems when required.
Zone Smoke Control
Zone smoke control requires that a building be divided into a number of smoke control zones. The
control zones are separated from one another by barriers, such as walls, floors, ceilings, and doors.
In high-rise buildings, each floor is a separate zone, or it may be subdivided into many smoke control
zones, or a smoke control zone consists of more than one floor.
In case of a building fire, the spread of smoke from the zone of fire origin (the smoke zone)
to adjacent zones is limited by pressure differences and airflows. The required pressure differential
depends on the pressure produced by the fire gases, i.e., the temperature of the building fire. To
form a pressure difference between the smoke zone and adjacent control zones, zone smoke
control provides smoke exhaust on the fire floor (smoke zone) by opening the smoke damper
connected to the smoke exhaust duct. At the same time, it supplies outdoor air to the floor or
floors above and below the fire floor (to control zones adjacent to the smoke zone) and pressurizes
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.31
them to prevent smoke contamination by operating the air-handling units (AHUs) on these
floors and closing the smoke dampers connected to the smoke exhaust duct, as shown in Fig. 22.7a.
During the smoke control mode, recirculation dampers in these AHUs are fully closed. Outdoor
air should be extracted directly through the coils, supply fans, and supply to the duct diffusers
to pressurize these control zones These mechanisms should be controlled by a DDC unit for fire
protection management overriding all other HVAC&R controls during a building fire.
Lougheed (1997) noted that the 0.05 in. WC (12.5 Pa) minimum design pressure difference
between the smoke zone and adjacent zones suggested by ANSI/NFPA 92A (1993) would be adequate
for zone smoke control applications in sprinklered office buildings. In most cases, a lower
pressure of approximately 0.028 in. WG (7 Pa) would still exceed the pressure produced by the fire
gases. Lougheed (1997) recommended a minimum required pressure difference of 0.1 in. WC (25
Pa) for an unsprinklered building with a ceiling height of 9 ft (2.74 m).
The supply and exhaust volume flow rates for zone smoke control are often matched with the
HVAC&R system in buildings, especially when an air economizer cycle is used. The exhaust air
from the AHU can be connected to a smoke exhaust duct, and the exhaust fan is generally located
on the rooftop.
In an air system operated only in recirculation mode, a smoke exhaust system for zone smoke
control should be installed. In such circumstances, the smoke exhaust volume flow rate can be determined
to equal about 6 air changes per hour (ach).
Design Considerations
The object of smoke control is to provide a smokefree escape route for occupants through the stairwell
to the outdoors during a building fire. This is a part of the building fire protection scheme. The
primary considerations are safety and reliability. Because the performance of stairwell pressurization
and the zone smoke control are related, they should be considered an integrated smoke control
system during system design.
Operation of both stairwell pressurization and zone smoke control systems must meet the
requirements of national and local codes and provide a smokefree escape route even under the following
conditions:
 A fire reaches a fire temperature of up to 1200°F (650°C) when the sprinkler system fails to
operate.
 Three or four stairwell doors, including the door on the fire floor, are opened simultaneously.
 Fire floor smoke exhaust is performed mainly by a zone mechanical smoke exhaust system.
 During summer operating conditions, the stack effect does not act as an additional assistant.
Recently, it has been recommended that smoke control technology be extended to elevator shafts
and lobbies used to evacuate disabled persons during a building fire. An elevator shaft and elevator
lobby pressurization system should then be installed.
During a building fire, HVAC&R should operate according to fire safety and smoke control
requirements. Smoke control systems are automatically actuated by a water flow indicator from
the automatic sprinkler system or from an area smoke detector. It is important to have feedback
mechanisms to verify system operation and performance, such as separate alarms to signal smoke
migration or burning fire, and adequate pressure level at key points during pressurization and
evacuation.
Volume Flow Rate
Volume flow rate is the primary determinant of the performance of stairwell pressurization and zone
smoke control systems. The volume flow rate of a zone smoke exhaust system from the fire floor is
22.32 CHAPTER TWENTY-TWO
usually 6 ach. This volume flow rate is used by many designers and has been verified as appropriate
in field tests by Tamura (1991) and Klote (1990, Part II). The volume flow rates of air supplied to
the floors immediately above and below normally should be the same as the supply volume flow
rate of the AHU or packaged unit for that floor.
Air discharged through the open stairwell doors and leaked through the closed stairwell doors
and walls can be summarized into the following types:
 Air discharged through an open stairwell door on the fire floor with a critical velocity vcrit
 Air discharged through the open first-floor exit door
 Air discharged through open stairwell doors on floors other than the fire floor and the first floor
 Air leaked through cracks around the stairwell door and in the stairwell wall
The volume flow rate supplied to a stairwell pressurization system then can be calculated as follows:
(22.11)
where vexit  average air velocity at first-floor exit door, fpm (m/ s)
Adoor  area of opened stairwell door on fire floor, ft2 (m2)
 volume flow rate discharged through open stairwell door on floors other than fire floor
and first floor, cfm (m3 /s)
No,d  number of open doors in stairwell other than fire floor and first floor
 air leakage through cracks across stairwell wall on floors with closed stairwell doors,
cfm (m3 /s)
Nc,d  number of closed doors in the stairwell
According to the experimental results in Tamura (1991), for a fire in a building with stairwell pressurization
and a smoke exhaust system to exhaust smoke from the fire floor, the critical velocity vcrit
can be taken as 300 fpm (1.5 m / s) for a fire temperature of 1200°F (650°C) and a mechanical exhaust
from the fire floor of 5.5 ach.
In stairwell pressurized systems serving up to 10 floors, No,d  3, including the door on the fire
floor. In systems serving 15 or more floors, No,d  4, including the door on the fire floor.
For safety, air velocity at the first-floor stairwell exit door vexit can be assumed to be equal to vcrit.
Discharge air velocity at open stairwell doors on floors other than the fire floor vo,d is less than vcrit,
in fpm (m / s). This is because the flow resistance of the cracks around doors and windows in floors
other than the fire floor is greater than that on the fire floor.
The air volume flow rate discharged through an open stairwell door other than that on the fire
floor can be calculated from Eq. (22.10) by using the effective area Ae of the airflow path instead of
the area of the door or opening A. Air leakage through cracks in the stairwell wall can also be calculated
from Eq. (22.10).
System Pressure Loss for Stairwell Pressurization System
The frictional loss of airflow per floor inside the stairwell pressurization system pf,s, in in. WC
(Pa), can be calculated by considering the stairwell as a rectangular duct, as follows:
(22.12)
where L  vertical height of each floor, ft (m)
gc  dimensional constant, 32.2 lbmft / lbf  s2
pf,s  K  1
5.19  L
De 
	a
2gc  V?
60Ao 2
V?
leak
V?
o,d
V?
s,p  (vcrit  vexit)Adoor  V?o,dNo,d  V?leakNc,d
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.33
	a  air density, lb/ ft3 (kg /m3)
 air supply volume flow rate, cfm (m3 /s)
Ao  cross-sectional area of free air passage in stairwell, ft2 (m2)
and the ratio R of the area of free flow passage (orifice) to the interior cross-sectional area of the
stairwell As, in ft2 (m2), can be calculated as
R  Ao /As (22.13)
In Eq. (22.12), De, in ft (m), indicates the circular equivalent of the stairwell, and it can be calculated as
(22.14)
where Ps  perimeter of the cross-sectional area of the stairwell, ft (m).
The pressure drop coefficient is K. It is mainly a function of the configuration of the stairwell
and the occupant density on the stairs. According to the experimental results of Achakji and Tamura
(1988), for a stairwell tested under the following conditions
 With a cross-sectional area of 134 ft2 (12.5 m2)
 A floor height of 8.5 ft (2.6 m)
 An occupant density of 0.18 person / ft2 (2.0 persons /m2)
the frictional loss of airflow per floor pf,s has the following values:
Here vo represents the air velocity calculated based on the free flow area, in fpm (m / s). The difference
in pf,s between top and bottom injection is small if other conditions remain the same.
When calculating the system pressure loss of a stairwell pressurization system with multiple injection
psy, in in. WC (Pa), as shown in Fig. 22.6, is calculated as
psy  pa  b  pb  o (22.15)
where pa  b, pb  o  pressure loss between points a and b and points b and o, respectively, in.
WC (Pa) Pressure loss pb  o usually varies between 0.10 and 0.40 in. WC (25 and 100 Pa) and is
only a small portion of the system pressure. However, it is difficult to calculate accurately because
of complicated airflow paths at various operating conditions and the influence of outdoor conditions.
Pressure loss pa  b should be calculated between points a and b according to the procedure
described in Chap. 17, including the velocity pressure at the supply inlet, and pb  o should be estimated
as 0.4 in. WC (100 Pa). The calculated system pressure psy should be multiplied by a
safety factor 1.2. The actual pb  o can be adjusted during acceptance testing.
In a stairwell pressurization system with a bottom single injection, as shown in Fig. 22.8a, the
maximum pressure difference between the stairwell and outdoors often occurs at the top of the
stairwell pc  o because the cross section of the stairwell is constant. The system pressure can
therefore be calculated as
psy  pa  b  pb  c  pc  o (22.16)
vo, fpm (m/ s) pf,s, in. WC (Pa)
Open-tread stair, bottom injection 879 (4.47) 0.128 (32)
659 (3.35) 0.070 (17.5)
439 (2.23) 0.032 (8)
Closed-tread stair, bottom injection 1055 (5.36) 0.200 (50)
848 (4.31) 0.116 (29)
527 (2.68) 0.05 (12.5)
De 
4As
Ps
V?
22.34 CHAPTER TWENTY-TWO
Here, for simplicity, pc  o again can be taken as 0.4 in. WC (100 Pa). Then
pb  c  pf,sNfl (22.17)
where Nfl  number of floors. Actually, part of the velocity pressure discharged from the bottom inlet
may be converted to static pressure, so that pb  c is smaller. However, more data are needed
before it can be calculated accurately.
In a stairwell pressurization system with top and bottom injection, as shown in Fig. 22.8b, the
system pressure loss can be calculated as it is in bottom injection, but pb  c may be smaller.
Computer-aided design software for zone smoke control systems are available to assist designers.
Example 22.1. In a smoke control system serving a 20-story high-rise building, each floor has the
following construction characteristics:
Floor area 10,000 ft2 (929 m2)
Area of the external wall 4800 ft2 (446 m2)
Vertical distance between floor and floor 10 ft (3m)
Area of the stairwell wall 600 ft2 (56 m2)
Elevator shaft (three elevators) wall area 1000 ft2 (93 m2)
Area of stairwell door 3 	 7  21 ft2 (2 m2)
Volume flow rate of supply air to conditioned
space through AHU on each floor 12,000 cfm (5663 L / s)
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.35
FIGURE 22.8 Bottom injection, and bottom and top injection: (a) bottom injection; (b) bottom and top
injection.
The pressure difference across the closed stairwell door is 0.1 in. WC (25 Pa). The pressure differences
between the stairwell and the outdoors on floors other than the fire floor whose stairwell
doors remain open are also 0.1 in. WC (25 Pa), because the pressure drop of the open stairwell door
is ignored. Determine the volume flow rates of the zone smoke control and stairwell pressurization
systems.
Solution
1. The volume flow rate of the zone smoke exhaust system from the fire floor at a rate of 6
ach is
The volume flow rate of air supplied to the floors immediately above and below the fire floor is
12,000 cfm (5663 L / s).
2. The volume flow rate of air that is discharged through the open stairwell door to the fire floor
, in cfm (L / s), and is exhausted through the zone smoke control system is
3. The volume flow rate discharged from the first-floor exit door is assumed to be the same as
that from the open stairwell door on the fire floor (6300 cfm, or 2973 L / s).
4. From the given data, the leakage area on the stairwell wall, including cracks around the stairwell
door, is
Aleak  0.25  0.00011 	 600  0.32 ft2
If the pressure difference across the closed stairwell door is 0.1 in. WC, from Eq. (22.10), the air
leakage rate leak through the cracks of the stairwell wall on each floor whose stairwell door is
closed can be calculated:
5. It is possible that the fire floor and the pressurized floors immediately above and below are all
in the lower 10 floors. Because of the stack effect in this 20-story building, air mixed with hot gas
from the burning fire is discharged from the upper 10 stories.
In a 20-story building, assume that there are four open stairwell doors. One is on the fire floor,
and another is the first-floor exit door. Then the remaining two opened stairwell doors should each
discharge a combined air path that may be a combination of six parallel paths, as shown in
Fig. 22.9.
a-b-o
a-c-d-o
a-c-e-f-o
a-g-h-o
a-g-i-j-o
a-k-l-o
6. The leakage area for the external wall for each floor is
Aleak  4800 	 0.00021  1.0 ft2
The leakage area for a floor area of 10,000 ft2 is
Aleak  10,000 	 0.000052  0.52 ft2
V?
leak  2610 A(p)1/2  2610 	 0.32(0.1)1 / 2  264 cfm (125 L / s)
V?
V?
fire  vcrit Adoor  300 	 3 	 7  6300 cfm (2973 L/s)
V?
fire
V?
ex  6 	 10,000 	
10
60
 10,000 cfm (4719 L / s)
V?
ex
22.36 CHAPTER TWENTY-TWO
The leakage area to the elevator shaft is
Aleak  3 	 0.25  1000 	 0.00084  1.59 ft2
If the total opening at the top of the elevator shaft is 5 ft2, assume that only 0.5 ft2 will be allocated
for the open stairwell door. For an airflow path a-b-o, if the pressure drop at the open door is negligible,
the effective area is
For airflow path a-c-d-o, the effective area is
For airflow path a-c-e-f-o, the effective area is
For airflow path a-k-l-o, the effective area is
The effective area of airflow path a-g-h-o is the same as that of a-c-d-o, that is, 0.46 ft2; and the
effective area for a-g-i-j-o is the same as that of a-c-e-f-o, that is, 0.35 ft2. Therefore, the total
Ae   1
1.592 
1
0.52 1 / 2
 0.48 ft2
  1
0.522 
1
0.522 
1
12 1 / 2
 0.35 ft2
Ae   1
A1 2 
1
A2 2 
1
A3 2 1 / 2
Ae   1
A1 2 
1
A2 2 1 / 2
  1
0.522 
1
12 1 / 2
 0.46 ft2
Ae   1
A1 2 1 / 2
 1 ft2
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.37
FIGURE 22.9 Air discharged from an open stairwell door through a
combined air path to the outdoors.
effective area of this combined airflow path is
Ae  2 	 1  2 	 0.46  2 	 0.35  0.48  4.1 ft2 (0.38 m2)
7. From Eq. (22.10), air discharged from this opened stairwell door is
8. From Eq. (22.11), the total supply volume flow rate of this stairwell pressurization system is
then
REFERENCES
Achakji, G. A., and Tamura, G. T., Pressure Drop Characteristics of Typical Stairshafts in High-Rise Buildings,
ASHRAE Transactions, 1988, Part I, pp. 1223–1237.
Alcorn, L. H., and Huber, P. J., Decoupling Supply and Return Fans for Increased Stability of VAV Systems,
ASHRAE Transactions, 1988, Part I, pp. 1484–1492.
Alley, R. L., Selecting and Sizing Outside and Return Air Dampers for VAV Economizer Systems, ASHRAE
Transactions, 1988, Part I, pp. 1457–1466.
ASHRAE, ASHRAE Handbook 1992, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1992.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
Avery, G., VAV Economizer Cycle: Don’t Use a Return Fan, Heating / Piping / Air Conditioning, no. 8, 1984,
pp. 91–94.
Clark, D. R., Hurley, C. W., and Hill, C. R., Dynamic Models for HVAC System Components, ASHRAE
Transactions, 1985, Part I, pp. 737–751.
Clark, J. A., and Harris, J. W., Stairwell Pressurization in a Cold Climate, ASHRAE Transactions, 1989, Part I,
pp. 847–851.
Grimm, N. R., and Rosaler, R. C., HVAC Systems and Components Handbook, 2d ed., McGraw-Hill, New York,
1998.
Kalasinsky, C. C., The Economics of Relief Fans vs. Return Fans in Variable Volume Systems with Economizer
Cycles, ASHRAE Transactions, 1988, Part I, pp. 1467–1476.
Kettler, J. P., Field Problems Associated with Return Fans on VAV Systems, ASHRAE Transactions, 1988, Part I,
pp. 1477–1483.
Klote, J. H., An Overview of Smoke Control Technology, ASHRAE Transactions, 1988, Part I, pp. 1211–1222.
Klote, J. H., Fire Experiments of Zoned Smoke Control at the Plaza Hotel in Washington, D.C., ASHRAE
Transactions, 1990, Part II, pp. 399–416.
Klote, J. H., Prediction of Smoke Movement in Atria: Part I—Physical Concepts and Part II—Application to
Smoke Management, ASHRAE Transactions, 1997, Part II, pp. 534–553.
Klote, J. H., What’s New in Atrium Smoke Management, HPAC, no. 4, 1999, pp. 28–31.
Klote, J. H., and Tamura, G. T., Design of Elevator Control Systems for Fire Evacuation, ASHRAE Transactions,
1991, Part II, pp. 634–642.
Kukla, M. D., Situations to Consider When Variable Air Volume Is an Option, ASHRAE Transactions, 1997, Part
II, pp. 823–829.
Lehr, V. A., Life Safety in Tall Buildings, Heating / Piping / Air Conditioning, April 1990, pp. 41–46.
Lougheed, G. D., Expected Size of Shielded Fires in Sprinklered Office Buildings, ASHRAE Transactions,
1997, Part I, pp. 395–410.
 (300  300)3 	 7  2 	 3384  264(20  4)  23,592 cfm (11,133 L / s)
V?
s,p  (vcrit  vexit)Adoor  V?o,dNo,d  V?leakNc,d
V?
 2610A(p)1 / 2  2610 	 4.1(0.10)1 / 2  3384 cfm (1597 L / s)
22.38 CHAPTER TWENTY-TWO
Mawhinney, J. R., and Tamura, G. T., Effect of Automatic Sprinkler Protection on Smoke Control Systems,
ASHRAE Transactions, 1994, Part I, pp. 494–513.
McGreal, M. P., Engineered Smoke Control Systems, HPAC, no. 4, 1997, pp. 69–72.
NFPA, Guide for Smoke and Heat Venting, ANSI/NFPA Standard, 204, Quincy, MA, 1988.
Peacock, R. D., Bukowski, R. W., Jones,W. W., and Forney, C. L., Technical Reference Guide for HAZARD I
Fire Hazard Assessment Method, NIST Handbook, National Institute of Standards and Technology, Gaithersburg,
MD, 1989, chap. 7, p. 146.
Schwartz, K. J., Jensen, R. H., and Antell, J., The Role of Dampers in a Total Fire Protection System Analysis,
ASHRAE Transactions, 1986, Part I B, pp. 566–576.
Shavit, G., Information-Based Smoke Control Systems, ASHRAE Transactions, 1988, Part I, pp. 1238–1252.
Tamura, G. T., Stair Pressurization Systems for Smoke Control: Design Considerations, ASHRAE Transactions,
1989, Part II, pp. 184–192.
Tamura, G. T., Field Tests of Stair Pressurization Systems with Overpressure Relief, ASHRAE Transactions,
1990a, Part I, pp. 951–958.
Tamura, G. T., Fire Tower Tests of Stair Pressurization Systems with Overpressure Relief, ASHRAE Transactions,
1990b, Part II, pp. 373–383.
Tamura, G. T., Fire Tower Tests of Stair Pressurization Systems with Mechanical Venting of the Fire Floor,
ASHRAE Transactions, 1990c, Part II, pp. 384–392.
Tamura, G. T., Determination of Critical Air Velocities to Prevent Smoke Backflow at a Stair Door Opening on
the Fire Floor, ASHRAE Transactions, 1991, Part II, pp. 627–633.
The Trane Company, Air Conditioning Fans, The Trane Company, La Crosse, WI, 1985.
Wang, S. K., Air Conditioning, vols. 2 and 4, Hong Kong Polytechnic, Hong Kong, 1987.
AIR SYSTEMS: FAN COMBINATION AND SMOKE CONTROL 22.39
CHAPTER 23
AIR SYSTEMS: MINIMUM
VENTILATION AND VAV
SYSTEM CONTROLS
23.1
23.1 MINIMUM VENTILATION
CONTROL 23.2
Basic Approach 23.2
Specific Controls in VAV Systems 23.2
Types of Ventilation Control 23.2
ASHRAE Standard 62-1999 23.3
Recirculation of Unused Outdoor Air in
Multizone Systems 23.4
Ventilation Controls for High-Occupancy
Areas 23.5
23.2 CO2-BASED DEMAND-CONTROLLED
VENTILATION 23.5
CO2-Based Demand-Controlled
Ventilation or Time-Based
Constant-Volume Control 23.5
CO2 Sensor or Mixed-Gases Sensor 23.7
Location of CO2 Sensor 23.7
Substantial Lag Time in Space CO2
Concentration Dilution Process 23.8
Base Ventilation and Purge Mode 23.9
A CO2-Based Demand-Controlled
Ventilation System 23.10
Application of CO2-Based Demand-
Controlled Ventilation System 23.11
23.3 MIXED PLENUM PRESSURE
CONTROLS 23.12
Basics 23.12
Case Study: Mixed Plenum Pressure
Control Monitoring Plenum
Pressure 23.12
Monitoring Pressure Drop of Louver
and Damper Controlling Mixed Plenum
Pressure 23.13
Applications 23.14
23.4 OUTDOOR AIR INJECTION FAN,
DIRECT MEASUREMENT, AND FAN
TRACKING SYSTEMS 23.14
Outdoor Air Injection Fan 23.14
Direct Measurement of Minimum
Outdoor Air Intake 23.15
Fan Tracking Systems 23.15
23.5 CONFERENCE ROOMS 23.16
23.6 SPACE PRESSURIZATION AND
RETURN VOLUME FLOW
CONTROLS 23.16
Characteristics of Space Pressure
Control 23.16
VAV Systems Return/Relief Fan
Volume Flow Control 23.17
23.7 DISCHARGE AIR TEMPERATURE
CONTROLS 23.18
Basics 23.18
System Description 23.19
Operation of Air Economizer
and Outdoor Air Intake—Case
Study 23.21
Discharge Air Temperature
Reset 23.22
23.8 DUCT STATIC PRESSURE AND
FAN CONTROLS 23.23
Duct Static Pressure Control 23.23
Set Point of Duct Static
Pressure and Sensor’s
Location 23.24
Comparison between
Adjustable-Frequency Drives
and Inlet Vanes 23.24
23.9 FUNCTIONAL CONTROLS FOR
VAV SYSTEMS 23.26
Nighttime Setback, Warm-Up, and
Cool-Down Control 23.26
Steam Humidifier Control 23.27
Dew Point Control 23.27
Diagnostics 23.28
23.10 RECOMMENDATIONS AND
INTERACTION BETWEEN
CONTROLS 23.28
Recommendations for VAV
Controls 23.28
Interaction between
Controls 23.29
Stability Problems 23.30
23.11 SEQUENCE OF OPERATIONS
OF A VAV REHEAT SYSTEM
WITH MINIMUM VENTILATION,
DISCHARGE AIR TEMPERATURE,
AND DUCT STATIC PRESSURE
CONTROLS—CASE STUDY 23.30
HVAC&R System 23.30
Sequence of Operations 23.30
Primary Considerations 23.34
REFERENCES 23.35
23.1 MINIMUM VENTILATION CONTROL
Basic Approach
As discussed in Sec. 4.10, indoor air quality (IAQ) is defined as an indication of harmful concentration
of indoor air contaminants, such as particulates, combustion products, volatile organic
compounds, nicotine, radon, and biological compounds that affects the health of the occupants or the
degree of satisfaction of a substantial majority of occupants exposed to such an indoor environment.
Contaminated source control, removal by air cleaners, and using outdoor ventilation air to dilute
the concentrations of indoor air contaminants are three basic strategies to improve IAQ. Emissions
and odors from the human body, volatile organic compounds, and many gaseous contaminants of
0.003- to 1-m minute particles can only be removed by expensive, high-efficiency air filters and
activated carbon filters. Use of adequate outdoor air to dilute the air contaminants has been proved
an essential, practical, and cost-effective means to improve IAQ.
Among the air systems, ventilation controls for constant-volume systems and dedicated ventilation
systems can be achieved either by fixing the outdoor and recirculating damper position during
the operating period or modulating the fan speed to match the variation of outdoor air supply. Also,
during the air economizer cycle, outdoor air intake in the AHU or PU is always greater than the
specified amount of outdoor air for ventilation required. However, the reduction in supply volume
flow rate at part-load operation of variable-air-volume (VAV) systems, a tighter building shell, and
the ineffective ventilation control may reduce the amount of outdoor air intake to less than the
requirement for ventilation. Minimum ventilation control (control in minimum outdoor air recirculating
mode) becomes one of the critical problems in VAV systems in commercial buildings. As
discussed in Sec. 4.10, ventilation means supplying and removing ventilation air, and ventilation air
consists of outdoor air plus treated recirculated air for acceptable indoor air quality.
Specific Controls in VAV Systems
In a multizone VAV system, zone temperature control directly affects the thermal comfort of the occupants
and is the basis of the VAV system control. During the minimum ventilation control, only
the amount of minimum outdoor air supply to a conditioned space is controlled; the space airflow
balance and space pressurization vary accordingly. There are interactions between ventilation and
space pressurization controls.
Zone temperature control is achieved by modulating the volume flow rate supplied to each
control zone. As the variation of zone load causes the change of damper positions and the airflow
passages as well as the zone volume flow in VAV boxes, the duct static pressure in the main supply
duct may rise or fall. Duct static pressure control then modulates the supply fan speed and thus the
system supply volume flow rate and fan total pressure to match the variation in zone loads.
Discharge air temperature control adjusts the cooling or heating capacity of a VAV system to
match the system load. It also resets the discharge air temperature and optimizes energy use. Therefore,
specific controls for a VAV system include
 Zone temperature controls
 Minimum ventilation control and space pressurization control
 Discharge air temperature control, including economizer, cooling capacity, mixed air, and heating
capacity controls in sequence, and low-temperature limit control
 Duct static pressure control
Types of Ventilation Control
In VAV systems, ventilation control includes air economizer control and minimum outdoor ventilation
air control during the recirculation mode. During the air economizer cycle including mixed air
23.2 CHAPTER TWENTY-THREE
control, a volume flow of outdoor air that exceeds the amount of minimum outdoor ventilation air is
always supplied to the conditioned space. Ventilation problems often occur during minimum outdoor
air recirculating mode, i.e., minimum ventilation controls. Currently used minimum ventilation
controls in VAV systems include the following:
CO2-based demand-controlled ventilation
Mixed plenum pressure controls
Outdoor air injection fan control
Fan tracking control
ASHRAE Standard 62-1999
As discussed in Sec. 4.10, ASHRAE Standard 62-1999, Ventilation for Acceptable Indoor Air Quality,
specifies two alternative procedures to obtain acceptable IAQ:
1. Ventilation rate procedure. In the ventilation rate procedure, acceptable indoor air quality is
achieved by providing ventilation air of specified quality and quantity to the conditioned space.
This procedure prescribes the following:
 The quality of the outdoor air shall be acceptable for ventilation, i.e., contaminants in outdoor air
do not exceed the concentrations listed in the Table of National Primary Ambient-Air Quality
Standards by EPA in Sec. 4.10.
 Outdoor air treatments are required when necessary.
 Ventilation rates for typical spaces are listed in Table 4.5. For details, refer to ASHRAE Standard
62-1999.
 Ventilation systems for spaces with intermittent or variable occupancy may have their outdoor air
quantity adjusted by damper or cycling of fan operation to provide sufficient dilution to maintain
contaminant concentrations within acceptable levels at all times.
 For intermittent occupancy, when contaminants are associated only with occupants or occupant
activities, do not result a health hazard because contaminants are dissipated during unoccupied
periods.
 When contaminants are independent of occupants or occupant acivities, the supply of outdoor air
should lead occupancy lag or lead time depending on the ventilation rate and air capacity per person
in the space.
 Where peak occupancies are less than 3 h in duration, the outdoor airflow rate may be determined
according to average occupancy for the duration of operation of the system provided the average
occupancy used is not less than half of the maximum.
2. Indoor air quality procedure. In the indoor air quality procedure, acceptable air quality is
achieved within the conditioned space by restricting the concentration of all known contaminants of
concern to some specified acceptable levels. The acceptable concentration level in Table of National
Primary Ambient-Air Quality Standards, as listed in Sec. 4.10, also applies indoors for the same exposure
time. The following limits apply for four other indoor contaminants according to Standard
62-1999:
Contaminant Concentration ppm Exposure time
Carbon dioxide CO2 1.8 g /m3 1000 Continuous
Chlordane 5 g/m3 0.0003 Continuous
Ozone 100 g/m3 0.05 Continuous
Radon 4 pCi /L Annual average
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.3
The new Standard 62-1999 contains the entire 1989 version, unchanged, along with the following
four new addenda:
 Addenda 62e removes the statement that ventilation rates in Table 2 accommodate a moderate
amount of smoking because of the dangerousness of second-hand tobacco smoke.
 Addenda 62c removes the requirement that the temperature and humidity conditions specified in
Standard 55 be maintained in all ventilated space, such as even in garages.
 Addenda 62d rewords the standard scope to state that compliance with the standard will not ensure
acceptable indoor air quality.
 Addenda 62f clarifies that carbon dioxide (CO2) is simply a useful indicator of the concentration
of human bioeffluents.
For details, refer to ASHRAE Standard 62-1999. Standard 62-1999 is under continuous maintenance
by a Standing Standard Project Committee for which the Standards Committee has established
a documented program for regular publication of addenda or revisions.
Recirculation of Unused Outdoor Air in Multizone Systems
For a multizone VAV system, the ratio of outdoor air to supply air required to meet the ventilation
and thermal load requirements is often different from zone to zone. The system ventilation air required,
in most cases, is actually the amount of outdoor air intake at the AHU or PU. The problem
is how the system-fixed ratio of outdoor air to supply air can satisfy the different ratios of outdoor
air to supply air in various control zones.
In a specific control zone, if the ratio of outdoor air to supply air is higher than is needed in this
zone, the result is an increase in unused outdoor air in the recirculating air (more truly, the concentration
of air contaminants is lower) which raises the dilution power of the supply air (including
outdoor and recirculating air) for other control zones.
Methods of calculating ventilation requirements using the unused outdoor air contained in recirculating
air were developed in Australia and are included in ASHRAE Standard 62-1999. The required
corrected fraction of outdoor air supply in system supply air Y, in cfm (L / s), considering the
outdoor requirement of the critical zone and the unused outdoor air contained in the recirculating
air can be calculated as
(23.1)
and
(23.2)
where  supply volume flow rate of air system, cfm (L / s)
 corrected outdoor air supply volume flow rate, considering critical zone and unused
outdoor air in the recirculating air, cfm (L / s)
o,sys  calculated system outdoor air volume flow rate, cfm (L / s)
o,cr  outdoor air volume flow rate required in critical zone, cfm (L / s)
s,cr  critical zone supply volume flow rate, cfm (L / s)
X  uncorrected fraction of outdoor air supply
Y  corrected fraction of outdoor air supply
Z  fraction of outdoor air in supply air in critical zone
V? V?
V?
V?
o,cor
V?
s
Z  V?o,cr / V?s,cr
X  V?o,sys / V?s
Y  V?o,cor / V?s
Y 
X
1  X  Z
23.4 CHAPTER TWENTY-THREE
The system outdoor air volume flow rate can be evaluated as
(23.3)
where Noc  number of occupants in conditioned area served by air system
o,req  ventilation air requirements as specified by ASHRAE Standard 62-1999, cfm/ person
(L/sperson)
Outdoor air volume flow rates o,cor, o,sys, and o,cr and supply volume flow rates , and s,cr all
must occur at the same time.
Ventilation Controls for High-Occupancy Areas
ASHRAE/IESNA Standard 90.1-1999 mandates that air systems with design outdoor ventilation
air intake greater than 3000 cfm (1420 L / s) serving areas having a design occupancy density exceeding
100 people per 1000 ft2 (93 m2) shall include means which automatically reduce outdoor
air intake below design volume flow rate when spaces are partially occupied. Outdoor ventilation
air control shall be in compliance with ASHRAE Standard 62 and local codes except for systems
with heat recovery.
23.2 CO2-BASED DEMAND-CONTROLLED VENTILATION
CO2-Based Demand-Controlled Ventilation or
Time-Based Constant-Volume Control
As discussed in Sec. 4.10, among the indoor contaminants, particulates should be removed by
medium- and high-efficiency filters; vented combustion products including CO do not normally exceed
the allowable indoor concentration levels in a conditioned space; volatile organic compounds
including formaldehyde may have above-allowable indoor concentration levels when the building is
new or after remodeling; smoking is prohibited in many public places and is often limited in a specific
area in many buildings; only 6 percent of U.S. homes in their annual average radon concentration
exceeds the EPA specified level and needs subslab depression system and others; and airborne
virus may become dangerous when any one of the occupants is sick and carries such a virus. Under
many circumstances, occupant-generated CO2, as a useful indicator of the concentration of body
bioeffluents, emissions, and related contaminants, may be used as one of the indicators of poor indoor
air quality.
Many buildings, such as auditoriums, assembly halls, airport terminals, retail stores, offices, and
classrooms, have a variable occupancy during working hours and thus a fluctuation of the amount
of indoor contaminants released. Figure 23.1 shows the CO2 concentration profiles measured in the
floor space and in the return air duct on the 18th floor with a 20 percent fixed outdoor air supply of
a 22-story building in Ottawa, Canada. On a weekday, the concentration of CO2 gradually rose from
400 ppm in the morning to about 800 ppm before the office workers went out for lunch. After
lunch, the concentration of CO2 fluctuated and reached a daily maximum at about 3 p.m. Then the
CO2 concentration dropped gradually to 450 ppm at 8 p.m. If the amount of ventilation air required
to dilute the concentration level of the indoor contaminants varied accordingly, energy can be
saved. More than the required amount of outdoor ventilation air intake means more than the required
amount of energy is used to cool and dehumidify the outdoor air during summer and to heat
it during winter.
Demand-controlled ventilation (DCV) means that when the number of occupants (occupant density)
in the conditioned space during the occupied period, and thus the demand, drops, less outdoor
air is taken into the AHU or PU and supplied to the conditioned space to meet the demand of dilution
of concentration levels of contaminants with nearly the same specified ventilation rate
V?
V?
s V?
V?
V?
V?
V?
o,sys  NocV?o,req
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.5
23.6 CHAPTER TWENTY- THREE
cfm/person (L/s-person). Time-based constant-volume ventilation control intends to supply
constant-volume conditioned outdoor air to the conditioned space for dilution during the occupied
period even if the occupant density drops.
Compared with the time-based constant-volume ventilation control, demand-controlled ventilation
has the following advantages:
.DCV supplies the demanded amount of outdoor air to the conditioned space to maintain an
acceptable air quality and, at the same time, saves heating and cooling energy to condition the
excessive amount of supplied outdoor air in the time-based constant-volume ventilation
control. According to the simulation results in Carpenter (1996), compared to a system without
DCV, l000-ppm CO2-based DCV reduces 25 to 30 percent of space heating energy use. With
1000-ppm CO2-based demand-controlled ventilation, the average ventilation rate is 0.1 to 0.3
ach (air changes per hour), 10 to 30 percent lower than that of the constant-volume ventilation
control-
.DCV actually has taken into account the dilution power of the infiltrated outdoor air, which resuIts
in a greater energy saving.
.DCV often uses CO2 or a group of mixed gaseous contaminants as an indicator (controlled variable).
The space concentration of CO2 itself provides feedback of the control system, whereas in a
constant-volume ventilation control system, the volume flow rate of the outdoor air is difficult to
measure exactly.
In a VAV system, DCV is nonnally activated only in minimum outdoor air recirculating mode.
During the air economizer cycle, the amount of outdoor air intake in the AHU or PU is always
greater than the system outdoor air volume flow rate ~.sys. The amount of outdoor air intake is determined
by the free-cooling requirement of the V A V system.
CO2 Sensor or Mixed-Gases Sensor
In DCV systems, two kinds of air sensors (air transmitter) are most often used as an indicator: CO2
sensors and a volatile organic compound (VOC, or mixed-gases) sensor, as discussed in Sec. 5.4.
Compared to a CO2 sensor, a mixed-gases sensor senses a variety of mixed gases, is less expensive,
and needs less maintenance. However, according to Stonier (1995), a CO2 sensor has the following
advantages:
 The concentration level of CO2 directly indicates the concentration of occupant-generated bioeffluents
in the occupied zone.
 The concentration of CO2 of outdoor air is usually rather steady, between 300 to 350 ppm in
many locations in the United States. If each occupant exhales 0.0106 cfm (0.3 L /min) of CO2,
from the balance of CO2 by using Eq. (4.26), the ventilation rate supplied to the occupied zone
can be roughly estimated. Or if the ventilation rate is known, then the number of occupants can be
approximately estimated.
 If the ventilation rate of outdoor air supplied to a specific control zone and its number of
occupants are known, then a higher-than-normal concentration level of CO2 may be due to a poor
ventilation effectiveness, because of the short circuit of supply air or improper location of the CO2
sensor.
ASHRAE Standard 62-1999 recommends that “Where only dilution ventilation is used to control
indoor air quality, an indoor to outdoor differential concentration not greater than about 700
ppm of CO2 indicates comfort (odor) criteria related to body bioeffluents are likely to be satisfied.”
Standard 62-1999 also points out that using CO2 as an indicator of bioeffluents does not eliminate
the need for consideration of other contaminants.
Steven Taylor, chair of the committee revising Standard 62-1989, was asked in IAQ and Energy
‘98 New Orleans, October 24 to 27, if CO2 is an appropriate indicator of indoor air quality. His response:
“The use of CO2 as a reliable indicator of IAQ has been overblown, . . . CO2 is not itself
a contaminant of concern at the concentrations found in most buildings. It can provide an indication
of the amount of outdoor air being supplied to the space per person, but it cannot be used to accurately
determine that rate unless the space is in nearly steady-state conditions. In an office, it typically
takes several hours to reach steady-state conditions, and people in most offices do not stay in
one place that long.”
Although CO2 is not itself an indoor contaminant or is a comprehensive indicator of IAQ, the indoor
CO2 concentration can provide an indication of outdoor air supplied to a space per person.
Reardon (1994) and many other field measurements and investigations had clearly shown
 The relationship between the concentration of space CO2 and the air change rate (as shown in
Fig. 23.2)
 The floor-space concentration profiles of CO2 or average concentration profiles of CO2 in a return
duct (as shown in Fig. 23.1)
The use of a CO2 sensor to measure the space CO2 concentration to control the amount of outdoor
air intake in a ventilation control system is often an appropriate choice unless there is a still better
sensor.
As the outdoor ventilation air dilution is the primary issue to improve IAQ, the space concentration
of CO2, to a certain degree, most often is an indirect indicator of IAQ.
Location of CO2 Sensor
Test results of Reardon (1994) showed that the CO2 concentrations measured in the occupied floor
spaces are more variable than data measured in the openings of the return air shaft on the same floor.
The concentration of CO2 of return air fairly well represents the space average CO2 concentration
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.7
data measured of that particular floor after air mixes in the return air plenum. On the other hand, the
floor space concentration data were consistently higher than the return air data. This may be due to
some degree of short-circuiting of outdoor air supply to the occupied zone in that floor.
Emmerich and Persily (1997) noted that air movements due to open doors or open windows
cause differences and instabilities in CO2 concentrations. Enermodal Engineering (1995) reported
that an air system with CO2 sensors in the return duct on each floor showed no significant difference
in IAQ compared to a system with a CO2 sensor in the central return. Central control with a set
point of 800 ppm provided similar performance but at a much reduced installation cost compared to
individual zone control with a set point of 1000 ppm. Ruud et al. (1991) found that concentrations
of CO2 measured at the wall and in the exhaust air were nearly identical with the wall-mounted sensor
having an additional 2-min delay compared to that located in the exhaust air.
Therefore, for most VAV systems, the suitable location to install a CO2 sensor lies in the return
main duct (or return air intake) after the return ceiling plenum and close to the AHU or PU for sensing
an average CO2 concentration. The operator should check the difference in CO2 concentrations between
the floor space and the return duct periodically during operation so that the CO2 set point can be
properly adjusted; e.g., the set point in the return duct is 100 to 200 ppm lower than the 1000-ppm target
concentration level. In a VAV system, a single centrally-controlled sensor is recommended.
For a dedicated ventilation and space recirculating system, a better location of a CO2 sensor is in
the exhaust air main duct if there is an exhaust air system for the ventilation air supply. A critical
zone or a representative floor space is also a candidate for the location of a CO2 sensor.
Substantial Lag Time in Space CO2 Concentration Dilution Process
When a certain amount of outdoor air is extracted into an AHU or a PU, it should thoroughly
mix with the recirculating air first and form an outdoor-recirculating mixture. The mixture is then
23.8 CHAPTER TWENTY-THREE
FIGURE 23.2 Measured daily peak and daily average CO2 concentrations in the return duct versus
air changes per hour. (Source: Reardon et al., ASHRAE Transactions, Part II, 1994)
supplied to a control zone in a VAV system, where it needs again to mix thoroughly with the space
air in that control zone and to distribute over the entire conditioned space, in order to dilute the concentration
of occupant-generated and other contaminants. It takes a substantial lag time to perform
such a mixing and dilution process.
Elovitz (1995) reported that there is a substantial lag time in the space CO2 concentration
dilution process. If CO2 rises, an increase in the outdoor air ventilation rate will not make it drop
instantly. The control continues calling for more and more dilution and opens the outdoor damper
wider and wider, until the drop in the CO2 concentration is sensed by the CO2. Recently a CO2
minimum ventilation control system in an insurance office building retrofit showed such a phenomenon.
As the concentration of CO2 raised above the set point, the control system kept cranking the
system up to 100 percent outdoor air intake until eventually the dilution effect showed up in the
return air; at that point, the control cranked the outdoor damper down until it was fully closed, as
the concentration of CO2 continued to fall long after the damper started to close.
The following are two methods that will solve the time-lag problem:
1. Determine the relationship between the sensed CO2 concentration and the required air changes
per hour of minimum outdoor air intake expressed in minimum outdoor damper setting in the
AHU or PU, as shown in Fig. 23.2 during commissioning. For each sensed CO2 concentration in
the return duct or floor space, the DDC controller resets the required minimum outdoor air
intake, in air changes per hour, as shown in Fig. 23.2, and the corresponding outdoor damper
setting according to the determined relationship between them during commissioning.
2. When the sensed CO2 concentration exceeds the set point, the outdoor damper will open an additional
5 percent. If, after a prespecified time interval, the sensed CO2 level still exceeds the set
point, the outdoor damper will open an additional 5 percent until the outdoor air intake in the
AHU or PU reaches the upper limit o,sys. The prespecified time interval should be determined
during commissioning.
Base Ventilation and Purge Mode
In a CO2-based demand-controlled ventilation system serving a typical office building, even if the
number of occupants drops to zero during the occupied period, the outdoor air intake should not
drop below a base ventilation. Base ventilation bv, in cfm (L / s), is the lower limit of minimum
outdoor ventilation air required to dilute mainly the nonoccupant-generated air contaminants, such
as building-generated volatile organic compounds (VOCs), to provide exhaust air in restrooms, and
to maintain a positive space pressurization. For a typical office building, the amount of base ventilation
required depends on interior building materials, building construction, and the tightness of the
building shell. The exhaust air in restrooms is often provided by the transfer air that also dilutes the
building-generated VOCs. The base ventilation bv is the sum of the outdoor air required for restroom
exhaust and exfiltration to maintain a space pressurization. Consider a one-story office building
of 20,000 ft2 (1859 m2) of floor area with an exterior wall area of 10,000 ft2 (929 m2):
 Two bathrooms have typically 12 water closets or urinals. According to ASHRAE Standard
62-1999, these two restrooms need about 50 	 12  600 cfm or 0.03 cfm / ft2 (0.15 L / s m2).
 As discussed in Sec. 20.2, the flow coefficient of an average leaky building is 1.0 cfm / ft2 (5.0
L/sm2) of exterior wall at 0.3 in. WC (75 Pa). From Eq. (20.2), the overall leakage rate is
From Eq. (20.1) the effective leakage area of this average leaky building is:
Ae,l  
V?
inf
4005?p

4572
4005?0.3
 2.08 ft2 (0.194 m2)
V?
leak  Cflow Aw (p)0.65  1.0(10,000)(0.3)0.65  4572 cfm (2157 L / s)
V? V?
V?
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.9
Under base ventilation, the amount of exfiltration required to maintain a space positive pressure of
0.005 in. WG (1.25 Pa) can be calculated as
The volume flow rate of base ventilation is then
or roughly 1200 cfm (566 L / s), that is, 1200/ 20,000  0.06 cfm / ft2 (0.30 L / sm2) of floor area,
which is about 43 percent of the value of the minimum outdoor air required .
Formaldehyde and other possible indoor air contaminants in the space
should be measured. If their concentrations are above allowable levels, or if the building shell is too
leaky, the base ventilation should be increased accordingly.
The amount of outdoor ventilation air provided for CO2-based demand-controlled ventilation in
a VAV system in an office building with a ceiling height of 9 ft (2.7 m) during various operating
modes is as follows:
The volume flow rate of minimum outdoor air during recirculating mode at design occupancy is the
calculated system outdoor air intake in the AHU or PU at summer design conditions o,sys by Eq.
(23.3).
It is recommended that a purge operation take place every weekday morning if the concentration
of the volatile organic compounds or other indoor contaminants is above the allowable level before
the building is occupied or if the difference between the indoor and outdoor CO2 concentrations is
greater than 200 or 300 ppm.
A CO2-Based Demand-Controlled Ventilation System
Figure 23.3 shows a CO2-based demand-controlled ventilation system in a VAV system. It consists
of a CO2 sensor, a DDC unit controller, a minimum outdoor damper, an economizer damper, a recirculating
damper, an exhaust damper, and damper motors. The CO2 sensor is located in the return
duct that connects the return ceiling plenum and the return fan inlet. During minimum outdoor air
recirculating mode:
 The supply fan and the minimum outdoor air damper of this VAV system must have the capacity
to provide the required amount of outdoor air at a volume flow rate of o,sys, in cfm (L / s), calculated
by Eq. (23.3).
 Outdoor air should be supplied to the occupied zone with a high ventilation effectiveness for the
dilution of the concentration of CO2 and other indoor contaminants.
 During commissioning, damper settings of required minimum outdoor air at design occupancy
, and base ventilation bv should be clearly marked, and the relationship between the measured
CO2 concentration and the minimum outdoor damper opening setting from bv up to o,sys
should be determined after tests.
 When a VAV system is in the occupied, minimum outdoor air recirculating mode, after supply
and return fans start, the minimum outdoor damper should be opened to an opening setting of
base ventilation.
V?
V?
V?
V?
o,sys
V? V?
cfm/ft2 (L/sm2) of floor area ach
Air economizer cycle, free-cooling mode 0.14–1.3 (0.7–6.5) 0.93–10
Minimum outdoor air recirculating mode
V ·
o,sys (at design occupancy) 0.14 (0.7) 0.93
Base ventilation V ·
bv (lower limit) 0.06 (0.25) 0.40
(V?bv  0.43 V?o, sys)
V?
bv  600  589  1189 cfm
V?inf  4005Ae,l ?p  4005 	 2.08 ?0.005  589 cfm (278 L / s)
23.10 CHAPTER TWENTY-THREE
 If the signal of CO2 concentration sensed by the CO2 sensor located in the return duct or a representative
floor space is sent to the DDC controller, the DDC controller resets the minimum outdoor
air intake o,sys to a corresponding minimum outdoor damper opening setting according to
the relationship between the measured return duct or space CO2 concentration and the damper
opening setting.
If the required amount of minimum outdoor air corresponding to the sensed CO2 level is less
than the base ventilation, the DDC controller resets the minimum outdoor air intake down to the
base ventilation bv.
 The recirculating damper is opened 100 percent.
 When the VAV system is in the off position, the minimum outdoor, economizer, and exhaust
dampers are closed, and the recirculating damper is opened fully.
During 100 percent outdoor air free-cooling mode, the air economizer control overrides the CO2-
based DCV control.
Application of CO2-Based Demand-Controlled Ventilation System
Chen and Demster (1996) and Emmerich and Persily (1997) recommended the use of CO2-based
DCV systems under the following circumstances:
 The existence of high occupancy density with considerable fluctuations in occupancy
 Low pollutant emissions from nonoccupant-generated sources
 In location with climate where heating or refrigeration equipment has a significant number of
working hours annually
V?
V?
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.11
Supply fan
DDC controller
Supply duct
Recirculating
damper
Minimum outdoor
damper
Return duct Return fan Exhaust
damper
Economizer
damper
CO2
CO2 CO2 sensor
FIGURE 23.3 A CO2-based demand-controlled ventilation control in a VAV system.
For a multizone VAV system that serves control zones with an obvious occupancy variation schedule,
such as in an office building where offices and conference rooms use a common AHU or PU,
when many occupants in the offices may go out for lunch, but a meeting is still taking place in a
conference room, remedies should be considered during design to supply additional outdoor air to
the conference room under such circumstances.
23.3 MIXED PLENUM PRESSURE CONTROLS
Basics
As discussed in Secs. 22.5 and 22.6, for VAV systems operating at minimum outdoor air recirculating
mode, when the supply volume flow reduces during part-load operation, the static pressure in
the mixed plenum (mixing box) where outdoor air mixes with the recirculating air tends to rise and
causes a small pressure difference between the mixed plenum and outdoors. The minimum outdoor
air intake in the AHU or PU reduces accordingly. Mixed plenum pressure control aims to maintain
a nearly constant static pressure in the mixed plenum and thus keeps a nearly constant pressure difference
between the mixed plenum and outdoors to extract a nearly constant amount of minimum
outdoor air in both design load and part-load operation.
Case Study: Mixed Plenum Pressure Control Monitoring Plenum Pressure
Graves (1995) developed a mixed plenum pressure control system in a VAV system by monitoring
the pressure difference between the mixed plenum pressure and outdoors atmospheric pressure and
modulating the recirculating damper opening to maintain a nearly constant mixed plenum pressure,
as shown in Fig. 23.4a.
23.12 CHAPTER TWENTY-THREE
DDC controllers
Supply fan
Outdoor
louver
Return fan
Recirculating
damper Recirculating
damper
Filters Minimum
outdoor air
damper
Minimum
outdoor air
damper
Economizer
damper
Economizer
damper
P
P Pressure sensor Mixing temperature element
P
M
Mixing
plenum
M
Mixing
plenum
Exhaust
damper
(a) (b)
FIGURE 23.4 Mixed plenum pressure controls: (a) monitoring mixed plenum–outdoors pressure differential;
(b) monitoring pressure drop across the outdoor damper and louver.
System Description. The VAV system includes two rooftop packaged units. Each has a capacity
of 8300 cfm (3917 L / s) with a variable-speed supply fan and a variable-speed return fan. A fixed
minimum oudoor damper supplies 1800 cfm (850 L / s) of outdoor air to the conditioned space, and
an economizer damper supplies the additional outdoor air during the air economizer cycle. When
the outdoor temperature is higher than 70°F (21.1°C), the VAV system is operated in minimum outdoor
air recirculating mode. A DDC unit closes the economizer damper and opens the recirculating
damper. Whenever the packaged unit is operating, the exhaust damper is opened 100 percent.
In the mixed plenum, outdoor air is drawn through a louver and damper on one side of the unit,
and exhaust air is discharged through the exhaust damper and louver on the opposite side. The
opening of the fixed minimum outdoor damper was set to provide 1800 cfm (3917 L / s) with a
plenum pressure of 0.50 in. WG (125 Pag). At design conditions, the airflow and pressure
drops across dampers and louvers are as follows:
Minimum Ventilation. A DDC control system is used for the controls of this rooftop unit including
a mixed plenum pressure control. A static pressure sensor located in the mixed plenum senses
the plenum pressure, and a shielded outdoor air pressure reference is located above the rooftop
packaged unit. The DDC system modulates the recirculating damper to maintain a static pressure
difference between the mixed plenum pressure and outdoors pm,p of 0.55 in. WC (137 Pa) 	 0.05
in. WC (12.5 Pa). If pm,p is too large, the recirculating damper is gradually opened until it reaches
the set point. If pm,p is too small, the recirculating damper is gradually closed until the set point is
maintained. According to field-measured results, the DDC mixed plenum pressure control system
has achieved a very stable control of pm,p.
Supply and Return Fans. The speed of the supply fan is modulated to maintain a duct static pressure
of 1.50 in. WG (375 Pa) at the location of the duct static pressure sensor in the supply main
duct during both design and part-load operations. The volume flow rate of the return fan is always
approximately 1800 cfm (850 L / s) less than the supply air volume flow. A linear relationship between
the supply fan speed and return fan speed was determined by testing the rooftop packaged
VAV system during commissioning. Roughly, when the supply fan was at 100 percent fan speed,
the return fan was at 90 percent fan speed. When the supply fan was operated at approximately 40
percent speed, the recirculating damper was closed and the return fan was turned off.
Actual operation of this system showed that the supply fan speed was always above 40 percent
because the minimum supply volume flow rate required by the VAV boxes was 35 percent, or about
2900 cfm (1369 L / s).
Monitoring Pressure Drop of Louver and Damper Controlling Mixed Plenum
Pressure
Mumma and Wong (1990) first introduced a kind of mixed plenum pressure control to modulate the
recirculating damper to maintain a predetermined pressure drop across an outdoor air louver and
damper combination, as shown in Fig. 23.4b. This mixed plenum pressure control consists of a
minimum outdoor damper, an economizer damper, a recirculating damper, a pressure sensor, a
DDC controller, damper motors, a variable-speed supply fan, and a variable-speed relief fan. It proceeds
as follows:
Airflow, cfm (L / s) p, in. WC (Pa)
Recirculating damper 6500 (3067) 0.55 (137)
Minimum outdoor damper 1800 (850) 0.30 (75)
Economizer damper 6500 (3067) 0.30 (75)
Exhaust damper 6500 (3067) 0.10 (25)
Outdoor air louver 8300 (3917) 0.25 (63)
Exhaust air louver 6500 (3067) 0.25 (63)
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.13
 The minimum outdoor damper is fixed at an opening setting which delivers minimum outdoor air
o,sys. This damper setting is determined during commissioning.
 When the supply fan reduces its speed at part-load operation, the pressure differential across the
outdoor louver and minimum outdoor damper combination drops. As this reduction is sensed by
the pressure transducer, a DDC unit closes the recirculating damper to maintain a preset pressure
drop across the combination, and therefore ensures that the amount of minimum outdoor air o,sys
intake is always extracted.
 If the supply fan increases its speed, the pressure drop across the outdoor louver-damper combination
rises. The DDC controller opens the recirculating damper to reduce the pressure drop
across the combination and maintains the required amount of minimum outdoor air o,sys intake.
Applications
In VAV systems, mixed plenum pressure controls ensure time-based constant volume of minimum
outdoor ventilation air supply to the conditioned space. The controls are simple and effective and
can be used for buildings where there is less variation in occupancy.
23.4 OUTDOOR AIR INJECTION FAN, DIRECT
MEASUREMENT, AND FAN TRACKING SYSTEMS
Outdoor Air Injection Fan
To provide a guaranteed amount of minimum outdoor air o,sys to a VAV system, recently an outdoor
air injection fan has sometimes been installed in parallel to the outdoor air passage leading to
the main supply fan, as shown in Fig. 23.5. An outdoor air injection fan has the following operating
charateristics:
V?
V?
V?
V?
23.14 CHAPTER TWENTY-THREE
Return
slot
Return
duct
Economizer
damper
Supply
fan
Return
fan
Mixed
plenum
Outdoor
air injection
fan Minimum
outdoor
air damper
Supply
duct
Ceiling
plenum
Slot
diffuser
ru
m
FIGURE 23.5 A VAV system with an outdoor air injection fan.
 The maximum supply volume flow rate of this outdoor air injection fan should be equal to the
amount of minimum outdoor air o,sys required. The fan total pressure of this injection fan is often
equal to the total pressure loss between the outdoor air intake and the mixed plenum inlet.
 The ductwork of the injection fan system is sized so that the velocity pressure is easily measured.
 An airflow measuring station with a velocity sensor sends signals to a DDC controller and adjusts
the fan speed as well as its volume flow through a variable-speed drive to correct for wind and
backpressure effects.
 The injection fan system is a constant-volume minimum outdoor air supply system and operates
whenever outdoor air is required.
 According to Avery (1996), if an outdoor air injection fan is installed with a supply and return fan
combination, the mixed plenum may be varied from negative pressure when the VAV system is
operating on 100 percent outdoor air or mixed air control to positive pressure during minimum
outdoor air cooling mode.
Outdoor injection fan systems have the advantages of easy determination and measurement of
the minimum outdoor airflow and control its required amount. They have been used in many applications
by designers. However, a small injection fan is often energy-inefficient. Also, an injection
fan needs extra fan and ductwork and results in a more complicated system pressure distribution
and control, especially for a retrofit in existing buildings.
Direct Measurement of Minimum Outdoor Air Intake
Minimum ventilation control by direct measurement of the minimum outdoor airflow during minimum
outdoor air recirculating mode needs a DDC controller to open or close the minumum outdoor
damper and recirculating damper according to the directly measured airflow signal, to provide
the required amount of minimum outdoor air for ventilation. The outdoor air at the intake is
usually designed according to a velocity of 500 fpm (2.5 m / s). During part-load operation, it may
reduce to 35 percent of the design value, that is, 175 fpm (0.87 m / s), or a velocity pressure of
0.002 in. WC (0.5 Pa). In such a velocity pressure, even a multiple-point Pitot-tube averaging
probe has difficulty sensing a 10 percent change in outdoor air volume flow rate. The direct measurement
of minimum outdoor air intake cannot accurately sense, measure, and provide minimum
ventilation control.
Fan Tracking Systems
In a VAV system with a supply and return fan combination using a fan tracking system for minimum
ventilation control, the volume flow rate of the supply fan is varied according to the system
load while at the same time maintaining a preset duct static pressure at the static pressure sensor in
the supply main duct. The volume flow rate of the return fan is controlled to “track” the supply fan
at a constant cfm (L / s) less than the supply volume flow rate. If the exhaust damper is tightly
closed and does not leak, if there is no infiltration or exfiltration, and if one ignores the difference in
densities between the outdoor and recirculating air, the volume flow rate of the supply fan minus
the volume flow rate of return fan sf  rf is the amount of outdoor air intake in the AHU or PU,
designated , in cfm (L / s).
Because outdoor air is often from one-fifth to one-seventh of the supply air, consider a supply
fan with a design volume flow sf of 10,000 cfm (4719 L / s) and a return fan with a design volume
flow rf of 8500 cfm (4011 L / s), which results in a design outdoor air intake of 10,000  8500 
1500 cfm (708 L / s). Due to the measuring and instrumental errors, sf reads 2 percent low and rf
reads 2 percent high. The controlled outdoor air intake is then 9800  8670  1130 cfm (533 L / s),
which is about 25 percent less than the design value. Field experience had proved that fan-tracking
minimum ventilation control failed to provide the required amount of outdoor air correctly and
energy-efficiently.
V?
V?
V?
V?
V?o
V?
V?
V?
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.15
However, for a VAV system with a supply and return fan combination, when a supply fan reduces
its supply volume flow rate as the space load drops, the volume flow of the return fan must
decrease accordingly, to meet the following requirements:
 Maintain a desirable space pressurization.
 Exhaust a specified amount of exhaust air and recirculate a specified amount of recirculating air
in the AHU or PU.
 Form a new airflow balance in the space.
The volume flow of the return fan tracks the volume flow of the supply fan in order to provide outdoor
air that exactly equals sf  rf but it failed to do so in actual practice. However, reduce the
speed of the return fan approximately according to sf  rf while the outdoor air intake is guaranteed
by DCV or mixed plenum pressure control, and this gives a successful minimum ventilation
control experience. This is discussed in a later section.
23.5 CONFERENCE ROOMS
Conference rooms have characteristics of high occupancy density (20 ft2 /person or 2 m2 /person)
and a variable occupancy with a schedule that may be quite different from private and general of-
fices. Conference rooms are often critical spaces to provide ventilation for acceptable indoor air
quality in office buildings. The following are recommendations of minimum ventilation control for
conference rooms in office buildings:
 Use a separate AHU or PU for the conference room with its own outdoor air intake and minimum
ventilation control, if possible.
 Increase the total supply airflow rate as well as the outdoor ventilation air rate. According to
ASHRAE Standard 62-1999, for a conference room with an occupancy density of 20 ft2 /person
(2 m2 /person) instead of 140 ft2 /person (13 m2 /person) in an office, the outdoor air requirement for
ventilation should be increased from 0.14 cfm / ft2 (0.7 L / sm2) to about 1.0 cfm / ft2 (5 L / sm2).
In a conference room, the supply air volume flow rate should be determined based on the outdoor
air requirement for minimum ventilation. Even the VAV system load and the supply volume
flow rate reduce during the occupied period; based on the ASHRAE Standard 62-1999 specified
outdoor air requirements, most of the conference rooms still have the required minimum outdoor air
to provide ventilation for acceptable indoor air quality.
23.6 SPACE PRESSURIZATION AND
RETURN VOLUME FLOW CONTROLS
As discussed in Sec. 20.3, for low-rise buildings except on stormy days, building pressure characteristics
are dominated by air systems (including mechanical ventilation systems) most of their operating time.
Characteristics of Space Pressure Control
The following are characteristics of space pressure control:
 According to the airflow balance, the mass flow rate of the outdoor air entering air system , in
lb / min (kg / min), is balanced with the mass flow rate leaving the system, or.
m?o  m?inf  m?ex,r  m?eu
m?o
V?
V?
V?
V?
23.16 CHAPTER TWENTY-THREE
where  mass flow rate of air exhausted, including air exhausted from the restroom and
from the AHU, lb/ min (kg / min). In most conditioned space, a space positive pressure pim is thus
formed because of a surplus of outdoor air entering the system, which is then exfiltrated
through the gaps and openings on the building shell to outdoors and can be calculated as
(23.4)
where  inf  volume flow rate of exfiltration, cfm (m3 /min)
Ae,l  effective leakage area on building shell, ft2 (m2)
r  space air density, lb/ ft3 (kg /m3)
pro  space and outdoor pressure difference, in. WC (Pa)
 As discussed in Sec. 20.3, the space and outdoor pressure difference often has a value between
 0.005 and 0.03 in. WC (1.25 and 7.5 Pa). If a pressure sensor in a space pressurization control
system needs to sense a 10 percent change in pim, that is, typically  0.0005 in. WG (0.13 Pa), it
would be difficult and expensive to specify such a level of accuracy for most comfort systems.
 For an average leaky one-story office building with an exterior wall area of 10,000 ft2 (929 m2)
and a floor area of 20,000 ft2 (1858 m2), as discussed in Sec. 23.2, from Eq. (20.2)
From Eq. (23.4), for this average leaky office building, the effective leakage area is
Also from Eq. (23.4) the outdoor air intake required to maintain various space positive pressures
and to exfiltrate them in this average leaky office building with 10,000 ft2 (929 m2) of exterior wall
is as follows:
VAV Systems Return/ Relief Fan Volume Flow Control
For a VAV system with either a supply and relief fan combination as shown in Fig. 22.3, or a supply
and return fan combination as shown in Figs. 22.1 and 22.4, if the volume flow rate of supply air reduces
during part-load, the amount of return air extracted along points ru should be reduced accordingly
in order to provide the required outdoor ventilation air and to maintain a desirable space pressure.
Modulating the speed of the return / relief fan with adjustable-frequency drives, and thus its
volume flow, to maintain space pressure within predetermined limits is a simple, direct, and effective
means of return / relief fan volume control.
Place a pressure sensor in a representative floorspace. The location of the indoor pressure sensor
must be away from the door and openings to the ouside, as well as away from the elevator lobbies.
The outdoor reference pressure should be typically 10 to 15 ft (3 to 4.5 m) above the building and
Required outdoor air intake
pro, in. WC (Pa) Exterior wall area, cfm/ ft2 (L/sm2) Floor area, cfm/ ft2 (L/sm2)
0.005 0.059 (0.30) 0.0295 (0.15)
0.01 0.083 (0.42) 0.0417 (0.21)
0.02 0.118 (0.60) 0.059 (0.30)
0.03 0.144 (0.73) 0.072 (0.36)
0.05 0.186 (0.95) 0.093 (0.47)
0.1 0.263 (1.33) 0.132 (0.67)
Ae,l 
V?inf
4005?pro

4572
4005?0.3
 2.08 ft2 (0.193 m2)
V?
leak  Cflow Aw(p)0.65  1.0(10,000)(0.3)0.65  4572 cfm (2157 L / s)
V?
V?inf  4005Ae,l ?pro
m?inf  V?inf r
 m?inf
m?ex,r, m?eu
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.17
well shielded from wind and air velocity effects. As a DDC unit controller reduces the speed of supply
fan to maintain a preset duct static pressure during part-load, another DDC unit controller also
modulates the speed of return fan to adjust the space pressure pr, floating within a differential or
dead band between an upper limit of 0.03 in. WG (7.5 Pag) and a lower limit of 0.005 in.
WG (1.25 Pag) using a floating control mode. The exhaust / relief damper is often opened between
0 and 30 percent during minimum outdoor ventilation recirculating mode and opened 100
percent during all outdoor air economizer mode.
If a VAV system using a supply and relief fan combination is operated during minimum outdoor
air ventilation recirculating mode with a constant-volume flow of minimum outdoor ventilation air
provided by mixing plenum pressure control, and if the building that this VAV system serves has an
air tightness between an average leaky and leaky building, then
 Either the relief fan is turned off and the relief damper is closed, so that all the outdoor air is
exhausted from the restrooms, other exhausts, and exfiltrates through the exterior leakage area.
 Or the relief fan is turned on at low speed and the relief damper exhausts only a volume flow less
than 30 percent of the designed volume flow. The remaining portion of the outdoor air is exhausted
outdoors through restrooms, other exhausts, and exfiltrates through the exterior leakage area.
During minimum ventilation recirculating mode, these exhausts and exfiltration, plus the
recirculating air extrated along points ru (recirculating / exhaust chamber) and m (mixing plenum)
by the supply fan are automatically balanced with the design and reduced supply volume flow rate
at part-load.
Another alternative of return / relief fan volume control in VAV systems is airflow tracking. Required
return / relief volume flow corresponding to various reduced supply airflow at part-load are
determined during commissioning by means of multipoint Pitot tube measurements. Return / relief
fan speed and volume flow can then be modulated at part-load accordingly. Airflow tracking is an
open-loop control. Supply fan volume flow must be adjusted to match with the volume flow of the
return fan during morning warm-up and cool-down mode.
23.7 DISCHARGE AIR TEMPERATURE CONTROLS
Basics
In a multizone VAV system, the discharge air temperature Tdis, in °F (°C), is the conditioned air
temperature leaving the supply outlet of the AHU or PU, where Tdis  Tcc  Tsf. Here Tcc represents
the temperature of the conditioned air leaving the coil, and Tsf indicates the temperature rise
across the supply fan, both in °F (°C). Typically, Tdis  55  2  57°F (13.9°C). The difference
between the supply temperature from the ceiling diffuser mounted on the ceiling plenum and the
discharge air temperature Ts  Tdis, in °F (°C), is the rise or drop in temperature resulting from the
duct heat gain or duct heat loss Tdu. For a typical multizone VAV system, Ts  Tdis is usually a
function of the temperature difference between the supply air and the ambient air surrounding the
supply duct and the supply volume flow rate. Its relationship can be expressed as Ts  Tdis 	 Tdu.
As discussed in Secs. 20.18 and 21.3, for a summer indoor design temperature of 75°F (23.9°C), a
lower conditioned air off-coil temperature Tcc and discharge air temperature Tdis result in the following:
 A greater supply temperature differential Ts  Tr  Ts, a lower space relative humidity, a
smaller supply volume flow rate , fan sizes, fan energy use, duct, terminal, and diffuser sizes
 A lower chilled water temperature entering and leaving the cooling coil Twe and Twl, a lower
evaporating temperature in the DX coil Tev, and therefore higher power input to the refrigeration
compressors
Excluding the the cold air distribution in the ice storage system, discussed in Chap. 31, a Ts
between 15 and 20°F (8.3 and 11.1°C) during summer design conditions is most widely used in air
V?
s
23.18 CHAPTER TWENTY-THREE
systems. The following list relates parameters when Ts  15 and 20°F except in an ice storage
system in which Ts may increase up to 30°F (16.7°C). All units are in °F (°C), except the space
relative humidity 
r is in percent:
System Description
Discharge air temperature control consists of the following control loops, as shown in Fig. 23.6a:
 Economizer control loop. For a temperature air economizer control, it consists of an outdoor temperature
sensor T2, a recirculating air temperature sensor T3, a DDC controller, outdoor damper
actuators D1 and D2, a recirculating damper actuator D3 and a relief damper actuator D4.
 Cooling coil control loop. It consists of a discharge air temperature sensor T1, a DDC controller,
and a cooling coil valve actuator V1.
 Mixed air control loop. It consists of a mixed air temperature sensor T4, a DDC controller, outdoor
damper actuators D1 and D2, and a recirculating damper actuator D3.
 Heating coil control loop. It consists of a discharge air temperature sensor T1, a DDC controller,
and a heating coil valve actuator V2.
 Low-limit control loop (coil freeze protection). It consists of a temperature sensor T4, a DDC controller,
and outdoor air damper actuators D1 and D2.
For VAV systems, discharge air temperature control is a kind of cooling and heating capacity
control so that conditioned air is discharged from an AHU or PU with a predetermined temperature
and relative humidity, such as 57°F (13.9°C) and 95 percent relative humidity, and can perform the
following requirements:
1. Provide a kind of conditioned air for zone temperature control by varying the supply volume
flow rate.
2. Based on the conditions of air leaving the coil sensed by sensor T1, provide economizer control
and modulate the cooling or heating capacity of the AHU or PU to maintain a predetermined Tdis.
3. In VAV systems, if an AHU or a PU serves both perimeter and interior zones, usually perimeter
heating is provided by the heating coil installed in the reheating VAV boxes or fan-powered VAV
boxes or by baseboard heating directly. In such a condition, Tdis should meet the free-cooling
requirement in interior zones even in winter.
4. According to the outdoor temperature or system load, reset Tdis to an optimum value for both
effective operation and energy saving.
For a cooling coil or heating coil in the AHU, the DDC controller modulates the coil valves continuously.
However, in a rooftop packaged unit, refrigeration capacity is often adjusted by cycling
the compressors on and off in the separate refrigerant circuits of the DX coils. The capacity control
of packaged units is discussed in Sec. 11.8.
Most VAV systems use the air economizer or water economizer as the first-stage cooling to save
energy. Cooling coil, mixed air, and heating coil control loops are operated in sequence so as to maintain
a predetermined discharge air temperature while preventing simultaneous cooling and heating.
For a typical AHU, the set point of the discharge air temperature at summer design conditions
may be 57°F (13.9°C). To avoid sudden repeated changes in operating mode among the cooling coil
and mixed air control loops, resulting in unstable operation, the set point for mixed air control is
better set at 55°F (12.8°C) as shown in Fig. 23.6b.
Ts Tcc Tsf Tdis Tdu Ts Tr 
r
15 (8.3) 55 (12.8) 2 (1.1) 57 (13.9) 3 (1.7) 60 (15.6) 75 (23.9) 50
20 (11.1) 50 (10.0) 2 (1.1) 52 (11.1) 3 (1.7) 55 (12.8) 75 (23.9) 45
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.19
23.20 CHAPTER TWENTY-THREE
Relief
fan
Heating
coil
Supply
fan
Cooling
coil
Recirculating
air damper
Minimum
outdoor
air damper
Relief
air damper
Economizer
damper
D4
D3
D2
D1
T2
T3
T3 Temperature sensor
Mixing
air Cooling coil
Min
100
100%
50
0
0
57
20
62
Capacity, %
D Damper actuator
F Flow measuring station
AC AC inverter
T4
V2 V1
D T0
T1
AO4
DO1
3 2
AO1
6 5 4 3 2
AI1
BI1
AC
(a)
(c) (b)
0 20 40 60 80
40
40
50
50
60
60
70
70 0.016
0.012
0.008
0.004
80 90
100
Tdis,F
T0,F
T,F
V0
Outdoor air temperature
. V0
.
r m
si
si Supply air for interior zone
sx Supply air for perimeter zone
sx
r ru
ru cc s
m
w
lb/lb
FIGURE 23.6 Discharge air temperature control for a typical AHU: (a) control diagram; (b) output diagram; (c) air
conditioning cycle.
The modulation of valve V1 in the cooling coil control loop of discharge air temperature control
can be performed in parallel with the control operations in 100 percent air economizer control. Although
mixed air control is part of air economizer control, it is also a part of the sequence control in
a discharge air temperature control. Mixed air control mixes the outdoor air with the recirculating
air by modulating the economizer damper and the recirculating damper as to provide a mixed air
temperature, such as 55°F (12.8°C) for the sake of free cooling in the interior zone. Mixed air control
functions only when To 55°F (12.8°C) and ends during minimum outdoor air heating mode.
For VAV reheat and fan-powered VAV systems, heating is often provided by a water heating coil or
an electirc heating coil in a terminal directly above the conditioned space. The modulation of the VAV
box and the control valve of the heating coil in sequence forms the zone cooling and heating control.
For a single-zone VAV system, a DDC unit actuates the economizer dampers, cooling coil valve,
economizer and recirculating dampers, and heating coil valves in sequence, based on the input
signal from a zone temperature sensor to maintain a predetermined zone temperature. The discharge
air temperature control is replaced by a zone temperature control.
Operation of Air Economizer and Outdoor Air Intake—Case Study
The outdoor air intake of a fixed dry-bulb air economizer in an AHU of a VAV system shown in
Fig. 23.7 is from Nabinger et al. (1994). In Fig. 23.7, the abscissa of the diagram is the indoor-outdoor
temperature difference of the conditioned space that the VAV system serves and the outdoor
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.21
FIGURE 23.7 The outdoor air intake of a temperature air economizer in a VAV system. (Modified from Nabinger
et al., ASHRAE Transactions, Part II, 1994. Reprinted by permission.)
air temperature, both in °F (°C). The ordinate is the measured outdoor air intake of a fixed dry-bulb
air economizer in a VAV system, expressed in supply flow density, cfm/ ft2 (L/sm2), and ach based
on a ceiling height of 9 ft (2.74 m). According to Nabinger et al. (1994), the test office building is
located in Overland, Missouri. The building consists of seven floors with a floor area of 378,000 ft2
(35,100 m2). Construction began in 1988, and occupancy began in late 1990.
The air conditioning system is a VAV system with fan-powered VAV boxes in the perimeter zone
and VAV boxes in interior zones. There are all together 15 AHUs with a total supply volume flow
rate of 267,000 cfm (126,000 L / s), and the design minimum outdoor air intake for this building is
50,600 cfm (23,900 L / s).
Outdoor air intake in the AHUs was measured by using the trace gas decay technique with SF6
as the trace gas. Trace gas was injected into each of the 15 AHUs and allowed a period of about 20
to 30 min to mix. Four decay tests were conducted each day. The tracer gas injection rates were calculated
based on a target concentration of 150 ppb (parts per billion). Tracer gas concentration was
monitored at 19 indoor locations and an outdoor location with the concentration measured at each
location every 10 min.
Operation of Air Economizer. When the outdoor air temperature To was below a set point value,
such as To 30°F (1.1°C), the AHU or PU was set to minimum outdoor intake to protect it from
freezing and to avoid heating more outdoor air than necessary. The AHU was operated in minimum
outdoor air heating mode, as shown in Fig. 23.7.
When the outdoor temperature 30°F  To  70°F (1.1°C  To  21.1 °C), the DDC controller
in mixed air control modulated the economizer and recirculating dampers to provide a mixture
of outdoor and recirculating air at the required mixed air temperature, such as 55°F (12.8°C).
The air economizer was operated in mixed air control. Here the outdoor air temperature from which
the mixed air control began To.mix  30°F (1.1°C) depended on the percentage of outdoor air
among the supply air, and on the load ratio.
When the outdoor air temperature To > 70°F (21.1°C), it was more economical to cool the recirculating
air than the outdoor air. The minimum outdoor and recirculating dampers opened, and the econimizer
damper was closed. The AHU or PU was operated in minimum outdoor air cooling mode.
In Fig. 23.7, the highest measured outdoor air supply flow density was out 0.52 cfm / ft2 (2.6
L/s ·m2), which was lower than the average building supply flow density 267,000 / 378,000  0.71
cfm/ft2 (3.5 L / s ·m2). The measured minimum outdoor air intake was lower than the ASHRAE
Standard 62-1999 specified minimum outdoor air intake of 0.93 ach, or 0.14 cfm / ft2 (0.7 L / s ·m2).
The average supply flow density for VAV systems at design flow for a typical floor (including
perimeter and interior zones) usually varies between 0.8 and 1.3 cfm / ft2 (4 and 6.5 L / s ·m2). The
reason for having lower flow density was a lower space load density at part load.
There were a certain number of minimum outdoor air intakes even lower than the base ventilation.
It is interesting to note that the amount of minimum outdoor air intake during heating mode
was higher than the amount of minimum outdoor air intake during cooling mode. This may be due
to the higher density of cold outdoor air in cold weather than the lower density of outdoor air in
warm weather.
Carbon Dioxide Concentrations. From November 1991 to November 1992, CO2 measurements
under full occupancy revealed that instantaneous daily peak values in the return ducts ranged from
450 to 750 ppm. In the space, they ranged from 500 to 850 ppm. On a few isolated occasions, CO2
concentrations in the return ducts were as high as 1100 ppm; in the space, they were as high as
1400 ppm when some of the AHUs were shut off.
Discharge Air Temperature Reset
Discharge air temperature or supply temperature reset saves energy in the following ways:
 It reduces the simultaneous cooling and heating if VAV reheating and dual-duct VAV systems are
used.
23.22 CHAPTER TWENTY-THREE
 It lowers the duct heat gain or loss.
 It raises the suction temperature of the refrigeration system.
 It increases the number of hours that the refrigeration system can be shut off during an air or water
economizer cycle.
 It increases the zone supply volume flow rate at low zone loads.
Discharge air temperature reset also has the following drawbacks:
 Lower fan energy savings because less volume flow can be reduced
 A higher space relative humidity during cooling mode operation
During part-load operation, because the reduction in space load and supply volume flow rate in the
perimeter zone is different from that in the interior zone, a separate discharge air temperaure reset
schedule for perimeter and interior zones is beneficial if separate AHUs or PUs for perimeter and
interior zones are adopted, or a dual-duct system is used.
For a conditioned space that requires strict relative humidity control, discharge air temperature
reset is not recommended. For a VAV system that uses the same AHU or PU to serve both perimeter
and interior zones, discharge air temperature reset must consider the higher space cooling load at
part load in the interior zone. In VAV systems using inlet vanes for duct static pressure and system
supply volume flow control, discharge air temperature reset is preferable because it may save more
energy. In VAV systems an analysis is often beneficial to determine whether discharge air temperature
reset saves more energy and is cost-effective.
23.8 DUCT STATIC PRESSURE AND FAN CONTROLS
Duct Static Pressure Control
In a VAV system, if more than one half of the VAV boxes close their dampers at the same time to reduce
the volume flow supplied to various zones at part-load operation, the static pressure of the supply
duct will be raised. Such a high static pressure may cause objectionable noise and volume flow
in some of the terminals and may result in unstable operation. At part load, the volume flow and the
fan total pressure of the supply fan as well as the return fan should be reduced to achieve a balanced
and optimized system operation if a supply and return fan combination is employed. The reduction
of the volume flow of the return fan is discussed in Sec. 23.6.
The purpose of supply duct static pressure control (or simply duct static pressure control) is to
maintain a nearly constant static pressure at a specific location in the main supply duct and, at the
same time, to save fan energy use. A duct static pressure control system consists mainly of a
static pressure sensor, a DDC controller, and an adjustable-frequency variable-speed drive, or an
inlet vanes actuator, as shown in Fig. 23.8. When the static pressure sensed by the static pressure
sensor at the predetermined location in the supply main duct is higher than a set point, the
DDC controller actuates the ac inverter or positions the inlet vanes to reduce the fan volume
flow and fan total pressure, so as to maintain a nearly constant static pressure at the location of
the sensor.
Duct static pressure shows a small capacitance lag, or time constant. Therefore, duct static pressure
responds quickly to changes in the supply fan output. Proportional-integral-derivative (PID)
control action with a wider proportional band is more appropriate.
A high-limit pressure sensor should be installed after the supply fan outlet to prevent damaging
the duct system by means of excessive pressure due to malfunction, or closing of the fire or smoke
dampers. When these circumstances arise, the excessive static pressure buildup due to the blockage
of the air passage can be sensed by the high-limit pressure sensor, and the DDC controller will send
a signal to turn off the fan.
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.23
Set Point of Duct Static Pressure and Sensor’s Location
The set point of the duct static pressure should be the minimum static pressure such that when added
to its associated velocity pressure, their sum—the total pressure—should overcome the maximum
total pressure loss in any of the branch takeoffs including VAV boxes, flexible ducts, diffusers, and
dynamic losses under all operating conditions. The duct static pressure is more stable than the duct
total pressure in a supply main duct when the airstream is flowing. More often, the set point of the
duct static pressure in VAV systems ranges between 0.8 and 1.5 in. WG (200 and 375 Pag)
The static pressure sensor should be located in the supply main duct in a position such that the
required supply volume flow rate is guaranteed for all VAV boxes with greatest energy savings during
part load. The nearer to the farthest branch takeoff, the greater the energy saving because pfix
is nearly minimum. For most VAV systems, the duct pressure sensor is often located in the main
duct near the farthest branch takeoff and at a location where steady static pressure can be measured.
Duct static pressure control is discussed also in Sec. 29.5.
Warren and Norford (1993) recommended that the duct static pressure set point be reset based
on messages from each VAV box collected by a poll manager control block during operation. Such
a DDC control system is complicated and expensive.
ASHRAE/IESNA Standard 90.1-1999 specifies that duct static pressure sensors used to control
variable-air-volume fan operating characteristics shall be placed in a position such that the controller
set point is no greater than one-third of the total design fan (supply) static pressure except for direct
digital control systems with zone reset capability where the sensor may be located at the fan discharge.
For a supply duct system with major duct splits, multiple sensors should be installed in each
major branch to ensure static pressure can be maintained in each of the major branches.
For air systems with direct digital control of individual zone VAV boxes connecting to the central
system controller, the duct static pressure set point shall be reset based on the zone requiring the
most pressure, that is, the set point is reset lower until one zone damper is nearly wide open.
Comparison between Adjustable-Frequency Drives and Inlet Vanes
In VAV systems, both adjustable-frequency variable-speed drives and inlet vanes are used to modulate
the volume flow and fan total pressure of the centrifugal supply, return, relief, and exhaust fans,
as described in Sec. 15.4. Generally, fans using adjustable-frequency variable-speed drive to vary
23.24 CHAPTER TWENTY-THREE
FIGURE 23.8 Duct static pressure control.
the fan speed are more energy-efficient at reduced volume flow than are positioning inlet vanes.
However, adjustable-frequency variable-speed drive is more expensive than inlet vanes.
Lorenzetti and Norford (1992) reported that adjustable-frequency variable-speed drives consumed
46 to 66 percent less supply fan energy than inlet vanes, based on the measured performance
of four AHUs during a retrofit project in two connected buildings. The brake horsepower of the supply
fan ranged from 21 to 36 hp. Only one of the AHUs operated 24 h/day, while the rest operated 8
to 15 h/day. From the results of energy simulation programs, other researchers also found that adjustable-
frequency variable-speed drives have an energy savings between 20 and 50 percent over inlet
vanes depending on the type and size of fan, annual operating hours, and characteristics and location
of the building.
Field measurements also showed that many fans installed with inlet vanes saved only 20 to 30
percent, even as little as 10 percent, when the volume flow rate was reduced to 0.3 of its design
value, if inlet vanes are mounted too far away from the impeller. Theoretically, the energy savings
should be greater than 50 percent.
Figure 23.9 compares the fan power consumption at fractions of the design volume flow rate
when adjustable-frequency variable-speed drives and inlet vanes are used. Adjustable-frequency
drives are often cost-effective under the following conditions:
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.25
FIGURE 23.9 Comparison of fan power consumption between adjustable-frequency, variable-speed drives and
inlet vanes.
 A variable supply volume flow is required during operation.
 The fan brake horsepower is 25 hp (19 kW) and greater.
 Backward-inclined and airfoil fans are the best candidates.
 The annual operating hours exceed 2500 h.
Mays (1998) suggested that the most common adjustable-frequency drive problem today is re-
flected wave voltages, which create two to three times overvoltage at the motor terminals with
pulse-width-modulated (PWM) drives as the rise time of the pulses becomes shorter. These cause
noise in motors. An external low-pass filter can be installed, or apply series reactors between the
PWM output terminals and the cable to the motor to slow the rise time or to minimize overvoltages.
The shorter the cable between the adjustable-frequency drive output and the motor, the lower will
be the peak voltage.
For a vane-axial supply fan, blade pitch controls are often used at reduced volume flow for fan
energy savings. Controllable blade pitch vane-axial fans have fan energy savings characteristics
comparable to those of adjustable-frequency variable-speed drives. But they are also expensive.
23.9 FUNCTIONAL CONTROLS FOR VAV SYSTEMS
Nighttime Setback, Warm-Up, or Cool-Down Control
Many VAV systems are shut down during the nighttime unoccupied period. Only the heating system
or the fan-powered VAV boxes are operated to maintain a nighttime setback temperature set point in
order to save energy. The purpose of warm-up control is to recover the setback temperature and
maintain an indoor environment between 70 and 72°F (21.1 and 22.2°C) during the occupied period
in winter.
The purpose of cool-down control is to cool and dehumidify the space air down to predetermined
limits following an increase in the space air temperature and humidity when the air system is
turned off during the unoccupied period in locations where the outdoor climate is hot and humid in
summer.
When the supply fan of the AHU or PU is turned on, if the difference between the space temperature
Tr sensed by the temperature sensor and the set point Tsp (Tr  Tsp) is less than or equal to a
predetermined limit, such as 3°F (1.7°C), then the air system enters the warm-up mode; and if
(Tr  Tsp)  a predetermined limit, such as 3°F (1.7°C), the air system enters the cool-down mode.
Both the warm-up and cool-down controls close the minimum outdoor, economizer, and relief / exhaust
dampers and open to 100 percent the recirculating damper. In addition, the warm-up control
opens fully the two-way valve of the heating coil or the gas valve of the furnace, and the cool-down
control opens fully the two-way valve of the cooling coil or starts all the refrigeration compressors
that serve the DX coil, until the space temperature drops within predetermined limits.
One problem in warm-up or cool-down control is to determine the optimum start time. The optimum
start time is affected by the following factors:
 Space temperature and relative humidity just prior to the warm-up or cool-down period
 Construction of the building envelope
 Capacity of the heating / cooling equipment
 Internal loads
 Outdoor temperature
These factors are used to calculate the thermal characteristics of a conditioned space served by an
AHU or PU, and start the equipment prior to occupancy to reach the desirable space temperature.
There are software packages available to calculate separate sets of heating and cooling rates for
23.26 CHAPTER TWENTY-THREE
spaces that have been unoccupied less than or greater than 24 h and provide an optimum start or
stop time of the air system.
A simple and effective way that is worthwhile to try is as follows: Record the outdoor air temperature,
space temperature, relative humidity, internal load, and equipment and damper operating
conditions at suitable time intervals, say, 10 min to 1 h, during the warm-up or cool-down period;
analyze and determine the optimum start time once sufficient data have been accumulated. Modify
the results until they are verified by experience.
Some manufacturers also provide cycling capacity during morning warm-up to maintain the
space temperature reached before the occupied period.
If both cool-down and purge mode or warm-up and purge mode operations are required, and the
outdoor air is significantly hot and humid, or colder than the space air during the occupied period,
then the purge operation must be performed first, and the cool-down or warm-up operation follows
until the space temperature is maintained within predetermined limits. During purge mode, the minimum
outdoor air, economizer, and exhaust dampers are opened 100 percent, and the recirculating
damper is closed. During warm-up or cool-down mode, the minimum outdoor air, economizer, and
exhaust dampers are closed, and the recirculating damper is opened 100 percent.
Only when the outdoor air can be used for free cooling, and to squeeze out the contaminated
indoor air at the same time, are the cool-down and purge operations combined into a single
operation.
Steam Humidifier Control
A steam humidifier is the most widely used humidifier when a warm air supply is required in winter in
an AHU or PU. A humidity sensor is often located in the supply duct just after the supply fan discharge,
to sense the relative humidity of the discharge air. The DDC controller receives the sensed signal and
positions the two-way valve to modulate the steam flow rate entering the steam humidifier, as shown in
Figs. 15.42 and 15.43, until the relative humidity of the discharge air meets the required set point.
When the fan is turned on, the humidifier control is energized. The steam valve closes when the
fan is turned off.
If a steam humidifier is installed, it is critical that the humidifying capacity not be exceeded, so
that a wet surface does not occur in the AHU as well as in the supply duct in order to maintain an
acceptable indoor air quality.
Dew Point Control
For air systems serving conditioned space that needed both close control of the space temperature
and relative humidity with a sensible ratio of the space conditioning line SHRs above 0.9, a
dew point control is often used. Dew point control involves using a water cooling coil, DX coil,
or air washer to cause a specific amount of conditioned air to cool and dehumidify or occasionally
cool and humidify to point cc, such as that on Fig. 23.10. Because of the high relative
humidity of the departing air, the temperature Tcc is approximately equal to the dew point on the
saturation curve.
When the indoor space temperature is to be maintained at 75°F (23.9°C), with a relative humidity
of 50 percent, Tcc  55°F (12.8°C), 
cc  97 percent, and SHRs  0.95, the state point of the
conditioned space r must move approximately along the horizontal dew point line. Either the sensible
cooling load is absorbed by the supply air or reheating is provided to the supply air by a reheating
coil to compensate the sensible load at part load, if the space temperature is maintained at 75°F
(23.9°C), the relative humidity must be at 50 percent. A temperature sensor senses the space temperature
Tr, and a DDC controller modulates the capacity of the reheating coil according to the
sensed space temperature to maintain Tr  75°F (23.9°C) with 
cc  50 percent.
Dew point control is simple and effective. Its main drawback is its need for reheating which
wastes energy. Dew point control is mainly used for industrial applications such as clean rooms.
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.27
Diagnostics
The system diagnostics automatically monitors the operation of all workstations, modems, LAN
connections, control panels, and controllers. The failure of any device should be displayed in a diagnostic
display module for which a legend is provided. In a recent model by a manufacturer, 49
different diagnostics can be read at the display module, and the last 20 diagnostics can be held in an
active history buffer log in the module. The diagnostic module is also capable of displaying outputs
of microprocessor-controlled run tests to verify operation of every thermistor, potentiometer, fan,
and compressor before the air system is started.
A smoke detector is usually located in the return air passage. Once the smoke detector detects smoke
in the return airstream, it sends a signal to the DDC controller and energizes the fire alarm system, as
shown in Fig. 21.3b. Smoke controls and stairwell pressurization systems are discussed in Chap. 22.
A pressure sensor is usually used to sense the pressure drop across the filter. As soon as the pressure
drop exceeds a predetermined limit, an alarm signal appears at the display module to indicate
the need for replacement or maintenance.
Safety controls are often part of the system components in HVAC&R equipment, as discussed
previously.
23.10 RECOMMENDATIONS AND INTERACTION
BETWEEN CONTROLS
Recommendations for VAV Controls
 Equipment, control valves, and air dampers should be properly selected and sized. Oversized
equipment and components degrade the control quality.
 Proportional-integral (PI) control mode is the widely used control mode in HVAC&R applications.
PI control eliminates the offset, resulting in a better indoor environment and a lower energy
23.28 CHAPTER TWENTY-THREE
FIGURE 23.10 Dew point control.
consumption than proportional control mode. A wider proportional band and properly selected
proportional and integral gains for the controller provide a stable and effective operation.
Cold discharge air temperature reset for the perimeter zone is energy-efficient when PI or proportional-
integral-derivative (PID) control mode is used.
 The use of an air economizer cycle is energy-efficient and provides adequate outdoor ventilation
air for occupants.
 A sensor should be located where the value it senses represents the controlled variable of the
whole zone to be controlled. A sensor should be exposed and located in a place that has sufficient
air movement. However, it should not be directly under a supply air jet, in which the air temperature
and relative humidity are quite different from those in the occupied zone. It also should not
be located in a stratified airstream.
 To measure the duct static pressure, if a long section of straight duct is available (say, a length of
duct section greater than or equal to 10 duct diameters), a single point, Pitot-tube type of duct
static pressure sensor can be used. Otherwise, a multipoint Pitot-tube array of flow-measuring
station with airflow straighteners should be used for velocity pressure or static pressure measurement.
The small holes used to measure static pressure in a Pitot tube should never be directly opposite
an airstream with a velocity pressure that can affect the reading. A reference pressure
should be picked up at a point with low air velocity outside the duct, at a point served by the same
air system, or in the ceiling plenum.
 A space pressure sensor should be located in an open area of the conditioned space where the air
velocity is less than 40 fpm (0.2 m / s) and where its reading is not affected by the opening of
doors. The reference pressure pickup is best located at the rooftop, at a level 10 ft (3.0 m) above
the building and shielded from the influence of wind gusts.
 If a simpler control system can do the same job as a more complex system, the simpler system is
always the first choice.
A sufficiently clear, well-followed operations manual and a well-implemented maintenance
schedule are key factors for an effective control system.
Interaction between Controls
In a typical multizone VAV system, each of the zone temperature control, discharge air temperature
control, minimum ventilation control, duct static pressure control, space pressurization control, and
warm-up or cool-down controls may consist of a single closed control loop, or several closed control
loops combined together. The interaction between these control loops can be categorized as follows:
 Sequence control. For this kind of control action, several individual control loops operate in sequence
in order to maintain a particular controlled variable within predetermined limits to prevent
simultaneous cooling and heating. For example, the discharge air temperature control in an AHU or
PU consists of an air economizer control loop as the first-stage cooling, a cooling coil control loop,
a mixed air temperature control loop, and a heating coil control loop in sequence. The two cooling
control loops can be actuated simultaneously, whereas the cooling coil, mixed air, and heating coil
control loops must be operated in sequence to prevent simultaneous cooling and heating.
 Override. If two control loops operate on the same control device, one control loop must override
the control action of the other control loop. One must have a priority to actuate the control device.
The predetermined precedence of one control signal over another in the DDC controller is called
override. For example, the minimum ventilation control and the air economizer control both
actuate the minimum outdoor air damper, economizer dampers, and recirculating dampers. When
the DDC controller fully opens the economizer and minimum outdoor dampers, the air
economizer control overrides the minimum ventilation control because the outdoor air intake
during the 100 percent air economizer cycle is far greater than the minimum outdoor air intake
during minimum ventilation control.
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.29
 Control loops are often operated independently of one another. For example, the cooling coil control
loop that modulates the two-way valve of the chilled water supply is operated independently
of the space pressurization control loop that modulates the speed of the return fan to maintain the
space pressure within preset limits.
 Control actions affect each other. The control action of one control loop somtimes may affect the
controlled variable of another control loop. For an example, the closing of the damper in a VAV box
of a zone temperature control loop reduces its volume flow and raises the static pressure of the duct
static pressure control loop. As the speed of the supply fan is reduced, the supply volume flow rate
and the fan total pressure are reduced accordingly. The duct static pressure is then maintained within
predetermined limits, and a new system balance is formed.
Stability Problems
 In VAV controls, instability may occur as the result of hunting because of a smaller proportional
band and a fast opening-closing controlled device.
 Avery (1992) noted that interaction problems may occur because the stroke speeds of the valve
actuator or damper actuator may be different, or the sensor response times may not be the same
for system components from different manufacturers.
 Control loop interaction between pneumatic operated VAV boxes and the electric inlet vane actuators
of the supply fan can be serious as the slow-operating electric inlet vane actuator cannot keep
pace with the changes in duct static pressure caused by fast-acting pneumatic actuators.
 Control loops having control actions affecting each other sometimes may destabilize the system.
23.11 SEQUENCE OF OPERATIONS OF A VAV REHEAT SYSTEM
WITH MINIMUM VENTILATION, DISCHARGE AIR TEMPERATURE,
AND DUCT STATIC PRESSURE CONTROLS—CASE STUDY
HVAC&R System
Consider a multizone VAV reheat system using an AHU to serve a typical floor in an office building
with perimeter and interior zones. The summer and winter design conditions are similar to those for
a VAV reheat system described in Sec. 21.3, except the winter design temperature is 30°F (1.1°C)
with a design outdoor humidity ratio of 0.003 lb / lb (0.003 kg / kg). Humidification is not required
in the AHU during the winter heating mode.
In minimum ventilation control, the sequence of operations of both CO2-based demand-controlled
ventilation and the alternative mixed plenum pressure control at the AHU intake are
included. The discharge air temperature control consists of a temperature air economizer cycle, a
water cooling coil control loop, a mixed air temperature control loop, and a water heating coil
control loop. The set point that changes over from minimum outdoor air recirculating mode to 100
percent outdoor air free-cooling mode is 75°F (23.9°C). There is a minimum outdoor damper, an
economizer damper, a recirculating damper, and an exhaust damper.
In duct static pressure control, supply and return fan pressure and volume flows are modulated
by varying the speeds of the supply and return fans.
Sequence of Operations
The recommended year-round sequence of operations of this VAV reheat system is as follows:
1. During the unoccupied period, the time-of-day clock shuts down the AHU and its associated
supply and return fans. Outdoor minimum damper D1, economizer damper D2, recirculating
23.30 CHAPTER TWENTY-THREE
damper D3, exhaust damper D4, cooling coil valve V1, and heating coil valve V2 are all closed.
Supply and return fan motors are off.
2. During the unoccupied period, an outdoor air temperature sensor (temperature transmitter) T1
measures the outdoor temperature To, and an indoor temperature sensor TT11 measures the indoor
temperature Tr. Also an outdoor CO2 sensor CO2 1 measures the outdoor CO2 concentration
CO2o, and an indoor CO2 sensor CO2 2 measures the concentrations of a representative floor
space CO2r.
In hot summer, if the summer indoor design temperature Tr,des  75°F (23.9°C) and the measured
representative zone temperature Trxn  3°F, a cool-down mode is required. If the measured
indoor-outdoor CO2 concentration difference (CO2r  CO2o  200 to 300 ppm, a purge mode is
also required. If both cool-down and purge modes are required and if To  78°F (25.6°C), the
purge mode must be performed first, then the cool-down mode.
 Based on the space temperature, space relative humidity, outdoor temperature, and temperature of
representative structural member, or according to the computer software, the DDC unit controller
starts supply and return fans at very low speed, such as 50 rpm through the adjustable-frequency
variable-speed drives.
 The minimum outdoor damper, economizer damper, and exhaust damper are then opened fully,
whereas the recirculating damper remains closed. The speeds of supply and return fans are gradually
increased to design speeds.
 Once the measured CO2r  CO2o 100 ppm, the purge mode is completed, and the DDC controller
closes the minimum outdoor, economizer, and exhaust dampers, and opens the recirulating
damper 100 percent.
 As To  75°F (23.9°C), the DDC controller sets the discharge air temperature set point at 52°F
(11.1°C) and opens 100 percent the cooling coil valve.
 When the difference between the measured zone temperature and the summer indoor design temperature
Tr  75 1°F (Tr  23.9 0.56°C), the cool-down mode is completed.
If both cool-down and purge modes are required, and if the outdoor temperature 52°F  To 
75°F (11.1°C  To  23.9°C), the DDC controller opens minimum outdoor, economizer, and
exhaust dampers all to 100 percent, closes the recirculating damper, and operates the air economizer
cycle in 100 percent outdoor air free-cooling mode. At the same time, it purges the conditioned
space with outdoor air. When Tr  75 1°F (Tr  23.9 0.56°C), the cool-down mode is
completed; and when CO2r  CO2o 100 ppm, the purge mode is completed.
3. As the temperature sensor measures the representative zone temperature Trxn  75°F
(23.9°C), the DDC controller calls for cooling. And if the outdoor temperature To  75°F (23.9°C),
the VAV system is in minimum outdoor air recirculating mode, as shown in Fig. 23.7.
Based on the discharge air temperature reset schedule for cooling listed below, as the outdoor
temperature To  75°F (23.9°C), the DDC controller resets the discharge air temperature Tdis 
52°F (11.1°C).
Here Tr indicates zone temperature, 
r zone relative humidity, Tcc off-coil temperature, Tmix
mixed air temperature, Tdis discharge air temperature, and Ts supply temperature. For a VAV system
using the same AHU that serves both perimeter and interior zones, Tdis only resets to 55°F
(12.8°C) when outdoor temperature To drops belows 75° F (23.9°C) because a high space sensi-
To, °F Tr, °F 
r, % Tcc, °F Tmix, °F Tdis, °F Ts, °F Tr  Ts, °F
76–95 75 42–45 50 50–80 52 55 20
59–75 75 43–46 53 53–78 55 59 16
58 75 or less 48–51 56 56–78 58 63 12
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.31
ble cooling load in the interior zone during part-load operation in an office building needs cold
air supply.
4. During minimum outdoor air recirculating mode, when the sensed zone temperature Trn exceeds
75°F (23.9°C), DDC terminal controllers modulate the opening of single-blade dampers of
the reheating VAV boxes in the perimeter zones and VAV boxes in the interior zones, to increase the
zone, volume flow of supply air to maintain a nearly constant zone temperature Trn  75°F
(23.9°C). During part-load operation, when Trn drops below 75°F (23.9°C), because of a reduction
in zone sensible cooling loads, DDC terminal controllers modulate the opening of single-blade
dampers of the reheating VAV boxes in the perimeter zones and VAV boxes in the interior zones, to
reduce the volume flow of the zone supply air to maintain a nearly constant zone temperature of
75°F (23.9°C), as discussed in Sec. 21.3.
5. During minimum outdoor air recirculating mode part-load operation, when reheating VAV
boxes and VAV boxes close their single-blade dampers to smaller openings, the pressure drop
across the dampers increases. This results in a higher static pressure in the supply main duct. As
such, an increase in duct static pressure is sensed by a pressure sensor located at a distance of
0.3Lmain from the farthest branch takeoff in the supply main duct, as discussed in Sec. 23.8; a DDC
controller modulates and reduces the speed of the supply fan through an adjustable-frequency
variable-speed drive to maintain a nearly constant duct static pressure set point, such as 1.2 in.
WG (300 Pag), at the location of the duct static pressure sensor. Therefore, the volume flow rate of
the supply fan is reduced to balance the reduction of zone supply volume flow rate during part
load.
6. During the minimum outdoor air recirculating mode, if CO2-based demand-controlled ventilation
is used, the relationship between the supply flow density of minimum outdoor air (as well as
the corresponding minimum outdoor damper opening) and the CO2 concentration in the return duct
(refer to Fig. 23.2) has already been determined during commissioning. Since as the supply fan is
modulated at a speed to maintain a preset duct static pressure of 1.2 in. WG (300 Pag), the DDC
controller opens the minimum outdoor damper to a setting to provide base ventilation bv first. According
to the measured CO2 concentration in the return duct, the DDC controller then resets the
minimum outdoor damper opening setting according to the determined relationship at a predetermined
time interval (such as 2 and 5 min).
The recirculating damper opens 100 percent, and the exhaust dampers open 30 percent.
For a VAV system with a supply and return fan combination, a pressure sensor is used to measure
the representative space pressure pr, in in. WG (Pag), and controls the space pressure in floating
control mode. Only when pr  0.03 in. WG (7.5 Pag) does the DDC controller increase the
speed of the return fan, which raises the amount of exhaust air through the exhaust damper eu and
maintains pr   0.03 in. WG (7.5 Pag). Also only when pr  0.005 in. WG will the DDC controller
reduce the speed of the return fan to increase space pressure to  0.005  pr   0.03 in.
WG (1.25  pr  7.5 Pag).
When the measured return duct CO2 concentration is  800 ppm, the minimum outdoor air intake
for a typical office building should be 0.14 cfm / ft2 (0.7 L / s m2). Minus the exhaust through
the restroom of 0.03 cfm / ft2 (0.15 L / s m2), the remaining part is 0.14  0.03  0.11 cfm / ft2
(0.55 L / s m2). Such an amount of outdoor air is sufficient to maintain a space pressure of  0.005
 pr   0.03 in. WG (0.005  pr   7.5 Pag) and an associated exfiltration from an average
leaky building as well as an amount of 0.015 cfm / ft2 (L/sm2) of exhaust air through the exhaust
damper.
V?
Measured CO2 concentration, ppm Required minimum outdoor air, cfm/ ft2 (L/sm2)
800 V ·
o,sys  0.14 (0.7)
400 V ·
bv  0.07 (0.35)
V?
23.32 CHAPTER TWENTY-THREE
Similarly, when the measured CO2 concentration is  400 ppm, the minimum outdoor intake
0.07 cfm / ft2 (0.35 L / s m2), and the remaining part is 0.07  0.03  0.04 cfm / ft2
(0.2 L / s m2) which can provide a space pressure around  0.005 in. WG (1.25 Pag).
7. During the minimum outdoor air recirculating mode, if a mixed plenum pressure control is
used, either a pressure sensor located in the mixed plenum measures the static pressure difference
between the mixed plenum and outdoors pm,p, as shown in Fig. 23.4a, or a pressure sensor measures
the pressure drop of the outdoor air damper and the louver, as shown in Fig. 23.4b. If the measured
static pressure difference or pressure drop decreases and is less than a set point, such as  0.4
to  0.6 in. WC (100 to  150 Pa), the DDC controller closes the recirculating damper until
pm,p equals the set point and an amount of minimum outdoor air of o,sys  0.14 cfm / ft2 (0.7
L/sm2) is extracted in the AHU. The minimum outdoor air damper is opened a fixed amount, such
as 100 percent.
Mixed plenum pressure control provides an approximately constant minimum outdoor air with an
amount of o,sys. During minimum outdoor air recirculating mode, for VAV systems with a supply
and return fan combination, the minimum outdoor damper opens 100 percent, the exhaust damper
opens 30 percent, and the economizer damper is closed. A space pressure sensor measures the representative
space pressure pr, in in. WG (Pag). If it exceeds  0.03 in. WG (7.5 Pag), a DDC controller
increases the speed of the return fan through a variable-speed drive, and thus reduces the amount of
minimum outdoor air intake as well as the space pressure and maintains pr at less than  0.03 in.
WG (7.5 Pag).
When the zone load reduces, the speed and volume flow of the supply fan reduce accordingly.
As the measured pr drops below 0.005 in. WG (1.25 Pag) because the return fan has not followed
the supply fan to reduce its volume flow appropriately, the DDC controller then reduces the speed
of the return fan, and therefore space pressure pr increases and is maintained at a value not less than
0.005 in. WG (1.25 Pag), as discussed in Sec. 23.6.
8. When the outdoor temperature 55°F  To  75°F (12.8°C  To  23.9°C), the AHU is operated
in 100 percent outdoor air economizer cycle, as shown in Fig. 23.7b. The DDC controller
opens 100 percent the minimum outdoor air damper, economizer damper, and exhaust damper. The
recirculating damper is closed. When 100 percent outdoor air is used for free cooling, it overrides
the minimum ventilation control.
The cooling and dehumidifying capacity of the chilled water cooling coil is also required at the
same time. The DDC controller modulates the opening of the two-way control valve of the cooling
coil to maintain a discharge air temperature Tdis  55°F (12.8°C).
9. When the outdoor temperature To,mix  To  55°F (12.8°C), the AHU is operated in the air
economizer cycle with mixed air control. Here To,mix is the outdoor air temperature at which the
AHU or PU starts the mixed air control (such as To,mix  30°F, or  1.1°C) to provide a mixture
of outdoor air and recirculating air at discharge air temperature Tdis directly, as discussed in Sec.
23.7. (Refer to Fig. 23.7.) So To,mix depends on the percentage of outdoor air among the supply
air and on the load ratio. In an air economizer cycle mixed air mode, a DDC controller opens 100
percent the minimum outdoor air damper and exhaust damper and modulates the economizer
damper and recirculating damper to maintain a mixed air temperature equal to Tdis  55°F
(12.8°C) directly.
Since the amount of outdoor air intake in the AHU during mixed air mode is always greater than
the amount of minimum outdoor air o,sys in recirculating mode, the mixed air control overrides the
minimum ventilation control.
To provide cold air supply for the interior zone even in winter, the DDC controller resets Tdis 
55°F (12.8°C).
10. In the perimeter zone
 If the perimeter zone temperature Trxn and interior zone temperature Trin drop below 75°F
(23.9°C), the terminal controller reduces the opening of the single-blade damper and thus the sup-
V?
V?
V?
V?
o,min  V?bv 
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.33
ply volume flow rate of cold conditioned air to the control zone and maintains a nearly constant
zone temperature Trxn  Trin 75°F (23.9°C).
 Further drop in the perimeter zone space sensible cooling load causes the damper in the reheating
VAV box to close to a smaller flow passage until it reaches its minimum setting, such as 30 percent
of the zone peak flow.
 When the temperature of perimeter zone n is 72°F  Trxn  75°F (22.2°C  Trxn  23.9°C),
zone n is operated in the deadband mode, as discussed in Sec. 21.3. The damper in the reheating
VAV box closes to its minimum setting, and neither cooling nor heating is provided except minimum
ventilation air.
 When the temperature of the perimeter zone n is Trxn  72°F (22.2°C), the zone temperature
control is operated in heating mode operation. As discussed in Sec. 21.3, the DDC terminal
controller actuates the reverse-action relay, and the cold primary air supply is still in minimum
setting to provide minimum ventilation. Also the terminal controller modulates the opening of
the two-way valve of the heating coil to maintain a preset zone temperature Trxn  72°F
(22.2°C). The zone relative humidity may be between 25 and 50 percent. At winter design conditions,
the warm air supply temperature should not exceed 87°F (30.6°C) in order to prevent the
buoyancy effect.
Primary Considerations
 The sequence of operations of a VAV reheat system can be divided into two categories: sequence
of operations for AHU or PU controls and for zone controls (including perimeter and interior
zones).
 Sequence of operations for AHU or PU is usually varied based on outdoor temperature To and can
be subdivided into
Minimum outdoor air recirculating mode, cooling
Air economizer cycle, 100 percent outdoor air free-cooling
Air economizer cycle, mixed air control
Minimum outdoor air recirculating mode, heating
 Sequence of operations for zone control is varied based on zone temperatures: perimeter zone
temperature Trxn and interior zone temperature Trin. Zone temperature control can be subdivided
into cooling, deadband, and heating mode operations.
 Minimum ventilation control is only valid to provide ventilation during minimum outdoor air
recirculating mode between base ventilation bv  0.07 cfm / ft2 (0.35 L / s m2) and minimum
outdoor air specified by ASHRAE Standard 62–1999, o,sys  0.14 cfm / ft2 (0.7 L / s m2) for a
typical office building. During air economizer mode, outdoor air intake in the AHU or PU is always
greater than o,sys; there is no need of minimum ventilation control.
 For a VAV system with supply and relief fan combination operated in minimum outdoor air recirculating
mode, if minimum outdoor air is provided by CO2-based demand-controlled ventilation
or mixed plenum pressure control, the space pressure is maintained between  0.005 and  0.03
in. WG (1.25 and 7.5 Pag) for a properly designed average leaky office building.
For a VAV system with supply and return fan combination, when the speed of the supply fan is
reduced during part-load operation, a DDC controller should reduce the speed of the return fan to
maintain a space pressure between  0.005 and  0.03 in. WG (1.25 and 7.5 Pag).
 If a humidifier is used, be sure that the humidifying capacity is effectively controlled and the
wet surface doesn’t occur inside the AHU, PU, or supply ducts, to create indoor air quality
problems.
V?
V?
V?
23.34 CHAPTER TWENTY-THREE
REFERENCES
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA 1999.
ASHRAE, ASHRAE Standard 62–1999, Ventilation for Acceptable Indoor Air Quality, Atlanta, GA 1990.
Avery, G., The Instability of VAV Systems, Heating / Piping / Air Conditioning, February 1992, pp. 47–50.
Avery, G., Selecting and Controlling Economizer Dampers, HPAC, no. 8, 1996, pp. 73–78.
Becelaere, R. V., Mixing Box Damper Testing, ASHRAE Transactions 1998, Part II, pp. 1226–1231.
Carpenter, S. C., Energy and IAQ Impacts of CO2-Based Demand-Controlled Ventilation, ASHRAE
Transactions, 1996, Part II, pp. 80–88.
Chen, S. Y. S., and Demster, S. J., Variable Air Volume Systems for Environmental Quality, McGraw-Hill, New
York, 1995.
Chen, S. Y. S.,Yu, H. C., and Hwang, D. D. W., Ventilation Analysis for a VAV System, Heating / Piping / Air
Conditioning, April 1992, pp. 36–41.
Ellis, R., and McKew, H., Back to Basics, Test 27—Central Air with Return Air System Advanced Energy-Effi-
ciency—Design, Engineered Systems, no. 8, 1997a, pp. 9–10.
Ellis, R., and McKew, H., Back to Basics, Test 29—Central Air with Return Air System Building Automation—
Design, Engineered Systems, no. 9, 1997b, pp. 11–12.
Elovitz, K. M., Variable-Speed Drives: What They Do and When to Use Them, Heaitng / Piping / Air Conditioning,
no. 12, 1993, pp. 67–75.
Elovitz, D. M., Minimum Outside Air Control Methods for VAV Systems, ASHRAE Transactions, 1995, Part II,
pp. 613–618.
Elovitz, D. M., Minimum Outside Air Ventilation in VAV Systems, Engineered Systems, no. 3, 1997, pp. 44–50.
Emmerich, S. J., and Persily, A. K., Literature Review on CO2-Based Demand-Controlled Ventilation, ASHRAE
Transactions, 1997, Part II, pp. 229–243.
Enermodal Engineering Ltd., An Evaluation of the Effect of CO2-Based Demand-Controlled Ventilation Strategies
on Energy Use and Occupant-Source Contaminant Concentrations, ASHRAE Research Project 740-TRP,
Waterloo, Ontario, Canada, 1995.
Englander, S. L., and Norford, L. K., Saving Fan Energy in VAV Systems—Part 1: Analysis of a Variable-
Speed-Drive Retrofit, ASHRAE Transactions, 1992, Part I, pp. 3–18.
Graves, L. R.,VAV Mixed Air Plenum Pressure Control, Heating/Piping/ Air Conditioning, no. 8, 1995, pp. 53–55.
Janssen, J. E., Ventilation for Acceptable Indoor Air Quality, ASHRAE Journal, October 1989, pp. 40–46.
Janu, G. J.,Wenger, J. D., and Nesler, C. G., Strategies for Outdoor Airflow Control from a System Perspective,
ASHRAE Transactions, 1995, Part II, pp. 631–643.
Kettler, J. P., Controlling Minimum Ventilation Volume in VAV Systems, ASHRAE Journal, no. 5, 1998, pp. 45–50.
Lo, L., VAV System with Inverter-Driven AHU for High-Rise Office Building in Tropical Climates—A Case
Study, ASHRAE Transactions, 1990, Part I, pp. 1209–1217.
Lorenzetti, D. M., and Norford, L. K., Measured Energy Consumption of Variable-Air-Volume Fans under Inlet
Vanes and Variable-Speed-Drive Control, ASHRAE Transactions, 1992, Part II, pp. 371–379.
Mays, M., Identifying Noise Problems in Adjustable Speed Drives, ASHRAE Journal, no. 10, 1998, pp. 57–60.
Mumma, S. A., and Wong,Y. M., Analytical Evaluation of Outdoor Airflow Rate Variation vs. Supply Airflow
Rate Variation in Variable-Air-Volume Systems When the Outdoor Damper Position Is Fixed, ASHRAE
Transactions, 1990, Part I, pp. 1197-1208.
Nabinger, S. J., Persily, A. K., and Dois,W. S., A Study of Ventilation and Carbon Dioxide in an Office
Building, ASHRAE Transactions, 1994, Part II, pp. 1264–1274.
Reardon, J. T., Air Change Rates and Carbon Dioxide Concentrations in a High-Rise Office Building, ASHRAE
Transactions, 1994, Part II, pp. 1251–1263.
Roberts, J. W., Outdoor Air and VAV Systems, ASHRAE Journal, September 1991, pp. 26–30.
Robinson, K. D., Damper Control Characteristics and Mixing Effectiveness of an Air-Handling Unit
Combination Mixing / Filter Box, ASHRAE Transactions, 1998, Part I A, pp. 629–637.
AIR SYSTEMS: MINIMUM VENTILATION AND VAV SYSTEM CONTROLS 23.35
Ruud, S. H., Fahlen, P., and Anderson, H., Demand Controlled Ventilation—Full Scale Tests in a Conference
Room, Proceeding of the 12th AIVC Conference Air Movement and Ventilation Control within Buildings, Air
Infiltration and Ventilation Center, Conventry, United Kingdom, 1991, pp. 187–200.
Stonier, R. T., CO2: Powerful IAQ Diagnostic Tool, Heating / Piping / Air Conditioning, no. 3, 1995 pp. 88–102.
Taylor, S., Is CO2 Appropriate Indicator of IAQ? ASHRAE Journal, no. 12, 1998, p. 8.
The Trane Company, Packaged Rooftop Air Conditioners (20 to 130 Tons), The Trane Company, Clarksville,
TN, February 1997.
Turpin, J. R., Driving into the Future, Engineered Systems, no. 6, 1997, pp. 34–39.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechnic, Hong Kong, 1987.
Waren, M., and Norford, L. K., Integrating VAV Zone Requirements with Supply Fan Operation, ASHRAE Journal,
no. 4, 1993, pp. 43–46.
Williams, V. A., VAV System Interactive Controls, ASHRAE Transactions, 1988, Part I, pp. 1493–1499.
23.36 CHAPTER TWENTY-THREE
CHAPTER 24
IMPROVING INDOOR AIR
QUALITY
24.1
24.1 IAQ PROBLEMS 24.1
24.2 VENTILATION, DILUTION
OF CONCENTRATION OF
CONTAMINANTS 24.2
Outdoor Air Requirements for
Ventilation 24.2
Importance of Time of Operation and
Ventilation Rate 24.2
Adopting an Air Economizer and Purge
Operation 24.2
Select an Appropriate Minimum
Ventilation Control System 24.3
Install a Minimum Outdoor Damper
and an Economizer Damper 24.3
24.3 ELIMINATING MICROBIAL
GROWTH 24.4
Basics 24.4
Contagious Respiratory
Microorganisms 24.4
Microbial Growth 24.4
Eliminate Water Leaks, Prevent
Dampened Surface and Material 24.5
Purge, Pressurization Control, and
Ultraviolet Germicidal Irradiation 24.5
24.4 FILTERS TO REMOVE
CONTAMINANTS 24.6
Remove Indoor Air Contaminants by Air
Filters 24.6
Filter Selection for IAQ 24.6
Service Life of Air Filters 24.7
Filter Installation 24.7
24.5 REMOVING GASEOUS
CONTAMINANTS BY ADSORBERS
AND CHEMISORBERS 24.8
Indoor Gaseous Contaminants 24.8
Activated Carbon Adsorbers 24.9
GAC Performance 24.9
Simulating GAC Applications 24.10
Chemisorption 24.11
Chemisorber Performance 24.11
24.6 MAINTENANCE TO
GUARANTEE IAQ 24.12
Inspection, Service, and Access 24.12
Monitoring of Operating Conditions
in Air Systems 24.12
Coils and Ductwork 24.12
24.7 SPACE PRESSURIZATION
CONTROL 24.13
24.8 LEGAL RESPONSIBILITY
FOR IAQ CASES 24.13
SBS or IAQ Cases 24.13
Who Is Legally Responsible 24.13
HVAC&R Engineer 24.14
REFERENCES 24.15
24.1 IAQ PROBLEMS
The sick building syndrome (SBS) received public attention from 1970s after the energy crisis as
a result of a tighter building and a reduced amount of outdoor ventilation air. Since Americans
are spending more and more of their time indoors, they need a comfortable and healthy indoor
environment and an acceptable indoor air quality. In an unhealthy building environment, uncomfortable
employees do not perform well and productivity declines. Worker illness due to a poor
indoor environment elevates absenteeism. A significant issue that is facing building owners, operating
mangers, architects, consulting engineers, and contractors today is the possibility of legal
suits filed by occupants or owners who feel that their health has been damaged by poor indoor air
quality. In the 1900s, indoor air quality (IAQ) has become one of the primary concerns in air conditining
(HVAC&R) system design, manufacturing, installation, and operation because of the following:
 IAQ is closely related to the health of the occupants inside a building, whether a building is a
healthy building or a sick building.
 IAQ and thermal control (zone temperature and relative humidity control) indicate primarily the
quality of the indoor environment in a building.
 IAQ and thermal control represent mainly the functional performance of an air conditioning
(HVAC&R) system.
As described in Sec. 4.10, the indoor air contaminants are the basic pollutants that affect IAQ.
According to the results of many field investigations and recent IAQ conferences, the primary IAQ
problems and the appropriate mitigations and improvements are based on the three basic strategies:
contaminated source control, removing contaminants by air cleaner, and ventilation air dilution,
which are summarized and discussed in the following sections.
24.2 VENTILATION, DILUTION OF
CONCENTRATION OF CONTAMINANTS
Outdoor Air Requirements for Ventilation
Persily (1989) took field measurements of the outdoor ventilation rates of 14 office buildings for approximately
one year in 1983 and found that they are typically between 0.6 and 1.2 ach, with a mean
value of 0.94 ach. Among these 14 office buildings 52 percent had a minimum level of outdoor air intake
that was lower than the design level, 45 percent had less than 20 cfm / person (10 L / sperson), 8
percent had less than 10 cfm / person (5 L / sperson), and 1 percent had less than 5 cfm / person (2.5
L/sperson). ASHRAE Standard 62-1999 specified outdoor air requirements for ventilation for each
occupant in various commercial facilities. Some of these requirements are listed in Table 4.5. Outdoor
air requirements for ventilation per person o,req, such as 20 cfm / person (10 L / s person) in an office
building, must be provided at both design flow and part-load reduced volume flow in VAV systems.
From Eqs. (23.2) and (23.3), the calculated system outdoor air volume flow rate o,sys  Noc o,req, in
cfm (L/ s). Here Noc indicates number of persons, and o,sys is the minimum outdoor air intake at design
flow in an AHU or a PU for the dilution of the concentrations of indoor air contaminants. For a
typical office building with a ceiling height of 9 ft (2.7 m) o,sys is 0.93 ach, as listed in Sec. 23.2.
According to Mendell (1993), a statistically significant correlation between outdoor ventilation
rate in offices and sick building syndrome complaints was found. Ventilation has been proved to be
the most important factor to improve IAQ.
Importance of Time of Operation and Ventilation Rate
An often discussed topic is whether the type of HVAC&R system is of any significance that affects
the aspect of ventilation and thus the SBS and IAQ. Sundell (1996), based on a study of 210 office
buildings in northern Sweden, noted that the type of air system—whether it was a modern new
HVAC&R system or an old simple system—had no significance on the aspect of ventilation and
IAQ. Instead, the important aspects of ventilation were the time of operation and the ventilation
rate. If the time of operation of the air system to provide ventilation was less than 10 h /day, the risk
of SBS was increased. As the ventilation rate cfm / person (L / s person) was reduced, the odds of
SBS symptoms were raised accordingly.
Adopting an Air Economizer and Purge Operation
Select an AHU or a PU with an air economizer and purge mode control. As discussed in Sec. 23.11,
if an air economizer is used, the amount of outdoor air intake in the AHU or PU during 100 percent
V? V?
V?
V?
V?
24.2 CHAPTER TWENTY-FOUR
air economizer cycle and mixed air control mode is far greater than the calculated minimum air intake
at design flow o,sys. A 100 percent air economizer cycle significantly improves the IAQ.
If the difference in concentration of CO2 between the conditioned space and the outdoors during
morning time unoccupied period  200 to 300 ppm, or the difference in concentration of a specific
indoor contaminant between outdoors and indoors is significantly greater than the normal operating
period, a purge mode operation to exhaust all the space air with higher concentration of indoor contaminants
and extract the outdoor air of lower concentrations of contaminants is beneficial to lower
the concentration of the indoor contaminant.
Select an Appropriate Minimum Ventilation Control System
As discussed in Chap. 23, for a multizone VAV system that serves space with an obvious occupancy
variation schedule and has low pollutants from nonoccupant-generated source, a CO2-based demand-
controlled ventilation (DCV) system should be used. In a DCV system:
 The amount of outdoor air intake during minimum outdoor air recirculating mode varies between
at design occupancy and the base ventilation bv when occupancy reduces.
 Outdoor air requirements for ventilation per person, in cfm / person (L / s person), meet the criteria
specified in ASHRAE Standard 62-1999.
 Part of the cooling and heating energy to condition outdoor air is saved during reduced occupancy.
For a VAV system that serves space with a schedule of less occupancy variation, a mixed plenum
pressure (MPP) minimum ventilation control system monitoring either plenum pressure or damperlouver
pressure drop should be used. In an MPP control system, a fixed amount of outdoor air o,sys
is extracted in an AHU or a PU for ventilation at both design flow and part-load volume flow during
minimum outdoor air recirculating mode.
During commissioning, it is important to determine the relationship between the measured
concentration of CO2 in the return air duct (or on a representative floor space) and the minimum
outdoor air intake expressed as the damper opening setting in a CO2-based demand-controlled ventilation
system, as discussed in Sec. 23.2.
Install a Minimum Outdoor Damper and an Economizer Damper
For better minimum ventilation control in a VAV system, it is recommended to install two outdoor
dampers: a minimum outdoor damper and an economizer damper. The minimum outdoor damper is
opened during minimum outdoor air recirculating mode. Its size is designed based on a volume flow
rate of o,sys, in cfm (L / s). The economizer damper is opened fully during 100 percent air economizer
cycle. Its size is based on a volume flow rate ec,d, in cfm (L / s), and can be calculated as
(24.1)
where  supply volume flow rate of the AHU or PU, cfm (L / s). The face velocity of either the
minimum outdoor damper or the economizer damper is usually 1000 to 3000 fpm (5 to 15 m / s),
and the ratio of the area of damper to the area of the duct of flow passage Adam /Ad is often 0.5 to 0.9,
as discussed in Sec. 5.7. The pressure drop of the damper when the damper is opened 100 percent
pdam can be calculated by Eq. (5.11) and often has a value between 0.2 and 0.5 in. WC (50 and
125 Pa) for minimum outdoor damper and economizer damper. During 100 percent air economizer
cycle, both the minimum outdoor damper and the economizer damper are fully opened, and the
pressure drops across the dampers are equal.
Outdoor air intakes should have wind shields and louvers, away from plumber stacks, and exhausts,
with air filters or cleaners if the outdoor air quality is poor, and consider prevailing winds,
as discussed in Sec. 16.2.
V?
s
V?
ec,d  V?s  V?o,sys
V?
V?
V?
V?
V?
o,sys
V?
IMPROVING INDOOR AIR QUALITY 24.3
24.3 ELIMINATING MICROBIAL GROWTH
Basics
Microorganisms or microbes include fungi, bacteria, and viruses. Fungi and some bacteria form
spores. Mold is minute fungi growing on moist organic substance. According to Kowalski and Bahnfleth
(1998), disease-causing microorganisms transmit respiratory irritation and other serious health
pathogens via the airborne route from two kinds of sources: Most contagious pathogens come from
human hosts, and most noncommunicable pathogens come from the environment—the HVAC&R
and building-related environment.
At the ASHRAE Journal Health Building /IAQ 1997 Conference, Dr. Bascom reported that “In
the United States and Australia, asthma rates have nearly doubled in the past 15 years.”
Contagious Respiratory Microorganisms
Kowalski and Bahnfleth (1998) noted that most contagious respiratory microorganisms induce their
human host to aerosolize large quantities of infectious bioaerosols by irritation, which causes cough
and sneezing. A single sneeze can generate 100,000 floating bioaerosols, and many carry viable microorganisms.
A single cough typically produces only 1 percent of these bioaerosols but are ten
times more frequent than sneezes. Keeping a distance from the infected person is a simple and useful
method to reduce the risk of being infected.
Microbial Growth
Spores are characteristically larger and more resistant to impacts that will destroy viruses and bacteria.
According to Kowalski and Bahnfleth (1998), noncommunicable diseases are almost entirely
due to fungal or actinomycete spores and environmental and agricultural bacteria. Spores form the
most important group in noncommunicable diseases. Microbes are commonly present in outdoor
air. Normally, indoor air spore levels tend to be from 10 to 100 percent of outdoor spore levels
when the species mix indoors reflects outdoors and are mostly less than 200 colony forming units
(CFU) per m3 because of transport from outdoors to indoors due to ventilation, infiltration, and people,
and because of microbial growth indoors. In the California Healthy Building Study from
Godish (1995), a naturally ventilated building had an indoor spore level of about 80 percent that of
outdoors, mechanical ventilated building about 65 percent, and an air conditioned building about 15
percent.
Spores germinate and grow in the presence of water and nutrients—dirt and debris. The key
factor is water. Water may exist inside a building or an HVAC&R system because of the following:
 Water leaks.
 Entrained water droplets.
 Wet coil face velocity exceeding 550 fpm (2.75 m / s).
 Wetted interior surface.
 Negative space pressure, due to improper airflow balance, extracts hot and humid outdoor air
which contacts with the indoor cold surfaces and causes surface condensation.
 Clogged condensate pan and drain system.
 Oversized humidifier or improperly designed direct evaporative cooler.
The result is a “damp building” which is characterized by dampened materials, damp surfaces,
mold, and microbial growth. Legionnella is a kind of pathogen transmitted via airborne route because
the cooling tower water system is not properly treated and maintained.
24.4 CHAPTER TWENTY-FOUR
Sundell (1996) noted that many studies have demonstrated a relationship between damp buildings
and respiratory and allergic diseases in children and adults, and estimated that approximately
20 percent of asthma in children in Sweden can be attributed to damp housing.
Eliminate Water Leaks, Prevent Dampened Surface and Material
Always remember that water and dirt (dust and debris) are two necessities for microbial growth.
Water leaks through the building envelope, especially through cracks and openings around outdoor
air ducts, must be prevented. Pooling of water on the roof should be prevented.
In an HVAC&R system without a humidifier, the condensate flows over part of the outer surface
of a water cooling coil or DX coil and collects in the condensate pan. This is often the only wetted
surface inside the system that contacts the conditioned supply air.
As discussed in Secs. 15.10 and 15.11, to prevent dampened surfaces and damp material, the
following is recommended:
 Choose a coil face velocity va 	 550 fpm (2.75 m / s) for corrugated fins to prevent condensate
droplets flying over.
 If the coil face has a vertical dimension exceeding 42 in. (1067 mm), an intermediate drain pan
must be added to prevent condensate carryover.
 Install an individual drain pan and drain pipes for each coil. Drain pan may extend 1.5 to 3 times
the coil depth in the direction of airflow.
 Condensate should be drained freely with a condensate trap, condensate collected in the condensate
pan should be completely drained, and cleaned periodically. There should be no dirt and debris
remaining in the condensate pan.
 Prevent surface condensation especially because of the infiltrated hot and humid outdoor air when
the space pressure is negative.
In case a humidifier must be installed in an air conditioning system, the following is recommended:
 For most comfort air conditioning systems, select a steam humidifier or a heating element humidifier,
instead of an atomizing or a supersonic humidifier, to prevent possible microbial growth.
 Modulation capacity control is essential. An accuracy 	  5 percent relative humidity is recommended.
Purge, Pressurization Control, and Ultraviolet Germicidal Irradiation
Kowalski and Bahnfleth (1998) recommended the following HVAC&R measures to eliminate microbial
growth and improve IAQ:
 If the concentration of a specific species of disease-causing microbe in outdoor air is substantially
lower than that in the indoor air, a 100 percent outdoor air purge at a flow rate of 1 ach lowers the
indoor concentration to 40 percent of the initial concentration level after 1 h of purge operation. If
the purging airflow rate is 4 ach, the indoor concentration level drops to a few percent of the initial
concentration level after 1 h of purge operation.
 Room pressurization control is often used in biohazard facilities and isolation rooms to prevent
migration of microbes from one room to another. Room pressure is difficult to maintain and control,
and is also expensive.
 The effectiveness of ultraviolet germicidal irradiation (UVGI) to inhibit fungal growth and kill the
spores depends on the air velocity of the airflow passing the UVGI, airflow patterns, degree of
IMPROVING INDOOR AIR QUALITY 24.5
maintenance, resistance of the microbes, and ambient humidity. A single-pass airflow through a
UVGI device may have only limited effect. However, recirculation airflow results in multiple
exposures or chronic dosing.
24.4 FILTERS TO REMOVE CONTAMINANTS
Physically, indoor airborne contaminants are in solid, liquid, or gaseous form. Solid particulate
contaminants include paper and fibrous particles, bacteria, fungus, spores, and viruses. Their size is
between 0.003 and 100 	m. Liquid contaminants include mists, water, paints, cleaning sprays, and
printing inks that typically have a size betweeen 1 and 50 	m. Gaseous contaminants such as CO,
CO2, and NO2 and volatile organic compounds are fine particulates. Their size typically ranges
from 0.003 to 0.006 	m.
Remove Indoor Air Contaminants by Air Filters
Basically, removing air contaminants by air filters, as discussed in Sec. 5.14, has the following primary
purposes:
 To provide protection from respirable particulates, bioaerosols, and toxic and nuisance contaminants
for occupants
 To protect the coils, ducts, and other air distribution flow passages from accumulated dust, dirt,
and microbial growth
 To reduce cleaning and maintenance expenses
Airborne particulate matter is typically in the size range of 0.01 to 100 	m.” Dust is defined as “an
air suspension of particles (aerosol) of any solid matter, usually with particle size less than 100 	m.”
However, particulates smaller than 10 	m in diameter may penetrate into the upper and lower
parts of the respiratory passage and lungs. The indoor concentration level of particulates less than 10
	m must follow National Ambient Air Quality Standards as listed in Sec. 4.10. In addition, size of
the particulates or indoor contaminants is an essential factor that affects the efficiency of air filters.
Filter Selection for IAQ
As discussed in Section 15.14, ANSI/ASHRAE Standard 52.2-1999 defines 16 minimun efficiency
reporting values (MERV) for coarse-, low-, medium-, and high-efficiency air filters based on particle
size efficiency E1 in size range 0.3 to 1.0 	m, E2 in size range 1.0 to 3.0 	m, E3 in size range
3.0 to 1.0 	m and arrestance efficiency. Filters should be selected primarily based on particle size
efficiency, the requirement of the conditioned space, pressure drop of the filter, service life, and thelife-
cycle cost analysis. Considering IAQ, Burroughs (1997) and Liu and Huza (1995) select the
following air filters to perform the specified duties:
 Employ medium-efficiency air filters MERV9 and higher with an E2 < 50 percent up to E2 > 95
percent in particle size 1 to 3 	m to control respirable particulates between 2 to 8 	m in size that
may be breathed deeply into the lungs.
 Employ minimum air filtration level of MERV7 with a particle size efficiency of 50 ? E3 < 70
percent for a size range of 3 to 10 	m in an air system for the protection of the coil and air distribution
system and to prevent the nutrition of microbial growth.
 Employ high-efficiency air filters MERV 13, 14, 15 with a low- or medium-efficiency prefilter;
employ ultrahigh-efficiency air filters, HEPA and ULPA filters, with a medium-efficiency prefilter
to extend service life of these filters.
24.6 CHAPTER TWENTY-FOUR
Gill (1996) recognized that the use of a single low-efficiency filter in an AHU or PU has given way
to a filter assembly with a prefilter and high-efficiency filter or at least medium-efficiency filters.
Kowalski and Bahnfleth (1998) recommended that HEPA filters be used to remove microbes from the
recirculating air. In a recirculating air system with an airflow rate of 1 ach, the microbial concentration
level drops to 40 percent of its initial level after 1 h of operation. If the airflow rate is 4 ach, the microbial
concentration drops to only few percent of the initial concentration level after 1 h of operation.
Service Life of Air Filters
When the media used in an air filter to remove air contaminants are clean, they are lower in effi-
ciency. As the filter is loaded with more dirt, its efficiency gradually increases to a maximum. Then
some particulates may have migrated through the media and be unloading back into the airstream
for low- and medium-efficiency filters. At the same time, the pressure drop across the dirt-loaded
air filter increases to a limit, and the amount of air passing through the air filter reduces to a point
such that the filter doesn’t pass through sufficient air for cooling or heating. This is often the end of
the service life of the air filter, and it should be replaced by a new one. It is difficult to choose an
optimum changing point between the increase of efficiency and the loss of airflow.
Ottney (1993) recommended the following service life of air filters as a reference for designers
and operators.
The following are useful measures to approach an optimum service life:
 As described in Sec. 23.9, a pressure sensor is used to measure the pressure drop pfil, in in. WC
(Pa), across each air filter. As soon as pfil exceeds a predetermined limit, an alarm signal appears
on the display module asking for replacement or maintenance. For a VAV system, pfil should be
measured when the airflow is approximately equal to the design airflow (such as the airflow during
morning cool-down or purge mode operation).
Sometimes the specified final pressure drop never reaches up because of (1) the blowout of the
filters, especially holes developed in bag filters, (2) leakages from the filter hardwares, (3) leakages
from the suction side of the supply fan, and (4) possible pressure sensor defects.
 Plan a preventive maintenance schedule, and change the air filters according to the planned schedule.
 Do not change the air filter just because it looks dirty on the surface. For panel filters, the filter
should be replaced when the air leaving side shows dirt.
Filter Installation
In Ottney (1993), the following is emphasized:
 Air filters should be installed in gasketed channels. A fabric pile or polypropylene-finned gasket
material provides a better seal. Poor-fitting and missing filler pieces are a primary factor that affects
leakage.
Max. pressure drop
Filter type Typical change time across air filter, in. WC (Pa)
Flat panel 30 to 60 days 0.5 (125)
2-in. (50-mm) pleated 3 to 6 months 0.9 (225)
4-in. (100-mm) pleated 10 to 14 months 0.9 (225)
6- to 12-in. (150 to 300-mm) cartridge 12 to 18 months* 1.5 (375)
21- to 36-in. bags 12 to 24 months* 1.5 (375)
HEPA 1 to 5 years* 2.0 (500)
*With prefilters
IMPROVING INDOOR AIR QUALITY 24.7
 Many draw-through AHUs or PUs form a negative pressure region between the mixed plenum
(including air filter) and the supply fan inlet. Any gaps in the casing, access panels, or doors
downstream from the air filter are pathways for unfiltered air to enter the air system. All joints of
the casing and access doors should be caulked or tape-sealed. Check all pipe and electric conduit
penetrations for leaks. Improper fit and loose seating that allows a substantial amount of air leakage
to bypass the filter are the most frequent problems observed.
 High-efficiency microfine fiberglass media of the bag filter have low tensile strength and are susceptible
to punctures. When the media contact with sheet-metal screws or static pressure tips, it
causes damage. If two adjacent pockets touch, abrasion results in a breach and they should be replaced.
The bottom row of the bag filter may come in contact with condensate water in some filter
banks. Wet filters grow microorganisms. These bags are either mechanically supported or
changed to rigid-style filter to prevent wetness.
 Rigid extended surface filters are pleated in boxed or cartridge style. They are usually damaged
by rough handling. The seal of the filter pack to the interior of its housing is essential.
 Panel-type, low-efficiency filters should be inspected to be sure that they are not installed backward.
The air entering side is more open and dry, and the air leaving side is more dense and oily.
 An inspection of particulate pathways or leaks should be conducted each year and the observations
recorded in a logbook.
24.5 REMOVING GASEOUS CONTAMINANTS
BY ADSORBERS AND CHEMISORBERS
Indoor Gaseous Contaminants
Among the indoor air contaminants as discussed in Sec. 4.10, combustion products, volatile organic
compounds (VOCs), smoke, objectable odors, and radon are all gaseous contaminants. Radon is an
odorless, colorless naturally occurring radiative gas. Only about 6 percent of U.S. homes have annual
average radon concentrations exceeding the acting level 147 Bq /m3 (4 pCi / L) set by the U.S. EPA.
According to ASHRAE Handbook 1997, Fundamentals, for homeowners, a short-term charcoal
canister is used as a screening technique to determine whether the long-term alpha track method is
necessary. If long-term measurement (3 months to 1 year) finds that the average indoor radon concentration
exceeding the acting level of 147 Bq /m3 (4pCi / L), actions to reduce exposure to indoor
radon concentrations must be taken. Exposures to indoor radon may be reduced by:
 Preventing radon entry into the building
 Removing or diluting radon and the radon progeny
An active subslab depressurization system is the most widely used effective method. This system
uses a fan to draw soil gas containing radon and its progeny from pipework and exhausts the soil
gas outdoors. Because of the depressurization of the soil gas containing radon beneath the floor
slab, the indoor radon concentration may be significantly reduced.
Tobacco smoke is the most common indoor gaseous contaminant. There are two types of smoke:
mainstream and sidestream. Mainstream smoke goes directly to the respiratory tract of the smoker.
Sidestream smoke is the part exhaled by the smoker and smoke produced between puffs. Both
mainstream and second-hand smoke may cause cancer. In many public spaces in the United States
smoking is prohibited, and in many buildings, smoking is limited to special areas. Outdoor air requirements
for ventilation in a smoking lounge are increased to 60 cfm / person (30 L / s person) in
ASHRAE Standard 62-1999.
Because gaseous indoor contaminants have very small molecules, gaseous adsorbers and
chemisorbers are often used
24.8 CHAPTER TWENTY-FOUR
 To eliminate or to reduce indoor gaseous contaminants such as combustion products, VOCs, and
objectable odors
 According to the ASHRAE Standard 62-1999, to reduce the concentration of the known and
specifiable contaminants in recirculating air to the required concentration level based on indoor
air quality procedures to achieve acceptable air quality
Activated Carbon Adsorbers
Brown et al. (1994) reported that mean concentrations of VOCs in various office buildings
ranged from 0.18 to 4.15 mg /m3 with a new building having a total VOC concentration as high as
39.3 mg /m3. For each individual compound, its mean concentration was about 0.05 mg /m3. Granular
activated carbon (GAC) is one of the most widely used adsorbents for indoor gaseous contaminants.
Adsorption (physical adsorption) is a surface phenomenona. In an adsorption process, the molecule
of the gaseous contaminant
 Must be transported from the carrier airstream across the boundary layer that surrounds the adsorber
granule
 Must diffuse into the pore of the adsorbent and occupy the surface
 Must be bound to the surface
Adsorption is different from absorption in which the gaseous contaminant is dissolved in or reacts
with the absorbing medium, which can be either a porous solid or a liquid.
GAC Performance
As discussed in Sec. 15.16, removal efficiency, adsorption capacity, and service life are important
performance parameters for GAC. Toluene has been used as a surrogate for indoor VOCs for the
testing of GAC performance. An increase in relative humidity generally reduces the adsorption of
VOC on GAC. This may affect some compounds more than others, and it becomes more significant
at low VOC concentrations. When the concentration of the contaminant downstream of the adsorber
rises until its concentration is the same as that of the upstream, the penetration at this time is called
breakthrough penetration.
In addition to the adsorption capacity and the removing efficiency discussed in Sec. 15.16, the
performance parameter breakthrough time tb is defined as the elapsed time between the start of
challenge and the time when the penetrating concentration reaches the specified breakthrough fraction,
10 percent breakthrough time t10% or 50 percent breakthrough time t50%.
Another performance parameter is called capacity at tb which is expressed as the ratio of the
mass of contaminant collected at a particular breakthrough time to the mass of the GAC bed, in
percentage.
Van Osdell and Sparks (1995) reported the performance of GAC challenged with toluene at various
concentrations at 77°F (25°C) and 50 percent relative humidity as follows:
Challenge concentration ppm 0.44 1.1 9.2 71.7
Challenge concentration, mg/m3 1.7 4.0 34 270
Carbon bed mass, g 24.1 23.1 23.4 25.8
t10%, h 625 344 72 11.9
t50%, h 750 422 88 15.0
Capacity at t10%, percent 6.6 9.2 16.2 18.8
Capacity at t50%, percent 7.9 11.3 19.7 23.8
IMPROVING INDOOR AIR QUALITY 24.9
Several test reports have estimated that the lifetime of GAC for indoor VOC control exceeds 1000 h
depending on its concentration. The retentivity is the maximum amount of contaminant that the
GAC can retain in clean air. The retentivity of GAC for toluene at 2860 mg /m3 is 17 percent.
Simulating GAC Applications
Van Osdell and Sparks (1995) did a study based on an IAQ simulated model utilizing a computer
program RISKBETA, which is a thoroughly mixed room model incorporating source / sink behavior
that can predict concentrations as a function of time. The simulated model is a copy/ storage room
of 9.6 ft 13.1 ft (3 m 4 m) in a building that has eight rooms. The air system serving this
building has a supply volume flow rate of 1177 cfm (2000 m3 / h) of 5 ach. It is assumed that 4.4 lb
(2 kg) of toluene is spilled in the copy/ storage room. One-half of the spilled solvent is cleaned up
as a liquid; the other half vaporizes and is carried throughout the building by the air system. There
is no other contaminative source in this building, and the outdoor air is assumed to be VOC-free. A
GAC adsorber is installed in the air system to remove VOCs from the recirculating air.
Because of the buildup of the concentration of VOC contaminants in the indoor air, they are adsorbed
into the sinks (paints, carpets, upholstery) and loaded into high levels. As the concentrations
of toluene of indoor air drops and decays to below the 0.1 mg /m3 level, the sinks become contaminative
sources and reemit the toluene at low but significant rates.
Figure 24.1 shows the predicted concentrations within a copy room and the average concentration
of the whole building with or without the GAC adsorbers based on the data and analysis of Van Osdell
and Sparks (1995). The time between the beginning of the decay of the concentration of toluene in the
copy room down to the 0.1 mg /m3 level without a GAC adsorber is less than 10 days. The difference
between the concentration decay curves with or without a GAC adsorber is comparatively small for
the first two days and then begins to diverge because the GAC adsorber removes more contaminants
than the sinks can reemit. Similar results can be found for the rest of the building.
As the concentration of contaminant continues to drop, the GAC will reach its retentivity limit
for the incoming contaminant concentration at around 0.2 mg /m3 and will cease to collect. In
24.10 CHAPTER TWENTY-FOUR
0
VOC concentration, mg/m3 0.01
0.1
40
Copier’s room
without adsorber
Copier’s room
with adsorber
80 120
Time, h
160 200
1
10
100
1000
Rest of
building
average
with adsorber
FIGURE 24.1 VOC concentration versus time decay curve for a simulated copy room.
Fig. 24.1, the sharp drop of the concentration decay curve will begin to parallel the curve that is
without GAC adsorbers.
Because the building average concentration remains above 5 mg /m3 for about 15 h, using GAC
adsorber alone does not mitigate the spill. Better cleaning, short-term increased ventilation, a separate
exhaust system for the copy room, or other similar measures should be taken to reach an
acceptable level more quickly and to reduce the loading of the sources / sinks in the building.
Chemisorption
Chemisorption is a type of adsoption that occurs when the molecules of a volatile contaminant are
first physically adsorbed onto the adsorbent. As soon as they are adsorbed, they react chemically
with the chemical impregnant added to the surface of the adsorbent. The chemical impregnant
makes the adsorbent more or less specific for a contaminant or a group of contaminants.
Chemisorption differs from the physical adsorption as follows:
 Chemisorption is generally irreversible whereas adsorption is reversible.
 Chemisorption improves when temperature increases; adsorption improves when temperature
decreases.
 Chemisorption does not produce heat, but requires heat.
 Water vapor often helps chemisorption but usually hinders adsorption.
Most chemisorptive media are formed from a high porous support. Potassium permanganate-impregnated
alumina (PIA) is the most widely used chemisorptive medium to remove gaseous contaminants.
GAC does effectively remove most hydrocarbons, many aldehydes and organic acids, and
nitrogen dioxides due to its high ratio of surface area to volume. PIA is effective against sulfur
oxides, formaldehyde, nitric oxide, hydrogen sulfide, and lower-molecular-weight aldehydes and
organic acids. PIA is often used in conjunction with GAC to cover a very broad spectrum of
gaseous contaminant removal.
Chemisorber Performance
A chemisorber is an adsorber with chemically impregnated media. Most chemisorbers consist of
chemisorptive media from a high porous support such as activated alumina or activated carbon
coated or impregnated with a chemical reactant.
Three types of chemisorber cartridges are currently available: V-bank of large-mesh carbon
trays, pleated dry composite media with fine-mesh carbon, and pleated nonwoven carbon-coated
fabric. Removal efficiency, service life, and pressure drop are essential performance parameters.
Removal Efficiency. (See Section 15.16.) Removal efficiency depends on chemical impregnates,
granular size, type of cartridge, and challenge concentration. Joffe (1996) reported that the removal
efficiency of NO2 measured at the beginning of a 2000 cfm (940 L / s) and 160-ppb challenge for a
pleated dry composite medium with fine-mesh carbon chemisorber dropped from about 85 percent
at the beginning to about 60 percent after 10 h of operation. For V-bank large-mesh carbon trays,
the average removal efficiency is only about 30 percent in a period of 10 h of operation.
High removal efficiency is important for toxic and corrosive gases and becomes critical in odor
control. According to ASHRAE Handbook 1997, Fundamentals, the human olfactory response S is
nonlinear. The relationship between perceived intensities and the contaminant concentration C can
be expressed as
S  k Cn (24.2)
where k  a multiplying factor
n  exponential factor, varies from 0.2 to about 0.7
IMPROVING INDOOR AIR QUALITY 24.11
This relationship indicates that the concentration of some odorants must reduce substantially to
make a perceived odor level change.
Upon exposure to contaminated air, the active surface is covered by absorbed pollutant
molecules, or the product of a chemical reaction. This results in a decrease of removal efficiency
over time.
Pressure Drop. There is always a compromise between higher performance and lower pressure
drop across a chemisorber. At a face velocity of 500 fpm (2.5 m / s), the pressure drop is typically
0.4 in. WC (100 Pa).
24.6 MAINTENANCE TO GUARANTEE IAQ
Inspection, Service, and Access
Inspection, service, access, and maintenance to prevent system failure and component deterioration
are essential factors to guarantee that air filters, coils, ducts, dampers, and other system components
are properly and efficiently operated. Keep all the surfaces clean that surround the airflow passage
and have contact with the outdoor air, recirculating air, and supply air, and without moisture except
cooling coils, condensate pans, and evaporative coolers.
Access is the critical factor for inspection, service, and maintenance for an HVAC&R system.
Many IAQ problems are the result of clogged coils or filters, blocked condensate pans, or carryover
of the condensate droplets. If there is no convenient access to inspect the coil bank and filters, unless
the AHU is partially dismantled, a small problem may become a disaster.
Hinged access doors to the filter section, fan sections, both sides of coil banks, condensate pans,
and mixed plenum are essential. Standing water, mold and microbial growth, accumulated dust, and
improperly operated sensors and instruments must be conveniently accessed, inspected directly and
regularly, and promptly serviced.
The more difficult the access, the later the problem will be discovered, and less chance that the
system component causing poor IAQ will be promptly found and improved.
Monitoring of Operating Conditions in Air Systems
The operating conditions that should be monitored and inspected include the following:
 Position of outdoor air dampers and minimum ventilation air controls
 Pressure drop across each filter
 Cleanliness of condensate drain pans in AHUs, PUs, and fancoils
 Cleanliness of coils
 Microbial growth in standing-water systems such as open-water system in condensate pans, evaporative
coolers, and cooling towers
 Minimum setting in VAV boxes
 Optimum operation of exhaust or relief air system
Coils and Ductwork
Dust accumulated in coils and on the surface of the duct inner liner is best prevented by installing a
medium-efficiency air filter (MERV7 with 50 ? E3 < 70 percent or higher) and without any air leakage
bypassing the air filter. Ottney (1993) noted many hospitals installed a prefilter and a highefficiency
final filter on the air leaving side of the AHU. The coils and ductwork between the prefilter
24.12 CHAPTER TWENTY-FOUR
and final filter in air systems with leak-free system components remained relatively free of visible
particulate accumulation.
For a wet coil with condensate, a face velocity 	 550 fpm (2.75 m / s) should be selected for
corrugated fins to prevent condensate carryover. If a fiberglass duct liner is used, a wet coil face velocity
	 500 fpm (2.5 m / s) should be selected to prevent microbial growth.
As discussed in Sec. 15.11, the condensate pan and condensate trap should be properly designed
and installed. Condensate in the pan should be drained completely. Dirt, scale, and debris must be
cleaned regularly.
For heating coils, if the coil leaks, the chemicals used in the hot water may be aerosolized and
enter the air directly, causing IAQ problems.
24.7 SPACE PRESSURIZATION CONTROL
As discussed in Sec. 4.13, the purpose of space pressurization control is to maintain a positive pressure
in an occupied space, a noncontaminated space, or a clean space to prevent untreated air, contaminated
air, air with objectable odors, or toxic air entering or infiltrating this occupied space, and
to maintain a negative pressure in contaminated or toxic area so that contaminated or toxic air
(treated if required) exhausts outdoors directly. Air always flows from a place with a higher pressure
to a place with a lower pressure. Space pressurization controls have been successfully applied
to many laboratories, clean spaces and rooms, and industrial applications.
24.8 LEGAL RESPONSIBILITY FOR IAQ CASES
SBS or IAQ Cases
According to Eisenstein (1992), the 1980s saw litigation involving significant claims based on
indoor air pollution, sick building syndrome (SBS), or IAQ. In the mid-1980s, there were homeowners
and occupants of commercial buildings filing lawsuits claiming injuries from formaldehyde
exposure. Cases filed by owners of mobile homes where large quantities of products containing
formaldehyde are used have drawn much attention. There were also cases filed claiming personal
injuries resulting from radon contamination and pesticide pollutants.
Who Is Legally Responsible
The most frequent cases of SBS or IAQ occurred in commercial office buildings with sealed
windows in which occupants rely on the HVAC&R system for their thermal comfort control and
ventilation requirements. A typical case is often caused by a combination of factors, including
different toxic indoor contaminants from building materials, inadequate ventilation air supply
due to defective ventilation control, and improper operation and maintenance of sophisticated air
systems.
Eisenstein (1992) recognized that SBS cases have the potential of spreading a web of liability
to a large number of professionals and industries. The prosecuting party, or plaintiff, is often an
individual or group of individuals claiming personal injuries as a result of exposure to toxic
substances in office suites. Another type of the SBS plaintiff is the commercial tenant who alleges
that he or she cannot conduct business in the leased space because of the poor IAQ, that employees
of the corporate tenant are frequently ill due to the poor IAQ, or that inadequate ventilation
air causes business interruption and lost profit. Many of the SBS cases are multimillion-dollar
lawsuits.
IMPROVING INDOOR AIR QUALITY 24.13
The list of defendants in SBS or IAQ cases varies with the type of buildings and may include
some or all of the following:
 Building owner/ managers
 Real estate developers
 Architects
 HVAC&R consultants, mechanical engineers, control engineers
 IAQ consultants
 General contractors, HVAC&R contractors
 Building and HVAC&R products manufacturers
 Leasing agents
Among them, the target defendants in SBS or IAQ cases are building owners who have the responsibility
to provide safe premises, the building managers who must ensure that the building and HVAC&R
system are maintained properly and the HVAC&R system is operated correctly, and all the entities that
participate in the design, construction, and installation of the ventilation /HVAC&R system.
The theories of liabilities used against defendants depend to some degree on the law in various
states. The following causes of actions have been successful in most states: contracts or breach of
lease; professional malpractice or negligence; strict liability; fraud, mispresentation, and punitive
damages.
HVAC&R Engineer
If the poor IAQ is determined to be due to defects in the ventilation /HVAC&R system, the mechanical
or HVAC&R engineer who designs the ventilation / air system will be vulnerable to claims of
negligence.
As discussed previously in this chapter, regarding HVAC&R, the primary tasks that must be
done to ensure good IAQ are as follows:
 Provide adequate ventilation to dilute the indoor air contaminants.
 Prevent microbial growth.
 Use medium- and high-efficiency and HEPA filters to remove particulates, bacteria, and viruses.
 Remove specific gaseous contaminants by adsorbers and chemisorbers.
 Adopt pressurization control so that only filtered, noncontaminated air of acceptable air quality is
supplied to the occupied zone, and contaminated air is exhausted directly from the contaminated
area.
 Implement scheduled and proper HVAC&R maintenance for effective air system operation.
It is extremely important that the engineers’ design follow the IAQ-related federal and local
codes and the ASHRAE IAQ-related standards, especially ASHRAE Standard 62-1999, Ventilation
for Acceptable Indoor Air Quality. If an engineer fails to design an air system in comformity with
appropriate ASHRAE standards, he or she will be unable to defeat a claim of negligence.
Although an engineer cannot be expected to foresee all the uses of the conditioned spaces during
the life of the building, the changes in the occupancy rates and the alteration in the use of the space
should usually be anticipated by the engineer, who must design a ventilation system to accommodate
an adequate ventilation rate for high contamination load during remodeling of suites.
Recently, the courts have ruled that a ventilation system in a building is a product. Under strict
liability, the designer, manufacturer, and installer of a defective product are considered liable
regardless of whether their conduct was negligent. This theory emphasizes the product, not any
individual, so that it is often easier for a plaintiff to succeed for a strict liability.
24.14 CHAPTER TWENTY-FOUR
From the HVAC&R side, proper and careful consideration of the design, construction, and operation
of an air system and the issues that may cause poor IAQ will greatly reduce the risk of SBSrelated
lawsuits. If an IAQ-related lawsuit still arises, the chance of a successful defense will be significantly
increased.
REFERENCES
ANSI/ASHRAE Standard 52.2-1999, Method of Testing General Ventilation Air-Cleaning Devices for Removal
Efficiency by Particle Size, ASHRAE Inc., Atlanta, GA, 1999.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, ASHRAE Inc., Atlanta, GA, 1997.
ASHRAE, ASHRAE Handbook 1999, Applications, ASHRAE Inc., Atlanta, GA, 1999.
ASHRAE Standard 62-1999, Ventilation for Acceptable Indoor Air Quality, ASHRAE Inc., Atlanta, GA, 1999.
Brown, S., et al., Concentrations of Volatile Organic Compounds in Indoor Air—A Review, Indoor Air,
Copenhagen, Denmark: Munksgaard International Publishers Ltd., vol. 4 no. 2, 1994, pp. 123–134.
Burge, H. A., The Fungi: How They Grow and Their Effects on Human Health, HPAC, no. 7, 1997 pp. 69–74.
Burroughs, H. E. B., Filtration: An Investment in IAQ, HPAC, no. 8, 1997, pp. 55–65.
Burroughs, H. E. B., The Art and Science of Air Filtration Management in Health Care, HPAC, no. 10, 1998,
pp. 79–86.
Collett, C. W., Ross J. A., and Sterling, E. M., Quality Assurance Strategies for Investigating IAQ Problems,
ASHRAE Journal, no. 6, 1994, pp. 42–50.
Eisenstein, H., IAQ: Who Is Legally Responsible? Heating / Piping / Air Conditioning, no. 8, 1992, pp. 43–47.
Gill, K. E., IAQ and Air Handling Unit Design, HPAC, no. 1, 1996, pp. 49–54.
Godish, T., Sick Building: Definition, Diagnosis, and Mitigation, Lewis Publishers, Boca Raton, FL, 1995.
Hays, S. M., and Ganick, N., How to Attack IAQ Problems, Heating / Piping / Air Conditioning, no. 4, 1992,
pp. 43–51.
Industrial News, New Approach to IAQ Urged, ASHRAE Journal, no. 12, 1997, pp. 8–10.
Joffe, M. A., Chemical Filtration of Indor Air: An Application Primer, ASHRAE Journal, no. 2, 1996,
pp. 42–49.
Kowalski,W. J., and Bahnfleth,W., Airborne Respiratory Diseases and Mechanical Systems for Control of
Microbes, HPAC, no. 7, 1998, pp. 34–48.
Liu, R. T., and Huza, M. A., Filtraton and Indoor Air Quality: A Practical Approach, ASHRAE Journal, no. 2,
1995, pp. 18–23.
Mendell, M. J., Non-specific Symptoms in Office Workers: A Review and Summary of the Epidemiologic Literature,
Indoor Air, vol. 3, 1993, pp. 227–236
Muller, C. O., and England,W. G., Achieving Your Indoor Air Quality Goals: Which Filtration System Works
Best? ASHRAE Journal, no. 2, 1995, pp. 24–32.
Ottney, T. C., Particle Management for HVAC Systems, ASHRAE Journal, no. 7, 1993, pp. 27–34.
Persily, A., Ventilation Rates in Office Buildings, ASHRAE Journal, no. 7, 1989, pp. 52–54.
Spicer, R. C., Microbial Growth Must Be Minimized, Maintenance Is a Must, Engineered Systems, no. 9, 1998,
p. 34.
Sundell, J.,What We Know, and Don’t Know about Sick Building Syndrome, ASHRAE Journal, no. 6, 1996,
pp. 51–57.
Turner,W. A., Controlling Ventilation during Renovation, HPAC, no. 11, 1998, pp. 49–52.
Van Osdell, D. W. and Sparks, L. E., Carbon Adsorption for Indoor Air Cleaning, ASHRAE Journal, no. 2, 1995,
pp. 34–40.
IMPROVING INDOOR AIR QUALITY 24.15
CHAPTER 25
ENERGY MANAGEMENT
AND GLOBAL WARMING
25.1
25.1 ENERGY EFFICIENCY AND UNIT
ENERGY RATE REDUCTION 25.1
Energy Use and Energy Efficiency
for HVAC&R Systems 25.1
Energy Efficiency during Design,
Construction, Commissioning,
and Operation 25.2
Reduction of Unit Energy Rates 25.2
25.2 GLOBAL WARMING 25.3
Kyoto Protocol 25.3
Total Equivalent Warming
Impact 25.3
Mitigating Measures 25.4
25.3 ENERGY EFFICIENCY 25.5
Federal Mandates 25.5
Energy Use Intensities 25.5
Energy Audits 25.6
Procedure for Energy Retrofit 25.6
Performance Contracting 25.7
Green Buildings 25.8
Energy Star 25.10
25.4 ENERGY CONSERVATION
MEASURES 25.10
25.5 CASE STUDY: ENERGY
CONSERVATION MEASURES
FOR AN OFFICE 25.12
25.6 RELATIONSHIP BETWEEN
HVAC&R SYSTEM CHARACTERISTICS
AND ENERGY USE 25.12
System Characteristics and Energy
Use Intensities 25.12
Energy Use of Heating-Cooling
Equipment and Fan 25.13
25.7 ELECTRICITY DEREGULATION 25.14
Electric Utilities prior to
Deregulation 25.14
Electric Deregulation 25.14
California Approach 25.15
Real-Time Pricing 25.15
Case Study: Automated Control
of RTP 25.16
25.8 SYSTEM SIMULATION 25.17
Energy Estimation and Energy
Simulation 25.17
Performance Equations 25.17
Physical Modeling 25.18
Steady-State and Dynamic
Simulation 25.18
Sequential and Simultaneous
Simulation 25.19
25.9 ENERGY SIMULATION OF
A CENTRIFUGAL CHILLER USING
PHYSICAL MODELING 25.19
System Model 25.19
Operating Parameters Affecting Chiller
Energy Performance 25.20
Simulation Methodology 25.20
Evaporator Model 25.20
Condenser Model 25.22
Cooling Tower Model 25.23
Centrifugal Compressor Model 25.23
25.10 ENERGY SIMULATION
SOFTWARE DOE-2.1E 25.25
Energy Simulation Software 25.25
Loads 25.26
Systems 25.26
Energy Efficiency Measures 25.27
Plant 25.27
25.11 ASHRAE/IESNA STANDARD
90.1-1999 ENERGY COST BUDGET
METHOD 25.28
REFERENCES 25.28
25.1 ENERGY EFFICIENCY AND UNIT
ENERGY RATE REDUCTION
Energy Use and Energy Efficiency for HVAC&R Systems
Energy use or energy consumption indicates the amount of energy used or consumed. Energy effi-
ciency indicates how efficiently energy is used. An energy-efficient HVAC&R system maintains a
comfortable indoor environment with an acceptable indoor air quality (IAQ) and consumes optimum
energy resources. According to the ASHRAE definition, energy conservation indicates that energy is
used efficiently. Energy management is the effort and measures taken to ensure that energy is used
efficiently, and the unit energy rates (electric and gas rates) are reasonably low for the sake of reduction
of energy cost. Energy management of HVAC&R systems consists of two areas: energy effi-
ciency (the reduction of energy use or energy conservation) and the reduction of the unit energy rate.
As discussed in Sec. 1.7, the estimate of annual U.S. energy use of the HVAC&R systems in
1992 was approximately one sixth of the total national energy use. In addition, energy use for
HVAC&R is closely related to the release of CO2 to the outdoor atmosphere which causes the
global warming effect. Energy efficiency is a challenge to every one of us in the HVAC&R industry
now and for many years to come in the future.
According to EIA Commercial Building Characteristics 1992, energy sources used for the heating
of the 67.8 billion ft2 of commercial buildings in the United States in 1992 were electricity, 22.8
percent; natural gas, 51.8 percent; fuel oil, 6.6 percent; and district heating, 7.3 percent. Energy
sources used for cooling U.S. commercial buildings in 1992 were electricity, 80.4 percent; natural
gas, 2.8 percent; and district cooling, 3.0 percent. Electricity and natural gas are the two primary
energy sources for HVAC&R systems.
Energy Efficiency during Design, Construction,
Commissioning, and Operation
Energy efficiency must be achieved in every stage of HVAC&R system construction and operation:
design, construction, commissioning, and daily operation. A well-designed and effectively functioning
energy management direct digital control (DDC) system is necessary for an energy-efficient
HVAC&R system.
 Design. Various alternatives should be compared and analyzed in terms of either payback years or
life-cycle cost to determine which is most energy-efficient and cost-effective.
 Construction and installation. Roof and external walls should be well insulated. The efficiency of
installed equipment should not be less than the minimum efficiencies specified by
ASHRAE/IESNA Standard 90.1-1999. The amount of air leakages from the ducts, and the duct
and pipe insulation all affect the energy use.
 Commissioning. The capacity of equipment, the air and water balance, and the coordination between
various components and control systems should be carefully measured, adjusted, and commissioned.
A poorly commissioned HVAC&R system will never function efficiently as specified.
 Operation. The energy use for chillers, compressors, fans, pumps, boilers, and furnaces should be
monitored, periodically checked, investigated, reduced, and improved.
For HVAC&R, to reduce the emissions of CO2 to the atmosphere by means of increasing energy ef-
ficiency in operation is the primary action to mitigate the global warming effect.
Reduction of Unit Energy Rates
Facility owners, facility managers, or the tenants as well as the designers and operators of the
HVAC&R systems all are concerned about the unit energy rate Er, as it affects the energy cost and
the operating cost even if there is no saving in HVAC&R system energy use. For electricity, energy
cost Ce is the product of the price of unit energy rate Er and energy use, in kWh or therms. The
lower Er, the smaller Ce. The reduction of unit energy rate Er is closely related to the following:
 The ratio of unit rate of kWh to therms which affects the designer in selecting electric cooling or
gas cooling
 Use of thermal storage to provide off-peak conditioning to reduce Er
25.2 CHAPTER TWENTY-FIVE
 Choice of most favorable and optimum Er rate from the electricity deregulation or gas deregulation
Energy cost is the dominant factor that affects the operating cost of an HVAC&R system.
25.2 GLOBAL WARMING
Kyoto Protocol
In December 1997, more than 150 countries met in Kyoto, Japan, and agreed to call for 38 developed
nations to reduce the emissions of the greenhouse gases (GHGs) to about 5.2 percent below
1990 levels by 2008 to 2012, in order to mitigate the potential risk of global warming. The agreement
is called the Kyoto Protocol and includes the following details:
 GHG emissions are to be reduced 8, 7, 6, and 6 percent below the 1990 level for the European
Union, United States, Japan, and Canada, respectively.
 Six greenhouse gases are covered by the Kyoto Protocol: carbon dioxide (CO2), methane (CH4),
nitrous oxide (N2O), perfluorocarbons (PFCs), sulfur hexafluoride (SF6), and hydrofluorocarbons
(HFCs). And CO2, CH4 , and N2O will use the 1990 emission as the base value, whereas PFCs,
SF6, and HFCs can be compared with either the 1990 or 1995 emission as the base value.
 Currently, reduced emission controls are only applicable to developed nations. However, future
negotiations will address the reduction of emissions and other issues for developing nations.
 Kyoto Protocol, using a basket approach, permits the limitation of emissions in any combination
as long as its CO2 -equivalent target is met. The Kyoto Protocol also permits counting of benefits
derived for “sinks” (forest, soil, and land used) for emissions trading. Details remain to be
worked out.
 The Kyoto Protocol will be opened for signature from March 16, 1998, for one year and will enter
into force after it has been ratified by at least 55 nations, representing 55 percent of the total
emissions from the developed nations.
U.S. Ambassador Stuart Eizenstat, who led the U.S. delegation in Kyoto, said in the 9th Annual
Energy Efficiency Forum, 1998, in Washington, “Kyoto is an insurance policy against the potentially
devastating impacts of global warming. If we act now, the premium on this policy will be far
more reasonable and less costly than if we delay and hope the problem goes away. The evidence
shows that it will not.”
In the 1998 ASHRAE winter meeting, an international panel of global climate experts concluded
that the threat of global warming can be reduced if the world acts now to implement energyconserving
technologies in buildings and to reduce energy use. Experts also agreed that “energy ef-
ficiency” is technologically possible and economically viable.
Total Equivalent Warming Impact
The assessments of the global warming effect should be based on total equivalent warming impact
(TEWI) concepts. In Baxter et al. (1998), total equivalent warming impact was defined as the combined
global warming effects corresponding with the CO2 released due to the indirect effect of energy
use over the lifetime of a system and the direct effect resulting from lifetime refrigerant emissions.
TEWI can be calculated as
TEWI  mrt(GWP)rt  mblow(GWP)blow  EannualL (25.1)
where mrt, mblow  mass of refrigerant and mass of blow agent released during its useful
lifetime, lb (kg)
ENERGY MANAGEMENT AND GLOBAL WARMING 25.3
(GWP)rt, (GWP)blow  equivalent global warming effect of CO2 equivalent emisssion per lb (kg) of
refrigerant or blow agent released, lb (kg)
 conversion factor to convert energy use to global warming effect
Eannual  annual energy use
L  equipment lifetime, years
It is difficult to calculate an absolute value of TEWI because of the uncertainties of annual refrigerant
leakage and GWP values. According to Baxter et al. (1998), for a typical supermarket in the
United States with low- and medium-temperature DX refrigeration systems, estimated emissions
due to refrigerant leakage are between 10 and 15 percent of the charge annually. If HFC-404A is
used as the refrigerant, TEWI is about 8.6 million lb CO2 equivalent. Of this, 57 percent is refrigerant
emissions. For centrifugal chillers, the refrigerant leakage is between 0.5 and 4 percent of the
charge annually. For a typical 1000-ton (3500-kW) centrifugal chiller in Atlanta, Georgia, where
the refrigerant leakage is 2 percent of the charge annually, if HCFC-123 is used as refrigerant, the
TEWI is about 5.1 million lb of CO2 equivalent; of this, about 1 percent is due to refrigerant emissions.
If HFC-134a is used as refrigerant, TEWI is about 5.9 million lb, and about 2.5 percent is due
to refrigerant emissions.
Mitigating Measures
CO2 Release due to Energy Use. David Gardner, asssistant administrator for the Environmental
Protection Agency, who attended the International Climate Change Conference in June 1997, in
Baltimore, reported that “85 percent of climate change problems result from combustion
processes.” For a combustion process using solid carbon C as the fuel, its reaction process and the
released heat of combustion can be evaluated as follows:
Reaction Heat of combustion
H (Qout), Btu/ lb (kJ / kg)
C  O2  CO2 14,100 Btu / lb (32,800 kJ / kg) (25.2)
1 lb  2.67 lb  3.67 lb
As expressed in Eq. (8.4), the combustion efficiency Ec, in percent, can be calculated as
where Qout, Qfuel  heat output and heat content rate fuel consumed, Btu /h (W). For any combustion
process and the successive energy transformation processes from heat to mechanical and then
to electrical, and so on, in Eq. (8.4), if you require a specified Qout for equipment in an HVAC&R
system, for the sake of reducing the CO2 in Eq. (25.2), if the fuel C must be reduced as well as the
Qfuel value in Eq. (8.4), there must be a corresponding higher combustion efficiency Ec. Not only
higher Ec but all higher energy efficiencies reduce Qfuel as well as the CO2 and other greenhouse
gases in the combustion process.
Refrigerant Emissions. HFCs are listed as one of the six greenhouse gases. Cox and Miro
(1998) predict, by 2050, HFCs may account for 2 to 3 percent of GHGs. The measures to reduce
the global warming effect because of the refrigerant emissions due to the leakage and purge are as
follows:
 Refrigerant leakage in supermarket medium- and low-temperature systems should equal or not
exceed the American Refrigeration Institute (ARI) target value of 6 percent of the charge annually.
For chillers, a refrigerant leakage of 1 to 3 percent of the annual charge is appropriate.
Ec 
100Qout
Qfuel
25.4 CHAPTER TWENTY-FIVE
 Secondary-loop systems using brine as the cooling medium and distributed systems in which
compressors are close to the display case have far less refrigerant leakage than DX systems.
 The use of greenhouse gases as blow agents should be minimized.
The selection of environmentally-friendly refrigerant with low ozone depletion and global
warming effect should be considered. From Table 9.1, compare the following refrigerants:
According to Baxter et al. (1998), the energy use of centrifugal chiller using HCFC-123a as refrigerant
is only about 87 percent of that using HFC-134a as refrigerant. More research is required to
determine a suitable alternative refrigerant for chillers when ODP, HGWP, and energy use are all
considered at the same time.
25.3 ENERGY EFFICIENCY
Federal Mandates
According to Cox and Miro (1998), the U.S. government is the single largest energy user. In late
1990s, the federal government provided energy to approximately 500,000 buildings with a floor
area of about 3.1 billion ft2 (288 million m2). Energy was needed in HVAC&R, lighting, and other
building services. Approximately 77 percent of the floor area was used for housing, office, storage,
and other purposes; the remaining 23 percent belonged to hospital, school, prison, and other uses.
Many federal buildings constructed before the energy crisis of 1973 are not energy efficient. The
president’s Executive Order 12759 in 1991 and the National Energy Policy Act of 1992 (EPAct) required
federal buildings to reduce enery use by 20 percent from 1985 levels in the year 2000. In addition,
Executive Order 12902 called for an energy efficiency improvement of 30 percent over the
1985 level by the year 2005.
The Federal Energy Management Program (FEMP) was established in 1974 to provide guidance
and assistance to improve energy efficiency in implementing energy management plans. It
has focused on the energy bill and potential for improvements. The Energy Policy Act established
the fiscal year (FY) 1995 goal of improving energy efficiencies in federal buildings by 10
percent from 1985 levels on a Btu / gross ft2 (W/m2) of floor area. This goal is exceeded in FY
1994 with a total reduction in federal buildings of 11.2 percent. In addition, energy costs were
$3.8 billion in FY 1994, and that was $1.5 billion less than in FY 1985. The federal government
accumulated an energy savings of $9.8 billion and reduced federal building petroleum-based fuel
use by 45.4 percent.
Energy Use Intensities
Energy use intensity (EUI) is the annual energy use per unit floor area, in MBtu/ ft2	 yr
(kWh/m2	 yr). The EUI depends mainly on the locations of buildings, building characteristics, operating
characteristics, and the HVAC&R system characteristics. According to DOE/ EIA (1992),
the abridged EUIs for commercial buildings in 1989 were as follows:
Ozone depletion potential Halocarbon global warming
(ODP) potential (HGWP)
HFC-134a 0 0.28
HFC-404A 0 0.95
HFC-407A 0 0.47
HCFC-123 0.016 0.02
ENERGY MANAGEMENT AND GLOBAL WARMING 25.5
Energy Audits
Energy audit refers to a month-by-month accounting, survey, and analysis of energy use in a building.
This energy use is checked against a budget or an energy estimate based on an hour-by-hour
detailed system simulation in order to identify energy efficiency opportunities. According to
ASHRAE Handbook 1999, HVAC Applications, following an energy use estimate, an energy analysis
can proceed on three levels:
 Walk-through assessment analyzes energy bills and a brief survey of the building for the sake of
identifying low-cost / or no-cost energy efficiency measures.
 Energy survey and analysis include a more detailed building survey and energy analysis with a
breakdown of energy use within the building. Actual energy use can be compared with EUI of
corresponding building types and locations. This level identifies and provides the savings and
cost analysis of all practical energy efficiency measures that meet the owner’s economic
criteria.
 Detailed analysis of capital intensive modifications focuses on potential capital-intensive modifications
and involves more detailed field data collection, system simulation, and engineering analysis.
In the detailed analysis level, an energy-efficient and cost-effective energy use estimate can be
calculated according to the ASHRAE/IESNA Standard 90.1-1999 Energy Cost Budget (ECB)
method. Refer to Sec. 25.11 for more information. The actual month-by-month energy bill can
be compared against the ECB and energy cost breakdowns on the building envelope, load calculations,
and HVAC&R system and components. Energy efficiency opportunities can then be
determined.
Procedure for Energy Retrofit
Energy retrofit is a project whose purpose is energy efficiency, i.e., to convert an existing system to
an energy-efficient system. Energy retrofit can often proceed as follows:
1. Organize a team. One individual should be chosen to be responsible for the team.
2. Draft a plan.
3. Make a field survey by walking through the building; talking with the operating manager, engineers,
operators, and electric utility representative; take field measurements and tests, if
necessary.
4. Proceed with an energy audit.
5. Justify the energy efficiency opportunities after cost analysis.
6. Implement the plan, starting with no-cost and low-cost opportunities.
MBtu/ ft2	 yr Electric, kWh/ ft2	 yr Natural gas, ft3 /ft2	 yr
Building Type Electric Gas Total NE MW SO WE NE MW SO WE
Assembly 27 39 64 7.1 9.5 7.3 57.5 25.9 45.8
Education 27 47 87 7.3 7.4 8.7 7.9 28.2 61.0 33.6 66.4
Health care 75 113 219 20.1 22.6 23.4 20.1 74.0 156.4 82.0 69.1
Mercantile / Serv. 44 46 85 15.1 12.5 12.1 13.4 62.9 62.1 29.4 31.1
Office 66 32 104 16.3 16.7 20.9 22.4 20.8 55.8 23.6 26.8
NE indicates northeast, MW midwest, SO south, and WE west.
Source: DOE/EIA (1992). Based on 4528 buildings. Reprinted by permission.
25.6 CHAPTER TWENTY-FIVE
7. After energy retrofit is complete, operate and control the HVAC&R system according to the requirements
of the proposed energy efficiency opportunities. Check the energy costs against
those from prior to the energy retrofit.
Performance Contracting
According to Mahoney and Weiss (1997), a performance contract legally guarantees energy and
operating savings based on the performance of the contractor over a specified period of 5 to 25
years. Performance contracting also allows a facility to complete a major upgrade in energy efficiency
funded by the saving of energy costs. The contractor must pay for savings that fail to
materialize.
Energy Service Companies. During performance contracting, most often an energy service company
(ESCO) initiates an HVAC&R system energy efficiency upgrade proposal to a facility owner tailored
to the facility’s needs. A performance contract includes design, installation, financing, project
management, maintenance, and monitoring. Most of the components of a performance contract are
the responsibility of the ESCO, and therefore, an experienced, technically competent person who has
a good relationship with a financial party is the key factor for sucessful performance contracting.
There are three kinds of ESCOs:
 Engineering ESCOs are companies that always specialized to provide engineering services, including
energy audits and system design. More often, engineering ESCOs also include project
management, financing, and performance guarantees.
 Equipment ESCOs are manufacturers of energy users or controlling equipment. They establish
their own energy efficiency service divisions to expand their marketing.
 Many utility ESCOs are owned by utilities which provide not only energy efficiency services, but
also electric and natural gas power. There are also ESCOs aligned between utilities and consolidated
HVAC&R contractors.
General Steps. Mahoney and Weiss (1997) noted that a performance contract works best when
ESCOs and local contractor are well coordinated. The ESCO is reponsible for engineering, project
management, guaranteed savings, and long-term project funding. The local contractor takes care
of the installation, maintenance, and other services. All involved parties, such as facility owner,
ESCO, local contractor, and equipment manufacturers, work well when the following general
steps are taken:
 Conduct an energy use estimate. This involves assessing current conditions, collecting data to estimate
potential savings, and determining factors affecting energy costs. Field surveys; interviews;
review of building and HVAC&R system drawings, energy use of electric, gas, or oil for a number
of years; occupancy schedules; and required indoor environmental parameters are important for
an energy estimate.
 Perform a detailed engineering analysis. Computer modeling and simulation can be used to evaluate
the current energy use as the baseline. Compare various alternatives, calculate and determine
each possible energy-saving measure.
 Design the energy-saving measures for retrofit with reasonable payback time. Facility owner
should approve an implementation plan prepared by the ESCO.
 Arrange project financing. After the approval of the implementation plan, the ESCO finds suitable
third-party financing for the payment of installation and other costs. As the transaction costs are
roughly the same for small, medium, and large loans and are expensive, a large or combination
project often helps.
ENERGY MANAGEMENT AND GLOBAL WARMING 25.7
 Install approved equipment. The facility owner selects the equipment type and brand and decides
whether a local contractor is needed or to use in-house staff for a small or medium-sized project.
 Commission the system and train operators. The ESCO verifies the operation characteristics of
the new or modified system components and the HVAC&R system. The facility’s staff learn how
to operate new equipment, and the new or upgraded HVAC&R system.
 Monitor energy savings. The operation and control of the building’s indoor environment and the
energy efficiency-upgraded HVAC&R system remain with the facility’s staff. The ESCO also
monitors the building’s indoor environment and the performance of the upgraded HVAC&R system
remotely. The facility manager is notified if an abnormal condition occurs. On many occasions,
the ESCO’s monitoring service to ensure that energy savings are achieved should be included
in the performance contract.
 Guarantee energy savings. If energy saving does not meet the projections, the ESCO pays the difference.
The ESCO monitors and reviews energy use throughout the life of the contract. The contract
also allows adjustments of the facility’s baseline energy use when there is a change in the facility’s
size, occupancy, or equipment.
 Performance contracting provides the energy efficiency upgraded cost for the HVAC&R facility
owners. Because the National Energy Policy Act of 1992 and Executive Orders 12759 and 12902
as well as utility companies offer fewer rebates, more and more federal building managers and
facility owners in the private sector will rely on performance contracting as a viable way to upgrade
the energy efficiency of HVAC&R systems. According to Mahoney and Weiss (1997),
“Today, performance contracting projects in the United States have generated $2.3 billion in
work and equipment. . . . The market is growing at a rate of 15 percent and has reached $750
million a year.” There are more than 100 companies providing performance contracts. One of
them, an engineering ESCO, has entered into more than 1000 agreements related to energy effi-
ciency since 1974.
Case Study: Performance Contracting. According to the Heating, Piping, and Air Conditioning
journal (January 1998, pp. 12–25), as the legislation permitted public school corporations to advertise
for self-funding energy efficiency proposals from qualified performance contractors in 1995,
the Evansville Vanderburg County School Corporation (EVCSC) in Evansville, Indiana, accepted a
performance contract proposal from the Energy System Group (ESG), a utility ESCO, for energy
efficiency upgrading of 35 school buildings. The ESG performance contract with EVCSC totaled
more than $35 million.
The performance contract includes fully automatic summer /winter changeover two-pipe system
in 19 buildings, pulse combustion gas boilers and DDC controls in 19 schools, and new lighting
systems, windows, and doors. Based on the actual energy use records:
 The new boilers lowered the natural gas consumption 45 to 55 percent in the first year of use.
 Based on the original proposal, the energy saving program will save $369,000 per year. After 15
months of operation, it will actually save about half a million dollars annually.
Green Buildings
Basics. A green building implies a building, including HVAC&R systems, that is energy-efficient
and environmentally friendly. It provides a healthy and comfortable indoor environment for its occupants,
is friendly to the outdoor environment by releasing minimum pollutants into the outdoor
atmosphere, and at the same time is considered sustainable, energy-efficient, and well maintained
(refer to Sec. 1.8). Sustainable design considers resources consumed in a way that balances the current
needs with those of future generations.
According to Cole and Larsson (1998), Green Building Challenge ‘98 (GBC ‘98) is a two-year
international collaborative process of developing and demonstrating an improved method to assess
25.8 CHAPTER TWENTY-FIVE
the enviromental performance of buildings. This process, initiated by Canada, has representatives
from 13 other countries.
GBA Tool. The second-generation assessment method developed during the GBC ‘98 process is
called the green building assessment (GBA) tool. GBA tool includes the following characteristics:
 It covers the following:
Resource consumption: energy, land, water, and materials
Enviromental pollutants: airborne emissions, solid waste, liquid waste, and others
Quality of indoor environment: IAQ, thermal quality (such as indoor temperature and humidity),
lighting quality, noise, and controllability of systems
Longevity: sustainability, adaptability, and maintenance of performance
Process: design, construction, commissioning, and operations planning
Climate factors: locations, outdoor climate, sun, daylight, wind impacts
 It has a core set of detailed criteria and subcriteria that can be modified to suit national, regional,
and building type variations. Assessments are based on applicable regulations or industrial norms
in various regions. The structure can be used at various levels of detail.
 The tool reflects a consistency in terminology and a scoring system. All criteria and subcriteria
are assessed on a 2 to 5 scale in which 0 is typical practice, 3 is best current practice, and 5 is
a demanding, attainable goal.
 It comprises a weighting system that can be modified to emphasize regional priorities.
Case Studies of Green Buildings. The following is a description of the essentials of two green
buildings: Gottfried et al. (1997) introduced San Diego’s Ridgehaven Building (RB), San Diego, California,
and Morrison (1998) reported on the Energy Resource Center (ERC), in Downey, California:
Additional energy-efficiency and IAQ measures are as follows. For the Ridgehaven Building,
 Most material removed from the building had to be reused.
 Exterior foil, metallic film-coated foam, and exhaust systems are used to minimize glass fiber,
VOC, and particulates entering the conditioned space.
 VAV systems and terminals are provided in variable-occupancy areas.
 Variable-speed drive for outdoor air fan and condenser water pumps are used.
 High-efficiency motors are utilized.
For the ERC:
 Direct-fired 30-ton double-effect absorption chiller-heaters
 Two dual-wheel desiccant dehumidification units
RB ERC
Building type Office Office-laboratory
Year constructed 1981 1950s, retrofit 1990s
Floor area, ft2 73,000 45,000
HVAC&R system WSHP (water-source Packaged system
heat pump)
Energy efficiency, kWh/ ft2 	 yr 9 22 percent less than code
IAQ OA, (outdoor air) cfm / person 20
Filter efficiency, dust spot 40 percent
Daylighting Maximized
ENERGY MANAGEMENT AND GLOBAL WARMING 25.9
 AHU with evaporative coolers
 Variable-speed drive to save fan energy use
 Nontoxic products such as interior paints containing no petroleum derivatives, VOCs, and nontoxic
floor sealers, padding, tile adhesives, and recycled carpeting material
 Linings in the ducts and plenums to prevent the growth of mold and bacteria
 Use of CO2 -based demand-controlled ventilation
Energy Star
The U.S. Department of Energy (DOE) and the U.S. Environmental Protection Agency (EPA) are
collaborating in rating to promote an energy-efficiency program for the recognization of
 EPA’s Energy Star Benchmarking Tool assesses a facility’s comparative energy performance
against standardized indoor criteria and compares with similar-use buildings in United States according
to Hoggard (2000). Each building is assigned a benchmark score from 0 to 100. All buildings
scoring 75 or better qualify for an Energy Star label for buildings. Currently, the Energy Star
label is only available for office buildings, and will be available for schools, retail stores, and other
buildings later on.
 Equipment of the top 25 percent of products such as computers, printers, copiers, and HVAC&R
equipment concerning energy efficiency with the Energy Star label.
Buildings with the Energy Star label are energy-efficient buildings, whereas green buildings are
both energy-efficient and environmentally friendly.
25.4 ENERGY CONSERVATION MEASURES
In principle, adopting an integrated design approach by considering the building as a whole, minimizing
heating and cooling loads, selecting high-efficiency equipment sized as closely to the design
load as possible, emphasizing commissioning, and optimizing operation promote energy efficiency
(energy conservation) in buildings. The following are possible energy conservation measures
(ECMs) for the HVAC&R system in buildings:
 Turn off electric lights, office appliances, and other equipment when they are not needed.
 Shut down AHUs, PUs, fan coils, VAV boxes, chillers, fans, and pumps when the space they serve
is not occupied or when they are not needed, except during warm-up, cool-down, or the purging
period prior to the morning occupied period.
 The time chosen to start or to stop the AHUs, PUs, chillers, and exhaust fans daily should be optimum.
 The set point of space temperature, relative humidity, cleanliness criteria, and indoor-outdoor
pressure differential should be optimum. Using different space temperature set points for summer
and winter seasons for comfort air conditioning is often energy-efficient.
 The discharge air temperature from the AHU or PU and the temperature of water leaving the
chiller or boiler should be reset for part-load operation based on the space temperature, system
load, or outdoor temperature.
 Reduce air leakages from ducts, dampers, equipment, and HVAC&R system components. Use
weather stripping to seal windows and revolving exterior doors to reduce infiltration. Ducts and
pipes should be well insulated. Carefully design the layout of ducts and pipes to minimize their
length, the number of duct and pipe fittings, as well as their pressure losses.
25.10 CHAPTER TWENTY-FIVE
 Use more energy-efficient cooling methods such as free cooling, evaporative cooling, or groundwater
cooling instead of refrigeration. Replace refrigeration with an air economizer, water economizer,
evaporative cooler, and even desiccant dehumidification if doing so provides the same
cooling results, is more energy-efficient and is cost-effective. An evaporative condenser is often
more energy-efficient than a water-cooled or air-cooled condenser in many U.S. locations.
 Use heat recovery systems, waste heat from gas-cooling engines, or heat pumps to provide winter
space heating when they are applicable and cost-effective.
 Use variable-air-volume systems instead of constant-volume systems, and variable-flow building
loop water systems instead of a constant-flow one if the airflow or the water flow should be reduced
during part-load operations. Large variable-speed aerofoil centrifugal fans are energyefficient
and often cost-effective in VAV systems.
 Use energy-efficient chillers, boilers, AHUs, PUs, and motors. Todesco (1996) recommended that
chillers have an energy use index of 0.50 to 0.55 kW / ton, condensing boilers of 90 percent effi-
ciency, and large supply and return fans of 65 percent total efficiency and higher.
ASHRAE/IESNA Standard mandates that minimum nominal full-load efficiency, in percent, for
electric motors shall comply with the requirements of Energy Policy Act of 1992 as follows:
 Adopt a direct digital control (DDC) energy management and control system for large and
medium-size HVAC&R new and retrofit projects.
 Adopt double-pane windows with low-emission coatings. Todesco (1996) recommended a U
value of 0.25 Btu /h	 ft2	 °F (1.4 W/m2	 °C) for exterior windows, also lower U values for external
wall and roofs. Reduce the heat gain in the summer and heat loss in the winter in locations where
the outdoor temperature is high in the summer and low in the winter.
 Use daylighting and controls for the perimeter zone. Also use energy-efficient fluorescent lamp and
electronic ballasts with a goal to achieve connected lighting loads of 0.75 W/ ft2 (8 W/m2) or better.
According to EIA’s Commercial Building Characteristics (1994), in commercial buildings, the
ratio of the floor area that had adopted the captioned energy conservation feature to the total conditioned
floor area, in percentage, in 1992 in the United States was as follows:
Off-hours heating reduction 68.1 percent
Off-hours cooling reduction 63.0 percent
Energy management and control system (EMCS) 21.1 percent
Energy audit 21.8 percent
Variable-air-volume (VAV) systems 20.5 percent
Economizer cycle 27.0 percent
HVAC&R maintenance 72.4 percent
Motor horsepower,
Open motors, rpm Enclosed motors, rpm
synchronous speed, rpm 3600 1800 1200 3600 1800 1200
1 82.5 80.0 75.5 82.5 80.0
1.5 82.5 84.0 84.0 82.5 84.0 85.5
2 84.0 84.0 85.5 84.0 84.0 86.5
5 85.5 87.5 87.5 87.5 87.5 87.5
10 88.5 89.5 90.2 89.5 89.5 89.5
20 90.2 91.0 91.0 90.2 91.0 90.2
50 92.4 93.0 93.0 92.4 93.0 93.0
100 93.0 94.1 94.1 93.6 94.5 94.1
200 94.5 95.0 94.5 95.0 95.0 95.0
ENERGY MANAGEMENT AND GLOBAL WARMING 25.11
25.5 CASE STUDY: ENERGY CONSERVATION
MEASURES FOR AN OFFICE
Parker et al. (1997) introduced an energy-efficient office building design for the Florida Solar Energy
Center (FSEC) in the hot and humid climate of Cocoa, Florida. This office building consists of
offices, a visitors’ center, and laboratories and has a floor area of 41,000 ft2 (3809 m2). Laboratories
were not included in this energy analysis. Eighteen energy conservation measures (ECMs) were
considered during the energy analysis. Some of the ECMs are not effective for the hot, humid climate
in Florida; e.g., the air economizer cycle showed little savings, and higher than code-mandated
levels of wall and roof insulation resulted in little advantage. Only the following ten ECMs
provided cost-effective energy savings.
 A high-efficiency lighting system of 0.9 W/ ft2 (9.7 W /m2) with T-8 fluorescent lamps in a reflective
troffer and with electronic ballasts saves energy use.
 Windows comprised of glazing units with high visible transmittance of 0.56 for daylighting and a
low shading coeffficient of 0.33 reduce unwanted solar heat gain. Also a low U value of
0.31 Btu /h 	 ft2 	 °F (1.76 W/m2 	 °C) reduces conductive heat gain.
 Daylighting perimeter illumination coupled with solar control system on the south facade enables
effective use of the perimeter office illumination. Dimming electronic ballasts are controlled by
photometric sensors to adjust ballast output combined with the available daylight to maintain a
constant desktop illumination level.
 Two high-efficiency screw chillers provide an IPLV of 0.65 kW / ton (COP  5.41) and of 0.60
kW/ ton (COP  5.86) at 50 percent part load.
 Energy star personal computers, printers, and copiers are used to save energy.
 A reflective white single-ply roof membrane is chosen instead of a gray or black one, to reduce
the solar irradiated heat gains through the roof.
 Use of occupancy sensors shuts off the VAV terminals when a room is not occupied.
 Using a DDC EMCS permits an increase in the cooling set point from 75 to 76°F (23.9 to 24.4°C)
because of a finer control tolerance as well as provides optimal start and stop capability.
 Using variable-speed fans and variable-speed building pumps reduces the fan speed or pump speed
during part load to save fan and pump energy use. Analysis showed that they are cost-effective.
 CO2 -based demand ventilation control is adopted for intermittently occupied spaces and zones.
In Florida, the average energy use of 160 state-owned office buildings in 1991 was 67
MBtu/h 	 ft2 	 yr (211 kWh/m2 	 yr). With these ten ECMs, the predicted energy use for the office
building of FSEC will drop to 27 MBtu /h 	 ft2 	 yr (85 kWh /m2 	 yr).
25.6 RELATIONSHIP BETWEEN HVAC&R SYSTEM
CHARACTERISTICS AND ENERGY USE
System Characteristics and Energy Use Intensities
As discussed in Sec. 25.3, for an HVAC&R (air conditioning) system, if the location of the building
and the construction of the building it serves as well as its operating parameters are nearly the
same, the energy use of this HVAC&R system varies as its system characteristics (types and configurations)
change. According to the DOE / EIA 1998 nonresidential buildings energy consumption
survey of commercial buildings’ consumption and expenditures in 1995, the energy use intensity
(EUI) of various HVAC&R systems is listed in Table 25.1.
25.12 CHAPTER TWENTY-FIVE
In Table 25.1, WSHP represents water-source heat pump. Heat pump systems that include the
air-to-air heat pumps and WSHP systems consumed the least energy because energy use in space
heating was the lowest, as shown in the breakdown of heating, cooling, ventilation, lighting, office
equipment, etc. below. Also, central systems consumed the most energy because the sensible cooling
load of lighting, office equipment, etc. was about 50 percent higher in central systems than in
packaged systems.
The energy consumption of cooling and ventilation in individual systems was the lowest. This indicated
that the fan energy use was lowest with a lower fan total pressure and a poorer outdoor ventilation
air supply than in other systems.
Energy Use of Heating-Cooling Equipment and Fan
As the heating energy used in an HVAC&R system mainly depends on the heating system characteristics,
the efficiency of the direct-fired gas furnace and of the gas boiler is around 90 percent
for the condensing furnace and boiler and 75 to 80 percent for conventional furnaces and
boilers. The energy use of cooling equipment mainly depends on the characteristics of the refrigeration
system.
For centrifugal chillers, it is assumed that energy-use for the centrifugal compressor is 0.55
kW/ ton (6.39 COP), and the energy use of water pumps and cooling tower fans is 0.2 kW / ton. For
packaged systems, the energy use of reciprocating or scroll compressors including the air cooled
condenser fans is from 1.1 to 1.2 kW / ton.
The energy use of ventilation depends mainly on the ventilation rate and the fan characteristics.
From Eq. (15.6), fan energy use P, in hp (kW), can be calculated as
(25.3)
where pf  system total pressure, in. WC (Pa)
 volume flow rate, cfm (L / s)
f m d  combined fan, motor, and drive efficiency
According to ASHRAE/ IES Standard 90.1-1989, typically pf and f m d have the following values:
V? Pf 
pf V?
6356 f m d
Energy consumption, kBtu/ ft2 	 yr
Heating Cooling Ventilation Lighting Office equipment Others
Heat pumps 16.1 9.0 2.8 23.4 7.6 6.6
Packaged systems 26.0 7.9 3.2 24.6 6.7 6.4
Individual systems 36.0 5.5 1.7 16.5 3.9 6.6
Central systems 29.0 10.0 5.7 34.0 10.5 11.2
ENERGY MANAGEMENT AND GLOBAL WARMING 25.13
TABLE 25.1 Energy Use of Various HVAC&R Systems
System characteristics EUI, kBtu/ ft2 	 yr (kWh / ft2 	 yr)
Heat pumps (including air-to-air and WSHPs) 28.1 (8.2)
Packaged systems (including VAV systems) 37.1 (10.9)
Evaporative systems 38.0 (11.1)
Individual systems 43.2 (12.7)
Single-zone, constant-volume packaged systems 44.0 (12.89)
Central systems (including VAV central systems) 44.7 (13.1)
For both fan-coil and water-source heat pump systems, for each ton of refrigeration capacity
there is about a corresponding 320 cfm (151 L / s) of volume flow for the small fan in the fan-coil
unit and 80 cfm (38 L / s) of volume flow in the dedicated outdoor ventilation system. The dedicated
outdoor ventilation system has often a total pressure loss of 5 in. WC (1250 Pa).
For VAV systems, the system total pressure often has a 1 in. WC (250 Pa) fixed pressure loss.
The variable part will drop to an average of its 65 percent design system total pressure loss during
part-load operation.
25.7 ELECTRICITY DEREGULATION
Electric Utilities prior to Deregulation
An electric utility is vertically integrated from three main components: generation, transmission,
and local distribution network, as shown in Fig. 12.11b. There are usually four kinds of generation
plants: high-efficiency coal and nuclear plants and inefficient diesel and gas-turbine plants.
Prior to the electricity deregulation, large facilities purchased electricity often based on the
time-of-use (TOU) electric rate schedule that divides a 24-h working day into two or three periods,
on-peak (such as from 9 a.m. to 10 p.m. weekdays) and off-peak periods (10 p.m. to 9 a.m. weekdays
plus weekends and holidays) or on-peak, partial-peak, and off-peak periods. Higher rates are
charged for on-peak hours than for off-peak hours. In addition, there is a demand charge and a
monthly service charge based on the largest on-peak electricity power demand. In the 1990s prior to
the electricity deregulation, unit electric rates ranged between $0.02 and $0.16 per kWh with an average
around $0.07 per kWh in the United States.
By using a time-of-use electric rate schedule, electric utilities tend to shift the electricity from
on-peak hours to off-peak hours, and they shift the daytime peak operation of inefficient diesel and
gas-turbine plants to nighttime base-load high-efficiency coal and nuclear plants to reduce operating
costs.
Electric Deregulation
Electric utilites were originally regulated to prevent duplicated investments in electricity generation,
transmission, and distribution networks to serve the same community. Electricity deregulation
encourages open markets and competition and is generally considered as a means to reduce
the cost of electricity and services. According to Warwick (1997), in the United States, deregulation
began 20 years ago with telecommunications, gas, banking, airlines, and trucking. For electric
deregulation, the Public Utility Regulatory Policies Act (PURPA) of 1978 required retail utilities
to buy power from small generation plants developed by independent producers. This formed a
competitive market in power generation and power wholesalers. In 1995, the Federal Energy Regulatory
Commission (FERC) required utilities with bulk power transmission lines to provide open
access to wholesale power buyers and sellers. As of 1997, two of the three major operations, generation
and transmission, are open to competition as the result of the federal reforms. The next
pf, in WC (Pa) f m d
Constant-volume single-zone packaged system 2.5 (625) 0.40
Multizone packaged system 4.0 (1000) 0.45
VAV reheat packaged system 4.0 (1000) 0.45
VAV reheat central system 5.0 (1250) 0.55
Dual-duct VAV central system 6.0 (1250) 0.55
Fan-coil system 0.5 (125) 0.25
Water-source heat pump system 0.5 (125) 0.25
25.14 CHAPTER TWENTY-FIVE
phase of electric deregulation requires direct access to retail markets based on the condition that a
competitive power market is provided.
California Approach
In late 1990s, several states followed California’s lead in having legislation in place by setting a
specific date for electric deregulation, such as Arizona in 1999, California in 1998, Massachusetts
in 1998, New Hampshire in 1998, Pennsylvania in 1999, and Rhode Island in 1997.
In 1999, according to Warwick (1998), Waintroob (1998), and Gottfried (1997), the following
holds:
 States in which electric utilities are deregulating typically adopt one of the following two approaches.
The first approach is to provide electric consumers with an immediate rate cut but to retain
regulatory control over power rates for the incumbent utility. Power-supplying competitors
have to offer power below the new, discounted rates. In most states, the new rate is accompanied
by a rate freeze or cap lasting for several years. The second approach is to separate the power cost
from the old utility rate and to allow electric users to buy from alternate competitors.
 Power supply is the major electric utility service for competition under electric utility deregulation.
California established a power exchange pool which acts as a power auction house and an
independent (transmission) system operator, or ISO, which manages all power transfers and settles
all power sales accounts when utilities continue to own and maintain their transmission lines.
The experience of direct access in New Hampshire using bilateral trading allows individuals to
negotiate for the best deal they can get. Those who do well have lower-cost good-quality electricity.
Bilateral trades are contracts with receive and delivery terms. The buyer has to arrange transmission,
if the contract is for power only.
 To implement competition without involving lawsuits by incumbent electric utilities and other jurisdictions
that rely on the tax dollars generated from power plant valuations, electric utilities
were offered means to recover their stranded costs within a certain period. In California, the
state’s degeneration bill AB 1980, enacted in September 1996, calls for a nonbypassable stranded
cost surcharge of approximately 2.5 to 6 cents per kWh payable over 4-year period until 2002 by
rate payers. In 2002, California consumers will realize the benefits of deregulated rate. Stranded
cost is the cost difference between a plant cost carried on its books and the real market value.
Real-Time Pricing
Recently, the use of the computer and data handling software, the improvements in telecommunications,
as well as the demand for better energy management following electric deregulation have
made it possible for electric utilities to offer to customers an electric rate structure that more precisely
represents the actual cost of providing the electric service. In addition, this electric rate structure
also depends on the climate and demand; varies hourly, 24 h daily and 8760 h annually; and is
known as real-time pricing (RTP). It began in 1985 in California. Many states allowed electric utilities
to offer RTP electric rate structure to commercial and industrial customers in the late 1990s.
Other rate structures will also be available in a deregulation environment as the power providers
maneuver to offer the most competitive plans possible.
According to Bynum (1998), the components of a typical bill, in percentage, in the late 1990s
are as follows:
Cost of generation 40 percent
Transmission, local distribution, service,
profit, etc. 60 percent
ENERGY MANAGEMENT AND GLOBAL WARMING 25.15
Figure 25.1 shows the daily range of RTP electric rates at summer and winter peaks. Because of the
use of the RTP electric rate structure instead of the TOU rate structure following electric deregulation,
Norford et al. (1998) noted the following:
 The need for such crude price signals as the demand charge is diminishing.
 Commercial and industrial facility operators have the opportunity to rationally evaluate the tradeoffs
and the possible reduction or shifts in building services, such as lights, heating, cooling, and
process loads. The general strategy of shifting loads from high-priced to low-priced hours is common
to both TOU and RTP approaches; however, RTP has a much greater range of price variation
than TOU.
 Knowledge-based system computer programs are used for RTP cost optimization control. RTP
control strategies can be summarized thus: use of thermal storage systems; utilization of fuel
switching of on-site auxiliary or emergency generation or cogeneration, as well as gas cooling
such as absorption chillers or gas-engine-driven chillers; general load shedding of lights and fans;
and batch scheduling of production processes.
Case Study: Automated Control of RTP
According to Gabel et al. (1998), the first automated RTP control was installed in the 51-story Marriott-
Marquis Hotel with 1.8 million ft2 (167, 200 m2) in New York City. The mechanical systems
include three 900-ton (3170-kW) centrifugal chillers and many constant-volume and VAV systems.
Heating is provided by purchased steam. The 1900 guest rooms were excluded from RTP control
for primary room comfort. About 70 air-handling, ventilation, and lighting systems for the remaining
55 percent of floor space were placed under RTP control.
25.16 CHAPTER TWENTY-FIVE
Unit electric rate, $/kwh
0
Summer
peak
Winter
peak
4 8 12
Hours
16 20 24
0.2
0
0.4
0.6
0.8
1.0
1.2
1.4
FIGURE 25.1 Real-time pricing (RTP) unit electric rate daily range at summer and winter peaks.
The control strategy for each system or equipment has criteria that must be met before its load
can be shed, reduced, or rescheduled. Here is an example of control logic. If the RTP price is
greater than $0.10 per kWh, if the AHU AC-2 is in an occupied period, and if the space temperature
AC-2 serves is not over 80°F (27°C), then command AHU AC-2 off. If the above control parameters
are not met, AC-2 continues to operate.
The electric utility provided day-ahead notice of upcoming hourly RTP rates downloaded via
dial-up phone line. The RTP control system responds to the rate schedule hour by hour, first by analyzing
its preprogrammed control strategies and then by executing the most appropriate events in
sequence. The control system also reevaluates the control logic for each controllable load each
hour, making adjustments if necessary. About 70 air system loads were configured for RTP control.
During the three most recent years with TOU electric rates, the electric energy use averaged approximately
38 million kWh per year, with a peak electric load of 6 MW. The electric bill averaged
$3 million per year. With automated control response to RTP rates:
 The electric bill saved $1 million in 4 years.
 Average energy use dropped more than 2 million kWh each year.
 Peak electric load reduced 1500 kW during high-price periods.
25.8 SYSTEM SIMULATION
Energy Estimation and Energy Simulation
Annual, seasonal, or short-term estimation of energy use in an air conditioning or HVAC&R system
is an important tool to assess energy performance and cost in both design and daily operations.
The annual energy use of various types of air, water, refrigeration, and heating systems can
be predicted and analyzed. Based on the results of energy estimation, the most energy-efficient
system can be selected.
For the daily operation of air conditioning systems in existing buildings, detailed simulation reveals
the system’s negative and positive effects on energy use and performance by comparing the
estimated evaluation against the actual measured results. The energy use of many alternative improvements
can be predicted and assessed. These estimates provide great potential for reducing the
energy use and operating cost of existing air conditioning systems.
The detailed simulation method uses a comprehensive computer program through a powerful
personal computer (PC) to simulate the thermal behavior of the building, and the performance of
the air, water, refrigeration and heating plants at various outdoor weather conditions, to estimate the
energy use of an air conditioning system. Because of the powerful capability of the computer program,
detailed simulation methods can include most of the important factors that affect system energy
use. The most widely used program is the year-round, hour-by-hour energy simulation software.
Short-term simulation is sometimes used in dynamic modeling to evaluate the influence of
the thermal storage effect of system and components on their performance.
Energy simulation is the representation of energy use of an actual system or component by a
model of analogous characteristics, and it is used to predict the performance and operating parameters
of a system or component. Mathematical equations are used to describe the operating characteristics
of the working substance and work and energy transfer of the system or component being
simulated. Energy use simulation can be represented by either performance equations or physical
modeling.
Performance Equations
Generally, a system component’s performance with regard to one or two independent operating
variables is often available from the manufacturer’s catalog. A polynomial expression is used to fit
ENERGY MANAGEMENT AND GLOBAL WARMING 25.17
the catalog data or actual performance into a regression equation, to mathematically relate the dependent
variable z to independent variable x in steady-state simulation as follows:
z  a0  a1x  a2x2  a3x3 (25.4)
To relate a dependent variable with two independent variables x and y, the polynomial expression is
z  b0  b1x  b2x2  b3y  b4y2  b5xy  b6x2y  b7xy2  b8x2y2 (25.5)
where a0, a1, . . . , a3, b0, b1, . . . , b8  coefficients. Computer programs are available to solve
the coefficients according to the data from the manufacturer or actual performance data.
Physical Modeling
Modeling. Setting up a component or a system model is the first step of energy simulation by
physical modeling. Modeling includes the following:
 Description of system or component configuration whether it is an air, water, refrigeration, or
heating system, or whether it consists of many components or only contains a single device.
 Description of the operating characteristics of the system or component, and the interaction between
system components, whether it can be simplified to a steady-state model or a dynamic
model. A simplification of the physical model that results in an error of only few percent of the fi-
nal result is recommended, in order to simplify the calculation and analysis.
Developing Mathematical Equations. Mathematical equations are developed to describe the operating
characteristics of the working substance, and work and energy transfer.
Solving for Outputs. Computer programs are used to solve equations simultaneously, in sequence,
or by iteration. The required operating parameters, the outputs, can thus be obtained. Sometimes performance
equations are used to link the required outputs with one or two operating parameters.
Verification. The results of predicted performance during energy simulation can be verified
against actual measured readings of similar models and operating conditions. According to Scientific
Computing (1997), using performance equations to simulate the energy use of an HVAC&R
system or component is simple, and lumps the capacity, energy use, or other required operating parameters
in one performance equation. The disadvantage is that the performance data must be provided
to create the performance equations. Physical modeling allows more freedom in configuring
the characteristics of the system or equipment.
Steady-State and Dynamic Simulation
According to its operating characteristics, energy simulation can be classified as steady-state or dynamic
simulation.
Steady-State Simulation. In steady-state simulation, the relationship between various operating
parameters within a certain time increment is described by mathematical equations independent of
time, such as
z  F(x, y, . . . ) (25.6)
where x, y, z  operating variables that are not a function of time. Steady-state simulation always
simulates relatively long-term system or component characteristics, such as annual or seasonal energy
simulation. A time increment of 1 h is typically used for analysis. Within the time increment,
the operating parameters are independent of time. However, the magnitude of the same operating
parameter may be different in successive hours.
25.18 CHAPTER TWENTY-FIVE
In steady-state simulation, the heat capacity of the working substance is often far greater than
the heat being absorbed and released from the equipment, pipes, ducts, and surroundings when
the temperature of the working substance fluctuates. Therefore, heat absorbed by the equipment,
pipes, and ducts and the heat transfer to or from the surroundings are often ignored. An hour-byhour,
year-round energy simulation computer program is an example of steady-state simulation.
Dynamic Simulation. In dynamic simulation, the relationship between various operating parameters
is described by a mathematical equation that is a function of time
z  F(x, y, t, . . .) (25.7)
where t  time variable. Dynamic simulation is mainly used to simulate short-term characteristics.
Its purpose is to analyze the interaction between the control actions and the dynamic response, and
the step change of an operating parameter on system performance. A time increment of minutes or
even seconds is required to analyze the rapid response of the system during simulation. In dynamic
simulation, the heat being absorbed and released from the equipment, instrument, pipe, and ducts is
often taken into consideration. The variation of zone temperature and supply volume flow caused
by a step change of the zone lighting load is an example of dynamic simulation.
Sequential and Simultaneous Simulation
When a system comprises many components linked together by a working substance and forms an
open circuit, the calculation can be started at a component whose output is the input of the next
linked component, and can progress in sequence to the component where the final result can be obtained.
Such an approach is called sequential simulation.
In sequential simulation, the output of a successive component usually does not affect the output
of the former component already calculated. If the output of a component does affect the output of
the former component already calculated, a computing loop is used until the calculated output is
equal to the assumed value.
The fuel input rate to a direct-fired gas heater is an example of sequential simulation. Calculation
starts from the heating requirement at the warm air heater. The next calculation is mass flow
and the temperature drop of the combustion gas, and finally the rate of fuel gas supplied to the
burner, which is the required result, is then calculated.
When the working substance that links the components together forms a recirculating closed loop
and many operating parameters are interrelated, all unknown operating parameters should be solved
simultaneously. A simultaneous simulation is required. The power input to a compressor in a reciprocating
compression refrigeration cycle is an example of closed-loop simultaneous simulation.
25.9 ENERGY SIMULATION OF A CENTRIFUGAL
CHILLER USING PHYSICAL MODELING
The following illustrates the detailed energy simulation processes of a centrifugal chiller for the sake
of calculating its power input and thus its accumulated annual energy use using physical modeling.
System Model
The centrifugal chiller has a two-stage compressor using HCFC-123 as refrigerant. Its refrigeration
cycle is the same as shown in Fig. 9.7. The centrifugal compressor is driven by a hermetic
motor of variable-speed drive. After compression, the refrigerant is discharged to a shell-and-tube
water-cooled condenser. A flash cooler or economizer is installed between the condenser and the
shell-and-tube flooded evaporator. Multiple orifice plates are used as the throttling device between
ENERGY MANAGEMENT AND GLOBAL WARMING 25.19
condensing pressure pcon and intermediate pressure pint, and between pint and evaporating pressure
pev. Water from the cooling tower is used as the condenser water. A proportional-integral (PI) DDC
panel is used to maintain a nearly constant temperature of chilled water as it leaves the evaporator.
During part-load operation, the DDC unit modulates the speed of the compressor to reduce both
the mass flow rate of refrigerant entering the compressor and its corresponding system head, to
create a new mass and energy balance.
The energy simulation of this centrifugal chiller is considered at steady state within a time increment
of 1 h. Sequential simulation, iterations, and perfomance equations are used to calculate
the final results. Many simplifications are adopted to reduce the calculations during simulation.
Operating Parameters Affecting Chiller Energy Performance
Operating parameters that affect the energy performance of this centrifugal chiller are as follows:
 Refrigeration load ratio Rload is defined as the ratio of the operating refrigeration load Qev to the
design full-load Qev,d. Ratio Rload affects the rate of heat transfer at the evaporator and condenser.
It also affects the efficiency of the two-stage compressor.
 Temperature of condenser water entering the condenser Tce is a function of outdoor wet-bulb temperature
To
	 and the performance of the cooling tower. Temperature Tce is closely related to the
condensing temperature Tcon in the condenser, and therefore the pressure lift p  pcon  pev, the
temperature lift T  Tcon  Tev, and the power input to the compressor Pin.
Simulation Methodology
From Eq. (13.5), power input to the compressor motor Pin,m, in hp, can be calculated as
(25.8)
where  mass flow rate of refrigerant, lb /h
cp mec mot  compression efficiency, mechanical efficiency, and motor efficiency of compressor
Work input Win or Wisen, in Btu / lb, is directly proportional to the enthalpy difference of the refrigerant
between condensing and evaporating pressure. To find Win, it is necessary first to determine the
condensing pressure pcon, the condensing temperature Tcon, the evaporating pressure pev, and the
evaporating temperature Tev. So the following information is required:
 An evaporator model is required to determine Tev at various operating load ratios Rload.
 A condenser model is required to determine Tcon at various Rload and outdoor wet-bulb temperatures
To
	.
 Calculation of isentropic work input Wisen to the compressor is required based on the two-stage refrigeration
cycle at various Rload and To
	 values.
 The actual power input to the compressor motor based on the compressor model and the annual
energy use is finally calculated.
Evaporator Model
From Eq. (13.7), the rate of heat transfer at the tube surface of the evaporator or the refrigeration
load Qev, in Btu / h (W), can be calculated as
Qev  AevUo,ev Tev (25.9)
m?r
Pin,m 
m?rWisen
2545 cp mec mot
m?r
25.20 CHAPTER TWENTY-FIVE
From Eq. (13.8), the overall heat transfer coefficient based on the outer surface area of the copper
tubes Uo,ev in Btu /h	 ft2 	 °F (W/m2 	 °C), is given as
(25.10)
The boiling coefficient for HCFC-123 is slightly lower than that for refrigerant CFC-11. From Eq.
(13.9), for a shell-and-tube flooded-type evaporator, the boiling coefficient of HCFC-123 can be
evaluated as
According to the experimental results in Webb and Pais (1991) and Jung and Radermacher (1991),
for HCFC-123 in copper tubes with integrated fins of 26 fins / in. (0.98 mm), constant Cb can be
taken as 2.5 and exponential index n is approximately 0.7. For enhanced surfaces, according to the
results of Webb and Pais (1991), ho is 35 percent higher.
The copper integrated-fin tubes currently used for evaporators usually have a fin spacing of 19 to
35 fins / in. (1.3 to 0.73 mm), typically 26 fins / in. (0.98 mm). The outside diameter of copper tubes
varies from to in. (15.9 to 19.1 mm). The ratio of outer surface area to inner surface area Ao /Ai
is often between 3 and 4.
For a closed-circuit chilled-water system for evaporators with conventional water treatments, the
fouling factor Rf can be taken as 0.00025 h 	 ft2 	 °F/ Btu (0.000044 m2 	 °C/W).
From Eq. (10.9), water-side heat transfer coefficient hi, in Btu /h 	 ft2 	 °F (W/m2 °C), can be calculated
as
(25.11)
From Eq. (13.11), the log-mean temperature difference between refrigerant and chilled water Tev
can be calculated as
(25.12)
The mass flow rate of chilled water , in lb / min (kg / min), flowing through the copper tubes
in the evaporator usually remains approximately constant during operation. Temperature Tel is
often set and reset according to the requirement of the air system and the energy efficiency of
the system.
From Eq. (13.13)
The evaporating temperature can then be determined as
(25.13)
At design load, Qev is equal to the design refrigeration load Qev,d, in Btu / h (kW). In part-load
operation,
Qev  RloadQev,d (25.14)
and the evaporating temperature at part load Tev,p can be similarly calculated.
Tev 
eB
Tel  Tee
eB  1
Tee  Tel
Qev / (AevUo,ev)
 B
m?w,ev
Tev 
Tee  Tev  (Tel  Tev)
ln [(Tee  Tev) / (Tel  Tev)]
NuD 
hiDh
k
 0.023 ReD
0.8Pr0.4
34
58
ho  Cb Qev
Aev n
Uo,ev 
1
1 / ( f ho)  AoRf / Ai  Ao / (Aihi)
ENERGY MANAGEMENT AND GLOBAL WARMING 25.21
Condenser Model
For a shell-and-tube water-cooled condenser, from Eqs. (10.16a) and (13.7), the rate of heat transfer
at the condenser Qrej, in Btu / h (kW), or the total heat rejection, can be calculated as
(25.15)
From Eq. (25.8) power input Pcom is known only when condensing temperature Tcon, work input
Win, and mass flow rate of the refrigerant all have been calculated. Therefore, from Eq. (10.17)
Qrej  FrejQev; assume a heat rejection factor Frej first. Then Tcon, Win, and Pcom can be calculated.
If the calculated Pcom and Qrej do not equal the assumed values, try another Frej until the assumed
and calculated values are equal to each other. For comfort air conditioning, Frej usually varies from
1.20 to 1.35.
As in the evaporator model, from Eq. (13.8), the overall heat-transfer coefficient based on the
outer surface area of the condenser Uo,con, in Btu /h 	 ft2	 °F (W/m2 	 °C), is
(25.16)
From Eq. (13.11), the log-mean temperature difference between the condensing refrigerant and the
condenser water Tcon, in °F (°C), is
(25.17)
From Eq. (13.10), the condensing coefficient hcon, in Btu /h 	 ft2 	 °F (W/m2 	 °C), is
In a filmwise condensation shell-and-tube water-cooled condenser having -in or -in.- (15.9- or
19.1-mm-) diameter copper tubes with integrated fins and using HCFC-123 as refrigerant, constant
Ccon can be taken as 10,500.
In a condenser using a well-maintained cooling tower water with proper water treatment, a fouling
factor Rf  0.00025 h 	 ft2	 °F/ Btu (0.000044 m2 	 °C/ W) is recommended. In industrial areas,
if a brush cleaning system is installed, Rf  0.0002 h 	 ft2 	 °F/ Btu (0.000035 m2 	 °C/W).
The operating pressure of HCFC-123, like CFC-11, is lower than atmospheric pressure, so air
and other noncondensable gases leak into the evaporator. The compressor transports them to a
higher level and then they accumulate in the condenser. Noncondensable gases reduce the condensing
area and raise the condensing pressure. Their effect is similar to that of a gas-side resistance Rg,
in h 	 ft2 	 °F/ Btu (m2 	 °C/W), at the condenser as follows:
Rg  0.00778  0.0173Rload  0.0114Rload
2 (25.18)
As in the evaporator model, C is calculated as follows:
Condensing temperature is therefore calculated as
(25.19) Tcon 
eC
Tcl  Tce
eC  1
Tcl  Tce
Qrej / (AconUo,con)
 C
34
58
hcon  Ccon 1
QrejAcon 1 / 3
Tcon 
Tcl  Tcon  (Tce  Tcon)
ln[(Tcl  Tcon) / (Tce  Tcon)]
Uo,con 
1
1 / ( fhcon)  AconRf / Ai  Rg  Acon / (Aihi)
m?r
 AconUo,con Tcon
Qrej  Qev 
2545Pcom
mot
25.22 CHAPTER TWENTY-FIVE
Cooling Tower Model
The approach of condenser water leaving the cooling tower Tce  To
	 is mainly influenced by the
load ratio Rload of the condenser, the configuration of the cooling tower, the outdoor wet-bulb temperature
To
	, and the number of transfer units (NTU) of the fill.
In a counterflow induced-draft cooling tower using PVC packing fill, with a condenser water
flow rate of 3 gpm per tonref (0.054 L / s 	 kW) of heat rejection and a water-air ratio of 1.2, if the
outdoor wet-bulb temperature 65°F To
	 78°F (18.3°C To
	 25.6°C), the temperature of
condenser water entering the condenser Tce, in °F (°C), can be roughly estimated by following performance
equations:
Tce  To
	  Tap,dKloadKwet
Kload  0.1  0.9Rload (25.20)
Kwet  4.8  0.0475To
	
where Tap,d  approach of cooling tower at design condition, °F (°C)
Kload  load factor
Kwet  factor considering drop of outdoor wet-bulb temperature
Centrifugal Compressor Model
From the refrigeration cycle shown in Fig. 9.7b and Eq. (9.33), isentropic work input Wisen  Win,
in Btu / lb (kJ / kg), for a two-stage compressor can be calculated as
Win  (1  x)(h2  h1)  h4  h3 (25.21)
From Eq. (9.35), the mass flow rate of the refrigerant at the condenser, in lb /h (kg / h), is given
as
(25.22)
From Eq. (9.30), the fraction of refrigerant evaporated in the flash cooler x can be evaluated as
(25.23)
On a pressure-enthalpy p-h diagram, the specific enthalpies of saturated liquid and vapor are functions
of temperature and pressure only. In centrifugal chillers used in comfort air conditioning systems,
the range between evaporating temperature Tev and condensing temperature Tcon is usually 20
to 120°F (11 to 66°C), and a simple polynomial representation can be used to calculate the specific
enthalpy from the known Tev and Tcon with acceptable accuracy.
At the flash cooler, because the saturated temperature and pressure are interrelated, for simplicity,
the intermediate saturated temperature between condensing and evaporating temperature can be
estimated as
(25.24)
According to the thermodynamic properties of HCFC-123 and its p-h diagram, the polynomial expression
to calculate the specific enthalpy of the saturated vapor of refrigerant HCFC-123 hv, between
20°F (11°C) and 70°F (21°C), in Btu / lb (kJ / kg), is
hv  89.7  0.145(Tev  20) (25.25)
Tint 
Tev  Tcon
2
x 
h5	  h8
h7  h8
m?r 
Qev
(h1  h9)(1  x)
m?r
ENERGY MANAGEMENT AND GLOBAL WARMING 25.23
The polynomial expression used to calculate the specific enthalpy of the saturated liquid of HCFC-
123 hl, between 60°F (15.6°C) and 110°F (43.3°C), in Btu / lb (kJ / kg), is
hl  20.2  0.24(Tl  60) (25.26)
Specific enthalpy differences h4  h3 and (1  x)(h2  h1) to determine the isentropic work can be
calculated more simply and with acceptable accuracy according to the corresponding saturated temperature
of the gaseous refrigerant at condensing, interstage, and evaporating pressures along the
constant-entropy lines of similar profiles in the superheated region. The polynomial expression used
to calculate the enthalpy difference of the gaseous refrigerant HCFC-123 between 60 and 120°F
(15.6 and 48.9°C) along the constant entropy line hs, in Btu / lb (kJ / kg), is
hs  0.135(Tint  Tev)  0.135(Tcon  Tint) (25.27)
where Tint  temperature of saturated gaseous refrigerant HCFC-123 at intermediate pressure, °F.
In Eq. (25.8) cp is the compression efficiency of the centrifugal compressor, which is defined as the
ratio of isentropic work to actual work delivered to the gaseous refrigerant during compression.
Also, mechanical efficiency mec is defined as the ratio of work delivered to the gaseous refrigerant
to the work input to the compressor shaft. The difference in these two work inputs is
mainly due to the loss in the bearings, and gear train, and during transportation of refrigerant in
the centrifugal chiller. For centrifugal compressors operated at a certain speed, mec is considered
a fixed value.
Mechanical efficiencies mec for centrifugal chillers manufactured after 1973 can be taken as
Without gear train 0.87
With gear train 0.85
Motor efficiency mot is a function of motor size and the load ratio of the chiller. Normally,
for motor size greater than 125 hp (93 kW), ASHRAE Standard 90.1-1999 recommends an efficiency
level equal to or greater than 93.0 percent as listed in Sec. 25.4. If the motor is to be operated
more than 750 h annually, a high-efficiency motor is most cost-effective. The efficiency
of a 200-hp (150-kW) motor should be 94.5 percent or greater. If a hermetic motor is used, another
2 to 4 percent of power input is required to provide the refrigeration capacity to cool the
hermetic motor.
Compression efficiency cp is often the most influential parameter during annual energy simulation
of a centrifugal compressor. Compression efficiency cp is a function of the volume flow
rate of refrigerant rf and the system head or pressure lift pt. Volume flow rf is closely related
to load ratio Rload, and system head is closely related to temperature lift T  Tcon  Tev. In
most centrifugal chillers, the chilled water leaving the chiller Tel is usually set at a constant
value when the outdoor wet-bulb temperature is high. As the outdoor wet-bulb temperature To
	
drops, the condenser water entering the condenser Tce and, therefore, the condensing temperature
Tcon fall accordingly. It is assumed that when To
	 drops below 70°F (21°C), the reset of Tel
offsets the fall of To
	.
Figure 25.2 shows the compressor map of a variable-speed centrifugal compressor (see Fig.
13.10). In Fig. 25.2, point A is the operating point of the centrifugal chiller at design load conditions
(Rload  1) and at a condenser entering water temperature Tce  85°F (29.4°C). If Tce remains
at 85°F and Rload drops to 0, then the operating point is at point B. At point B, temperature lift T is
calculated as follows:
(Tcon  Tev)B  (Tcon  Tev)A  (Tel  Tev)A  (Tcon  Tcl)A  (Tcl  Tce)A (25.28)
In a typical centrifugal chiller, if (Tel  Tev)A  (Tcon  Tcl)A  8°F, (Tcl  Tce)A  10°F, and a typical
(Tcon  Tev)A  67°F (similar to scheme A in Sec. 13.6), the operating curves, or required system
head, of the compressor at other Tce values can then be plotted on the compression map, as
shown in Fig. 25.2.
V?
V?
25.24 CHAPTER TWENTY-FIVE
The compression efficiency cp can therefore be calculated by a polynomial regression (performance
equation) based on two variables, Rload and Tce, as follows:
cp  C1  C2Rload  C3R2
load  C4Tce  C5T2
ce  C6RloadTce  C7R2
loadTce
 C8RloadT2
ce  C9R2
loadT2
ce (25.29)
where C1, C2, . . . ,C9  coefficients. From the manufacturer’s compressor map and the added
operating curves at various outdoor wet-bulb temperatures To
	, the coefficients can be determined
by a computer program. In a two-stage centrifugal compressor, the mean value of compression effi-
ciency of two individual stages is assumed to equal the overall compression efficiency cp for simplification.
25.10 ENERGY SIMULATION SOFTWARE DOE-2.1E
Energy Simulation Software
For energy audit, energy efficiency improvements, and the selection of energy-efficient HVAC&R
systems, Lawrence Berkeley Laboratory and Hirsch Associates, sponsored by the Department of
Energy (DOE) collaboratively developed energy simulation computer program DOE-2.1E. It is the
most widely used because it includes various types of systems, equipment, and energy-efficiency
measures and is friendly to the user. DOE2.1E is an hour-by-hour energy simulation software. The
other two widely used energy simulation software programs are TRACE 600, developed by The
Trane Company, and HAP E 20-II, developed by Carrier Corporation. Building loads analysis and
systems thermodynamics (BLAST) is often used in energy research works.
ENERGY MANAGEMENT AND GLOBAL WARMING 25.25
FIGURE 25.2 Compressor map of a centrifugal chiller at various load ratios Rload and
Tce.
According to Scientific Computing (1996), another DOE-sponsored energy simulation software
DOE-2.1E 100 or DOE-2.2 has similar system and equipment capacities to DOE-2.1E. The following
table shows their differences:
Loads
DOE-2.1E has the following primary simulation programs: loads, systems, energy efficiency measures,
and plant. The LOADS programs calculate the sensible and latent components of the hourly
heating or cooling loads for each designated conditioned space (zone) in a building under the following
conditions:
 The software uses 8760 h of actual weather data annually.
 Indoor temperature is assumed to be kept at a constant value within the hour.
 The LOADS program is reponsive to weather and solar conditions; to schedules of people, lighting,
and equipment; to infiltration; to the heat storage effect of massive walls and roofs; and to the
effect of building shade due to solar radiation.
 The LOADS program considers both external and internal loads. External loads are due to heat
transfer through walls, heat transfer through windows, infiltration through walls and windows,
and solar heat gain through windows. Internal loads are due to heat gains from people, lights, and
equipment inside the conditioned space.
 The LOADS program performs calculations in a hierarchical order: building (system), space
(zone), wall, window, or door.The program calculates the external loads for all windows and
doors on a wall, then for all walls in a space. For the space, LOADS calculates the internal loads,
combined with external loads, and gives the total space loads. When all space (zones) loads are
summed at a consistent time, this gives the heating or cooling load for the hour of the building
(system). During heating and cooling load calculations, the typical floor of a building is usually
divided into many zones, such as five or nine zones; each often has external walls and windows
facing one, two, or even three orientations except the interior zone.
 The LOADS program performs load calculations by using weighting factors (transfer function coefficients).
Custom weighting factors for various room configurations are also used as an option
instead of a preset group of weighting factors. Detailed calculations of each external or internal
load using weighting factors (transfer function method) can be found in Sec. 6.6.
Systems
The energy use simulation of the following air and water systems is included:
 Single-zone constant-volume central systems
 Multizone constant-volume reheat central systems
 Multizone variable-air-volume (VAV) reheat central systems
 Multizone fan-powered VAV central systems
DOE-2.1E DOE-2.2
Equipment time step 1 h 15 min
Sequential-iterative Sequential Sequential-iterative
Zone volume flow automatically calculated Supply Outside
Heat recovery Sensible Sensible and latent
Heat recovery variations during cooling As a function of To and To
	
25.26 CHAPTER TWENTY-FIVE
 Multizone dual-duct VAV central systems
 Multizone VAV reheat packaged systems
 Multizone fan-powered VAV packaged systems
 Desiccant dehumidification packaged systems
 Evaporative cooling systems
 Two-pipe fan-coil system
 Four-pipe fan-coil systems
 Packaged terminal air conditioner
 Water-source heat pump systems
 Ground-source heat pump systems
 Gas-engine-driven heat pump systems
 Thermal storage systems
Energy Efficiency Measures
Based on Scientific Computing (1996), the energy use is calculated and analyzed with the following
energy efficiency measures and controls:
 Control of the air leaving coil temperature. Constant, coldest /warmest, reset schedule based on
outdoor temperature, and user-defined reset temperature schedule.
 Air economizer control. No economizer, air economizer actuated by a fixed temperature, air economizer
actuated by comparing outdoor air enthalpy with return air, air economizer actuated by
comparing outdoor air temperature with a fixed temperature, and supply air contains fixed amount
of outdoor ventilation air.
 Type of air-to-air heat recovery. Heat exchanger, rotary wheel, and run-around coils.
 Fan operating schedule and control. Fan is operating according to user-defined schedule; fan control
includes constant-volume, variable-speed fan inlet vanes, discharge damper, cycles on and
off, high and low speeds, as a function of part-load ratio.
Plant
Energy simulation of the following heating and cooling plant equipment can be performed either by
performance equations or by physical modeling:
 Boilers include multistage resistance-type electric boilers, gas / oil-fired hot water boilers, and
gas / oil-fired steam boilers. Boilers can be operated only when load is passed from the system, or
in standby operation. Boilers can also be operated at a modified energy input / heat output as a
function of the part-load ratio.
 Single-effect or double-effect direct-fired absorption chiller is used in which the cooling capacity
can be modified as a function of refrigeration load required, or as a function of chilled water temperature
entering the chiller.
 Open and hermetic centrifugal chillers modify the capacity and electric power input as a function
of chilled water temperature entering the chiller, or their part-load ratio.
 Centrifugal chillers are used with double-bundle condensers in which condenser water is used as
the hot water for winter heating in perimeter zone.
 Open and hermetic reciprocating chillers are used with the type of condenser specified, including
air condenser or evaporatively-cooled condenser.
ENERGY MANAGEMENT AND GLOBAL WARMING 25.27
 Reciprocating chiller driven by a diesel engine has its COP modified as a function of the part-load
ratio when the engine is at minimum speed, and COP is modified as a function of chilled water
temperature entering the chiller when the engine is operated above minimum speed at part load.
 Cooling towers use the following controls of the condenser water temperature leaving the tower:
fixed set point; or wet-bulb temperature reset as outdoor wet-bulb temperature drops; or control of
the number of cells operating, either actual number of cells needed or all cells. The control of the
temperature of condenser water leaving the tower comprises the following devices and setups:
Fluid bypass, a three-way valve to bypass condenser water around the cooling tower
Cycling, cycling on and off of the tower fans
Two-speed fans, that cause the tower fans to cycle between off, low, and high speeds
Variable-speed fans to vary the speed of the tower fans
25.11 ASHRAE/IESNA STANDARD 90.1-1999 ENERGY
COST BUDGET METHOD
The energy cost budget method is an alternative to the prescriptive provisions in Standard 90.1-
1999 (which is similar to the building envelope trade-off option to the prescriptive building envelope
option in Sec. 3.13). Using the energy cost budget method compliance will be achieved if
 All mandatory provisions are met
 The design energy cost does not exceed the energy cost budget when evaluated in accordance
with specified provisions
 The energy efficiency level of components specified in the building design meet or exceed the ef-
ficiency level used to calculate the design energy cost
Compliance shall be documented and submitted to the authority having jurisdiction as follows:
1. The energy budget for the budget building design and the design energy cost for the proposed
design
2. A list of energy-related features that are included in the design
3. The input and output reports from the simulation program including a breakdown of energy use
of components
4. An explanation of any error messages noted in the simulation program output
The simulation program shall be a computer-based program to analyze the energy consumption in
buildings (such as, DOE-2, BLAST). The simulation shall be performed using hour values. The simulation
program shall include the calculation methodologies of the building components being modeled.
For details, including calculation of design energy cost, calculation of the energy cost budget, and
HVAC&R systems, refer to Standard 90.1-1999.
REFERENCES
Akbari, H., Konopacki, S. J., Lister, L. D., and Debaillie, L. P., Energy End-Use Characterization at Fort Hoot,
Texas, ASHRAE Transactions, 1996, Part II, pp. 724–733.
ASHRAE, ASHSRAE Handbook 1999, HVAC Applications, ASHRAE Inc., Atlanta, GA, 1999.
ASHRAE/ IES, ASHRAE/ IES Standard 90.1-1989, Energy Efficient Design of New Buildings Except New
Low-Rise Residential Buildings, ASHRAE Inc., Atlanta, GA, 1989.
25.28 CHAPTER TWENTY-FIVE
ASHRAE/IESNA Standard 90.1-1999 Energy Standard for Buildings Except Low-Rise Residential Buildings,
ASHRAE Inc., Atlanta, GA, 1999.
Baxter, V., Fischer, S., and Sand, J. R., Global Warming Implications of Replacing Ozone-Depleting Refrigerants,
ASHRAE Journal, no. 9, 1998, pp. 23–30.
Better Late Than Never: Energy Efficiency Is a Federal Priority, Air Conditioning, Heating & Refrigeration
News, Aug. 18, 1997, p. 16.
Bobenhausen,W., and Lahiri, D., HVAC Design for Green Buildings, HPAC, no. 2, 1999, pp. 43–50.
Bynum, R., The Current Situation, Engineered Systems, no. 12, 1998, pp. 32–36.
Clark, D. R., Hurley, C. W., and Hill, C. R., Dynamic Models for HVAC System Components, ASHRAE Transactions,
1985, Part I, pp. 737–751.
Cole, R. J., and Larsson, N., Green Building Challenge ‘98, ASHRAE Journal, no. 5, 1998, pp. 20–23.
Cox, J. E., and Miro, C. R., Reducing the Federal Energy Bill, ASHRAE Journal, no. 12, 1998, p. 24.
Cox, J. E. and Miro, C. R., Montreal and Kyoto Protocols’ Relevance to HCFs, ASHRAE Journal, no. 1, 1999,
p. 16.
DOE/ EIA Nonresidential Building Energy Consumption Survey: Commercial Building Consumption and Expenditures
1989, DOE/ EIA 0318 (89), 1992.
Dolan,W. H., Gas Cooling for the Commercial Sector—Present and Future Perspective, ASHRAE
Transactions, 1989, Part I, pp. 968–971.
Gabel, S. D., Carmichael, L., and Shavit, G., Automated Control in Response to Real-Time Pricing of Electricity,
ASHRAE Journal, no. 11, 1998, pp. 26–29.
Ginsberg, M., and Parker, S., Meeting Federal Mandates, Engineered Systems, no. 9, 1996, pp. 52–60.
Gottfried, D. A., Implications of U.S. Electricity Deregulation, HPAC, no. 5, 1997, pp. 55–58.
Gottfried, D. A., Schoichet, E. A., and Hart, M., Green Building Environmental Control: A Case Study, HPAC,
no. 2, 1997, pp. 71–78.
Hansen, S. J., Performance Contracting: Fantasy or Nightmare? HPAC, no. 11, 1998, pp. 71–76.
Hicks, T. W., and Clough, D. W., Building Performance with the ENERGY STAR Label, HPAC, no. 10, 1997,
pp. 49–54.
Hoggard, J., Shoot for the Star, Engineered Systems, no. 1, 2000, p. 38.
Jung, D. S., and Radermacher, R., Prediction of Heat Transfer Coefficients of Various Refrigerants during Evaporation,
ASHRAE Transactions, 1991, Part II, pp. 48–53.
Lawrence Berkeley Laboratory, DOE-2 Users Guide, Version 2.1, DOE National Technical Information Service,
Springfield, VA, 1980.
Mahoney, J. W., and Weiss, D. W., Performance Contracting: A Guaranteed Solution, HPAC, no. 3, 1997, pp.
61–65.
Miller, D. E., A Simulation to Study HVAC Process Dynamics, ASHRAE Transactions, 1982, Part II, pp. 809–825.
Morrison, D., Green Design Focus: Energy Efficiency and Environmental Responsibility, HPAC, no. 2, 1998,
pp. 66–76.
Norford, L. K., Englander, S. L., and Wiseley, B. J., Demonstration Knowledge Base to Aid Building Operators
in Responding to Real-Time-Pricing Electricity Rates, ASHRAE Transactions, 1998, Part I A, pp. 91–103.
O’Neal, D., and Kondepudi, S. N., Demonstrating HVAC System Performance through System Simulation,
ASHRAE Transactions, 1986, Part II B, pp. 116–129.
Park, C., Bushby, S. T., and Kelly, G. E., Simulation of a Large Office Building System Using the HVACSIM 
Program, ASHRAE Transactions, 1989, Part I, pp. 642–651.
Parker, D. S., Fairey, P. W., and McIlvaine, J. E. R., Energy Efficient Office Building Design for Florida’s Hot
and Humid Climate, ASHRAE Journal, no. 4, 1997, pp. 49–57.
Quiriconi, M. A., Dean, M. L., and Litven, N. A., Failing Grade for Ailing Equipment, Engineered Systems, no.
12, 1997, pp. 72–78.
Schiess, K., RTP  TES  ? Engineered Systems, no. 10, 1998, pp. 102–110.
School Saves Nearly $500,000 Annually on Utility Bills, HPAC, no. 1, 1998, pp. 12–19.
Scientific Computing, HVAC System Design: Energy Simulation Software Review, Engineered Systems, no. 2,
1996, pp. 42–64.
ENERGY MANAGEMENT AND GLOBAL WARMING 25.29
Scientific Computing, Up for Review (Again), Engineered Systems, no. 1, 1998, pp. 76–84.
Silver, S. C., Jones, J. W., Peterson, J. L., and Hunn, B. D., CBS/ ICE: A Computer Program for Simulation of
Ice Storage Systems, ASHRAE Transactions, 1989, Part I, pp. 1206–1213.
Todesco, G., Super-Efficient Buildings: How Low Can You Go? ASHRAE Journal, no. 12, 1996, pp. 35–40.
Waintroob, D., Retail Electric Competition, Engineered Systems, no. 5, 1998, p. 48.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechnic, Hong Kong, 1987.
Warwick,W. M., Into the Looking Glasses: Utility Industry Restructuring and You, HPAC, no. 5, 1997, pp.
47–53.
Warwick,W. M., Top 10 Lessons from Competitive Power Purchases, HPAC, no. 8, 1998, pp. 77–82.
Webb, R. L., and Pais, C., Pool Boiling Data for Five Refrigerants on Three Tube Geometries, ASHRAE Transactions,
1991, Part I, pp. 72–78.
Wong, S. P. W., and Wang, S. K., System Simulation of the Performance of a Centrifugal Chiller Using a Shelland-
Tube Type Water-Cooled Condenser and R-11 as Refrigerant, ASHRAE Transactions, 1989, Part I, pp.
445–454.
Zaidi, J. H., and Howell, R. H., Energy Use and Heat Recovery in Water-Loop Heat Pump, Variable-Air-
Volume, Four-Pipe Fan-Coil, and Reheat HVAC Systems: Part 1 and Part 2, ASHRAE Transactions, 1993, Part
II, pp. 13–39.
25.30 CHAPTER TWENTY-FIVE
CHAPTER 26
AIR CONDITIONING
SYSTEMS: SYSTEM
CLASSIFICATION,
SELECTION, AND
INDIVIDUAL SYSTEMS
26.1
26.1 CLASSIFICATION OF AIR
CONDITIONING SYSTEMS 26.1
Basic Approach 26.1
Air Conditioning Systems
Classification 26.2
Air, Heating, and Cooling
Systems Designation 26.2
26.2 AIR CONDITIONING
SYSTEM, SUBSYSTEM,
AND MAIN COMPONENTS
SELECTION 26.3
Selection Levels 26.3
Requirements Fulfilled during
Selection 26.4
Applications and Building
Occupancies 26.4
System Capacity 26.5
Indoor Air Quality 26.5
Zone Thermal Control and
Sound Problems 26.6
Energy Efficiency 26.7
Fire Safety and Smoke
Control 26.7
Space Limitations 26.8
Maintenance 26.8
Initial Costs 26.8
26.3 INDIVIDUAL AIR CONDITIONING
SYSTEMS 26.8
Basics 26.8
Advantages and Disadvantages 26.9
26.4 ROOM AIR CONDITIONING
SYSTEMS 26.9
Equipment Used in Room Air
Conditioning Systems 26.9
Configuration and Cooling Mode
Operation 26.11
Energy Performance and Energy
Use Intensities 26.11
Controls 26.12
Features 26.12
System Characteristics 26.12
26.5 PACKAGED TERMINAL AIR
CONDITIONING SYSTEMS 26.13
Equipment Used in Packaged Terminal
Air Conditioning Systems 26.13
Heating and Cooling Mode
Operation 26.13
Minimum Efficiency Requirements 26.14
System Characteristics 26.15
Applications 26.15
REFERENCES 26.15
26.1 CLASSIFICATION OF AIR CONDITIONING SYSTEMS
Basic Approach
Because of the variation in building occupancies as well as the outdoor weather in various locations,
and because of different operating requirements from facility developers, owners, and users,
air conditioning or HVAC&R systems are usually designed, installed, and operated in different
types and configurations, with different system characteristics to meet these requirements.
The purpose of classifying air conditioning systems is to distinguish one type from another and to
provide a background for the selection of an optimum air conditioning system based on requirements.
As each air condition system consists of air, water, heating, and refrigeration systems, the classification
of air conditioning systems and is often mixed with the classification of air systems and water and
refrigeration systems. If a designer cannot properly classify an air conditioning system and distinguish
it from others, it will be difficult for him or her to select an appropriate system for the client.
During the classification of air conditioning systems, the following points should be considered:
 The classification of air conditioning systems, or HVAC&R systems, should include the primary
aspects of air systems as well as heating and cooling systems if possible, as the air system directly
affects the control of the indoor thermal environment and indoor air quality.
 The system and the primary equipment used should be compatible with each other. For example,
the primary equipment in a unitary packaged system is the packaged unit.
 System classification should mainly be based on practical applications. For example, as the indoor
air quality becomes one of the primary criteria used to select an air conditioning system,
it must be considered whether an “all-water system” without outdoor ventilation air supply can
exist.
 System classification should be simple, and each of the air conditioning systems should be clearly
different from the others.
Air Conditioning Systems Classification
As discused in Chap. 1, air conditioning systems can be classified currently into eight categories according
to their configurations and operating characteristics:
 Individual room air conditioning systems or simply individual systems
 Evaporative cooling air conditioning systems
 Desiccant-based air conditioning systems or simply desiccant systems
 Thermal storage air conditioning systems or simply thermal storage systems
 Clean room air conditioning systems or simply clean room systems
 Space conditioning air conditioning systems or simply space systems
 Unitary packaged air conditioning systems or simply packaged systems
 Central hydronic air conditioning systems or simply central systems
First, as discussed in Sec. 1.5, the individual, space, packaged, and central systems together had
more than 98 percent of floor area both in commercial buildings in 1992 and in air conditioned
homes in 1991 in the United States. Clean room and desiccant-based air conditioning systems are
often processed air conditioned systems. Except where deep-well water or an air economizer is
available, evaporative cooling systems are the cheapest cooling systems and are widely used in arid
southwestern areas of the United States. Thermal storage systems have a quite different water system
and operating characteristics from a central system. Second, there are many specific air conditioning
systems used in many industrial applications such as air conditioning used for textile mills
with air washers, and also new air conditioning technology will be invented and developed. As soon
as their importance and their distinct system characteristics are recognized by the HVAC&R industry,
a new category of air conditioning system should be added.
Air, Heating, and Cooling Systems Designation
The title of air conditioning systems listed in later sections of this chapter and succeeding chapters,
such as VAV reheat rooftop packaged systems, also designates the air, heating, and cooling systems:
26.2 CHAPTER TWENTY-SIX
 A packaged system always has a cooling system that uses the DX coil to cool air directly. A desiccant-
based system usually uses the DX coil as supplementary cooling. An individual system
also uses a small, self-contained, factory-assembled refrigeration system that has a DX coil to
cool air.
 A central system has a cooling system that uses chilled water as a cooling medium to cool air indirectly.
A thermal storage system is always a central system that uses chilled water or brine as
the cooling medium.
 As discussed in Sec. 9.18, for DX coils in individual systems, rotary compressors are the most
widely used. For DX coils in packaged systems, reciprocating and scroll compressors are most
widely used. Scroll compressors are gradually replacing the reciprocating compressors in packaged
systems because of their higher energy efficiency. For chillers in central and space systems,
centrifugal and screw compressors are widely used in large chillers whose cooling capacities are
greater than 75 tons (264 kW); and screw, scroll, and reciprocating compressors are widely used
in medium-size and small chillers. Screw and scroll types are also gradually replacing the reciprocating
compressors in small and medium-size chillers.
 As discussed in Sec. 8.1, hot water, heat pump, and direct-fired warm air furnace heating systems
were installed in commercial and residential buildings with a floor area of both about 75 percent
in the early 1990s. In packaged systems, direct-fired warm air furnaces and heat pumps are often
used. For individual systems, either heat pumps or electric resistance heaters are often used. In
central and clean space systems, hot water heating systems are most widely used.
 As discussed in Sec. 20.9, air systems can be classified into three categories:
Constant-volume (CV) outdoor recirculating air mixing systems
Variable-air-volume (VAV) outdoor recirculating air mixing systems
Dedicated ventilation and space recirculating systems
Only air systems of the space air conditioning systems are dedicated ventilation and space circulating
systems. Individual, packaged systems in residentials and clean room air conditioning systems
are mostly CV air systems. All the other air conditioning systems may be either CV or VAV
systems and will be so designated.
26.2 AIR CONDITIONING SYSTEM, SUBSYSTEM,
AND MAIN COMPONENTS SELECTION
Selection Levels
During the design of a new or retrofit system, air conditioning system, subsystems, and main components
selection is performed mainly on three levels:
 Air conditioning system level. This level deals with such aspects as whether a VAV packaged air
conditioning system or a fan-coil air conditioning system should be selected.
 Air system, water system, central plant cooling and heating system, and control system level. For
a VAV packaged system, it must be determined whether the system is to be a single-zone VAV
air mixing system, a multizone VAV reheat air mixing system, or a fan-powered VAV air mixing
system. In a packaged system, its cooling system is always a DX system with either reciprocating
or scroll compressors. Its heating system is either a direct-fired warm air heating system or, in locations
where the unit electric rate is low, a heat pump system or an electric heating system.
For a space air conditioning system, its air system is always a dedicated ventilation and space
recirculating system. For the water system, the choice is between a plant-building loop with
variable flow in the building loop and a plant-through-building loop. Either a hot water heating
system or an electric heating system can be selected for a fan-coil system. A fan-coil system
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.3
requiring large tonnage of refrigeration load often uses centrifugal or screw chillers, and a smalltonnage
fan-coil system usually uses screw and scroll chillers.
Except for room air conditioning systems and packaged systems of small size, usually an energy
management and control system (EMCS) with direct digital control (DDC) is cost-effective.
For a specific project, the control zones, control points, and selection of generic and specific control
functions should be optimum.
 Main system components level. For air systems this involves the selection of main components,
such as:
A forward-curved centrifugal, an airfoil centrifugal, and a vane-axial supply fan.
A supply-relief fan combination or a supply and return fan combination.
A CO2-based demand control ventilation, or a mixed plenum pressure ventilation control.
An air economizer or a water economizer, a thermal wheel, run-around coils, or a heat pipe air-toair
heat recovery system.
An adjustable frequency variable-speed drive, inlet vanes, or an inlet cone duct-static pressure control.
A mixing flow, stratified mixing flow, or displacement flow pattern and supply outlets and return
inlets layouts.
For water systems, the main system components level involve selection of
A plant-building loop or a plant-through-building loop
In a plant-building loop, whether balance valves are necessary
In a plant-through-building loop, whether a bypass throttling flow, distribution pumping, or variable
flow
An open expansion tank or a diaphragm expansion tank
For refrigeration systems, main system components level involves the following:
For heat rejection systems, whether an air-cooled, water-cooled, or evaporatively cooled condenser
is selected
For a centrifugal chiller, whether a two-stage or a three-stage impeller, and whether a variablespeed
drive or an inlet vane’s capacity modulation is selected
For reciprocating or scroll refrigeration systems, whether a variable-speed, two-speed, cylinder
unloader, or on /off capacity control is selected
Requirements Fulfilled during Selection
To properly, effectively, and energy-efficiently design an air conditioning (HVAC&R) system, the
designer, facility owner, or developer should collaborate to select the system according to the following
requirements:
Applications and Building Occupancies
When air conditioning systems are used for different applications or building occupancies, they
need different design criteria, operating hours, and different system characteristics. For example, a
constant-volume central system is always used for a class 10 clean room to fabricate semiconductor
wafer. When the clean room is in operation, adequate clean air must be provided to maintain unidirectional
flow to prevent the contamination of semiconductor wafers by submicrometer-size particulates.
A case study of a clean-space air conditioning system for a class 10 clean room is discussed
in Chap. 30.
26.4 CHAPTER TWENTY-SIX
A four-pipe fan-coil system is the most widely used air conditioning system for guest rooms
in luxury hotels. This is because a four-pipe fan-coil system can provide individual temperature
and fan speed controls as well as a positive supply of adequate outdoor ventilation air. The fan
coil can be turned off conveniently when the room is not occupied. It is isolated acoustically
from adjacent rooms. The most annoying space maintenance tasks, such as changing the filters
and periodic inspection and maintenance of the fan coil, can be done when the guest room is
unoccupied.
Specific design criteria are always related to applications. Specific design criteria usually dictate
the type of air conditioning system that should be selected. For instance, a high-precision constanttemperature
room maintained at a temperature of 70  0.2°F (21  0.1°C) within a limited working
space needs a constant-volume central system with electric terminal reheat. Electric reheat can
be modulated more precisely than any other type of heating system. A central system can maintain
a more uniform discharge air temperature at the AHU than DX coil packaged units. Usually, such a
high-precision constant-temperature room is a clean space. A constant-volume system is always
preferable to a VAV system for such rooms.
System Capacity
System capacity is closely related to the selection of an air conditioning system. For a single-story
small retail shop, a constant-volume packaged system is often chosen. If the conditioned space is a
large indoor stadium with a seating capacity of 70,000 spectators, a single-zone VAV central system
with many AHUs using peripheral nozzles at different levels to supply conditioned air is often
selected.
Indoor Air Quality
Indoor air quality (IAQ) depends mainly on the minimum ventilation control; the removal of bacteria,
particulates, irritating vapors, and toxic gases; preventing dampness; and proper maintenance of
the HVAC&R system. Among various minimum ventilation controls, the dedicated ventilation system
with CO2-based demand-controlled ventilation guarantees the minimum outdoor ventilation
rate specified by ASHRAE Standard 62-1999 and is energy-efficient.
The dedicated ventilation system, the mixed plenum pressure control, and the outdoor air injection
fan are good minimum ventilation control systems with satisfactory performance. All guarantee
the specified constant-volume flow of minimum ventilation rate at the outdoor air intake in the
AHU or PU.
The supply and return fan tracking systems, and the direct measurement of outdoor air intake in
the AHU and PU are considered as poor minimum ventilation control systems. As the volume flow
rate of the outdoor ventilation air is usually less than one-fourth that of the total supply air, it is
difficult to measure it correctly by measuring the volume flow rate of supply and return air. This
may result in an insufficient outdoor ventilation intake, especially during part-load operation in
VAV systems.
Air economizer cycle and purging operation prior to the occupied period on weekday mornings
both improve IAQ because of the introduction of all outdoor air to the conditioned space.
As discussed in Secs. 15.16 and 24.5, to remove particulates, bacteria, viruses, irritating vapors,
and toxic gases from the conditioned space, air filtration and gaseous adsorption and chemisorption
in the AHUs or PU can be rated into the following four grades:
 Low-efficiency air filters are used to remove dusts of size 3 to 10 mm, such as, spores, pollens,
and textile fibers. Panel filters made of wire mesh and synthetic and glass fibers are often used as
the filtration medium.
 Medium-efficiency filters are used to remove dusts of size 1 to 3 mm, such as, bacteria, automotive-
emissions, and fine dust for the protection of coils and air distribution systems and to prevent
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.5
the nutrition of microbial growth. Pleated filters with synthetic and glass fibers as the filter’s media,
are also used.
 High-efficiency filters are used to remove dusts of size 0.3 to 1 mm, such as, bacteria, cooking oil
fumes, and tobacco smoke. Bag or cartridge filters with synthetic and glass fibers of submicrometer
diameter are used as filter media.
 In ultrahigh-efficiency filtration, HEPA has a DOP efficiency of 99.97 percent, and ULPA filters
have a DOP efficiency of 99.999 percent. HEPA and ULPA filters are used to remove viruses, carbon
dust, combustion smoke, and radon progeny. They are widely used in clean rooms and clean
space. To remove objectionable odors, irritating vapors, and toxic gases of molecular size between
0.003 and 0.006 m, activated carbon adsorbers and chemisorbers are often used. When
high-efficiency filters, HEPA and ULPA filters, activated carbon adsorbers, and chemisorbers are
used, they must be protected by medium-efficiency filters to extend their service life.
Zone Thermal Control and Sound Problems
Zone thermal control comprises the temperature and humidity control of the control zone, which affects
directly the thermal comfort of the occupants in the conditioned space. In addition to the heating
and cooling capacities needed to meet the space load requirements, the scope of control and
control quality can be assessed in the following three areas:
Scope. Today zone control is the basis of thermal individual control. Usually, a room or a conditioned
space with a floor area often between 100 and 1000 ft2 (10 and 100 m2) is considered as a
controlled zone. For each control zone, there is often a sensor, a controller, and a corresponding terminal
device (a VAV box, a fan coil, or a water-source heat pump), to control the space temperature
and space humidity according to the requirements.
For a large open office, additions of cubical partitions, and use of supply outlets that can adjust
the volume flow or the flow direction to regulate are means to control the space temperature and humidity
that cover a smaller control zone.
Recently, a desktop task conditioning system has been developed from an underfloor upward
flow space air diffusion system which can provide zone control for an individual worker in the conditioned
space, as discussed in Sec. 18.9.
Volume Flow Control and Supply Temperature Control. For cooling, the zone control quality depends
on the volume flow control of cold supply air of a VAV box to meet the space sensible load reduction
at part-load operation or the reduction of the chilled water flow to a cooling coil. For heating, the
zone control quality depends on the capacity control of a reheating system or a perimeter heating system.
Modulating the volume flow of cold supply air during part-load operation only slightly reduces
the space relative humidity. However, reducing the chilled water flow to a cooling coil may signifi-
cantly increase the space relative humidity especially when the fan coil is oversized.
Control Modes. More often, the zone control mode is a DDC proportional-integral (PI) modulation
control without offset. It can be also a two- or three-fan speed control (small fan in fan coil or
in water-source heat pump) plus on /off control of the water flow. A DDC with PI or proportionalintegral-
derivative (PID) control always has a better control quality because it has no offset except
due to instrument inaccuracy.
Sound problems often are the result of improperly designed systems and poorly selected fans,
pump, and compressors. Noise always is very annoying. Some air conditioning systems such as individual
systems and space conditioning systems have inherent sound problems because the room
air conditioners and water-source heat pumps are often installed directly in the conditioned space.
Many fan-coil units are hung in the ceiling plenum directly above the conditioned space where the
recirculating air inlet provides a short and convenient sound transmission path. The estimated sound
levels in the occupied zone for various air conditioned systems are as follows:
26.6 CHAPTER TWENTY-SIX
Energy Efficiency
As discussed in Sec. 25.6, each type of air conditioning system has its own system characteristics
and energy use intensity (EUI), expressed in kBtu /h  ft2  °F (kWh/m2  °C), as shown in Table
25.1. Boilers, furnaces, compressors, fans, and pumps are energy users. To properly select an air
conditioning system, the designer should know the energy efficiency of main energy users as well
as the estimated EUI of the selected air conditioning system.
Clearly distinguish the type and configuration of energy users in an air conditioning system:
Heating equipment
Direct-fired warm air furnace
Hot water boilers, or condensing hot water boilers
Refrigeration compressors: reciprocating, scroll, screw, or centrifugal
Water pumps: centrifugal hot water pumps, centrifugal chilled water pumps, or centrifugal condenser
water pumps
Fans: forward-curved/ aerofoil centrifugal supply fans, centrifugal / plug / axial return fans, or axial
relief fans, air-cooled axial condenser fans, evaporatively-cooled centrifugal condenser fans,
or propeller cooling tower fans, and exhaust fans
Motors: high-efficiency motors
Estimate the combined efficiency of various energy users and the EUI of the selected system.
Consider energy conservation measures (ECMs), such as VAV systems, air or water economizers,
heat recovery devices, variable-speed drives, and occupancy sensors, for the selected air conditioning
system to improve its energy efficiency.
Fire Safety and Smoke Control
In case of a building fire in highrise buildings, as discussed in Sec. 22.12, the water flow indicator
of the automatic sprinkler system actuates smoke control systems and stairwell pressurization systems,
and the building automation system (BAS) coordinates various air systems to provide a
smokefree escape route through the emergency exits and stairwells to the outdoors, as follows:
 Stairwell pressurized systems are actuated.
 Smoke is exhausted to outdoors from the fire floor (smoke zone) through the smoke exhaust
system.
 Supply outdoor air to floor or floors above and below the fire floor or adjacent to the smoke zone
to pressurize them to prevent smoke contamination.
For air systems with air economizers, the exhausted volume flow from the fire floor and the volume
flow to pressurize floors or control zones adjacent to the smoke zone are approximately equal to the
supply volume flow rate of the air economizer cycle.
For air systems with dedicated outdoor ventilation systems, or with a local outdoor ventilation air
intake (such as air systems in the fan-coil, water-source heat pump systems, or room air conditioning
Condition Air conditioning system Equipment Estimate
Poor Room air conditioning system Room air conditioner 45 to 50 dBA
Acceptable Four-pipe fan-coil system Fan-coil unit (400 cfm, 189 L / s) 35 to 40 dBA
Good VAV reheat central system AHU, VAV-box 25 to 35 NC
Excellent Single-zone VAV central system AHU 15 to 20 NC
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.7
systems), a smoke exhaust system of about 6 ach (air changes per hour) should be installed; and the
dedicated outdoor ventilation system can be used to pressurize the adjacent control zones according
to local and federal fire code requirements.
Space Limitations
Space limitations specified by the architect or facility owner also influence the selection of the air
conditioning system. For instance, for a high-rise building of more than 30 stories, if rooftop space
is not available for the penthouse of AHUs and other mechanical equipment, or if there is no space
left for supply and return duct shafts, a floor-by-floor AHU central system may be the practical
choice. Because of the available ceiling space, low-velocity duct systems are often used in highceiling
industrial buildings to save air-transporting energy.
Maintenance
A central system with AHUs, a few water-cooled centrifugal chillers, and cooling towers needs less
maintenance work than a packaged system with many rooftop air-cooled reciprocating packaged
units. A VAV reheat central system needs less maintenace work in the fan and plant rooms than a
fan-coil system, which often requires much maintenance work in the ceiling space directly above
the conditioned space.
Initial Costs
Initial cost and operating costs (mainly energy cost) are always primary factors that affect the selection
of an air conditioning system, especialy for a developer who sells buildings with installed air
conditioning systems. The initial cost of the air conditioning (HVAC&R) system in a building, expressed
in $ / ft2 ($ /m2), depends on the building occupancies, system configurations, size of the
building, and capabilities of specific systems. According to Gladstone and Humphreys (1995), the
initial costs of an HVAC&R system for school buildings are as follows:
Generally, the more complex an air conditioning system becomes and the more features it has, the higher
will be the initial cost. Among all these, the Class 1 clean room central system is the most expensive.
26.3 INDIVIDUAL AIR CONDITIONING SYSTEMS
Basics
As discussed in Sec. 1.3, an individual room air conditioning system or simply an individual system
uses a self-contained, factory-made packaged air conditioner to serve an individual room. It is ready
to use after electric cable and necessary water drainage are connected. Individual systems always
use a DX coil to cool the air directly. Individual systems can be subdivided into the two following
air conditioning systems:
System type $/ ft2
Single-zone, constant-volume packaged systems 5.25
Multizone reheat constant-volume central system 6.60
Multizone dual-duct central systems 8.25
26.8 CHAPTER TWENTY-SIX
 Room air conditioning (RAC) systems
 Packaged terminal air conditioning (PTAC) systems
As discussed in Sec. 1.3, individual air conditioning systems always use a DX coil for cooling. They
are normally used in perimeter zones for the convenience of local outdoor ventilation air intake.
Advantages and Disadvantages
Individual air conditioning systems have the following advantages:
 There are no supply, return, or exhaust ducts.
 Individual air conditioning systems are the most compact, flexible, and lower in initial cost than
others, except portable air conditioning units.
 Building space is saved for mechanical rooms and duct shafts.
 It is easier to match the requirements of an individual control zone.
 They are quick to install.
Individual systems have the following disadvantages:
 Temperature control is usually on /off, resulting in space temperature swing.
 Air filters are limited to coarse or low-efficiency filters.
 Local outdoor ventilation air intake is often affected by wind speed and wind direction.
 Noise level is not suitable for critical applications.
 More regular maintenance of coils and filters is required than for packaged and central systems.
Individual room and unitary packaged air conditioning systems are both self-contained, factorymade,
packaged systems using DX coils for cooling. However, individual systems are single-zone
units, whereas a packaged system can be either a single-zone or multizone unit. Individual systems
have no supply and return ducts. Also, individual systems provide far smaller heating-cooling capacities,
poorer outdoor ventilation intake, lower-efficiency filters, simpler control system, and far
less ECMs than packaged systems. Room air conditioners are often window-mounted. Most of the
packaged units are rooftop units.
As discussed in Sec. 1.5, individual air conditioning systems were found in about 26 percent of
66 million air conditioned homes and packaged systems in 74 percent in 1991 in the United States.
Also, individual systems were found in 22 percent of commercial buildings and packaged systems
in 48 percent in 1992 in the United States.
26.4 ROOM AIR CONDITIONING SYSTEMS
Equipment Used in Room Air Conditioning Systems
Room air conditioning systems are the most widely used individual air conditioning system in the
United States. Window-mounted or through-the-wall types of room air conditioners; room heat
pumps (RHPs) to provide both heating in winter and cooling in summer; and split air conditioners
(SACs) with an outdoor condensing unit and an indoor air handler, for the sake of greater flexibility
in location of the air handler as well as isolation of compressor noise outdoors are three kinds of
equipment used in room air conditioning systems. A room heat pump has a similar configuration to a
room air conditioner except a four-way reversing valve is added to change from cooling mode to
heating mode, and vice versa. According to Mahoney (1996), annual shipments of window-mounted
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.9
and through-the-wall room air conditioners increased from 2.8 million units in 1991 to 4.3 million in
1995 in the United States.
The cooling capacity of currently available room air conditioners ranges between 5000 and
34,000 Btu /h (1.5 and 10 kW). Refrigerant HCFC-22 and alternative refrigerants HFC-407C and
HFC-410A are the refrigerants used now and after 2020. Electric resistance heating is often used in
room heat pumps to supplement the winter heating when the outside weather is cold. According to
Rosenquist (1999), the average service life of a room air conditioner is 12.5 years.
26.10 CHAPTER TWENTY-SIX
Rotary
compressor
Condensing
coil
Cooling
air
Outdoor
ventilation
air intake
Motor
Return air
Return air
Coarse
air filter
DX coil
(a)
(b)
Cooling
air
Outdoor
fan
Indoor
fan
Supply
air
FIGURE 26.1 A window-mounted room air conditioner: (a) Sectional view; (b) front view.
Configuration and Cooling Mode Operation
Figure 26.1 shows a window-mounted room air conditioner. The cabinet is divided into indoor and
outdoor compartments which are separated by insulated wall to reduce the heat transfer. The DX
coil and indoor fan are in the indoor compartment. The outdoor compartment contains the compressors,
condensers, outdoor fan, capillary tube, and fan motor. The fan motor often has a doubleended
shaft which drives both fans.
Return air from the conditioned space flows through a coarse air filter, is cooled and dehumidi-
fied in a DX coil, and then enters the inlet of the indoor fan. In a room air conditioner, the indoor
fan is a forward-curved centrifugal fan. The conditioned air is pressurized in the impeller and
forced through the air passage that leads to the supply grille. The conditioned air is then supplied to
the conditioned space to offset the space cooling load.
Outdoor air is extracted by the propeller fan (outdoor fan) and forced through the condensing
coils, in which hot gaseous refrigerant is condensed to liquid refrigerant. During condensation, condensing
heat is released to the outside through the cooling air. A portion of outdoor ventilation air is
extracted by the indoor fan and mixed with the return air. The opening of the outdoor ventilation air
intake is adjustable.
According to ASHRAE Handbook 1996, HVAC Systems and Equipment, RAC is rated under the
following standard conditions:
Evaporating temperature 45°F (7.2°C)
Compressor suction temperature 55°F (12.8°C)
Condensing temperature 130°F (54.4°C)
Liquid temperature 115°F (46.1°C)
Ambient temperature 95°F (35°C)
The schematic diagram and refrigeration cycle of a room heat pump system are shown in Fig. 12.1.
The heating and cooling modes operation are similar to those in a rooftop heat pump packaged system,
discussed in Sec. 12.2 and shown in Fig. 12.3.
Energy Performance and Energy Use Intensities
According to Rosenquist (1999), the room air conditioners built in the early 1980s often had an
EER of 7.3 and greater. In 1999, the maximum rotary compressor efficiencies range from 10.7 to
11.1 EER. Many RAC manufacturers are using slit-type aluminum fins and grooved or rifled copper
refrigerant tubing in the heat exchange coils. Incorporating a large surface area to capacity ratio, it
is able to yield rotary compressor efficiencies of 11.1 to 11.3 EER (3.25 to 3.31 W/W).
Permanent split capacitor (PSC) motors are the predominant fan motor used in RACs. PSC motors
have an efficiency between 55 and 70 percent. Some RAC fan motors use shaded pole motors
with an efficiency of 30 to 40 percent.
ASHRAE/IESNA Standard 90.1-1999 specifies the minimum efficiency requirement for room air
conditioners with louvered sides and room air conditioner heat pumps with louvered sides as follows:
Efficiency as of
Equipment type Capacity Qrc Btu/h Minimum efficiency 10/29/2001
Room air conditioner Qrc  6000 8.0 EER 9.7 EER
6000  Qrc  8000 8.5 EER 9.7 EER
8000  Qrc  14,000 9.0 EER 9.8 EER
14,000  Qrc  20,000 8.8 EER 9.7 EER
Qrc  20,000 8.2 EER 8.5 EER
Room air conditioner heat Qrc  20,000 8.5 EER 9.0 EER
pump with louvered sides Qrc 	 20,000 8.5 EER 8.5 EER
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.11
For other types of room air conditioners and heat pumps, refer to Standard 90.1-1999.
According to DOE/ EIA Commercial Buildings Consumption and Expenditures 1995, individual
systems using mainly room air conditioners in commercial buildings in 1995 in the United
States had an annual energy consumption of 43.2 kBtu/ ft2  yr (12.7 kWh/ ft2  yr or 137
kWh/m2  yr), which is only less than the EUI for central systems and single-zone constant-volume
packaged systems, as shown in Table 25.1. Higher energy use in RAC systems is mainly due to the
lower efficiency of the compressor and fan and the motor and drive combined efficiencies.
Controls
RACs usually use a two-stage thermostat to separate cooling mode and heating mode operations.
One of the RAC controls selects the operation mode. Another control adjusts the set point and cycles
the compressor on and off to maintain the required set point. The third one changes the fan
speed if required.
There is another alternative. The thermostat reduces the indoor fan speed when the space temperature
approaches the set point by using a two- or three-stage speed operation. If the space temperature
drops further, it cycles the compressor off. If the space temperature drops still further, the
indoor fan is finally cycled off.
For a room heat pump system, a two-stage thermostat often selects the heating or cooling mode
operation manually. Then it cycles the compressor and evaporator fan separately or simultaneously
to maintain a space temperature set point. Usually, there is a 2 to 5°F (1.1 to 2.8°C) difference between
heating and cooling set points for room air conditioning systems.
If room air conditioners use a two-speed (high-low) or three-speed (high-medium-low) indoor
fan and if the fan speed is regulated only a few times a day manually, then the supply volume flow
of this RAC is often steady within an hour, and its air system is constant-volume in nature. If the
thermostat in an RAC reduces the fan speed when the space temperature approaches its set point
automatically, the air system is actually a VAV system.
Features
Many RACs have the following features:
 Most RACs are designed to extract outdoor ventilation air and mix it with the recirculating air.
Many RACs also can exhaust a portion of the return air through the condenser fan.
 Many RAC condenser fans have a slinger-ring which splashs the condensate that has been collected
in the pan under the DX coil where it flows onto the condenser. The water drops hit the
outside surface of the air-cooled condenser, evaporate, and increase the air-side heat-transfer coef-
ficient of the air-cooled condenser. However, in coastal areas, the corrosive effect of water with
dissolved salts must be considered.
 The supply velocity from most RACs ranges from 300 to 1000 fpm (1.5 to 5 m / s). RACs have
adjustable vanes mounted on the supply outlet to alter the direction of the air jet as well as its
throw, to prevent drafts.
 Some RACs incorporate subcoolers to subcool the hot liquid refrigerant discharge from the condenser.
Subcoolers are often added after the condenser outlet and submerged in the condensate
before the capillary tube inlet. Subcoolers improve the refrigeration effect and thus the COP of
the RAC system.
System Characteristics
System characteristics define system capabilities, functions, and the differences between a specific system
and other systems. System characteristics of room air conditioning systems are listed in Table 26.1.
26.12 CHAPTER TWENTY-SIX
26.5 PACKAGED TERMINAL AIR CONDITIONING SYSTEMS
Equipment Used in Packaged Terminal Air Conditioning Systems
Packaged terminal air conditioners (PTACs) and packaged terminal heat pumps (PTHPs) are two
main types of equipment used in packaged terminal air conditioning systems. Packaged terminal
heat pumps add a four-way reversing valve and provide winter heating through the release of condensing
heat from the outdoor coil and supplementary electric heater.
Figure 26.2 shows a packaged terminal air conditioner. The main chassis contains an air conditioner
to provide heating and cooling. A cabinet sleeve is used to slide the chassis through the wall.
A front panel with a return air grille provides a neat appearance. There is also an exterior louver to
show a more attractive appearance on the outer building facade. Heating can be provided by heat
pumps, electric heating, gas heater, or hot water heating system, especially an existing hot water
system. According to ASHRAE Handbook 1996, all-electric PTACs are dominating the current market.
In the mid-1990s, PTACs with electric resistance had a share of 49 percent of packaged terminal
air conditioning systems, and PTHPs a share of 45 percent. PTACs are available with cooling
capacity of 6000 to 18,000 Btu / h (1.8 to 5.3 kW).
Heating and Cooling Mode Operation
In a typical PTHP, the return air enters from the return grille on the front panel. It then flows
through a coarse filter and an indoor coil (DX coil) in which it is cooled and dehumidified during
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.13
TABLE 26.1 System Characterisics for RAC, PTAC, and Their Heat Pumps
RAC, RHP, SAC PTAC, PTHP
Zone thermal and sound control
Control zone Single Single
Control methods Electric, two-stage thermostat or Electric, two-stage thermostat
DDC, HI-LO, or HI-MED-LO fan speed or DDC
Control modes On-off On-off
Heating-cooling mode changeover Manual Manual, automatic
Sound control 45 to 50 dBA 45 to 50 dBA
Indoor air quality (IAQ)
Minimum ventilation air control Constant-volume flow, affected Constant-volume flow, affected
by wind direction and speed by wind direction and speed
Filters Coarse or low-efficiency filters Coarse or low-efficiency filters
Air systems
Types Constant-volume or VAV air mixing Constant-volume air mixing
Indoor fan Forward-curved centrifugal Forward-curved centrifugal
Indoor fan (IF) total pressure 0.6 in. WC 0.6 in. WC
Combined IF-motor drive efficiency 25% 25%
Volume flow control HI-LO or HI-MED-LO fan speed Single-speed or HI-LO speed
Outdoor fan Propeller Propeller
Cooling systems
Refrigeration compressor Rotary Rotary and reciprocating
Refrigerants HCFC-22, HFC-407C, HFC-410A HCFC-22, HFC-407C, HFC-410A
Evaporator DX coil DX coil
Condenser Air-cooled Air-cooled
Refrigerant flow control Capillary tube, RHP with Capillary tube, PTHP with four-way
four-way reversing valve reversing valve
Heating systems
Type Heat pump, or electric heating Hot water, electric heating, or heat pump
Energy use, cooling 8.0 to 10.0 EER 8.5 to 10.0 EER
cooling mode operation. The conditioned air is then mixed with the outdoor ventilation air extracted
by the indoor fan and discharged through the adjustable supply grilles upward. A thermostat
senses the zone temperature and controls the refrigerant supplied to the DX coil in on /off mode to
maintain a preset zone temperature.
During heating mode operation, the indoor coil acts as a condenser. The return air is then heated
by the indoor coil to maintain the required zone temperature.
As in an RAC, spray the condensate on the outdoor coil (condensing coil) where it evaporates
and enhances the release of condensing heat. In coastal areas, if corrosive saltwater is found in the
condensate sump, consider installing a condensate draining system.
Minimum Efficiency Requirements
ASHRAE/IESNA Standard 90.1-1999 specifies that the minimum efficiency requirements for new
installed packaged terminal air conditioners and packaged terminal heat pumps in cooling mode operation
can be calculated as:
(26.1)
Efficiency as of October 29, 2001 for packaged terminal air conditioners:
(26.2) EER  12.5 	
0.213Qrc
1000
EER  10.0 	
0.16Qrc
1000
26.14 CHAPTER TWENTY-SIX
Exterior
louver
DXcoil
Cabinet
sleeve
Return
air grille
Supply
grille
Filter
Chassis
FIGURE 26.2 A packaged terminal air conditioner.
Efficiency as of October 29, 2001 for packaged terminal heat pumps:
(26.3)
For new installed packaged terminal heat pumps in heating mode operation, their minimum effi-
ciency requirements can be calculated as:
(26.4)
Efficiency as of October 29, 2001:
where Qrc  cooling capacity, Btu / h. For factory-labeled replacement of packaged terminal air
conditioners and heat pumps, only their required minimum EER or COP will be slighly lower as of
October 29, 2001. Refer to ASHRAE/IESNA Standard 90.1-1999 for details.
System Characteristics
System characteristics of PTACs and PTHPs are listed in Table 26.1. Compare the system characteristics
of RACs with PTACs, also RHPs with PTHPs; they are quite similar, except for the following
differences:
 RACs and RHPs are often window-mounted, whereas PTACs and PTHPs are mounted through
the wall.
 PTACs and PTHPs have a better appearance on the building facade outdoors as well as on the
front panel indoors.
 Some PTACs and PTHPs offer more control functions such as automatic changeover from cooling
mode operation to heating mode operation, and vice versa, evaporator freeze-up protection,
limited operation when zone temperature limits are exceeded, thermostat set points adjustments
during the unoccupied period, and fault detection and diagnostics.
Applications
According to Air Conditioning, Heating & Refrigeration News on April 15, 1996, PTAC shipments
in 1994 in the United States comprised 75,000 units and PTHPs 100,000 units. PTACs and PTHPs
are widely used in commercial buildings. They are mostly adopted in relatively small zones in the
perimeter zones of buildings such as hotels, motels, apartments, nursing homes, and offices.
REFERENCES
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
ASHRAE, ASHRAE/ IES Standard 90.1-1999 Energy Standard for Buildings Except New Low-Rise Residential
Buildings, Atlanta, GA, 1999.
DOE/ EIA, 1998 Nonresidential Building Energy Consumption Survey: Commercial Buildings Consumption
and Expenditure 1995, DOE/ EIA-0318 (95).
Gladstone, J., and Humphreys, K. K., Mechanical Estimating Guidebook, McGraw-Hill, New York, 1995.
COP  3.2 	
0.026Qrc
1000
COP  2.9 	
0.026Qrc
1000
EER  12.3 	
0.213Qrc
1000
AC SYSTEMS: SYSTEM CLASSIFICATION, SELECTION 26.15
Harold, R. G., Sound and Vibrations Consideration in Rooftop Installations, ASHRAE Transactions, 1991, Part
I, pp. 445–453.
Lindford, R. G., and Taylor, S. T., HVAC Systems: Central vs. Floor-by-Floor, Heating / Piping / Air Conditioning,
July 1989, pp. 43–57.
Mahoney, T. A., 5 Million Unitary Shipments May Become Industry Yardstick, Air Conditioning, Heating and
Refrigeration News, April 15, 1996.
McGreal, M. P., Engineered Smoke Control Systems, HPAC, no. 4, 1997, pp. 69–72.
Rosenquist, G.,Window-Type Room Air Conditioner, ASHRAE Journal, no. 1, 1999, pp. 31–36.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechnic, Hong Kong, 1986.
26.16 CHAPTER TWENTY-SIX
CHAPTER 27
AIR CONDITIONING
SYSTEMS: EVAPORATIVE
COOLING SYSTEMS AND
EVAPORATIVE COOLERS
27.1
27.1 EVAPORATIVE COOLING AND
EVAPORATIVE COOLING SYSTEMS 27.1
27.2 DIRECT EVAPORATIVE COOLING
AND DIRECT EVAPORATIVE
COOLERS 27.2
Direct Evaporative Cooling Process 27.2
Saturation Efficiency 27.2
Direct Evaporative Coolers 27.4
Operating Characteristics 27.6
27.3 INDIRECT EVAPORATIVE COOLING
AND INDIRECT EVAPORATIVE
COOLERS 27.6
Indirect Evaporative Cooling
Process 27.6
Indirect Evaporative Coolers 27.7
Heat-Transfer Process 27.7
Effectiveness 27.10
Operating Characteristics 27.11
Part-Load Operation and Control 27.12
27.4 INDIRECT-DIRECT TWO-STAGE
EVAPORATIVE COOLING
SYSTEMS 27.13
Indirect-Direct Two-Stage Evaporative
Cooler 27.13
Operating Characteristics Using Outdoor
Air as Cooled and Wet Air 27.15
Operating Characteristics Using Return
Air as Wet Air and Outdoor-Return Air
Mixture as Cooled Air 27.15
Energy Efficiency Ratio and Energy Use
Intensities 27.16
Case Study: A Two-Stage Evaporative
Cooling System in Nevada’s
College 27.16
System Characteristics 27.18
27.5 ADD-ON EVAPORATIVE
COOLERS 27.18
Add-on Indirect-Direct Evaporative Cooler
to a DX Packaged System 27.18
Tower Coil and Rotary Wheel
Combination 27.20
Tower, Plate-and-Frame Heat Exchanger,
and Coil Combination 27.22
Plate-and-Frame Heat Exchanger 27.23
27.6 DESIGN CONSIDERATIONS 27.24
Scope of Applications 27.24
Beware of Dampness, Sump
Maintenance, and Water Leakage 27.24
Selection of Summer Outdoor Design
Conditions 27.24
REFERENCES 27.26
27.1 EVAPORATIVE COOLING AND EVAPORATIVE COOLING
SYSTEMS
Evaporative cooling is an air conditioning process that uses the evaporation of liquid water to cool
an airstream directly or indirectly so that the final dry-bulb or dry- and wet-bulb temperatures of the
airstream being cooled are lower than those before undergoing the evaporative process.
An evaporative cooling air conditioning system, or simply evaporative cooling system, is an air
conditioning system in which more than 50 percent of the total cooling provided annually is evaporatively
cooled. It consists of mainly evaporative coolers, fans, pumps, filters, heat recovery devices,
heat exchangers, a mixing box, dampers, controls, and other components. An evaporative cooler is a
piece of equipment in which the evaporative cooling process proceeds. A central or a packaged air
conditioning system may be comprised of an add-on evaporative cooler which provides 50 percent
or less of the total cooling annually.
There are three types of evaporative cooling processes: (1) direct evaporative cooling, (2) indirect
evaporative cooling, and (3) indirect-direct evaporative cooling, as shown in Fig. 27.1.
27.2 DIRECT EVAPORATIVE COOLING AND DIRECT
EVAPORATIVE COOLERS
Direct Evaporative Cooling Process
In a direct evaporative cooling process, the airstream to be cooled comes directly into contact with
the wetted medium or water spray. The direct evaporative cooler in which the direct evaporative
cooling process proceeds is shown in Fig. 27.1a. Air enters the direct evaporative cooler at point 1
and leaves at point 2. The release of the latent heat of evaporation from the directly cooled
airstream lowers the airstream temperature; the airstream’s humidity ratio increases because of the
added water vapor.
Saturation Efficiency
Saturation efficiency is an index used to assess the performance of a direct evaporative cooler. As
discussed in Sec. 15.21, saturation efficiency sat is defined in Eq. (15.54) as
(27.1)
where Tae, Tal  temperature of air entering and leaving direct evaporative cooler, °F (°C)
thermodynamic wet-bulb temperature of entering air, °F (°C)
For a direct evaporative cooler, if Tae and sat are known, Tal can be determined from a psychrometric
chart. The value of sat depends on the following factors:
1. Face velocity va of the air flowing through the direct evaporative cooler, in fpm [m/ (60 s)].
For a specific cooler (water dipping or water spraying) with fixed face area Aa, in ft2 (m2), and a
given water flow rate , in gpm (L / s), a higher va yields the following results:
 Higher volume flow rate of the cooled air, in cfm (m3/s)
 Greater evaporative cooling effect Qev,c, in Btu / h (W), which can be calculated as follows:
Qev,c  60vaAa acpa(Tae  Tal) (27.2)
where a  air density, lb/ ft3 (kg /m3)
cpa  specific heat of moist air, Btu / lb °F (J /kg°C)
 Lower saturation efficiency sat, mainly due to a smaller water flow rate for each ft3 (m3) of cooled
air.
For most direct evaporative coolers used for comfort air conditioning, the face velocity should
usually be no greater than 600 fpm (3.0 m / s) in order to prevent the carryover of water droplets.
Otherwise, a water eliminator should be installed, which significantly increases the air-side pressure
drop.
V?
a
V?
w
T*ae 
sat 
Tae  Tal
Tae  T*ae
27.2 CHAPTER TWENTY-SEVEN
AC SYSTEMS: EVAPORATIVE COOLING 27.3
FIGURE 27.1 Types of evaporative systems: (a) direct evaporative cooling; (b) indirect evaporative cooling; (c) indirect-
direct evaporative cooling.
2. Water-air ratio . This is the ratio of the mass flow rate of spraying water to the mass
flow rate of cooled air, both in lb / min (kg / min). A greater indicates a comparatively greater
contact area between air and water and, therefore, a higher sat.
3. Configuration of the wetted surface. Wetted media that provide a greater contact surface and a
longer contact time between the air and water yield higher value of sat.
When a direct evaporative cooler is used to supply cooled air to maintain a space temperature of
80°F (26.7°C) during the summer, it is important to realize that a higher sat always means a point
nearer to the saturation curve, point 2. For a predetermined space temperature Tr  80°F (26.7°C),
it means a higher space relative humidity 	r, as shown in Fig. 27.1a.
In a direct evaporative cooling process, recirculating water is usually used in order to save water,
which is often more economical. The temperature of the recirculating water always approaches the
wet-bulb temperature of the cooled air.
Because the air is sprayed or in contact with dipped water, direct evaporative cooling provides a
certain degree of air cleaning. However, if the cooled air contains a great deal of dirt or particulate
matter, an additional filter should be used to prevent clogging of the wetted medium or nozzles.
Other parameters should be considered to assess the performance of a direct evaporative cooler
including
 Use of freshwater or makeup water, usually expressed in gal /h per 1000 cfm (L /h per m3) of
cooled air
 Air-side pressure drop, in. WC (Pa)
Direct Evaporative Coolers
A stand-alone self-contained direct evaporative cooler that can provide cooled air for a conditioned
space independently consists of mainly the following: a wetted medium, a fan (which is usually a
centrifugal fan to provide the required total pressure loss and a lower noise level), dampers, a control
system, and a sump at the bottom. For water-spraying systems, a circulating pump and piping
connection are needed to distribute water evenly. To drip water on the medium from the top (except
for rotary evaporative coolers), air filters, dampers, and an outer casing are necessary. Provisions
should be made to bleed off the water in order to prevent mineral buildup.
Direct evaporative coolers can be categorized according to the characteristics of the wetted
medium as follows:
1. Air washers. An air washer, or water-spraying chamber, is itself a direct evaporative cooler.
The characteristics and performance of air washers are discussed in detail in Sec. 15.21.
2. Evaporative pads. These media are generally made of 2-in.- (50-mm-) thick aspen wood
fibers with necessary chemical treatment and additives to increase wettability and to prevent the
growth of microorganisms, as shown in Fig. 27.2a. Evaporative pads are mounted in removable galvanized
steel or plastic frames. Because evaporative pads require comparatively lower face velocities,
in a self-contained direct evaporative cooler integrated with a centrifugal fan, three sides of the
fan cabinet are often mounted with evaporative pads to increase the surface area.
3. Rigid media. These are sheets of rigid and corrugated material made from plastic, impregnated
cellulose, or fiberglass, as shown in Fig. 27.2b. Air and water typically flow in a crossflow
arrangement so that horizontal channels for airflow and vertical channels for water flow meet
between two corrugated sheets. The depth of the rigid medium is typically 12 in. (300 mm) in
the direction of airflow but may vary from 8 to 16 in. (200 to 400 mm). Rigid media need no
supporting frame. They have lower air pressure drops and can easily be cleaned by water flushing.
4. Rotary wheel. A wetted medium in the shape of a rotary wheel is made of corrosion-resistant
materials such as plastic, impregnated cellulose, fiberglass, or copper alloy, as shown in Fig. 27.2c.
m?w /m?a
m?w /m?a
27.4 CHAPTER TWENTY-SEVEN
The depth of the wheel along the airstream direction is from 6 to 10 in. (150 to 250 mm). The rotary
wheel is often driven by a motor and gearbox and rotates slowly at a speed of 1 to 2 r / min. The
bottom of the wheel is submerged in a water tank. Air flows through various channels of the
medium in the direction along the depth of the rotary wheel.
AC SYSTEMS: EVAPORATIVE COOLING 27.5
FIGURE 27.2 Wetted media for direct evaporative coolers: (a) evaporative pad; (b) rigid media; (c) rotary
wheel.
Operating Characteristics
Table 27.1 lists the operating characteristics of direct evaporative coolers with various types of
wetted media. The evaporative pad is the traditional type of wetted medium widely used in
residential and small commercial buildings. It has a low initial cost and is easy to operate and
maintain.
Rigid media such as impregnated cellulose need no support structure, do not emit debris, and
have a service life as long as that of aspen pads. They withstand a comparatively higher face velocity,
provide a lower air pressure drop, and have a slightly greater saturation efficiency than aspen
pads.
The rotary wheel has a more complicated structure. However, it has no water-recirculating
system. It is easier to connect in series with other refrigeration coolers or desiccant dryers as
an add-on direct evaporative cooler in an AHU or PU for energy-efficient and cost-effective
operation.
An air washer is a large-capacity, bulky, and expensive direct evaporative cooler. It is usually
used for both humidification and evaporative cooling in industrial applications. The saturation effi-
ciency values for direct evaporative coolers usually range from 0.75 to 0.95.
Among these direct evaporative coolers, only the cooled air discharged from the air washer is
often at an oversaturation state. It carries numerous minute liquid water droplets. Oversaturation
causes wetted surfaces and dampness along the evaporatively cooled air distribution passage. Wetted
debris are the nutrients for molds and other microorganisms. To prevent serious IAQ problems,
it is recommended that evaporative pad, rigid media, and rotary wheel be selected as the evaporative
wetted media in comfort air conditioning. If an air washer must be used in process air conditioning,
refer to Sec. 15.23 for necessary remedies.
27.3 INDIRECT EVAPORATIVE COOLING AND INDIRECT
EVAPORATIVE COOLERS
Indirect Evaporative Cooling Process
In an indirect evaporative cooling process, the primary airstream to be cooled is separated from a
wetted surface by a flat plate or a tube wall, as shown in Fig. 27.1b, and is called cooled air. The
cooled air does not directly contact the evaporating liquid. A secondary airstream flows over the
wetted surface so that the liquid water will evaporate and extract heat from the primary airstream
through the flat plate or tube wall. This wet secondary airstream is known as wet air. The purpose
of the secondary airstream is to cool the wetted surface, evaporatively approaching the wet-bulb
temperature, and to absorb the evaporated water vapor.
In an indirect evaporative process, the cooled airstream’s humidity ratio remains constant because
the air to be cooled does not contact the evaporating liquid. This process is represented by
horizontal line 1-2 on the psychrometric chart shown in Fig. 27.1b.
27.6 CHAPTER TWENTY-SEVEN
TABLE 27.1 Operating Characteristics of Direct Evaporative Coolers with Various Wetted Media
Saturation Air-side Water-air
Type of efficiency Face velocity, pressure ratio Water usage,
medium sat fpm drop, in. WC gal /h1000 cfm Remarks
Air washer 0.80–0.90 400–800 0.2–0.5 0.1–0.4
Evaporative pad 0.80 100–300 0.1 1.3 Pad thickness of 2 in.
Rigid media 0.75–0.95 200–400 0.05–0.1 Thickness of 8–12 in.
Rotary wheel 100–600 0.5
m?w / m?a
Indirect Evaporative Coolers
Figure 27.3 shows a typical indirect evaporative cooler made by an Australian manufacturer. The
main components of this cooler are a plate heat exchanger, a water spray and recirculating system,
an outdoor air intake with filters, a supply fan and exhaust fan connected by the same vertical shaft,
and a fiberglass or stainless-steel casing to prevent corrosion.
The core part of this indirect evaporative cooler is the plate heat exchanger. The plates are made
from dimpled, thin polyvinyl chloride plastic. These plates are spaced from 0.08 to 0.12 in. (2 to 3
mm) apart and form alternate horizontal and vertical passages (i.e., the air to be cooled flows horizontally,
and the air that is sprayed flows vertically). Because the plates are only about 0.01 in.
(0.25 mm) thick, the thermal resistance of each plastic plate is very small, although the thermal
conductivity of the plastic is low.
Hot, dry outdoor air at point o enters the intake and filters and is extracted by the supply fan.
It then enters the back of the exchanger and is forced through the horizontal passages, in which
it releases its heat through the plastic plates to the adjacent wetted surfaces of the vertical
passages. The cooled air at point s flows out the front to the conditioned space, as shown in
Fig. 27.3a.
Water sprays over the vertical passages at the top of the plate heat exchanger, and forms both
wetted surfaces and water droplets. Evaporation from these wetted surfaces and droplets absorbs
heat released from the air flowing horizontally through the plastic plates. Excess water drops to a
sump, which recirculates it to spraying nozzles by means of a pump. Makeup water is supplied
from the city water supply to account for the evaporation and carryover. Water is periodically bled
off to prevent the buildup of solid matter.
Return air from the conditioned space at point r is drawn through the vertical passages between
the plastic plates. It absorbs the evaporated water vapor, and its humidity ratio increases. The higher
the velocity of the wet air, the greater the wet surface heat-transfer coefficient hwet, the larger the enthalpy
difference 	hs,w between the saturated air film at the wetted surface and the wet airstream,
and the higher the pressure drop of the wet airstream. Wet air is then forced through the exhaust fan
and discharged to the outdoor atmospheric at point ex.
Other types of indirect evaporative coolers may use absorbent-lined vertical passages to drip water
from the top through distributing troughs instead of using water sprays. Propeller fans may be
used instead of centrifugal fans for wet air exhaust. Dampers may be used to extract outdoor air
from outdoors or return air from the conditioned space as the wet air depends on which of them has
a lower entering wet-bulb temperature.
Heat-Transfer Process
There are three fluid streams in a plate heat exchanger: cooled air, wet air, and water films along the
vertical passages. Because the temperature of the saturated air film above the wetted surface is
nearly equal to the wet-bulb temperature of the wet air flowing over the surface, the heat from the
airstream to be cooled on the other side of the plastic plate is transferred to the wetted surface to
evaporate liquid water. The heat transfer process in an indirect evaporative cooler takes place
mainly between the cooled and wet airstreams.
On the cooled air side, the amount of water vapor that permeates the plastic plate is very small
and can be ignored; therefore, the entering air is sensibly cooled from point o to point s at a constant
humidity ratio, i.e., along a horizontal line os on the psychrometric chart shown in
Fig. 27.3d. According to Dowdy and Karabash (1987), the heat-transfer coefficient hair on the
air side, in Btu / h ft2  °F (W/m2  °C), is a function hair  f (Reo
0.8 Pr0.33) and can therefore be
calculated as
(27.3) hair  0.023 ka
Dh ReD
0.8Pr0.3
AC SYSTEMS: EVAPORATIVE COOLING 27.7
27.8 CHAPTER TWENTY-SEVEN
FIGURE 27.3 A typical indirect evaporative cooler: (a) schematic diagram; (b) airstream flowing through the
passages; (c) heat transfer through plastic plates; (d ) process on psychrometric chart.
where ka  thermal conductivity of air, Btu /h ft °F (W/m2 °C)
Dh  hydraulic diameter of cooled air passage, ft (m)
In Eq. (27.3) the hydraulic diameter Dh is given as
(27.4)
where Aca  area of cooled air passage, ft2 (m2)
Pca  perimeter of cooled air passage, ft (m)
On the wet air side, water is sprayed onto the wet airstream; however, because of the heat transfer
from the cooled air through the plastic plate to the wet air, the saturation process is no longer
adiabatic. Return air from the conditioned space (which becomes the wet airstream) is humidified
from point r to ex, as shown in Fig. 27.3d. According to Wu and Yellot (1987), the relative humidity
of the air exhausted from the indirect evaporative cooler is about 95 percent, and the change in the
dry-bulb temperature is rather small. Consequently, there is an increase in wet-bulb temperature and
air enthalpy due to the increase in latent heat.
In the plate heat exchanger, the cooled air and the wet air are in a crossflow arrangement. The
temperature of the saturated air film on the wet air side depends on the wet-bulb temperature of the
local wet airstream, and the wet-bulb temperature of the wet air gradually increases during the humidifying
process. The increase in the wet-bulb temperature of the wet air can be determined from
its enthalpy increase 	hwet, in Btu / lb (J / kg), and can be calculated as
(27.5) 	hwet  hex  hr 
V?
ca cacpa(To  Ts )
V?
wet wet
Dh 
4Aca
Pca
AC SYSTEMS: EVAPORATIVE COOLING 27.9
50 60 70 80
80
90%
80%
60%
50%
40%
70
60
50
90 100
0.04
0.08
0.012
0.016
0.020
op
o
r rp1
s
sp
rp2
(d)
ex
FIGURE 27.3 (Continued)
where hex, hr  enthalpy of wet air at points ex and r, Btu / lb (J / kg)
To, Ts  temperature of cooled air at points o and s, °F (°C)
volume flow rate of cooled air and wet air, cfm (m3 /s)
ca, wet  cooled air and wet air density, lb/ ft3 (kg /m3)
cpa  specific heat of moist air, Btu / lb°F (J /kg°C)
The average temperature of the saturated film on the wet air side Ts,a, in °F (°C), is approximately
equal to the average water temperature in the water sump Tw,s, in °F (°C). According to actual
observation, Tw,s is about 3°F (1.7°C) higher than the wet-bulb temperature of the return air for this
indirect evaporative cooler. As with Eq. (10.21) in Sec. 10.8, the surface heat-transfer coefficient on
the wet air side hwet can be calculated as
(27.6)
In Eq. (27.6), hdry indicates the sensible heat-transfer coefficient from the wetted surface when it is
dry; it can be calculated as in Eq. (27.3):
(27.7)
where kwet  thermal conductivity of wet air, Btu /h ft °F (W/m°C)
Dh,w  hydraulic diameter of wet air passage, ft (m)
Lwet  length of wet air passage, ft (m)
Effectiveness
As the heat flow in an indirect evaporative cooler is often in a crossflow pattern, the performance of
an indirect evaporative cooler is mainly determined by its effectiveness. Usually, the cooled air has
a smaller heat capacity rate than the wet air. Therefore, the indirect evaporative cooler effectiveness
in is defined as follows:
(27.8)
where Tca,e, Tca,l  temperature of air to be cooled entering and leaving indirect evaporative cooler,
°F (°C)
wet-bulb temperature of wet air entering indirect evaporative cooler, °F (°C)
Peterson and Hunn (1992) and Peterson (1993), based on experiments in the performance of indirect
evaporative coolers, recommended that the effectiveness of an indirect evaporative cooler in
either outdoor air or mixing air applications be represented by the following relationships:
(27.9)
and from Eqs. (15.36) and (15.37)
(27.10)
(27.11) Cwet  60 V?wet wetcsat
Cca  60 V?ca cacpa
C 
Cmin
Cmax

Cca
Cwet
in 
1
1 Cmin / Cmax
T
wet,e 
in 
Tca,e  Tca,l
Tca,e  T
wet,e
hdry  0.023 kwet
Dh,w Re0.8Pr0.4
hwet   m
cpa hdry
V?
ca,V?wet 
27.10 CHAPTER TWENTY-SEVEN
where Cca, Cwet  heat capacity rate of cooled air and wet air, Btu /h°F (W/ °C)
volume flow rate of cooled air and wet air, cfm [m3 / (60 s)]
ca, wet  density of cooled air and wet air, lb/ ft3 (kg /m3)
cpa  specific heat of moist air, Btu / lb°F (J /kg°C)
In Peterson (1993), when cooled and wet airstreams are all outdoor air, the theoretically calculated
effectiveness in,c and the measured in based on tests are as follows:
Wang (1996) conducted tests on the surface wettability effect on heat exchanger plates. The base
test used 0.063-in.- (1.6-mm-) thick aluminum plates with various coatings 0.004 in. (0.1 mm)
thick. The cooled air was set at a volume flow rate of 26.5 cfm (0.75 m3 / s), and the ratio of wet air
to cooled air was set at 1.09. Also, the inlet temperature of spraying water was controlled at 77°F
(25°C). For a typical test (ANC plates), when the water-air ratio by mass increased from 0.067 (10
mL/ min) to 0.27 (40 mL /min), the mean value of effectiveness in of six types of coated aluminum
plates increased from 0.38 to 0.63 as follows:
As the water-air ratio exceeds 0.044, any further increase has an insignificant effect on effectiveness
in. In Eq. (27.11), csat indicates the saturation specific heat per degree of wet-bulb temperature of
the wet air at constant pressure, in Btu / lb °F (J /kg°C); as discussed in Sec. 10.8,
(27.12)
where hs  enthalpy difference along the saturation curve, Btu / lb (J / kg). In Peterson’s (1993)
tests, csat varied between 0.825 and 0.899 Btu / lb°F (3.46 and 3.77 kJ /kg°C).
The total cooling capacity of the indirect evaporative cooler Qc, in Btu / h (W), can then be calculated
as
(27.13)
Operating Characteristics
Stand-alone, self-contained indirect evaporative coolers are made in sizes to handle volume flow
rates of 1060, 2600, and 3200 cfm (0.53, 1.3, and 1.6 m3 / s). The size of the indirect evaporative
cooler shown in Fig. 27.3 is 2600 cfm (1.3 m3 /s). Based on observation, its maximum power
Qc  60V?ca cacpain(Tca,e  T
wet,e)
m  csat   dhs
dT
 
Water spraying rate, lb/min (mL/ min) Water-air ratio Average in
0.022 (10) 0.011 0.38
0.044 (20) 0.022 0.57
0.088 (40) 0.044 0.63
0.11 (50) 0.055 0.63
Test Cmin /Cmax in,c in
1 0.295 0.77 0.75
2 0.322 0.76 0.76
3 0.363 0.73 0.73
4 0.411 0.71 0.71
5 0.467 0.68 0.68
6 0.543 0.65 0.66
7 0.680 0.60 0.60
8 0.868 0.54 0.55
V?
ca, V?wet 
AC SYSTEMS: EVAPORATIVE COOLING 27.11
consumption on hot summer days is 1.68 hp (1250 W), and the centrifugal fan’s total pressure is
from 1 to 1.3 in. WC (250 to 325 Pa).
In indirect evaporative coolers, either outside air or return air from the conditioned space can
be used as the air to be cooled or the wet air. It depends on which can provide better indirect evaporative
cooling results and indoor temperature and relative humidity. In Phoenix, Arizona, the
cooling design wet-bulb temperature corresponding to 1 percent annual cumulative frequency of
occurrence is 75°F (23.9°C). For a summer space temperature of 80°F (26.7°C) and a relative
humidity of 50 percent, the wet-bulb temperature of return air is only 66.5°F (19.2°C). However,
when the return air is used as the cooled air (primary airstream), a specified amount of outdoor
ventilation air must be mixed with return air, and also there is always a space latent load; therefore,
the space relative humidty of the space air wr using the return-outdoor mixture is always
significantly greater than that when only the outdoor air is used. Refer to the discussion in the next
section for more details.
According to Peterson and Hunn (1992), a seasonal energy efficiency ratio (SEER) of 17.7
Btu/ Wh (5.2 COP) is expected from an indirect evaporative cooler in Dallas, Texas. This is 70 percent
higher than an air conditioner with refrigeration.
The operating characteristics of indirect evaporative coolers are affected by the flow rates and
the pressure drops on the cooled air and wet air sides. For a specific cooler, the greater the volume
flow, the greater the heat-transfer coefficients, the higher the air velocity flowing through the passages
in the plate heat exchanger, and the higher the pressure drop.
The air velocity of the cooled air flowing through the passages is usually from 400 to 1000 fpm
(2 to 5 m / s). It is important to limit the air velocity of the wet air in order to prevent carryover of
water droplets. The indirect evaporative cooler effectiveness in usually ranges from 0.60 at an airside
pressure drop of about 0.2 in. WC (50 Pa) to an effectiveness of 0.80 at a pressure drop of
about 1 in. WC (250 Pa).
Usually, the cooled air-side pressure drop of indirect evaporative coolers ranges from 0.2 to 1.5
in. WC (50 to 375 Pa) depending on the air velocities in the heat exchanger and in the distributing
duct. The wet air-side pressure drop varies from 0.5 to 1 in. WC (125 to 250 Pa). The volume flow
ratio of cooled air to wet air changes from 0.85 to 1.67. This ratio is the key parameter that affects
the heat capacity ratio C and, therefore, the indirect cooler’s effectiveness, as shown in Eq. (27.9).
The altitude of the unit’s location also has a significant effect on its air density and, therefore, its
performance.
Part-Load Operation and Control
For a constant-airflow unit, if the conditions of the outdoor air and the sensible heat ratio of the space
conditioning line both remain constant, when there is a reduction in the space cooling load, the space
conditioning line sr, as shown in Fig. 27.3d, tends to extend a shorter distance from point s. The
space temperature Trp1 drops, and space relative humidity 	rp1 is slightly higher at part-load operation,
as shown by point rp1 on the psychrometric chart. If the space cooling load remains the same
and the condition of outdoor air changes from point o to point op, with lower outdoor dry-bulb and
wet-bulb temperatures, then the supply temperature Tsp, space air temperature Trp2, and space relative
humidity 	rp2 will all be lower at part load.
For a small, stand-alone, self-contained indirect evaporative cooler, the fans automatically cycle
on and off by means of a control system according to the space temperature at part-load operation.
In large coolers, a multispeed fan motor can often be modulated at part-load operation when the
space temperature drops below a predetermined limit. Serial combination of two indirect evaporative
coolers has the same effect of increasing the water-air ratio as well as contact area and is usually
not cost-effective.
Example 27.1. An indirect evaporative cooler is used in a retail store in Denver, Colorado. The 1
percent summer design wet-bulb temperature for Denver is 63°F (17.2°C), and the coincident
dry-bulb temperature is 80°F (26.7°C). If the summer indoor design temperature for this retail store
27.12 CHAPTER TWENTY-SEVEN
is 77°F (25°C) and the supply temperature differential is 10°F (5.6 °C), calculate the effectiveness
required for this indirect evaporative cooler.
Solution If outdoor air is used as the cooled air, from Eq. (27.8),
A cooling effectiveness of 0.76 is required for this indirect evaporative cooler.
27.4 INDIRECT-DIRECT TWO-STAGE EVAPORATIVE
COOLING SYSTEMS
Indirect-Direct Two-Stage Evaporative Cooler
When the cooled air leaves an indirect evaporative cooler during a hot summer, its dry-bulb temperature
may be still above 70°F (21.1°C) with a relative humidity between 60 and 80 percent. It
is beneficial to add a direct evaporative cooler so that the temperature of the cooled air can be reduced
further with an increase in relative humidity. A higher relative humidity of cooled supply air
is often acceptable when the supply air absorbs the space sensible load and maintains a desirable
space relative humidity during hot summer. Compared with a single-stage indirect evaporative
cooler, an indirect-direct two-stage evaporative cooler significantly lowers the space temperature
it serves.
In an indirect-direct two-stage evaporative cooling system (TSECS), a direct evaporative cooler
is always connected in series after an indirect evaporative cooler to form an indirect-direct evaporative
cooler. If an indirect evaporative cooler is connected in series after a direct evaporative
cooler, such a direct-indirect evaporative cooler is actually a single-stage direct evaporative cooler
with a considerably higher discharge cooled air temperature than that of an indirect-direct cooler.
An indirect-direct two-stage cooler can provide cooled air with a dry-bulb temperature of 67°F
(19.4°C) and a relative humidity of about 95 percent when the entering outdoor air is at a dry-bulb
temperature of 93°F (33.9°C) and a wet-bulb of 70°F (21.1°C). The temperature of this cooled air
may increase to an indoor temperature of 78° F (25.5°C) with an indoor relative humidity of 66 percent
after absorbing the space cooling load during a hot summer. Indirect and direct evaporative
cooling processes connected in series are illustrated in Figs. 27.1c and 27.4a along with a psychrometric
chart.
For an indirect-direct two-stage cooler at part-load operation, one of the stages, either direct or
indirect, can be turned off based on the indoor and outdoor conditions. In an indirect-direct twostage
cooler, plate heat exchangers are widely used in the indirect cooler. On the wet air side, either
a water spray or a wetted surface treated with water-absorbent material is used. In the direct cooler,
a rigid medium with impregnated cellulose is often used.
Compared with central systems using a chilled water cooling coil or a DX packaged system using
a DX coil, an indirect-direct two-stage evaporative cooling system with or without a DX coil refrigeration
has the following advantages:
 Evaporative cooling replaces all the refrigeration power required and maintains an indoor thermal
environment within the ASHRAE summer comfort zone when the outdoor air condition is favorable.
 More outdoor ventilation air can be extracted so that IAQ will be improved.
 Better indoor relative humidity control is possible in winter heating when the indoor relative
humidity is very low.
in 
Tca,e  T
ca,l
Tca,e  T
wet,e

80  (77  10)
80  63
 0.76
AC SYSTEMS: EVAPORATIVE COOLING 27.13
27.14
20%
40%
50%
60%
80%
90%
80
70
60
78/66%
75/45%
93/70
93/60
50
40
100 90 80 70 60 50 40
0.020
0.016
0.012
0.008
0.004
100 90
20%
30%
40%
50%
60%
70%
80%
90%
80 70 60
13.0
13.5
14.0
50
50
60
70
80 90 100
0.004
0.008
0.012
0.016
(b)
(a)
w, lb/lb
w, lb/lb
ca, l
ca, l
ca, l ca, e
s
s
r
o
s
ss
r
r
o
o
FIGURE 27.4 Operating characteristics of an indirect-direct two-stage evaporative cooling system: (a) using outdoor
air as cooled air and wet air; (b) using 30 percent outdoor air and 70 percent recirculating air mixture as cooled air and
return air as wet air.
The disadvantages include more maintenace for evaporative coolers because of water sprays and
wetted media.
Operating Characteristics Using Outdoor Air as Cooled and Wet Air
According to the analysis in ASHRAE Handbook 1995, HVAC Applications, for an indirect-direct
evaporative cooler that uses all outdoor air for its cooled and wet air, if the effectiveness of its
indirect evporative cooler is taken as 60 percent, then the saturation efficiency of its direct evaporative
cooler is 90 percent. Also, its cooled air discharged from the indirect-direct evaporative
cooler can be controlled to maintain a space air condition within ASHRAE summer comfort
zones after absorbing the space cooling load at a temperature of 78°F (25.6°C) or 75°F (23.9°C),
with varying space relative humidity as the condition of outdoor air and the discharged cooled air
change. If the space conditioning line has a sensible heat ratio of 0.90, the relationships between
outdoor air state point o, discharge air points, and indoor space conditions point r are shown in
Fig. 27.4a.
In Fig. 27.4a upper diagram, when the outdoor air enters the indirect-direct two-stage evaporative
cooler at a 93°F (33.9°C) dry-bulb temperature and 70°F (21.1°C) wet-bulb temperature,
it is evaporatively cooled indirectly to point ca,l with a dry-bulb temperature of 78°F (25.6°C)
and a wet-bulb temperature of 66°F (18.9°C), and then evaporatively cooled directly to 66.5°F
(19.2°C) with a relative humidity of 95 percent. Cooled air absorbs the system heat gain and the
space cooling load, and is controlled to maintain a space temperature of 78°F (25.6°C) with
a relative humidity of about 66 percent. This outdoor air condition (93°F dry / 70°F wet or
33.9°C dry / 21.1 °C wet) is the upper limit of the space air condition within the ASHRAE summer
comfort zone that a properly designed indirect-direct two-stage evaporative cooling system
can provide.
In Fig. 27.4a lower diagram, when the outdoor temperature enters the two-stage evaporative
cooler at a dry-bulb value of 93°F (33.°C) and a 60°F (15.6°C) wet-bulb temperature, the cooled air
is discharged from the two-stage evaporative cooler at 54°F (12.2°C) dry-bulb and 52°F (11.1°C)
wet-bulb temperatures. The space condition can then be controlled and maintained at 75°F (23.9°C)
with a relative humidity of 45 percent.
When a two-stage evaporative cooling system uses all outdoor air to provide an indoor environment
within the ASHRAE comfort zones, the key parameter that affects the final indoor temperature
and relative humidity is the entering outdoor wet-bulb temperature To
.
Operating Characteristics Using Return Air as Wet Air and Outdoor-Return
Air Mixture as Cooled Air
Consider an indirect-direct two-stage evaporative cooling system using return air as the wet air.
Suppose the following conditions hold:
 The outdoor air dry-bulb temperature is 93°F (33.9°C), and the outdoor wet-bulb temperature is
70°F (21.1°C).
 Thirty percent of outdoor air is mixed with the return air, and the mixture is used as the cooled air.
 The effectiveness of the indirect evaporative cooler is 0.60, and the saturation efficiency of the
direct evaporative cooler is 90 percent.
 Space temperature is controlled and maintained at 78°F (25.6°C).
For a mixture of 30 percent outdoor air and 70 percent return air, the temperature of the mixture Tm,
in °F (°C), can be approximately calculated as
Tm  0.3(93) 0.7(78)  27.9 54.6  82.5°F (28.1°C)
AC SYSTEMS: EVAPORATIVE COOLING 27.15
Assume that the wet-bulb temperature of the space air is 72°F (22.2°C). From Eq. (27.8), the
cooled air temperature leaving the indirect cooler Tca,l, in °F (°C), can be calculated as
From Fig. 27.4b, draw line ca,l-ss from point ca,l that parallels the thermodynamic wet-bulb temperature
line. As the saturation efficiency is 0.9, point s, the condition of cooled air leaving the
direct evaporative cooler, can then be determined; Ts is 70.5°F (21.4°C). Draw line sr from point s
with a sensible heat ratio of 0.9. Point r, which represents the condition of the space air, intersects
line ro at r. The space air temperature is 78°F (25.6°C), space relative humidity is 75 percent, and
the space wet-bulb temperature is 72°F (22.2°C), which is approximately equal to the assumed
value.
From the above analysis:
 Because of the addition of the space latent load, the space humidity ratio and space relative
humidity at point r must be higher than the humidity ratio and the relative humidity of the outdoor-
return air mixture point ca,e in Fig. 27.4b. The space humidity ratio and relative humidity
are also higher when the outdoor-return air mixture is used instead of all outdoor air as shown in
Fig. 27.4a upper diagram.
 All outdoor air supplied to the conditioned space significantly improves the IAQ more than when
the outdoor-return air mixture is used.
Energy Efficiency Ratio and Energy Use Intensities
According to ASHRAE Handbook 1999, HVAC Applications, the energy efficiency ratio (EER) for
an indirect-direct, two-stage evaporative cooling system, including the energy use of fan and pumps
in indirect-direct two-stage coolers, ranges between 12.9 and 38.7. Cities like Austin, Texas, have
the lowest EER; and Denver, Colorado, and Albuquerque, New Mexico, both have the highest EER.
Compared with a conventional refrigeration system whose EER  10, an indirect-direct two-stage
evaporative system installed in the Los Angeles area only needs 35 percent of the energy use of a
conventional refrigeration system.
However, a stand-alone evaporative cooling system needs ducts, filters, circulating pumps, and
supply, recirculating, and exhaust fans. According to DOE/ EIA Commercial Buildings Consumption
and Expenditures 1995, the energy consumption of evaporative cooling systems in commercial
buildings in 1995 in the United States was 38.0 kBtu/ ft2 yr (11.1 kWh/ ft2 yr, or 119
kWh/m2  yr), as shown in Table 25.1, which is less than that of other air conditioning systems
except for heat pump systems.
Case Study: A Two-Stage Evaporative Cooling System in Nevada’s College
Scofield (1998) reported an indirect-direct two-stage evaporating cooling system with DX coil refrigeration
which was used in a new three-story building at Western Nevada Community College, at
an elevation of 5100 ft (1555 m), in Carson City, Nevada. This new classroom and laboratory complex
has a floor area of about 70,000 ft2 (6500 m2).
HVAC&R System. An indirect-direct two-stage evaporative cooling-VAV packaged system (ECVAVPS)
with DX coil and air-cooled condenser was compared with a fan-powered VAV central system
with water-side economizer. When only an indirect-direct two-stage evaporative cooler is
adopted without any refrigeration as the final stage of cooling, occupants have complained of
higher indoor relative humidity. The final decision was made to install an EC-VAVPS with heat
recovery by heat-pipe heat exchangers.
Tca,l  Tca,e  in(Tca,e  T 
 wet,e)  82.5  0.6(82.5  72)  76.2F (24.6C)
27.16 CHAPTER TWENTY-SEVEN
The 55°F (12.8°C) all outdoor cooled air from the two-stage evaporative cooler provides cooling
for 95 percent of the cooling hours. However, this cooled air temperature is again lowered to
50°F (10.0°C) by the DX coil so that the space relative humidity in summer can be maintained at
40 to 45 percent. This also reduces the supply volume flow as the air density is increased.
Winter Heating. During winter heating, the heat-pipe heat exchanger uses exhaust heat to raise
the outdoor air temperature from 0°F (17.8°C) up to 42°F (5.6°C). As the supply volume flow
rate is turned down 50 percent in winter, a DDC controller positions the outdoor and return themselves
to introduce 76 percent outdoor air and 24 percent recirculating air. A 10 percent difference
AC SYSTEMS: EVAPORATIVE COOLING 27.17
TABLE 27.2 System Characteristics of Indirect-Direct Two-Stage Evaporative Cooling Systems
System structure TSECS with or without DX refrigeration
Zone thermal and sound controls
Control zones Single zones or multiple zones
Control method VAV or CV
Control mode DDC, PI control mode, or on-off
Sound control NC 30 to 45
Indoor air quality
Minimum ventilation control All outdoor air, or 20% to 30% fixed amount of outdoor
ventilation air
Filters Medium efficiency filters (MER 9 to 12)
Air systems
Types Two-stage evaporative VAV packaged system
(VAV reheat)
Supply fan, types Centrifugal, unhoused or plug fans
Supply fan total pressure 3.0 to 4.0 in. WC
Combined supply fan and motor drive efficiency 0.5
Supply fan volume flow control Inlet vanes, variable-speed drives
Return / relief fan Axial fan, centrifugal, unhoused or plug fans
Return / relief fan total pressure 0.6 to 1.2 in. WC
Return fan volume flow control
Supply-relief fan combination Automatically balanced
Supply-return fan combination Space pressurization control
Heat recovery Heat pipe heat exchanger, 0.65 effectiveness
Operating modes Cooling, heating, nighttime setback, purging,
warm-up, optimum start and stop
Cooling systems
Indirect evaporative cooler Effectiveness 0.6
Direct evaporative cooler 4- to 12-in. rigid media, saturation efficiency 90%
Refrigeration systems
Compressors Scroll, or reciprocating
Energy use EER 10 to 12
Capacity control Multiple units, staging on /off
Evaporator DX coil
Condensers Air-cooled or evaporatively-cooled
Heating systems
Type Hot water reheat
Boiler Gas and electric boilers
Maintenance Sumps equipped with side-stream centrifugal slit
separators; periodical flush of sumps
in supply and return / exhaust volume flow is used to maintain a positive space pressure. When the
outdoor air temperature is raised above 10°F (12.2°C), the heat-pipe heat exchanger can furnish
100 percent outdoor air.
During the dry winter season another alternative is to overheat the outdoor air by using a hot water
heating coil. At the same time, use the 4-in.- (100-mm-) deep wetted media direct evaporative
cooler to add necessary humidification and still maintain the required 50°F (10°C) supply air temperature.
When the outdoor air temperature is between 30 and 40°F (1.1 and 4.4°C), it is expected
to maintain the indoor relative humidity at 30 percent.
Cooling Mode Performance. The operating parameters of this indirect-direct two-stage evaporative
cooling VAV packaged system with DX coil refrigeration are as follows:
 Outdoor air enters the indirect evaporative cooler at 96°F (35.6°C) dry-bulb and 62°F (16.7°C)
wet-bulb temperatures.
 It is evaporatively cooled indirectly with an effectiveness of 0.66 to a dry-bulb temperature of
73.5°F (23.0°C) and a wet-bulb temperature of 54.3°F (12.4°C).
 It is sensibly cooled in a DX coil by refrigeration to a dry-bulb temperature of 59.2°F (15.1°C)
and a wet-bulb temperature of 48.5° F (9.2°C).
 It is evaporatively cooled in a direct evaporative cooler to a dry-bulb temperature of 50°F
(10.0°C) and a wet-bulb temperature of 48.5°F (9.2°C) and is then supplied to the conditioned
space with a saturation efficiency of 90 percent.
 The cooled supply air absorbs the fan and duct heat, as well as the space cooling load, and finally
becomes the space air at a dry-bulb temperature of 75°F (23.9°C) and a wet-bulb temperature of
60°F (15.6°C) with a space relative humidity of 45 percent.
Energy and Initial Cost Saved. For a duty cycle of 7 a.m. to 7 p.m., 6 d / week, there is a 78 percent
reduction in annual ton-hour (kWh) consumption when an indirect-direct two-stage evaporative
cooling VAV packaged system with DX coil refrigeration is used compared with a conventional
VAV central system. The installation cost for the two-stage evaporative cooling VAV packaged system
is $857,183 and for the conventional VAV central system is $971,016.
System Characteristics
System characteristics of an indirect-direct two-stage evaporative cooling system (TSECS) are
listed in Table 27.2.
27.5 ADD-ON EVAPORATIVE COOLERS
Add-on Indirect-Direct Evaporative Cooler to a DX Packaged System
In locations where outdoor air has a higher wet-bulb, temperature, evaporative cooling alone cannot
provide the required cooling to maintain a desirable indoor environment. In many applications, an
add-on evaporative cooler to a packaged system, as shown in Fig. 27.5a either with a component
sequence: indirect cooler, DX coil, and direct cooler or with a sequence: indirect cooler, direct
cooler, and DX coil, is often more economical than a packaged system alone.
Anderson (1986) has compared a packaged system with an add-on indirect-direct cooler and a
packaged system using a DX coil only:
27.18 CHAPTER TWENTY-SEVEN
Outdoor conditions
Supply air
Space temperature
100°F (37.8°C) dry, 70°F (21.1°C) wet
57.5°F (14.2°C) dry, 56.5°F (13.6°C) wet
78°F (25.6°C)
AC SYSTEMS: EVAPORATIVE COOLING 27.19
50 percent
2 in. WC (500 Pa)
2.75 in. WC (688 Pa)
0.85 in WC (213 Pa)
0.6
0.7
0.9
9 Btu /Wh (COP 2.64)
15 percent
2.25
1.10 kW/ ton (COP 3.20)
1.79 kW/ ton (COP 1.96)
Space relative humidity
System pressure drop
DX coil only
Evaporative cooling and DX coil
Wet air
Fan efficiency
Indirect cooler effectiveness
Direct cooler saturation efficiency
EER refrigeration
Outdoor ventilation air
Ratio of installation cost of evaporative coolers
and DX coil to DX coil only
Energy use (air and refrigeration side)
Evaporative coolers and DX coil
DX coil only
FIGURE 27.5 Evaporative cooler and DX coil versus DX coil only: (a) Evaporative cooler and DX coil; (b) DX coil
only.
In the evaporative cooler and DX coil combination, return air is used for the wet air, and outdoor
air is used only for the cooled air. For cooling capacity control of such a combination, components
are usually energized in the following sequence: indirect cooler first, then direct cooler, and finally
the DX coil.
If the electric power rate is $0.05 /kWh, the number of payback hours for an indirect-direct
evaporative cooler added onto a DX coil packaged system is 7948; if the rate is $0.075 /kWh, the
number of payback hours is 5295; and if the rate is $0.1 /kWh, the number of payback hours is
3974 h.
Tower Coil and Rotary Wheel Combination
When a cooling tower is connected to a water cooling / heating coil as shown in Fig. 27.6a, the
tower coil becomes an indirect evaporative cooler. During the cooling season, the condenser water
from the tower is forced through the coil to cool the air flowing over it. Such a cooling coil is often
used as a precooling coil, because there may be another cooling coil downstream. In the heating
season, hot water from the condenser may flow through the coil and heat the air.
Using water from the cooling tower to cool the air by means of a precooling coil (to replace all
or part of the refrigeration) is often called a water-economizing process. The combination of a
cooling tower and the connected water cooling coil is called a water economizer and is shown in
Fig. 27.6a.
If the tower coil is in series with a direct evaporative cooler using a rotary wheel, the combination
is actually an indirect-direct cooler which can maintain space conditions in summer similar to
those achieved with refrigeration for areas where the outdoor wet-bulb temperature is below 65°C
(18.3°C), as discussed in Sec. 27.4. Water returns from the precooling coil, typically at a temperature
of 78°F (25.6°C), enters the cooling tower, and is evaporatively cooled to about 70°F (21.1°C).
Water is then drawn through the precooling coil, where it absorbs the heat from the air flowing over
the coil. It is then pumped back to the tower at a temperature of about 78°F (25.6°C) to be evaporatively
cooled again.
Outdoor air at a dry-bulb temperature of 100°F (37.8°C) and a wet-bulb temperature of 65°F
(18.3°C) is drawn through the precooling coil by the supply fan and is sensibly cooled to 75°F
(23.9°C). It then flows through a rotary wheel type of direct cooler and is evaporatively cooled,
typically to 57.5°F (14.2°C) dry-bulb and 56°F (13.6°C) wet-bulb temperatures. After that, air is
supplied to the conditioned space. Recirculating air may be used instead of outdoor air when it is
more beneficial.
In an effective tower coil, the approach of the cooling tower should be around 5°F (2.8°C). In
order to have such an approach, the cooling tower must be 60 percent larger than a tower with a
10°F (5.6°C) approach. The row depth and fin spacing of the precooling coil should be selected to
sensibly cool the outdoor air from 100°F (37.8°C) to 75°F (23.9°C) at an entering water temperature
of 70°F (21.1°C).
Field experience and tests have shown that a low face velocity through the wetted media of the
rotary wheel results in a saturation efficiency above 0.90 and prevents carryover. In these tests, the
air velocity was about 700 fpm (3.5 m / s) with a pressure drop of 0.25 in. WC (63 Pa). The rotary
wheel revolved at a rate of approximately 1.5 r / min.
In a system with an add-on evaporative cooler, the horsepower and number of operating hours of
the fan always far exceed those of the water pump. To save energy and reduce operating costs for a
large unit, a two-speed (or even three-speed) fan is often economical. The efficiency of the cooler
that consists of cooling tower, cooling coil, and rotary wheel depends largely on the conditions of
the fills in the tower, the inner surface of the cooling coil, and the wetted media in the rotary wheel.
If the wetted surfaces are clogged with dirt and scale, the efficiency will decrease proportionally to
the resulting drop in evaporation and airflow. Periodic bleedoff and other necessary water treatments
are essential for good performance.
Another type of spraying coil–rotary wheel combination is shown in Fig. 27.6b. When a spraying
coil is connected to a cooling coil, water that has been evaporatively cooled in the former can be
pumped to the cooling coil to absorb heat from the ambient air. Such a spraying coil-coil combination
is actually an indirect cooler. If this indirect cooler is connected with a rotary wheel or other
type of direct cooler, the resulting combination has a system performance similar to that of a tower-
27.20 CHAPTER TWENTY-SEVEN
AC SYSTEMS: EVAPORATIVE COOLING 27.21
FIGURE 27.6 (a) Tower coil and rotary wheel combination; (b) spraying coil-coil and rotary wheel combination; (c)
evaporative cooling process.
coil and rotary wheel combination. The heat transfer between two airstreams by means of a spraying
coil-coil combination is always higher than that of a coil-to-coil arrangement, which is commonly
located inside the outdoor air and exhaust airstreams or between makeup air and exhaust
airstreams, and is often called a runaround system.
Tower, Plate-and-Frame Heat Exchanger, and Coil Combination
When the outdoor wet-bulb temperature drops below a certain value (about 40°F, or 4.4°C), it is
possible to use the condenser water from the cooling tower to cool the air and thus replace all or
part of the chilled water from the refrigeration plant, as shown in Fig. 27.7. To prevent the dirt and
solid matter contained in the open-circuit tower condenser water from scaling the inner surface of
the chilled water coil, a plate-and-frame heat exchanger is often used so that the tower condenser
water does not enter the coil directly.
In Fig. 27.7, the condenser water from the tower at 45°F (7.2°C) enters the plate-and-frame heat
exchanger. It cools all or part of the chilled water returning from the cooling coil to 50°F (10.0°C).
Chilled water from the plate-and-frame heat exchanger is then mixed with the water from the centrifugal
chiller and enters the chilled water cooling coil at 45°F (7.2°C).
To cool the chilled water at the heat exchanger when the outdoor wet-bulb temperature is 40°F
(4.4°C), the approach of the cooling tower should be no greater than 5°F (2.8°C) and the temperature
difference at the heat exchanger between the condenser water from the tower and the chilled
27.22 CHAPTER TWENTY-SEVEN
FIGURE 27.7 Tower, plate-and-frame heat exchanger, and coil combination.
water from the coil also should not exceed 5°F (2.8°C). Cost analysis is required to determine the
optimum size of the cooling tower and heat exchanger and the flow rates of the condenser water and
chilled water.
Plate-and-Frame Heat Exchanger
A plate-and-frame heat exchanger, as shown in Fig. 27.8, is a liquid-to-liquid heat exchanger. It
consists of a number of corrugated metal plates that are usually made of stainless steel, titanium,
or aluminum-brass alloy. Corrugated plates and gaskets form alternate passages that separate two
different fluids. Gaskets are usually made of elastomers. Warm fluid flows downward on one side
of the corrugated plate, and cold fluid flows upward on the other side in a counterflow arrangement.
The corrugated plates are compressed together by a fixed-end frame and a movable-end
frame with clamping bolts. Proper selection of the gasket material and proper operating conditions
are important to prevent fluid leakage. The plates and movable-end frame are suspended
from an upper carrying bar and lower guide bar. Fluid connections are located in the fixed-end
frame.
The gap between two adjacent plates is rather small, usually ranging from 0.1 to 0.2 in. (2.5 to 5
mm). Because the plates are corrugated, a high degree of turbulence is produced, which results in a
high heat-transfer coefficient. The following equation may be used for the calculation of the heat
transfer:
(27.14)
where k  thermal conductivity of fluids, Btu/h ft °F (W/m°C)
Lg  spacing of gap, ft (m)
The water velocity is usually between 60 and 200 fpm (0.3 and 1 m / s). The higher the fluid velocity,
the higher the heat-transfer coefficient and the greater the pressure drop of the liquid flowing
through the heat exchanger. For a typical plate-and-frame heat exchanger, the overall heat-transfer
coefficient may be 740 Btu /h ft2 °F (4200 W/m2 °C) through the corrugated plate between the
h  0.2536  k
Lg ReL
0.65Pr0.4
AC SYSTEMS: EVAPORATIVE COOLING 27.23
FIGURE 27.8 Plate-and-frame heat exchanger.
warm and cold liquids at a pressure drop of about 14.7 psi (101 kPa). The maximum working pressure
for a high pressure model can be as high as 350 psig (2412 kPa).
The plate-and-frame heat exchanger has a high heat-transfer coefficient on the surfaces of its
corrugated plates, a smaller temperature difference between the fluids on the two sides of each
plate, and a compact size. It is easily dismantled for cleaning and routine maintenance.
27.7 DESIGN CONSIDERATIONS
Scope of Applications
In locations where the outdoor dry-bulb temperature To  95°F (35°C) and outdoor wet-bulb temperature
To
  62°F (16.7°C), a stand-alone direct-indirect two-stage evaporative cooling system
plus DX coil refrigeration can provide a comfortable indoor environment for occupants at a space
temperature of 75°F (23.9°C) and a relative humidity of 45 to 50 percent, using evaporating cooling
most of the time in the cooling season.
In locations where the outdoor dry-bulb temperature To  95°F (35°C) and outdoor wet-bulb temperature
To
  65°F (18.3°C), a stand-alone indirect-direct two-stage evaporating cooling system can
maintain a comfortable indoor environment for occupants at a space temperature of 75°F (23.9°C) and
a relative humidity of 55 to 65 percent. In locations where the outdoor dry-bulb temperature To 
95°F (35°C) and the outdoor wet-bulb temperature To
  70°F (21.1°C), a stand-alone indirect-direct
two-stage evaporating system can maintain an indoor environment within the summer comfort zones
for occupants at a space temperature of 78°F (25.6°C) and a relative humidity of 55 to 66 percent.
In locations where the outdoor wet-bulb temperature To
  72°F (22.2°C), an add-on evaporative
cooler or an add-on evaporative cooler plus DX coil refrigeration save energy and is often economical.
Life-cycle cost analysis is recommended, especially for small and medium-size systems or
systems with a low number of operating hours during the cooling season.
In addition to the comfort cooling in residential and commercial buildings, evaporative cooling
has been used in process cooling in industrial applications such as spot cooling, cooling of large
motors, cooling of gas turbines and generators, textile mills, cooling of wood and paper products,
laundries, animal barns and houses, and product storage cooling.
Evaporative cooling is best suited to applications where both summer cooling and winter humidification
are required.
Beware of Dampness, Sump Maintenance, and Water Leakage
In direct evaporative coolers, cooled air contacts directly with liquid water. Select rigid wetted
media instead of atomizing humidifiers in which water droplets may cause dampness or wetted
surfaces at a certain distance downstream of the atomizing humidifiers. Sumps must be investigated
and cleaned regularly. Water bleedoffs and necessary water treatments are essential to maintain a
clean and efficient evaporative cooling system. The quantity of bleedoff should be no greater than
the rate of evaporation. The water sump must be properly sealed so that water leaks are prevented.
These considerations should be taken into account during system design.
Selection of Summer Outdoor Design Conditions
During the design of an evaporating cooling system, outdoor conditions should be selected to calculate
the performance of the evaporative coolers; dry-bulb, wet-bulb,or dew point temperature; and
whether this corresponds to 0.4, 1, or 2.5 percent annual accumulative frequency of occurrence and
the mean coincident dry-bulb, wet-bulb, or dew point temperature.
27.24 CHAPTER TWENTY-SEVEN
Figure 27.9 shows the performance of an indirect-direct two-stage evaporating cooling VAV system
if 1 percent dry-bulb and mean coincident wet-bulb (1 percent DB /MWB) temperatures, 1 percent
wet-bulb and mean coincident dry-bulb (1 percent WB/ MDB) temperatures, and 1 percent
dew point and mean coincident dry-bulb (1 percent DP / MDB) temperatures of Denver, Colorado,
are selected.
From ASHRAE Handbook 1997, Fundamentals, in Denver the outdoor design condition based
on 1 percent wet-bulb temperature and the mean coincident dry-bulb temperature To

 has a dry-bulb
temperature of 80°F (26.7°C) and 63°F (17.2°C). In an indirect evaporative cooler, if its effectiveness
is 0.60, the temperature of cooled air discharged from the cooler is
Draw a line from point ca,l
 that parallels the thermodynamic wet-bulb temperature line. With the
saturation efficiency of the direct evaporating cooler sat  0.9, the cooled air leaves this direct
evaporating cooler with a dry-bulb temperature of 60.3°F (15.7°C) and a relative humidity of 95
percent.
If the space temperature is maintained at 75°F (23.9°C) and if the sensible heat ratio of the
space conditioning line is 0.9, the space relative humidity can be determined from the psychrometric
chart to be 60 percent. The supply temperature differential used to calculate the supply volume
flow rate is 	Ts  75  60.3  14.7°F (8.2°C).
Similarly, the performance of the indirect-direct two-stage evaporating cooling system using 1
percent dry-bulb temperature and the mean coincident wet-bulb temperature, also 1 percent dew
point temperature and the mean coincident dry-bulb temperature can be calculated as follows:
T
ca,l  80  0.6(80  63)  80  10.2  69.8F(21.0C)
AC SYSTEMS: EVAPORATIVE COOLING 27.25
100 90 80 70 60 50
50
60
70
100 90 80
0.04
0.08
0.012
0.016
w, lb/lb
r

ca,l

ca,l
ca,l
r
r
To

To, F
To
To
s

s
s
80%
60%
50%
40%
30%
20%
FIGURE 27.9 Comparison of 1 percent dry-bulb, 1 percent wet-bulb, and 1 percent dew point temperatures in Denver,
Colorado.
From these analyses:
 If 1 percent DB / MWB is adopted as the outdoor air design condition, the design supply differential
is 21.5°F (11.9°C). When outdoor air of 1 percent WB/MDB occurs, the volume flow rate
must increase 21.5 / 14.7  1.46, or 46 percent, to maintain a space temperature of 75°F (23.9°C).
Otherwise,the space temperature Tr may increase to 75 21.5  14.7  81.8°F (27.7°C).
 From Fig. 27.9, the temperature of supply cooled air is lower than others if 1 percent DB /MWB
is adopted as the outdoor air design condition. The 1 percent DB / MWB does not include all the
outdoor air at 1 percent WB/ MDB and 1 percent DP /MDB.
The annual cumulative frequency of occurrence is no longer 1 percent and is greater than 1 percent,
or 1.5 percent, or even more. If 1 percent WB/ MDB is adopted as the outdoor air design condition,
from Fig. 27.9, it includes more outdoor air at 1 percent DB /MWB and 1 percent DP/MDB
than 1 percent DB /MWB, and the annual cumulative frequency of occurrence will be less than 1
percent DB /MWB.
The performance of an evaporative cooling system is more closely related to the outdoor wetbulb
temperature To
 than to other outdoor parameters. The value of To
 not only determines the temperature
of cooled air leaving the direct cooler, but also has a decisive effect on the lowest possible
wetted surface temperature in an indirect cooler. Outdoor air at a high dry-bulb temperature To and
a low wet-bulb temperature To
 can easily be cooled in a direct cooler. Outdoor air with a high wetbulb
temperature often needs refrigeration.
For a stand-alone indirect-direct two-stage evaporating cooling system with or without DX
refrigeration that serves a more demanding project, a wet-bulb temperature corresponding to 0.4
percent annual cumulative frequency of occurrence and the mean coincident dry-bulb temperature
(0.4 percent WB/ MDB) are recommended for the design outdoor air condition.
For a stand-alone two-stage evaporative cooling system or an add-on evaporative cooler, a wetbulb
temperature corresponding to 1 percent annual cumulative frequency of occurrence and the
mean coincident dry-bulb temperature (1 percent WB/ MDB) are recommended for the design outdoor
air condition.
REFERENCES
Anderson,W. M., Three-Stage Evaporative Air Conditioning versus Conventional Mechanical Refrigeration,
ASHRAE Transactions, 1986, Part IB, pp. 358–370.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
Beaudin, D., Evaporative Cooling System for Remote Medical Center, ASHRAE Journal, no. 5, 1996, pp. 35–38.
Brown,W. K., Fundamental Concepts Integrating Evaporative Techniques in HVAC Systems, ASHRAE
Transactions, 1990, Part I, pp. 1227–1235.
Brown,W. K., Application of Evaporative Cooling to Large HVAC Systems, ASHRAE Transactions, 1996, Part
I, pp. 895–907.
Colvin, T. D., Office Tower Reduces Operating Costs with Two-Stage Evaporative Cooling Systems, ASHRAE
Journal, no. 3, 1995, pp. 23–24.
Description 1 percent DB /MWB 1 percent WB/MDB 1 percent DP /MDB
Supply cooled air temperature Ts, °F 53.5 60.3 56.5
Space relative humidity, percent 44 60 52
Supply air differential 	Ts, °F 21.5 14.7 18.7
27.26 CHAPTER TWENTY-SEVEN
DOE/ EIA, 1998 Nonresidential Building Energy Consumption Survey: Commercial Buildings Consumption
and Expenditures 1995, DOE/ EIA-0318 (95).
Dombroski, L., and W. I. Nelson, Two-Stage Evaporative Cooling, Heating / PipingAir / Conditioning, May
1984, pp. 87–92.
Dowdy, J. A., and N. S. Karabash, Experimental Determination of Heat and Mass Transfer Coefficients in Rigid
Impregnated Cellulose Evaporative Media, ASHRAE Transactions, 1987, Part II, pp. 382–395.
Dowdy, J. A., R. L. Reid, and E. T. Handy, Experimental Determination of Mass-Transfer Coefficients in Aspen
Pads, ASHRAE Transactions, 1986, Part II, pp. 60–70.
McClellan, C. H., Estimated Temperature Performance for Evaporative Cooling Systems in Five Locations in
the United States, ASHRAE Transactions, 1988, Part II, pp. 1071–1090.
McDonald, G. W., M. H. Turietta, and R. E. Foster, Modeling Evaporative Cooling Systems with DOE-2.1D,
ASHRAE Transactions, 1990, Part I, pp. 1236–1240.
Meyer, J. R., Evaporative Cooling for Energy Conservation, Heating / Piping / Air Conditioning, September
1983, pp. 111–118.
Mumma, S. A., C. Cheng, and F. Hamilton, A Design Procedure to Optimize the Selection of the Water-Side
Free Cooling Components, ASHRAE Transactions, 1990, Part I, pp. 1250–1254.
Peterson, J. L., An Effective Model for Indirect Evaporative Coolers, ASHRAE Transactions, 1993, Part II,
pp. 392–399.
Peterson, J. L., and B. D. Hunn, Experimental Performance of an Indirect Evaporative Cooler, ASHRAE
Transactions, 1992, Part II, pp. 15–23.
Scofield, C. M., and N. H. DesChamos, Indirect Evaporative Cooling Using Plate-Type Heat Exchangers,
ASHRAE Transactions, 1984, Part IB, pp. 148–153.
Scofield, M., Savings out of Thin Air, Eng. Systems, no. 9, 1998, pp. 98–125.
Sun, T. Y., Design Experience with Indirect Evaporative Cooling, Heating / Piping / Air Conditioning, January
1988, pp. 149–155.
Supple, R. G., and D. R. Broughton, Indirect Evaporative Cooling—Mechanical Cooling Design, ASHRAE
Transactions, 1985, Part IB, pp. 319–328.
Wang, T. A., and R. L. Reid, Surface Wettability Effect on an Indirect Evaporative Cooling System, ASHRAE
Transactions, 1996, Part I, pp. 427–433.
Watt, J. R., Nationwide Evaporative Cooling Is Here! ASHRAE Transactions, 1987, Part I, pp. 1237–1251.
Wu, H., Performance Monitoring of a Two-Stage Evaporative Cooler, ASHRAE Transactions, 1989, Part I,
pp. 718–725.
Wu, H., and J. I. Yellott, Investigation of a Plate-Type Indirect Evaporative Cooling System for Residences in
Hot and Arid Climates, ASHRAE Transactions, 1987, Part I, pp. 1252–1260.
Yellott, J. I., and J. Gamero, Indirect Evaporative Air Coolers for Hot, Dry Climates, ASHRAE Transactions,
1984, Part IB, pp. 139–147.
AC SYSTEMS: EVAPORATIVE COOLING 27.27
CHAPTER 28
AIR CONDITIONING
SYSTEMS: SPACE
CONDITIONING SYSTEMS
28.1
28.1 SPACE AIR CONDITIONING
SYSTEMS 28.1
Applications 28.1
Advantages and Disadvantages 28.2
Induction Systems 28.3
28.2 FAN-COIL SYSTEMS AND FAN-COIL
UNITS 28.3
System Description 28.3
Operating Characteristics 28.3
Fan-Coil Units 28.5
Volume Flow Rate 28.7
Cooling and Dehumidifying 28.8
Heating Capacity 28.9
Sound Power Level of Fan-Coil
Units 28.9
28.3 FOUR-PIPE FAN-COIL SYSTEMS 28.9
General Description 28.9
Dedicated Ventilation System 28.10
Space Recirculation Systems 28.11
Temperature of Chilled Water Supplied to
Coils 28.11
Exhaust Air to Balance Outdoor
Ventilation Air Intake 28.12
Part-Load Operation 28.13
Zone Temperature Control and Sequence
of Operations 28.13
System Characteristics 28.14
Calculation of Operational
Parameters 28.14
Applications 28.20
28.4 TWO-PIPE FAN-COIL SYSTEMS 28.20
Two-Pipe Systems 28.20
Nonchangeover Two-Pipe Systems 28.20
Changeover Two-Pipe Systems 28.23
System Characteristics 28.24
Applications 28.24
28.5 WATER-SOURCE HEAT PUMP
SYSTEMS 28.24
System Description 28.24
Operating Characteristics 28.24
Loop Temperatures 28.25
Water-Source Heat Pumps 28.26
Energy Performance and Energy Use
Intensity of WSHPs 28.27
Closed-Circuit Evaporative Water
Cooler 28.27
Water Heater 28.29
Storage Tanks 28.29
Air Systems and Maintenance 28.29
Controls 28.30
System Characteristics 28.31
Case Studies: Water-Source Heat Pump
Systems 28.31
Design Considerations 28.32
28.6 PANEL HEATING AND
COOLING 28.33
REFERENCES 28.33
28.1 SPACE AIR CONDITIONING SYSTEMS
Applications
As discussed in Sec. 1.5, space air conditioning systems, or simply space conditioning systems
are one of the four major air conditioning systems (individual, packaged, space, and central) that
comprised about 8 percent of total floor area of commercial buildings in 1992 in the United
States. According to the survey of the Census Bureau, the shipments of fan-coil units comprised
215,000 units, and the shipment of water-source heat pumps comprised 99,000 units in 1994 in
the United States. Space conditioning systems can be subdivided into the following air conditioning
systems:
 Four-pipe fan-coil air conditioning systems, or simply four-pipe fan-coil systems
 Two-pipe fan-coil air conditioning systems, or simply two-pipe fan-coil systems
 Water-source heat pump systems
Advantages and Disadvantages
Space conditioning systems have the following advantages compared to central and packaged systems:
 Both heating and cooling devices in the form of terminals, such as fan coils or water-source heat
pumps, are installed directly above or within the conditioned space, or very near to it. There is no
return duct in a space conditioning system except the large core water-source heat pumps. Only a
delivery outlet or a short recirculating duct supplies conditioned air to the perimeter zone.
 As discussed in Sec. 20.9, a separately dedicated outdoor ventilation system is always required to
provide ventilation air for occupants as well as the necessary exhaust air. At the same time, a
space recirculating system is used to condition the recirculating air for fan-coil and water-source
heat pump systems. The required amount of ventilation is always guaranteed, and a dedicated outdoor
air ventilation system facilitates the adoption of demand-controlled ventilation. The air economizer
cycle for free cooling is only limited to the volume flow of the outdoor ventilation air.
In space conditioning systems except the large core water-source heat pumps, the outdoor ventilation
air duct is probably the only main duct with comparatively less headroom required than in
the supply duct that crosses under the beams in the ceiling plenum.
 Both four-pipe fan coils and console water-source heat pumps provide individual zone control
with each fan coil or each console water-source heat pump serving an individual zone. There is
often an additional high-low speed or high-medium-low speed control on the volume flow of the
supply fan. The four-pipe fan coil or the water-source heat pump can be changed from cooling
mode to heating mode and vice versa automatically or manually, and it adjusts its capacity to
meet the variation of the zone load during part-load operation.
 There is less cross-contamination between rooms and control zones.
 According to Zaidi and Howell (1993), the energy use of a four-pipe fan-coil system was only 73
percent compared with a VAV reheat system, and the energy use of a water-source heat pump was
only 63 percent that of a VAV system. Anantapantula and Sauer (1994) found that the energy use
of a four-pipe fan-coil system compared to a VAV reheat system was even lower. Lower energy
use in fan-coil and water-source heat pump systems is due to the following:
First, there is far less energy required to transport the conditioned air and the recirculating air;
also, no energy is needed to transport the return air.
Second, a fan-coil system can transfer excess heat from the interior zone of a building to the
perimeter zone through a heat recovery system, as discussed in Sec. 13.3. A water-source heat
pump operated in heating mode can absorb the heat rejected by another water-source heat pump
that is operated in cooling mode and recovers it for heating in the perimeter zone.
Space conditioning systems have the following disadvantages:
 Only low-efficiency filters are used in fan coils and water-source heat pumps. Low efficiency air
filters are unable to remove particulates of size  3 m and are ineffective at proving an acceptable
IAQ for occupants as well as protecting the coils and components in the air distribution system
from dust and contaminants.
 Fan-coil systems reduce the chilled water flow rate to decrease their cooling capacity during part
load. Therefore, this results in a significantly higher zone relative humidity at part-load operation.
The zone relative humidity may increase to 70 percent at part load, especially when the fan coil is
oversized.
 Because fan coils and water source heat pumps are often mounted directly above or near to the
conditioned space, they may create a noise problem, as discussed in Sec. 26.2, which is always
28.2 CHAPTER TWENTY-EIGHT
annoying. A receiver may receive a sound pressure level between 35 and 40 dBA in a conditioned
space produced by a typical fan coil of 400 cfm (189 L / s).
 Fan coils and water-source heat pumps are scattered terminals. Both have moving parts, filters,
and condensate pans and require site maintenance directly above, near, or within the conditioned
space, which causes difficulties, inconvenience, and possible leaks of water and condensate.
In addition, because only low-efficiency or coarse filters are used in the fan coils and water-source
heat pumps, more cleaning work is required for the coils, ducts, and air distribution components.
 The volume flow rate of outdoor ventilation air supplied to the conditioned space is not sufficient
to pressurize the floor immediately above or below the fire floor, to prevent smoke contamination
in case of a building fire.
Induction Systems
In an induction system, recirculating air extracted through the coil is due to a negative pressure
formed adjacent to the high-velocity supply nozzle. An induction system sees a considerably
greater energy use than fan-coil systems and is not recommended in new and retrofit projects.
28.2 FAN-COIL SYSTEMS AND FAN-COIL UNITS
System Description
A four-pipe or a two-pipe fan-coil system includes boilers and chillers in the central plant, water
system supplying chilled or hot water to the fan coils, a space recirculating system using fan coils
to condition the space recirculating air, and a dedicated outdoor ventilation air system using an outdoor
air AHU to condition the outdoor air, as shown in Fig. 28.1. Both the space recirculating and
dedicated outdoor air systems have ducts, diffusers, inlets, controls, and accessories.
Outdoor air is often cooled and dehumidified, heated, or sometimes even humidified in a separate
outdoor air or makeup AHU. The outdoor air is then transported to the fan coils where outdoor
air is mixed with the recirculated air and conditioned. If the conditioned outdoor air is supplied directly
to the conditioned space via supply ducts in the dedicated ventilation system, then it is mixed
with the conditioned recirculating air from the fan coil of the space recirculating system in the
space.
Outdoor (primary) air mixing with recirculating air in the outdoor air AHU is not economical
because recirculating air must be transported back to the outdoor air AHU and conditioned there instead
of in the fan coil in or near the conditioned space. Normally, there are no return fans or return
ducts in four-pipe or two-pipe fan-coil systems.
Chilled water or hot water is supplied to the fan coil through a two-pipe or four-pipe water system.
In locations with a moderate winter, or where electric rates are favorable in winter, an electric
heating coil in each fan-coil unit is sometimes used for winter heating. This greatly simplifies the
operation of fan-coil systems.
Four-pipe and two-pipe fan-coil systems and water-source heat pump systems are the air conditioning
systems that use dedicated ventilation and space recirculating air systems.
Operating Characteristics
During summer cooling mode operation, outdoor air at point o in Fig. 28.1 is cooled and dehumidi-
fied, and filtered in the outdoor air AHU. The outdoor air leaves the cooling coil at point pc, flows
through the supply fan outlet at point pf in the dedicated ventilation system, and is transported by
the supply duct to the fan coil in the space recirculating systems or directly to the conditioned space
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.3
28.4
FIGURE 28.1 A four-pipe fan-coil system with outdoor air supplied to the mixing plenum of fan-coil unit:
(a) schematic diagram; (b) air conditioning cycle.
at point ps. At ps, the dedicated ventilation system is combined with the space recirculating system
and forms the air system of the four-pipe or two-pipe fan-coil system. The outdoor air can be mixed
with the recirculating air in three ways:
 Outdoor air at point ps is supplied to the mixing plenum of the fan coil. It is then mixed with the
recirculating air, which is extracted by the fan in the fan-coil unit through the filter and forms the
mixture m. Air pressure at the mixing plenum is slightly less than the space pressure, so that space
air can be extracted into the plenum. The mixture then flows through the fan and coil in the fancoil
unit, cools and dehumidifies at the coil, and leaves the coil at point fc. Air that has been conditioned
in the fan coil is supplied to the conditioned space to offset the space load. State point fc
may or may not be on line pc-ps, but will be near point pf.
 Outdoor air at point ps is supplied to the conditioned space directly. In this case, space air is
filtered, and cooled and dehumidified (or heated), in the fan coil. The conditioned space air discharged
from the fan-coil unit at point fc is mixed with the outdoor air in the space. It is assumed
that the mixture at point m offsets the space load. In such an arrangement, the recirculating
airstream will recirculate along its own circuit.
 Outdoor air at point ps is supplied just before the supply outlet of the fan-coil unit. It mixes with
the conditioned air from the fan coil at point fc. The psychrometric cycle of this arrangement is
approximately the same as that of outdoor air supplied directly to the conditioned space.
Outdoor air supplied to the mixing plenum or before the supply outlet of the fan-coil unit provides
a balanced distribution of outdoor air in the conditioned space. On the other hand, outdoor air
supplied directly to the conditioned space has a shorter outdoor air supply duct, and the conditioned
air is better distributed when one of the fan-coil units is turned off during capacity control at partload
operation, as multiple fan-coil units are installed in a large room.
The temperature increase from fan power heat gain in the fan-coil unit is about 0.5°F (0.3°C) for
a permanent-slit capacitor fan motor and about 0.8°F (0.5°C) for a shaded-pole fan motor. The supply
duct in the ceiling plenum after the fan-coil unit is usually very short, so such a duct heat gain
or loss is negligible.
During winter heating mode operation, in extremely cold weather, outdoor air may be preheated
to point ph by the heating coil in the outdoor air AHU. After absorbing the supply fan
power to point pf and releasing the outdoor air supply duct heat loss to point ps in the dedicated
ventilation system, the outdoor air is mixed with recirculating air extracted by the fan coil at
point r, and forms a mixture m, as shown by line ps-m-r in Fig. 28.1b. The mixture is heated
in the fan coil to point cf by the fan power heat gain and then to point s by the heat released
from the coil in the space recirculating system, and the mixture is supplied to the conditioned
space.
Outdoor air can also be mixed with heated air from the fan-coil unit when outdoor air is supplied
directly to the conditioned space. In locations with a mild winter, only outdoor air is heated in
the outdoor air AHU.
Fan-Coil Units
A fan-coil unit, or a fan coil, is a terminal unit installed directly inside the conditioned space or in
the ceiling plenum just above the conditioned space. A fan-coil unit includes a small motor driven
centrifugal fan or two small centrifugal fans connected in parallel, a finned coil, a filter, an outer
casing, and controls. Sometimes, a cooling coil and a heating coil may be connected in series along
the airflow, as shown in Fig. 28.2.
A fan-coil unit can be a horizontal unit installed inside the ceiling plenum (Fig. 28.2a and b) or
a vertical unit mounted on the floor under the windowsill (Fig. 28.2c) or a stack unit installed vertically
along the two sides of the window. Vertical and stack units are usually used to offset the cold
draft on the inner surface of window glass or on the external wall during cold weather. Cold draft
often flows downward along the glass because it is heavier than the surrounding air. Warm air
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.5
discharged from a fan coil during winter heating raises the inner surface temperature of window
glass. Horizontal and vertical fan-coil units are shown in Fig. 28.2. Fan-coil units are available in
standard sizes 02, 03, 04, 06, 08, 10, 12, 16, and 20. Size 02 means a nominal flow rate of 200 cfm
(0.1 m3 / s), 04 means 400 cfm (0.2 m3 / s), and so on.
Fan. Double-width, double-inlet, forward-curved centrifugal fans are usually used because of
their compact size and lower noise level. The fan wheels are usually made of aluminum or galvanized
steel with a diameter less than 10 in. (250 mm) in most cases. Fan housings are die-formed
with integral scrolls and inlets.
28.6 CHAPTER TWENTY-EIGHT
FIGURE 28.2 Fan-coil units: (a) horizontal unit; (b) horizontal unit with inner lined plenum; (c) vertical unit.
Fan Motor. Permanent-split capacitor (PSC) motors and shaded-pole (SP) motors are used. As in
Sec. 26.4, PSC motors have an efficiency of 55 to 70 percent which is considerably higher than that
of SP motors, which have 30 to 40 percent efficiency.
Two-speed high-low switches or three-speed high-medium-low switches are used to vary the fan
speed manually or automatically (by a DDC controller). Fan motors are generally protected by a
thermal overload protector. Periodic oiling of the bearing (twice per year) is required.
Coils. Coils are usually made from copper tubes and aluminum fins. Cooling/ heating coils usually
have two, three, or four rows of fins, depending on the coil’s cooling capacity and the sensible
heat ratio of the cooling and dehumidifying process SHRc. Two-row coils or three-row coils are
widely used. Four-row coils have a greater dehumidifying capacity.
Usually, there is only one coil for both heating and cooling. A separate electric heating coil is
sometimes used with two-stage step control in locations where the heating season is short or energy
rates are low in winter. Manual air vents are installed to prevent the formation of air pockets inside
the water circuit. A galvanized-steel pan with an insulating liner is often used to drain the condensate
during dehumidification and to prevent outer surface condensation on the fan-coil unit.
To reduce the air-side pressure loss, the face velocity of the air flowing through the coil is usually
from 200 fpm (1 m / s) to 300 fpm (1.5 m/ s).
Filters. Usually low-efficiency, low-pressure-drop permanent filters are used. They are easy to
clean and replace periodically. Sometimes disposable, low-efficiency fiberglass filters are used.
Casing. The external cabinet is usually made of 18-gauge (1.3-mm) galvanized-steel sheet with a
corrosion-resistant surface coating. The cabinet is lined with insulation to prevent outer surface
condensation.
Although there are fans, water coils, and filters in both fan-coil units and air-handling units, a
fan-coil unit is distinguished from an AHU by the following characteristics:
Volume Flow Rate
The volume flow rate of a fan-coil unit is affected by the following factors:
 The external pressure drop because of the filters, the supply duct, and the diffusers in the space recirculating
system
 The position of the fan switch, whether is high-low or high-medium-low
 The elevation of the fan coil above sea level
Fan coil AHU
Classification of A terminal unit Basic equipment in
equipment air system
Location Under window sill Fan room
or in ceiling plenum
Volume flow, cfm (m3 / s)  2000 (1) 1200–50,000 (0.6–25)
Fan total pressure,  0.6 (150)  6 (1500)
in. WC (Pa)
Sound power level, dB Lower Higher
Filter efficiency Low Medium or high
Coil row depth 2, 3, or 4 rows 2, 3, 4, 6, or 8 rows
Fin spacing Fixed Custom-made
Selection of unit Based on cooling capacity Based on volume flow rate
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.7
The nominal flow rate is the flow rate of a fan coil whose external pressure drop is at a specific value
when the fan switch is in the high position and the fan coil is at sea level. The higher the external pressure
drop, the lower the volume flow rate. If the external pressure drop increases from 0.06 to 0.3 in.
WC (15 to 75 Pa), the volume flow rate of the fan coil may decrease to 55 percent of its nominal value.
For a typical fan coil, the volume flow rate of fan switch position medium is about 80 percent of
the high value, and low is only 70 percent of the high value.
Cooling and Dehumidifying
Because the face velocity of a specific fan-coil unit at nominal volume flow rate is nearly the same
for various sizes, and because the outer surface area of a water cooling coil is divided into dry and
wet parts during cooling and dehumidifying, the coil capacity and the sensible heat ratio of the
cooling and dehumidifying process SHRc depend on the following factors:
 Dry- and wet-bulb temperatures of entering air
 Entering water temperature
 Water temperature rise in the coil
 The surface area and number of rows, including both pipe surface and fin area, of the coil
For a 04 fan-coil unit with an entering air temperature of 80°F (26.7°C) dry-bulb and 67°F
(19.4°C) wet-bulb, and an entering water temperature of 45°F (7.2°C), if the water temperature rise
is 10°F (5.6°C), then its cooling and dehumidifying capacity Qcc varies from 11 to 14 MBtu /h
(3.22 to 4.1 kW), and SHRc varies from 0.65 to 0.80.
If the volume flow rate of a fan-coil unit deviates from the nominal value because of a greater
external pressure drop or a higher altitude, its total cooling capacity should be corrected as
follows:
Qc,c  CpCaQc,r (28.1)
where Qc,c  corrected total cooling capacity, MBtu/ h (kW)
Qc,r  manufacturer’s catalog listed cooling capacity at a specific nominal volume flow rate,
MBtu/ h (kW)
Cp, Ca  total cooling capacity correction factor for excessive external pressure drop and high
altitude, respectively
The corrected sensible cooling capacity of the fan coil Qcs,c, in MBtu/ h (kW), is
Qcs,c  CpsCasqcs,r (28.2)
where Qcs,r  manufacturer’s catalog-listed sensible cooling capacity at nominal volume flow rate,
MBtu/ h (kW)
Cps, Cas  sensible cooling capacity correction factor for excessive external pressure and high
altitude, respectively
Correction factors Cp, Cps, Ca, and Cas can be found in the manufacturer’s catalog. If these data are
not available, the following values can then be used for Cp and Cps:
External pressure drop, in. WC (Pa) Cp Cps
0.06 (15) 1 1
0.1 (25) 0.96 0.96
0.2 (50) 0.84 0.83
0.25 (63) 0.76 0.73
28.8 CHAPTER TWENTY-EIGHT
If the sensible heat factor of the cooling and dehumidification process 0.5  SHRc  0.95, for each
1000 ft (328 m) higher than sea level up to an altitude of 10,000 ft (3280 m);
(28.3)
Heating Capacity
The heating capacity of the selected fan coil can be found in the manufacturer’s catalog. Usually,
for the same fan coil, a greater heating capacity can be provided at winter design conditions because
the temperature difference between the hot water and heated air is higher than that between
chilled water and cooled air.
Sound Power Level of Fan-Coil Units
Because the fan-coil unit is usually located inside the ceiling plenum or directly under the window
sill, the room effect and the short supply duct are often not sufficient to attenuate fan noise in the
fan-coil unit. Therefore, sound power level is often an important factor to consider during the selection
of a fan-coil unit.
For a typical fan-coil unit of size between 02 and 12, the sound power level Lp rating, measured
in a reverberant room according to ARI Standard 443-70 for an octave band with a middle frequency
of 1000 Hz, varies from 45.5 to 52 dB. For a fan-coil unit of size 04, because Lp  47.5 dB
for an octave band with a middle frequency of 1000 Hz, if the room effect is 7.5 dB, the space NC
level is about 40. The greater the size of the fan-coil unit, the higher the NC level. Refer to the manufacturer’s
catalog for details.
28.3 FOUR-PIPE FAN-COIL SYSTEMS
General Description
The water systems used for the fan coils can be classified into two categories: two-pipe systems and
four-pipe systems. The three-pipe system was discontinued because of the energy loss in the common
return pipe.
A four-pipe fan-coil system is equipped with two supply mains: a chilled water supply and a
hot water supply. There are also two return mains: a chilled water return and a hot water return, as
shown in Fig. 28.3. The finned coil may be a common coil, as shown in Fig. 28.3a, or two separate
coils, a cooling coil of two or three rows, and a heating coil of one row, as shown in Fig. 28.3b. In
a common coil, the DDC controller admits the chilled water and hot water in sequence to two
three-way valves and modulates their water flow at part-load operation to maintain a preset space
temperature. The chilled water stream never mixes with the hot water in these three-way valves.
When two separate coils are used, a DDC controller admits chilled water to the cooling coil and
hot water to the heating coil in sequence and modulates their water flow at part load. Hot water,
steam, or electric energy can be used as the heat source.
A drainage pipe for condensate in the fan-coil unit is needed as the air is often cooled and dehumidified
in the fan coil, especially where conditioned space may have hot and humid infiltrated outdoor
air.
A four-pipe system is more flexible, easy to operate because there are no troublesome
Cas  1 
0.03(altitude, ft)
1000
Ca  1 
0.01(altitude, ft)
1000
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.9
changeover problems, and lower in operating cost than a two-pipe system. On the other hand, it has
a higher initial cost; and if reverse return pipe is used, many pipes must be squeezed into the ceiling
plenum. Four-pipe fan-coil systems are most widely used.
Dedicated Ventilation System
The dedicated ventilaton system in a four-pipe fan-coil system must provide minimum outdoor air
, in cfm (L / s), specified for occupants according to ASHRAE Standard 62-1999, served by each
fan coil. The estimated cooling capacity Qcpn, in Btu / h (W), provided by the minimum outdoor air
to offset the space cooling load can be calculated as
(28.4) Qcpn  60V?p 	ps(hr  hps)
V?
p
28.10 CHAPTER TWENTY-EIGHT
FIGURE 28.3 Four-pipe fan-coil unit system: (a) common coil; (b) separate coil.
where 	ps  air density of outdoor air, lb/ ft3 (kg /m3)
hr, hps  enthalpy of space air and outdoor air, Btu / lb (J / kg)
The estimated sensible cooling capacity of the minimum outdoor air Qs,pn, in Btu / h (W), that offsets
the space sensible cooling load can be calculated as
(28.5)
where Tr, Ts  temperature of space air and outdoor air, °F (°C). The total cooling capacity Qcp and
the total sensible cooling capacity Qsp of the cooling coil in the outdoor air AHU, both in Btu / h, are
the sum of the cooling capacities required to condition the outdoor air supplied to each of the fan
coils served by this outdoor air AHU, and they can be calculated as
(28.6)
where n  number of fan coils served by the outdoor air AHU.
Space Recirculation Systems
Each fan-coil unit forms a space recirculation system. Its system total pressure varies between 0.06
and 0.6 in. WC (15 and 150 Pa).
Estimated Cooling Capacity. The size of a fan-coil unit is selected to meet the required cooling
capacity and sensible cooling capacity to offset the space cooling load Qrc and sensible cooling load
Qrs, both in Btu / h (W). As the conditioned outdoor air supplied to a specific fan coil or to the conditioned
space served by the fan coil offsets part of the space load, the estimated total cooling capacity
Qc,fc and sensible cooling capacity Qs,fc of a fan coil, both in Btu / h (W), for a specific control
zone or conditioned area it serves can be calculated as
(28.7)
where hr, hps  enthapy of zone air and conditioned outdoor supply air, Btu / lb (J / kg)
Tr, Tps  temperature of zone air and conditioned outdoor supply air, °F (°C)
 volume flow rate of outdoor (primary) air supplied to each fan coil, cfm [m3 / (60 s)]
Sensible Heat Ratios. When a four-pipe fan-coil system is operated under steady equilibrium, the
sensible heat ratio of the space conditioning line SHRs is usually approximately equal to the sensible
heat ratio of the cooling and dehumidifying process of the fan coil SHRc,f because the cooling
and dehumidifying process in the fan coil cf-fc and the space conditioning line fc-r form a cycle.
During summer design conditions, the temperature of chilled water supplied to the fan coils and
to the outdoor air AHU is usually the same, Tw,p  Tw,f, both in °F (°C). Also, the humidity ratios of
the conditioned outdoor air at point ps and the air leaving the fan-coil point fc are approximately
equal to each other, wps  wfc, both in lb / lb (kg / kg). For fan-coil units serving the interior zone,
more often sensible ratios SHRs  SHRc,f  0.8. For fan-coil units serving a perimeter zone,
SHRs  SHRc,f  0.95, as shown in Fig. 28.1b.
Temperature of Chilled Water Supplied to Coils
In a four-pipe fan-coil system, the space or zone temperature is a specified design criterion within
the ASHRAE comfort zones. In summer, it is usually between 75 and 78°F (23.9 and 25.6°C). The
humidity ratios of the supply air from the outdoor air AHU wps and the supply air from the fan coil
wfc, as well as the humidity ratio of the space air wr depend on the temperature of the chilled water
V?
p
Qs,fc  Qrs  Qsp  60V?p 	pscpa(Tr  Tps)
Qc,fc  Qrc  Qc,p  60V?p 	ps(hr  hps)
Qcp  n
1
Qcpn Qsp  n
1
Qspn
Qspn  60V?p 	pscpa(Tr  Ts)
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.11
supplied to the fan coils Tw,f and to the outdoor air AHU Tw,p. Also Tw,f and Tw,p have a direct influence
on the cooling capacity of the fan coils and the outdoor air AHU.
If Tw,p  Tw,f, then wps  wfc, and wps  wr. A considerable amount of both sensible and latent
load is undertaken by the outdoor air during cooling mode operation. The required cooling capacities
of the fan-coil units are considerably less than those of the space cooling load.
If Tw,p  Tw,f, wps  wfc, and wps  wr, there is a certain amount of sensible load and a comparatively
less amount of latent load undertaken by the outdoor air supplied from the outdoor air AHU
during cooling mode operation. The required sensible cooling capacity of the fan coils is smaller
than the space sensible cooling load.
In the early days, the chilled water temperature supplied to the fan coils Tw,f was deliberately
equal to or higher than the dew point of the entering air to the fan coil Tf,en, so that there was no
condensate in the fan coil. The coil was easier to maintain and clean. Only dehumidified outdoor air
was used to offset the space latent load. However, many four-pipe fan-coil systems are now designed
so that Tw,p  Tw,f , as shown in Fig. 28.1b, because of the following:
 In locations where outdoor climate is hot and humid, the infiltrated air most probably forms condensate
in the fan coil
 It is less expensive as well as more convenient to control and operate when Tw,p  Tw,f.
Exhaust Air to Balance Outdoor Ventilation Air Intake
First, as discussed in Sec. 20.3 and expressed in Eq. (20.8), at a steady state, the mass flow rate of
entering air must be equal to the mass flow rate of leaving air, and
(28.8)
where ,  mass flow rate of outdoor air intake and recirculating air, lb/ min (kg / min), and
. If the designed entering air is exceeds the leaving air, the space will be pressurized
to a higher space pressure, a portion of space air is then exfiltrated, and a new balance is formed
with an additional term  .
Second, for a typical office building, the required outdoor ventilation air intake is 0.14
cfm/ft2 (0.70 L / s 	m2) of floor area. The exhaust in the rest rooms is about 0.03 cfm / ft (0.15
L/s	m2), which results in an excessive outdoor ventilation air of 0.14  0.03  0.11 cfm / ft2 (0.56
L/s	m2) of floor area. If there is no exhaust system to exhaust this excessive amount of outdoor air
to outdoors, then all the excessive outdoor air will be squeezed out through the leakage area on the
exterior walls at a pressurized space-outdoor pressure difference pro.
For an average leaky office building that has an exterior wall area of 10,000 ft2 (929 m2) and a
floor area of 20,000 ft2 (1858 m2), as discussed in Sec. 23.2, its effective leakage area is 2.08 ft2
(0.193 m2). From Eq. (23.4), the space and outdoor pressure differences to exfiltrate all the excessive
outdoor ventilation air are as follows:
Therefore, the following is recommended when a four-pipe fan coil system is used:
 For leaky buildings, outdoor ventilation air exhaust is not required.
 For average leaky buildings, the space pressurization due to excessive outdoor ventilation air should
be calculated and analyzed. If the pressurized space-outdoor pressure difference pro  0.03 in. WC
(7.5 Pa), outdoor ventilation exhaust due to excessive outdoor ventilation air is not recommended.
Ae,l, ft2 (m2) pro, in. WC (Pa)
Average leaky 2.08 (0.193) 0.07 (17)
Leaky 4.16 (0.387) 0.017 (4.3)
V?
o,sys
m?inf
m?s  m?o 
 m?rec
m?rec m?o
m?o 
 m?rec 
 m?inf  m?rt 
 m?ex
28.12 CHAPTER TWENTY-EIGHT
Part-Load Operation
At cooling mode part-load operation as the zone sensible cooling load decreases, the zone temperature
decreases accordingly. In a four-pipe fan-coil system, when the zone temperature sensor senses
this temperature decrease, a DDC controller closes the two-way valve in a fan coil, reducing the
chilled water flow rate entering the coil. As the chilled water flow reduces, the temperature of
chilled water leaving the fan coil Twfl rises accordingly, which raises both the temperature and humidity
ratio of the conditioned (cooled and dehumidified) air leaving the fan coil Tfcp and wfcp as
well as the humidity ratio of the zone air wrp, as shown in Fig. 28.1b. As soon as the reduction in
the sensible cooling effect of the conditioned air due to the increase of Tfcp in Qs,fcp  60 	scpa
(Tr  Tfcp) is just equal to the reduced sensible zone cooling load, a new balance is formed and the
zone temperature is maintained at the indoor design temperture Tr.
At cooling mode part-load operation, the chilled water temperature entering the fan coil and the
cooling coil of the outdoor air AHU Twfe is better reset 5 to 7°F (2.8 to 3.9°C) higher, according to
either the outdoor air temperature or the system load. Chilled water temperature reset can save compression
energy and reduces the sensible cooling effect of the conditioned air. On the contrary, it
also increases the zone relative humidty. Analyses should be done to select an appropriate value.
During cooling mode part-load operation, reducing the chilled water flow rate entering a cooling
coil to match the reduction in zone or space load always causes a higher zone relative humidty. But
reducing the supply air volume flow rate to match the reduction in zone or space load in VAV systems
causes the zone relative humidity to remain approximately the same.
Zone Temperature Control and Sequence of Operations
The zone temperature control of a four-pipe fan-coil system includes a fan-coil unit FC-1, using a
separate cooling coil and a heating coil that serve a control zone rx1 in the perimeter zone, and the
corresponding outdoor air AHU which supplies the conditioned outdoor ventilation air to the fan coil
FC-1. The operating mode can be divided into cooling mode, deadband mode, and heating mode.
1. When the time-of-day clock signals the four-pipe fan-coil system in the off position, and the
outdoor air AHU and fan-coil unit serving control zone rx1 in the perimeter zone is shut off, the
two-way valves for both cooling and heating coils in the fan coil FC-1 should be closed. Also the
outdoor air damper and the supply fan in the outdoor air AHU (OAAHU) are shut off.
2. When the time-of-day clock signals the four-pipe fan-coil system to the on position, and the
perimeter zone temperature sensor senses a zone temperature Trx  75°F (23.9°C), both the outdoor
air AHU and fan coil FC-1 call for cooling. A DDC system controller then opens the two-way
valves, resets the discharge air temperature Tdis, and starts the supply fan motor of the OAAHU with
the outdoor damper closed. After a time delay, such as 30 s, the outdoor damper of the OAAHU is
fully opened.
3. A DDC terminal controller next opens the two-way valve of the cooling coil of the fan coil,
starts the fan motor and the small fan, and modulates the valve opening to maintain a zone temperature
of 75°F (23.9°C). The two-way valve of the heating coil remains closed.
4. During cooling mode part-load operation, the zone sensible cooling load reduces. When the
drop in zone temperature Trx1p is sensed by the temperature sensor of control zone rx1, a DDC terminal
controller reduces and modulates the opening of the two-way valve of the cooling coil in the
fan coil, to maintain a preset zone temperature of 75°F (23.9°C).
Because of the reduction in the chilled water flow in the cooling coil, the result is an increase in
the temperature of chilled water leaving the cooling coil as well as in the temperature and humidity
ratio of the cooled conditioned air leaving the coil, which results in a higher zone relative humidity
rp, as shown in Fig. 28.1b.
Usually, the sensible heat ratio of the space conditioning line (SHRsp) at part load is smaller than
that at summer design conditions. This further increases the zone relative humidity.
V?
fc
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.13
5. When the space cooling load is reduced to 50 percent of the design load, a DDC system controller
resets the chilled water temperature in linear proportion from 45 to 50°F (7.2 to 10.0°C) according
to a preset range of outdoor temperature To. Also, a DDC terminal controller reduces the
fan speed in the fan coil from high to low.
6. When the zone temperature drops within the range 72.0°F  Trx1p  75°F (22.2°C  Trx1p 
23.9°C), the fan coil FC-1 is operated in deadband mode. In deadband mode, the DDC terminal
controller actuates the following:
 Both two-way valves for cooling and heating coils in the fan-coil are closed.
 The fan is still operating at low speed to provide air movement and filtration.
7. When the zone temperature Trx1  72.0°F (22.2°C), the fan coil FC-1 is operating in heating
mode. In heating mode operation:
 The modulation of the hot water flow by the DDC terminal controller should be reverse-acting;
i.e., the higher the sensed zone temperature Trx1, the smaller the output; and the lower the sensed
Trx1, the greater the output.
 The DDC controller modulates the two-way valve opening of the heating coil to maintain a zone
temperature Trx1 of 72°F (22.2°C).
 The fan in the fan coil is operating at low speed.
8. In heating mode, the heating process in the fan coil is a horizontal line ms with a constant humidity
ratio (as shown in Fig. 28.1b). The discharge temperature of the OAAHU To,dis is often reset
between 80 and 100°F (26.7 and 37.8°C) according to outdoor temperature To through a DDC system
controller. The lower the outdoor temperature To, the higher To,dis. When the outdoor temperature
To drops to a preset value, the fan in the fan coil raises its speed to high speed.
During heating mode part-load operation, if the outdoor humidity ratio wo is higher than the
winter design conditions, the humidity ratio of the mixture of outdoor and recirculating air wm is
raised accordingly. The result is a higher zone relative humidity which is more comfortable to occupants
in winter
Systems Characteristics
System characteristics of four-pipe fan-coil (4PFC) systems are listed in Table 28.1.
Calculation of Operational Parameters
Consider a large office in an office building using a four-pipe fan-coil system. The outdoor ventilation
air from the outdoor-air AHU is supplied to the mixing plenum of the fan-coil unit. At summer
and winter design loads, the operating parameters for this four-pipe fan-coil system in this office
room are as follows:
Summer Winter
Indoor space temperature, °F 78 72
Space relative humidity, percent (wet-bulb temperature) 45 (63.5°F)
Space cooling or heating load, Btu /h 29,000 9500
Space sensible cooling load, Btu /h 26,200
Space latent load, Btu /h 2700
Part-load, operation, cooling load, Btu /h 15,900
Part-load operation, sensible cooling load, Btu /h 13,100
Outdoor temperature, °F 90 20
Outdoor relative humidity, percent 50 80
Outdoor air, cfm 200 200
28.14 CHAPTER TWENTY-EIGHT
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.15
TABLE 28.1 System Characteristics of Space Conditioning Systems
4PFC 2PFC WSHP
Zone thermal and sound controls
Control zone Multizone Multizone Multizone
Control methods DDC or electric DDC or electric DDC or electric
Control modes PI or on /off PI or on /off PI or on /off
Heating-cooling changeover Automatic Nonchangeover; Automatic
changeover, auto
matic or manual
Sound control 35–45 dBA 35–45 dBA 40–45 dBA
Indoor air quality
Minimum ventilation air control Guarantee minimum outdoor ventilation air supply
Filters: FC or WSHPs Low efficiency Low efficiency Low efficiency
Outdoor-air AHU Medium to high efficiency Medium to high efficiency
Air systems
Space recirculating system Fan coil Fan coil Water-source heat pump
Type of supply fan Forward-curved Forward-curved Forward-curved
centrifugal centrifugal centrifugal
Combined fan-motor-drive 25% 25% Console, 25%
efficiency Core, 30 to 40%
Volume flow control High-low or high-medium- High-low or high-medium- Console, multi speed
low fan speed low fan speed Core, constant speed
Space recirculating system total 0.06–0.6 0.06–0.6 Console, 0.06–0.6
pressure, in. WC Core,  3.0
Dedicated outdoor ventilation system Constant-volume or DCV Constant-volume or DCV Constant-volume or DCV
Type of supply fan Centrifugal Centrifugal Centrifugal
Combined fan-motor drive efficiency 45% 45% 45%
Volume flow control Constant /variable-speed Constant /variable-speed Constant /variable-speed
Outdoor ventilation system total 3–4 3–4 3–4
pressure, in. WC
Refrigeration systems
Type Centrifugal or screw Centrifugal or screw Reciprocating, rotary, or scroll
Refrigerants HCFC-123, HCFC-123, HCFC-22, HFC-407C,
HCFC-22,or HCFC-22,or or HFC-410A
HFC-134a HFC-134a
Evaporator Chiller Chiller DX coil
Condenser Water-cooled Water-cooled Water-cooled
Cooling tower Open-circuit Open-circuit Open-circuit, or
closed-circuit
Refrigerant flow control Orifice, float,or Orifice, float,or Capillary tube
expansion valve expansion valve
Minimum performance, 0.5–0.7 0.5–0.7 10–15 EER
compressor, kW/ ton
Heating system
Type Hot water or Hot water or WSHP
electric heating electric heating
Gas-fired boiler efficiency AFUE 80 to 93% AFUE 80 to 93% AFUE 80 to 93%
Maintenance More site maintenance More site maintenance More site maintenance
Fault detection and diagnostics Chiller Chiller Core WSHPs
1. Assume that the temperature of outdoor air supplied from the OAAHU Tps is 61°F (16.1°C) and
the relative humidity is 81 percent. From the psychrometric chart, hps  24.8 Btu / lb (57.7 kJ /kg). At
summer design conditions, the enthalpy of space air is 28.8 Btu / lb (67.0 kJ / kg), and outdoor air density
	ps  0.075 lb / ft3 (1.2 kg /m3); from Eq. (28.7) the estimated fan-coil cooling capacity is
Qc,fc  Qrc  60 p 	ps(hr  hps)
 29,000  60(200)(0.075)(28.8  24.8)  25,400 Btu /h (7442 W)
The estimated fan-coil sensible cooling capacity is
Qs,fc  Qrs  60 p 	pscpa(Tr  Tps)
 26,200  60(200)(0.075)(0.243)(78  61)  22,482 Btu /h (6587 W)
If the chilled water supplied to the fan coils and the outdoor air AHU are both at 45°F (7.2°C) during
summer design conditions, for an entering air of 78°F (25.6°C) dry bulb and 63.5°F (17.5°C)
wet bulb temperatures and a chilled water temperature rise of 10°F (5.6°C), one manufacturer’s catalog
gives the following capacities for a size 03 fan-coil unit:
V?
V?
28.16 CHAPTER TWENTY-EIGHT
7.2 MBtu /h (2.1 kW)
6.3 MBtu /h (1.8 kW)
Total cooling capacity
Sensible cooling capacity
Because the external pressure drop in the supply duct after the fan coil is 0.06 in. WC (15 Pa) and
the fan coils are installed at sea level, no volume flow corrections are needed.
Four 03 fan-coil units are selected, so that the total cooling capacity is
Qc,fc  4  7200  28,800 Btu /h (8438 W)
which is greater than the estimated cooling capacity of 25,400 Btu / h (7442 W) at summer design
load. The sensible cooling capacity of fan coils is
Qs,fc  4  6300  25,200 Btu /h (7384 W)
which is also greater than the estimated sensible cooling capacity of 22,482 Btu / h at summer design
load.
2. Draw line r-fc from the conditioned zone point r with known SHRs  26,200 / 29,000  0.9,
and a 	fc  1/ 13.7  0.73 lb / ft3. The temperature of supply air leaving the fan-coil unit Tfc, in °F
(°C), can be calculated as
 78  26,200 /60  4  300  0.073  0.243  57.5°F (14.2°C) (28.9)
where corrected volume flow rate of fan coil, cfm (L / s)
	fc  air density at fan outlet in fan coil, lb / ft3 (kg /m3)
Point fc can then be determined. From the psychrometric chart, hfc  23.3 Btu / lb (54.2 kJ / kg).
3. Let Twp be the water temperature rise in the cooling coil of the outdoor air AHU. It can
be assumed that the temperature of air leaving the cooling coil of the outdoor air AHU Tpc,
in°F (°C), is
Tpc  Tw,p 
 Twp (28.10)
For Twp  10°F, Tpc  45 
 10  55°F (12.8°C).
V?
fc 
Tfc  Tr 
Qrs
60V?fc 	fccpa
Generally, the cooling coil may have six or eight rows. The relative humidity of air leaving the
cooling coil pc  98 percent, and the humidity ratio wpc can thus be determined. For Tpc  55°F
(12.8°C) and pc  98 percent, wpc  0.0092 lb / lb (0.0092 kg / kg) and hpc  23.2 lb / lb (54.0
kJ / kg).
For simplicity, the fan power heat gain for the OAAHU can be assumed as 2°F, and the duct heat
gain can be estimated at 4°F, so the supply temperature of the OAAHU Tps, in °F, is
Tps  Tpc 
 2 
 4  61°F (16.1°C) (28.11)
Because wps  wpc, state point ps can also be determined.
4. For a fan coil whose outdoor air is mixed with recirculating air in the mixing plenum of the
fan coil, draw line r-ps. From Fig. 28.1b,
(28.12)
Point m can thus be determined. In this case, Tm  75.0°F (23.9°C), and wm  0.0092 lb / lb
(0.0092 kg / kg). Because Tcf  Tm 
 0.5  75 
 0.5  75.5°F (24.2°C), and wcf  wm  0.0092
lb / lb (0.0092 kg / kg), point cf can also be determined. From the psychrometric chart, hcf  28.2
Btu/ lb (65.6 kJ / kg).
5. If 	cf  1/ 13.7  0.073 lb / ft3 (1.168 kg /m3), the cooling coil load Qc,fc and sensible cooling
coil load Qs,fc of the fan-coil unit, both in Btu / h, can be calculated as
Qc,fc  60 fc 	cf (hcf  hfc) (28.13)
 60  4  300  0.073(28.2  23.3)  25,754 Btu /h (7546 W)
Qs,fc  60 fc 	cfcpa(Tcf  Tfc) (28.14)
 60  4  300  0.073  0.243(75.5  57.5)  22,990 Btu /h (6736 W)
where hcf,hfc  enthalpy of air at fan outlet and supply outlet, as shown in Fig. 28.1a, Btu / lb (J / kg)
Tcf,Tfc  temperature of air at fan outlet and supply outlet, °F (°C)
Check the calculated cooling coil load and sensible load against the selected fan coil’s cooling
and sensible capacity. The calculated coil load Qc,fc of 25,754 Btu /h (7546 W) is very near to the
estimated capacity of 25,400 Btu / h (7442 W), and both are less than the selected capacity Qc,fc of
28,800 Btu /h (8438 W). Similarly, the calculated sensible coil load Qs,fc is 22,990 Btu /h (6736 W).
It is slightly greater than the estimated 22,482 Btu /h (6587 W) because of the fan power heat gain.
Both are less than the selected capacity Qs,fc of 24,000 Btu /h (7032 W).
6. Calculate the cooling coil load in the outdoor air AHU Qcp, in Btu / h (W). From the psychrometric
chart, To  90°F, o  50 percent, and ho  38.5 Btu / lb, so
Qcp  60 p 	ps(ho  hps)
 60  200  0.075(38.5  23.2)  13,770 Btu /h (4035 W) (28.15)
The cooling coil load due to the outdoor air in the OAAHU Qcp is about one-half of the fan coil’s
cooling coil load. This includes the cooling and dehumidifying load of outdoor air, the system heat
gain of the dedicated outdoor ventilation air system, and a portion of the zone cooling load.
7. During cooling mode part-load operation, the zone temperature is still maintained at 78°F
(25.6°C); that is, Trp  Tr. The fan in the fan coil can be switched from high speed to low speed
while chilled water is reset to a higher temperature. If the zone sensible load ratio is reduced to 50
percent of the full-load value and the volume flow rate of the fan at low speed is only 70 percent of
the high speed, then the temperature of air leaving the fan coil at part load Tfc,p, in °F (°C), can be
V?
V?
V?
m-r
ps-r

V?
p
V?
fc

200
4  300
 0.17
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.17
calculated as
(28.16)
where Qrsp  zone sensible cooling load at part load, Btu /h (W)
Trp  zone air temperature at part load, °F (°C)
At part-load operation, the relative humidity of supply air leaving the fan coil fc,p may vary between
80 and 90 percent. If fc,p  88 percent, point fcp can be plotted on the psychrometric chart
as shown in Fig. 28.1b. From the chart, hfcp  27.2 Btu / lb (63.3 kJ / kg).
8. During cooling mode part-load operation, the sensible heat ratio of space conditioning line
SHRs  13,100 / 15,900  0.82. Draw line fcp-rp from fcp with a sensible heat ratio of the space
conditioning line at part load SHRsp  0.82. Line fcp-rp intersects the zone temperature line Trp at
rp. Therefore, point rp can be determined. From the psychrometric chart, rp  57.5 percent.
9. When the space cooling load is reduced to 50 percent of design load, if the temperature of
chilled water supplied to fan coils Tw,f and that to the outdoor air AHU Tw,p are reset to 50°F (10°C)
at part-load operation, and the chilled water temperature increase in the cooling coil of the outdoorair
AHU is still 10°F, then the condition of supply air from the outdoor air AHU at part-load point
psp can be determined as in full-load operation, Tpsp  60 
 2 
 4  66°F (18.9°C), and wpsp 
0.0110 lb / lb (0.0110 kg / kg).
Draw line psp-rp. The condition of the mixture at part load, point mp, can be determined from
the following relationship:
(28.17)
From the psychrometric chart, Tmp  75.3°F (24.1°C), and wmp  0.0116 lb / lb (0.0116 kg / kg).
Because Tcfp  Tmp 
 0.5°F  75.3 
 0.5  75.8°F, and wcfp  wmp  0.0116 lb / lb, from the
psychrometric chart, hcfp  31.0 Btu / lb (72.1 kJ / kg).
10. Fan-coil cooling coil load at part load Qcfcp and sensible coil load at part load Qsfcp, both in
Btu/ h (W), can be calculated as
(28.18)
 60(4)(300)(0.7)(0.073)(31.0  27.2)  13,980 Btu /h (4096 W)
(28.19)
 60(4)(300)(0.7)(0.073)(0.243)(75.8  63.3)  11,176 Btu /h (3275 W)
11. At winter heating mode design load operation, if only two fan-coil units are operated and
hfg,o  1061 Btu / lb (2468 kJ / kg), the humidity ratio difference between zone air and the fan-coil
supply air is
(28.20)

2700
60  2  300  0.073  1061
 0.00097 lb / lb (0.00097 kg /kg)
wr  ws 
Qrl
60V?fc 	hfhfg,o
Qsfcp  60V?fc 	cf cpa(Tcfp  Tfcp)
Qcfcp  60V?fc 	cf (hcfp  hfcp)
rp-mp
rp-ps

V?
p
V?
fc

200
4  300  0.7
 0.24
 78 
0.5(26,200)
60  4  300  0.7  0.073  0.243
 63.3F(17.4C)
Tfc,p  Trp 
Qrsp
60  0.7V?fc 	cf cpa
28.18 CHAPTER TWENTY-EIGHT
where ws  humidity ratio of fan-coil supply air, lb/ lb (kg / kg). As shown in Fig. 28.1b, ws  wm
 wfc, and from the psychrometric chart, the humidity ratio of outdoor air wo  0.0017 lb / lb.
Therefore,
(28.21)
So
and ws  wm  wr  0.00097  0.0046  0.00097  0.0036 lb / lb (0.0036 kg / kg).
12. Because at heating mode operation Tr  72°F (22.2°C), from the psychrometric chart, the
relative humidity at winter design condition r  28 percent.
13. If air leaves the preheating coil of the outdoor air AHU at Tph  83°F and Tpf  Tph 
 fan
temperature rise  83 
 2  85°F, because wo  wpf  0.0017 lb / lb, from the psychrometric
chart, air density at the fan supply outlet in the outdoor-air AHU 	pf  1/ 13.75  0.073 lb / ft3.
The preheating coil load Qcph, in Btu / h (W), can be calculated as
(28.22)
14. If the heat loss of outdoor air supply duct is 0.5°F, Tps  85  0.5  84.5°F, and wps  wpf
 0.0017 lb / lb, then point ps can be determined. Draw line ps-r. Because wm  0.0036 lb / lb
(0.0036 kg / kg), point m can be plotted on the psychrometric chart. From the chart, Tm  76.2°F
(24.5°C).
15. The temperature of the fan-coil supply air at winter design load Ts, in °F (°C), can be calculated
as
(28.23)
As ws  wm  0.0036 lb / lb (0.0036 kg / kg), point s can be determined, and lines r-s and m-s can
be drawn. The temperature difference Ts  Tr  86.9  72  14.9°F (8.3°C), which is smaller
than the supply air temperature difference for warm air during heating mode 15°F (8.3°C)
specified by ASHRAE handbooks to prevent an excessive buoyancy effect, as discussed in
Sec. 18.6.
16. The air temperature at the fan outlet in the fan coil Tcf  Tm 
 0.5  76.2 
 0.5  76.7°F.
The heating coil load in the fan-coil unit Qch, in Btu / h (W), can be calculated as
(28.24)
 60  2  300  0.073  0.243(86.9  76.7)  7152 Btu /h (2095 W)
17. At winter heating mode part-load operation, the warm supply air from the fan coil, point sp,
moves along the horizontal line m-sp and is supplied at a lower temperature Tsp in order to maintain
the required zone temperature Trp.
Qch  60V?fc 	cf cpa(Ts  Tcf)
 72 

9500
60  2  300  0.073  0.243
 72 
 14.9  86.9F (30.5C)
Ts  Tr 

Qrh
60V?fc 	cf cpa
 60  200  0.073  0.243(83  20)  13,411 Btu/h (3929 W)
Qcph  60V?p 	pf cpa(Tph  To )
 0.0046 lb / lb (0.0046 kg /kg)
wr  wo 

wr  ws
0.33
 0.0017 

0.00097
0.33
wr  ws
wr  wo

V?
p
V?
fc

200
600
 0.33
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.19
Applications
Four-pipe fan-coil systems with a dedicated ventilation and space recirculating air system are
widely used in hotels and motels where the exhaust air system for bathrooms can easily balance
part of the outdoor ventilation air supply. They are also used in hospitals, schools, and offices.
28.4 TWO-PIPE FAN-COIL SYSTEMS
Two-Pipe Systems
A two-pipe system equipped with a supply pipe and a return pipe is shown in Fig. 28.4. In such a
system, chilled water is supplied to the coil to cool and dehumidify the air during cooling mode operation.
In heating mode operation, chilled water is changed over to hot water and then supplied to
the coil to heat the air.
A two-way valve is usually installed before the coil inlet because it costs less, is easier to install,
and saves pump power at part-load operation when the water flow rate is reduced. A DDC controller
is often used to modulate the water flow at part load.
Nonchangeover Two-Pipe Systems
When a two-pipe fan-coil system is used to serve a perimeter zone in a building, changeover from
chilled water to hot water or vice versa is a troublesome process and may take several hours. Therefore,
in locations where winter weather is moderate, a nonchangeover two-pipe fan-coil system may
be used.
In a nonchangeover two-pipe system, chilled water is supplied to the fan coil throughout the
year when the space is occupied. Warm outdoor air is supplied to the fan coil in winter to offset the
space heating load in the perimeter zone, and in spring and fall when space heating is required. In
such an arrangement, various zones at different orientations in the building that need cooling and
heating simultaneously during spring and fall can be served, as shown in Fig. 28.5a.
28.20 CHAPTER TWENTY-EIGHT
FIGURE 28.4 Supply and return mains in a two-pipe fan-coil system.
In a typical nonchangeover two-pipe fan-coil system, as shown in Fig. 28.5a, chilled water Tel
enters the coil at a temperature of about 45°F (7.2°C) at summer design conditions, whereas the
temperature of outdoor air supply Tps may be maintained at 62°F (16.7°C). At cooling mode partload
operation, Tel supply for both fan-coil units and the outdoor air AHU can be reset gradually, up
to 52°F (11.1°C), when the space sensible cooling load drops to 30 percent of the design value.
When outdoor temperature To drops below 70°F (21.1°C), the outdoor air supply temperature Tps
begin to rise. The lower To, the higher Tps.
If the space heating load of any room or control zone in the perimeter zone is offset entirely by
the heated outdoor air in a nonchangeover two-pipe fan-coil system, the required volume flow rate
of heated outdoor air , in cfm [m3 / (60 s)], can be calculated as
(28.25)
where Qrh  room heating load, Btu /h (W)
Aex  total area of building shell in that room, ft2(m2)

(AexUm 
 60V?inf 	ocpa)(Tr  To)
60 	pscpa(Tps  Tr)
V?
p 
Qrh
60 	pscpa (Tps  Tr)
V?
p
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.21
FIGURE 28.5 Operating parameters of air and water in a two-pipe fan-coil system: (a) nonchangeover;
(b) changeover.
28.22 CHAfYfER TWENTY-EIGHT
140
120
~
0
~~
'§
&
§
.. "~
~
-g
'"
..
"~
~
E
"t:
~
100
80
60
40
0 20 60 80 100
Outdoor temperature 1;, , of
(b)
FIGURE 28.5 (Continued)
Urn = weighted average of overall heat-transfer coefficient of building shell in that room,
Btulh .ft2.oF (W Im2. OC)
Vinf = volume flow rate of infiltrated air, cfm [m3 1(60 s)]
PO' Pps = density of outdoor air and outdoor air supply, lblft3 (kglm3)
T r = zone air temperature, oF (OC)
T ps = supply temperature of conditioned outdoor air, oF (OC)
The air transmission ratio is defined as the ratio of outdoor air volume flow rate supplied to a room
in the perimeter zone Vp, in cfm [m31(60 s)], to the transmission loss per degree of outdoor-indoor
temperature difference ~T tran' Btulh .oF (W 1°C). According to Eq. (28.25), it can be calculated as
Vp Vp -=
~Ttran AexUm + 60VinfPoCpa
Tr- To
(28.26)
60ppscpa(Tps -Tr)
For all the rooms in the perimeter zone supplied by the same outdoor ventilation air system in a
two-pipe fan-coil system, the temperature difference between room air and outdoor air Tr  To is a
constant at any time during operation. If the temperature drop due to the heat loss of the outdoor air
supply duct is ignored, for the same outdoor ventilation air system, the supply air temperature difference
Tps  Tr is again a constant. Because air density 	ps and specific heat of supply air cpa can
also be considered constant, and if the zone heating load in the perimeter zone is offset entirely by
the conditioned outdoor air, the air transmission ratio for all the control zones in the
perimeter zone in the same outdoor ventilation air system for a two-pipe nonchangeover fan-coil
system must be a constant.
The current trend is to have an adequate amount of outdoor air supply for a better indoor air
quality. Because the conditioned outdoor air supply must be higher than the outdoor ventilation air
requirement, this often means that a nonchangeover two-pipe fan-coil system has sufficient conditioned
outdoor air for winter heating. In massively constructed buildings, the air transmission ratio
can be smaller, such as 0.7 , because of the heat storage effect of the building shell.
One of the primary drawbacks of a nonchangeover two-pipe fan-coil system is the waste of energy
during simultaneous heating and cooling in some control zones during winter heating. In a
cold winter in the northern hemisphere, for a control zone facing south, warm outdoor air heating,
solar radiation, and some internal loads often combine to result in a higher zone temperature than
that for a control zone facing north in the same outdoor ventilation air system. Chilled water is then
admitted to the fan coil to cool the zone air simultaneously. Simultaneous heating and cooling for
the same control zone must be avoided. The waste of energy of the conditioned outdoor ventilation
air system in a nonchangeover two-pipe fan-coil system is similar to the waste of energy in an air
skin system, discussed in Sec. 21.3.
Changeover Two-Pipe Systems
During the cooling mode operation of a changeover two-pipe fan-coil system, the space cooling
load is offset by the combined cooling effect of the cooled outdoor ventilation air and chilled water
supplied to the fan coil, as shown in Fig. 28.5b. As discussed in Sec. 7.6, ASHRAE/IESNA Standard
90.1-1999 specifies that two-pipe fan-coil changeover systems are acceptable when the design
dead band width between changeover from one mode to the other is at least 15°F (8.3°C) outdoor
air temperature. It also will allow operation in one mode at least four hours before changeover to
the other mode. At the changeover point, reset controls allow heating and cooling supply temperatures
to be no more than 30°F (16.7°C) apart.
During the fall season, before the changeover from cooling mode operation to heating mode operation,
when outdoor temperature To is higher than changeover temperature Tco, the supply temperature
of outdoor ventilation air increases according to To when To falls below 70°F (21.1°C). The
heating effect of the conditioned outdoor ventilation air is able to offset the individual zone heating
load in the perimeter zone. Chilled water is used to provide cooling and dehumidifying when
needed. Meanwhile, the chilled water supplied temperature Tel should be reset between 45 and 52°F
(7.2 and 11.1°C) at part-load operation before changeover.
During changeover, the chilled water may be reset to 52°F (11.1°C), and it is changed over to
hot water at a temperature higher than 80°F (26.7°C). Meanwhile, the warm air supply is changed
over to cold air supply, typically at a temperature of 50°F (10.0°C). The changeover process was
discussed in Sec. 7.6, and the outdoor changeover temperature Tco can be calculated from Eq. (7.5).
Changeover may require several hours. To prevent two to three changeovers within the same day,
operation in one mode lasts at least four hours before changeover to the other mode, and a tolerance
of about  2°F (1.1°C) is often used. There are many factors that influence the changeover temperature
Tco in actual practice. Therfore, calculated Tco should be modified according to actual operating
experience.
When To  Tco, a changeover two-pipe fan-coil system is in heating mode operation. The hot
water temperature is reset according to outdoor temperature To. Cold outdoor air is supplied, typically
at 50°F (10.0°C). Any zone cooling load of an individual control zone in the perimeter zone
can be offset by the combined effect of the cold outdoor supply air and the transmission and
V?
p /Ttran
V?
p /Ttran
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.23
infiltration loss. Zone heating loads are offset by the hot water supplied to the fan coils. During
heating mode operation, the refrigeration system that serves the perimeter zone is usually shut
down. However, the changeover two-pipe fan-coil system still operates with a cold source of outdoor
air and a heat source of hot water simultaneously.
In spring, heating mode operation changes over to cooling mode operation at outdoor temperature
Tco only when the combined effect of outdoor ventilation air cooling and the transmission and
infiltration loss does not offset the cooling load of every individual control zone in the perimeter
zone.
In a two-pipe fan-coil system, except for the changeover from heating mode to cooling mode
operation or vice versa, operating characteristics of the fan-coil unit and the space recirculating system,
exhaust air balance, and the zone temperature control are similar to those in a four-pipe fancoil
system.
System Characteristics
System characteristics of a two-pipe fan-coil system (2PFC) are listed in Table 28.1.
Applications
Because of the waste of energy in nonchangeover two-pipe fan-coil systems and the difficulties in
changeover operations, the applications of two-pipe fan-coil systems are comparatively less.
28.5 WATER-SOURCE HEAT PUMP SYSTEMS
System Description
In a water-source heat pump system, some water-source heat pumps located in the shady side of the
perimeter zone of a building may extract heat from a water loop to heat the supply air, and other
water-source heat pumps in the core part of the building may reject heat to the water loop to cool
the supply air. Excess heat is therefore transferred from the core of the building to the perimeter
zone of the building. A water-source heat pump system conserves more energy than many other air
conditioning systems only when simultaneous heating and cooling occur in a building and therefore
excessive heat from the core part is recovered and transferred to the perimeter zone of the building.
A typical closed-circuit water-source heat pump system is illustrated in Fig. 28.6. It consists of
water-source heat pumps, an evaporative water cooler, a boiler or water heater, two water circulating
pumps, an expansion tank, piping, necessary accessories, and controls. One of the water circulating
pumps is the lead pump, the other is a standby. A storage tank is optional.
Operating Characteristics
During hot summer, when the outdoor wet-bulb temperature is 78°F (25.6°C), all the water-source
heat pumps are operting in cooling mode. The condenser water leaves the closed-circuit evaporative
water cooler at 95°F (35.0°C) and enters the water loop. After all heat is rejected from water-source
heat pumps, the condenser water returns to the evaporative water cooler at a temperature of 105°F
(40.6°C) and cools to 95°F (35.0°C) again.
During moderate weather, water-source heat pumps serving the shady side of the building are in
heating mode, whereas those serving the sunny side of the building are in cooling mode. The rise or
fall in condenser water temperature depends on the ratio of the number of cooling units to the total
number of units, or, more exactly, the ratio of heat rejection to heat extraction. If heat rejected from
the cooling units is greater than heat extracted by the heating units, the average temperature of the
28.24 CHAPTER TWENTY-EIGHT
water loop tends to rise, and vice versa. When the condenser water temperature entering the evaporative
water cooler Te,ev rises to 90°F (32.2°C), the cooler starts to operate and brings the condenser
water temperature down to a lower value (say, 87°F, or 30.6°C).
During cold weather, some water-source heat pumps serving the core of the building still operate
in cooling mode. If the heat extracted by the units operating in heating mode is greater than the
heat rejected, the average temperature of the water loop drops. When the condenser water temperature
Te,ev falls below 60°F (10.6°C), the boiler (or water heater) is energized. Heat is added to the
water loop to raise Te,ev to 63°F (17.2°C) or higher.
A water-source heat pump system recovers the heat rejected to the water loop by the core units
or by the units on the sunny side and supplies it to the units serving the perimeter zone in which
heating is required to offset heat losses.
ASHRAE/IESNA Standard 90.1-1999 specifies that in a water-source heat pump system, watersource
heat pumps are connected to a common water loop with central devices of heat rejection
(such as a cooling tower) and heat addition (such as a boiler). Such a system shall have controls to
provide heat pump water supply temperature with a dead band of at least 20°F (11.1°C) between
initiation of heat rejection and heat addition by the cooling tower and boiler (for example, 90 
60  30°F or 16.7°C).
Loop Temperatures
Water-source heat pumps (WSHPs) are rated according to Air-Conditioning and Refrigeration Institute
(ARI) Standard 320 at the following entering condenser water temperatures: for cooling mode
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.25
FIGURE 28.6 A typical closed-circuit water-source heat pump system: (a) schematic diagram; (b) water-source heat
pump.
performance 85°F (29°C) and for heating mode performance 70°F (21°C). Pietch (1991) listed the
performance of WSHPs of a WSHP system at various loop temperatures:
Here EERc,ra indicates the energy efficiency ratio of the WSHP at the cooling mode rated conditions,
ARIrated c cap indicates the ARI-rated cooling capacity, and COPh,ra indicates the COP of the
WSHP at heating mode rated conditions.
WSHPs at cooling mode operation require lower condenser water temperature (CWT) to reduce
energy use, and WSHPs at heating mode need higher CWT for a higher COP. Optimizing loop temperatures
depend on the condition of zone loads, the number of WSHPs in cooling and heating
mode operations, and the outdoor weather. Because a water loop may have WSHPs in both cooling
and heating mode operations, the water loop temperature operating limit is usually maintained between
60 and 90°F (15.6 and 32.2°C).
Gottfried et al. (1997) suggested when the CWT has been lowered to a rather lower value in a
late afternoon, this allows the CWT to rise to a level that maintains WSHP efficiencies and at the
same time keep evaporative cooler (cooling tower) energy use to a minimum during the last working
hours of the day.
Water-Source Heat Pumps
Three types of WSHPs are widely used in WSHP systems:
 Console type for perimeter zones. Most have a cooling capacity of 1 ton (12,000 Btu / h, or 3.5
kW) or less.
 Vertical WSHPs often installed in mechanical rooms. They are used to serve the core (interior
zone) of the buildings and are connected with ducts. Vertical units often have cooling capacity
from 10 up to 24 tons.
 Ceiling-mounted horizontal WSHPs, often for special rooms. Ceiling units have a cooling capacity,
usually between 0.5 and 5 tons.
Most WSHPs have a heating capacity about 10 percent higher than their cooling capacity.
A water-source heat pump usually consists of an air coil that is a finned coil to condition the air;
a double-tube water coil to reject or extract heat from the water loop; a forward-curved centrifugal
fan, which is often located downstream of the air coil; single or twin hermetic compressors; a short
capillary tube; a reversing valve; an outer casing; controls; and accessories. A two-speed fan motor
is often used for better capacity control. A typical water-source heat pump is shown in Fig. 28.6b.
In a vertical water-source heat pump, the centrifugal fan is usually located at the top outlet, and the
hermetic compressor is often mounted in the ottom of the unit. In a horizontal unit, the centrifugal
fan is usually located at the end of the unit.
During cooling mode operation, the air coil acts as an evaporator and the water coil as a condenser.
Air in a typical water-source heat pump is cooled and dehumidified at the air coil from an
entering dry-bulb temperature of 80°F (26.7°C) and wet-bulb temperature of 67°F (19.4°C) to a
Loop tempera- Relative cooling
ture, °F (°C) Rating mode capacity EER/EERc,ra Heating capacity /ARIrated c cap COP/COPh,ra
90 (32.2) 0.98 0.94 1.50 1.08
85 (29.4) Cooling 1.00 1.00 1.45 1.06
80 (26.7) 1.02 1.06 1.40 1.04
75 (23.9) 1.04 1.12 1.35 1.02
70 (21.1) Heating 1.06 1.18 1.30 1.00
65 (18.3) 1.08 1.24 1.25 0.98
60 (15.6) 1.10 1.30 1.20 0.96
28.26 CHAPTER TWENTY-EIGHT
leaving condition of 60°F (15.6°C) dry-bulb and 57°F (13.9°C) wet-bulb temperatures. The suction
temperature is about 43°F (6.1°C). On the other hand, in heating mode operation, the air coil acts as
a condenser and the water coil as an evaporator. During cold weather air is heated at the air coil
from 70°F (21.1°C) to an off-coil temperature of 100°F (37.8°C). The suction tmperature in winter
is about 45°F (7.2°C). Sometimes, electric resistance heating, in steps, is added between the air coil
and the centrifugal fan instead of using a centralized water heater.
Energy Performance and Energy Use Intensity of WSHPs
ASHRAE Standard 90.1-1999 mandatorily specifies the minimum efficiency requirements for water-
source heat pumps as follows:
Gottfried et al. (1997) reported a new high-efficiency WSHP with an EER of 14.9 (4.33 COP with a
cooling capacity from 21,000 ro 57,000 Btu /h, or 6.2 to 16.7 kW) and a high-efficiency motor with
a minimum efficiency of 90 percent.
According to DOE/EIA Commercial Buildings Consumption and Expenditures 1995, the energy
use intensity (EUI) for heat pumps (including mainly water-source heat pumps and air-source heat
pumps) in commercial buildings in 1995 in the United States was 28.1 kBtu / ft2 	yr (8.2 kWh / ft2 	yr
or 88 kWh /m2 	 yr) which is the smallest of all the air conditioning systems.
Closed-Circuit Evaporative Water Cooler
A closed-circuit evaporative water cooler (CCEWC), sometimes called a closed-circuit cooling
tower, resembles a cooling tower. However, condenser water in a cooling tower is an open-circuit
system, whereas in an evaporative water cooler the condenser water is a closed-circuit system (see
Fig. 28.6a). Condenser water flows through a coil onto which recirculating water is sprayed. Heat is
rejected from the condenser water through the tube wall and is absorbed by the vaporized liquid at
the outer surface of the coil. A closed-circuit evaporative water cooler is usually located outdoors,
most probably on the rooftop.
Both centrifugal fans and propeller fans can be used in a closed-circuit evaporative water cooler.
The capacity of a centrifugal fan is more easily controlled with a damper and makes less noise;
therefore, it is more frequently used.
The capacity of a closed-circuit evaporative water cooler can be modulated by the following
methods:
 Convective cooling—opening of the top damper only
 Natural-draft evaporative cooling—spraying of recirculating water and opening of the top
damper
 Forced-draft evaporative cooling—operating the fan and the water spraying
 Forced-draft evaporative cooling with damper modulation
Operation Capacity, Btu /h Rating conditions Minimum efficiency Efficiency as of 10 /29/2001
Cooling Qrc  65,000 85°F entering water 9.3 EER
86°F entering water 11.2 EER
17,000  Qrc  65,000 85°F entering water 9.3 EER
86°F entering water 12.0 EER
65,000  Qrc  135,000 85°F entering water 10.5 EER
86°F entering water 12.0 EER
Heating Qh  135,000 (heating capacity) 70°F entering water 3.8 COP
68°F entering water 4.2 COP
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.27
The motive for a closed-loop water system is to prevent fouling of the water coils in water-source
heat pumps. Closed-circuit evaporative water coolers were widely used in the early years of the application
of WSHP systems. However, CCEWCs have the following disadvantages:
1. According to Cooper (1994), the maintenance of spray nozzles, pumps, and heat exchangers of a
CCEWC required twice amount of service of an open-circuit cooling tower.
2. According to Kush (1990), field experience in an installation in Stamford, Connecticut, showed
that the heat and pumping energy losses from the outdoor evaporative water cooler and piping
work were about one-half of the energy input to the electric boiler in January 1988.
3. For locations where the outdoor temperature drops below 32°F (0°C), freeze protection of the
outdoor portion of the water loop must be considered.
The current trend is as follows:
 Divide the water loop of a WSHP system into two portions: closed-circuit and open-circuit. A
plate-and-frame heat exchanger is used to connect these two part and transfers heat energy from
the closed circuit to open circuit, as shown in Fig. 28.7.
 Use a cooling tower (open-circuit) instead of a closed-circuit evaporative water cooler, and two
sets of water circulating pumps, one for the closed-circuit and the other for the open-circuit cooling
tower and accessories. Set only the cooling towers and the necessary pipework and accessories
in the open circuit outdoors.
 Properly insulate the outdoor pipework and equipment. Provide freeze protection for the outdoor
open-circuit water loop, including the addition of a certain percentage of ethylene glycol solution
to the water loop. Ethylene glycol is expensive. Inhibitors must be added at the same time to resist
corrosion. Ethylene glycol solution reduces the heat-transfer coefficient as well as the heating and
cooling capacities of the equipment.
ASHRAE/IESNA Standard 90.1-1999 specifies that for climates with HDD65 greater than 1800
hours annually, if a closed-circuit tower is used, then either an automatic valve shall be installed to
bypass all but a minimal volume flow of water around the tower for freezing protection or low-leakage
positive closure air dampers shall be provided. If an open-circuit tower is used directly in the
28.28 CHAPTER TWENTY-EIGHT
T3
T1
T2
Optional
storage
tank
3
Gas-fired
water heater
Cooling tower
pump
Plate-and-frame
heat exchanger
Cooling tower
(open-circuit)
Closed-circuit
water-loop pumps
2 1
Vertical
WSHP
T4
4
T5
5
Roof
Vertical
WSHP
Console
WSHP
Console
WSHP
FIGURE 28.7 A water-source heat pump system using an open-circuit cooling tower.
heat pump water loop, then an automatic valve shall be installed to bypass all heat pump water flow
around the tower. If an open-circuit tower is used in conjunction with a separate heat exchanger to
isolate the tower from the heat pump water loop, then the heat loss shall be controlled by shutting
down the circulating pump on the outdoor cooling tower loop during cold winter when the tower is
not in use.
Water Heater
A water heater or a hot water boiler is used to provide heat energy to the water loop to maintain a
preset operating temperature. Electric boiler is most widely used. It is easiest to install and maintain.
In locations where electric rates are high, electric boilers require higher operating costs than
gas- or oil-fired boilers.
For better capacity control, electric load in an electric boiler is often divided into stages, as
discussed in Sec. 8.4. The heating capacity is controlled in steps by sensing the water temperature
entering the electric boiler.
Gas- or oil-fired heaters are combustion devices containing combustion chambers. Water at a
temperature of 60°F (21.1°C) from the water loop cannot directly contact one side of the combustion
chamber and a flame on the opposite side. It will cause condensation and deteriorate the water
heater. Water contacting one side of the combustion chamber must be recirculated at a temperature
of 140°F (60.0°C) or above. Loop water can then be heated by mixing.
Storage Tanks
Whether a storage tank should be used in a WSHP system depends on the local electric rate structures
and a careful analysis of initial costs and operating cost. The principle of thermal storage is
covered in Chap. 31. Because of the electric deregulation as well as the application of real-time
pricing (RTP) rate structures, the provision of a storage tank in a WSHP system is often cost-effective
in many applications. Heat rejected to the water loop at a temperature between 60°F (15.6°C)
and 90°F (32.2°C) can be stored in the storage tank to offset nighttime heat losses if the water loop
operates at night with a setback space temperature.
Many utilities offer lower electric rates at off-peak times. A cooling tower or evaporative cooler
can be used to cool the water to a lower temperature. It can then be stored in the tank for use during
on-peak hours in summer to reduce costs. An electric boiler also may be energized at night to raise
the water temperature to 180°F (82.2°C) or higher. High-temperature stored water can be used in
winter peak hours to reduce electricity costs.
Air Systems and Maintenance
A WSHP system must have a dedicated outdoor ventilation system to provide ventilation air for occupants.
Outdoor ventilation air can either be supply to the return plenum mixing with the recirculating
air or supply to the conditioned space directly. The centrifugal supply fan in a WSHP itself
forms a space recirculating system. Console WSHPs have no ducts. The operating characteristics of
the air systems in a WSHP system are similar to those in a four-pipe fan-coil system.
According to Cooper (1994), maintenance in a WSHP system consists of mainly cleaning and
changing filters and replacing belts and seal bearings. Sometimes a WSHP is removed and the
condenser is acid-cleaned. There is also maintenance required for cooling towers, pumps, and water
heaters. For the 3 million ft2 (278,800 m2) of commercial office buildings that have been maintained,
records showed that on WSHPs wih a cooling capacity of 1 ton (3.5 kW) and less, the
replacement percentage is less than 0.5 percent per year. For WSHPs of 2 to 5 tons (7 to 17.6 kW),
the rate is slightly higher, for WSHPs of 6 to 10 tons (21.1 to 35.0 kW), slightly higher still.
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.29
Controls
Water-Loop Temperature Control. In a WSHP system with an open-circuit cooling tower and a
plate-and-frame heat-exchanger, the water temperature leaving the plate-and-frame heat exchanger
is controlled between 60 and 90°F (15.6 and 32.2°C) when the WSHP system is in operation. For a
typical WSHP system, a temperature sensor T1 may be located at the exit of the plate-and-frame
heat exchanger, as shown in Fig. 28.7. When the microprocessor-based DDC controller receives a
signal from T1, it actuates the following:
28.30 CHAPTER TWENTY-EIGHT
Alert by alarm. WSHP, cooling tower, water
heater, and pumps all shut down. Start the water-
loop water pumps.
Energize the water heater.
Start the cooling tower water circulating pumps.
Open the damper of the cooling tower.
Turn on the first-stage tower fans.
Turn on the second-stage tower fans.
Alert by alarm. System will shut down.
At 57°F (13.9°C)
At 60 to 68°F (15.6 to 20.0°C)
At 84°F (28.9°C)
At 86°F (30.0°C)
At 88°F (31.1°C)
At 90°F (32.2°C)
At 105°F
An outdoor air sensor T3 resets the water heater energizing temperature T2 in such a manner that
when the outdoor temperature To drops from 60 to 0°F (15.6 to 17.8°C), T2 increases linearly
from 60 to 68°F (15.6 to 20.0°C).
To maintain space air at 75°F (23.9°C), 50 percent relative humidity, and a dew point of 55°F
(12.8°C) in summer, it is critical to maintain a minimum water-loop temperature Twl  57°F
(13.9°C). If Twl  57°F (13.9°C), condensate may form on the outer surface of uninsulated pipes in
the conditioned space and cause damage.
The actuating temperatures for turning off the second-stage and first-stage tower fans, shutting
off the cooling tower water pump, and closing the damper in the cooling tower as the water temperature
decreases should be successively 2°F (1.1°C) lower to prevent short cycling [e.g., turn off the
second-stage tower fans at 90°F (32.2°C), turn off the first-stage tower fans at 88°F (31.1°C)].
Capacity Control of Water Heater. If the water heater is a gas-fired hot water boiler or a hot water
heat exchanger, a temperature sensor T2 senses the water temperature leaving the water heater. The
heating capacity can be controlled by modulating the flow rate of gas or hot water by means of a
DDC controller. Because most hot water boilers contain a limited amount of water, if the temperature
sensor is located at the exit side of the boiler, the temperature response to modulation is fast.
Therefore, a broadband DDC controller should be used to prevent short cycling.
If an electric boiler is used, step control in appropriate stages can be performed through a sensor
located at the inlet to sense the temperature of entering water. Step control often causes temperature
fluctuation at the leaving side.
Safety Control. In addition to the high and low operating limit control of the water-loop temperature,
a flow switch or pressure differential sensor should be installed across the water-loop pump. In
case of pump failure, the entire system, including the water-source heat pumps, cooling tower, and
circulating pump, or water heater shuts down. After a delay of 10 to 15 s, the standby pump is energized.
At the same time, an alarm sounds with an indicating light. When water flow is resumed, the
cooling tower and circulating pumps, or the water heater and water-source heat pumps, are
restarted, with a delay between each stage to limit any sudden increase in starting current.
Water-Source Heat Pump Control. When the pushbutton is engaged, the electric circuit is energized
and the outdoor air damper is opened. When the zone temperature rises above the cooling
mode set point, the temperature sensor T4 (Fig. 28.7) calls for cooling. The DDC controller then
starts the fan. For small console WSHPs used in the perimeter zone, a multispeed fan is usually
used. Fan is at either the high or low position, which is set manually. After a 30- to 50-s delay, if the
high-pressure and low-pressure safety control circuits are closed, the compressor is started. For
large vertical or horizontal WSHPs, the fan is belt-driven. Fan motor starts the fan after safety control
circuits are closed. If there are multiple compressors, the leading compressor is started. WSHP
is now in cooling mode operation. When the zone temperature drops below the cooling mode set
point and is sensed by T4, the DDC controller then stops the compressor, or one of the compressors
is deenergized. The fan should be operated continuously when the space is occupied, to provide
outdoor ventilation air to the space.
When the zone temperature drops below the heating mode set point and is sensed by the temperature
sensor T5, the DDC controller starts the compressor. WSHP is operating at heating mode.
When the zone temperature rises above the heating mode set point, the DDC controller stops the
compressor. There is always a deadband of 1 to 3°F (0.6 to 1.7°C) if the cooling mode is automatically
changed to heating mode or vice versa.
In either heating or cooling mode operation, if the discharge pressure rises above the high-pressure
limit, or the suction pressure drops below the low-pressure limit, the high- or low-pressure
control opens the electric circuit, lights an alarm lamp, and stops the compressor.
ASHRAE/IESNA Standard 90.1-1999 specifies that each water-source heat pump shall have a
two-position valve for a water system having a total pump system power exceeding 10 hp (7.5 kW).
System Characteristics
System characteristics of water-source heat pump systems are listed in Table 28.1.
Case Studies: Water-Source Heat Pump Systems
Gottfried et al. (1997) introduced a water-source heat pump system with an open-circuit cooling
tower in Ridgehaven Building, San Diego (RBSD), California, as discussed in Sec. 25.3. RBSD is
an office building of 73,000 ft2 (6784 m2). In 1995, the new WSHP system installed in each of the
two wings of the Ridgehaven Building included the following primary components:
 72 high efficiency WSHPs having an EER of 14.9 (4.33 COP)
 100-ton (open-circuit) cooling tower
 400,000 Btu /h (117 kW) gas-fired hot water boiler
 Condenser water isolation valve for each WSHP to cut off condenser water flow when the WSHP
is shut off
Ridgehaven Building provides 20 cfm (10 L/s) outdoor ventilation air for each occupant. Using
specified high-efficiency filters for outdoor-air AHU, the RBSD target energy use is 9 kWh/ ft2 	yr
(97 kWh/m2 	 yr). The first month’s operating data showed that the annual energy use for a WSHP
system with an open-circuit cooling tower prior to commissioning is approximately 11 kWh / ft2 	yr
(118 kWh/m2 	 yr).
For an office building in Stamford, Connecticut, Kush (1990) analyzed and reported on the actual
performance of a closed-circuit water-source heat pump system that served a total area of
72,000 ft2 (6691 m2) of office space on three floors. This system has 140 one-ton perim (console)
WSHPs, twelve 10-ton core WSHPs, a closed-circuit evaporaive cooler, and a 300-kW electric
boiler. Stamford has a 97.5 percent winter design temperature of 9°F (12.8°C) and 5617 heating
degree-days. In summer, 2.5 percent design dry-bulb temperature is 84°F (28.9°C) with a mean coincident
wet-bulb temperature of 71°F (21.7°C).
According to the results measured between November 1987 and October 1988, the breakdown
of annual HVAC&R energy use was as follows:
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.31
The energy use for heating in core WSHPs is mainly caused by excessive outdoor ventilation air.
The monthly HVAC&R energy use of this building between November 1987 and October 1988
is shown in Fig. 28.8. The highest HVAC&R energy use, 110,050 kWh, occurred in January 1988.
Of this, 66 percent was consumed by the electric boiler. The highest in summer months was that for
August 1988—65,715 kWh. About one-half of August’s energy was used by core WSHPs. Yearround,
the lowest energy use, 32,330 kWh, occurred in May 1988, when the core WSHPs consumed
40 percent of the energy input.
Design Considerations
 Cooper (1994) recommended the following rules for occupant satisfaction:
Provide core WSHPs from mechanical rooms.
Provide console WSHP with individual control for control zones in perimeter zone.
Provide adequate outdoor ventilation air.
Give the tenant means to use the HVAC system in off hours.
Decentralize to the points free of major shutdowns.
Adjust tenant space conveniently.
28.32 CHAPTER TWENTY-EIGHT
32 percent
4 percent
11 percent
18 percent
8 percent
15 percent
2 percent
9 percent
Electric boiler
Evaporative water cooler
Water-loop pumps
Perimeter WSHPs
Core WHPs
Fan
Cooling
Heating
Others
FIGURE 28.8 HVAC&R monthly energy use of a water-source heat
pump system.
 If ceiling units are used, good design of ducting, return plenum, vibration hangers, and piping flex is
mandatory and requires careful consideration for noise attenuation. When console WSHPs are used,
they should not oversized. An oversized WSHP creates more noise and is not energy-efficient.
 A water flow rate of 2.5 to 3.5 gpm (0.16 to 0.22 L / s), typically 3 gpm (0.19 L / s) per cooling
ton, is appropriate for water-source heat pump systems.
 Outdoor ventilation air rates should be adequate and should not be greater than that specified in
ASHRAE Standard 62-1999, as listed in Table 5.9. Outdoor air damper should be closed when
the conditioned space is not occupied.
 Fan power in water-source heat pumps has a certain influence on energy use. Both supply volume
flow rate and total pressure loss should be carefully calculated in order to select the proper unit.
 Variable-speed water-loop pumps and cooling tower fans are often cost-effective in large WSHP
systems.
28.6 PANEL HEATING AND COOLING
Panel heating and cooling involves the control of the mean radiant temperature of the conditioned
space, which is closely related to the thermal comfort of the occupants. Floor-panel heating systems
have been used by many residences in Europe since the 1980s, as discussed in Sec. 8.8. However,
panel heating and cooling is expensive and has a very limited application in the United States. If the
conditioned space has a large glaze area, metal ceiling panel should be used. A dedicated ventilation
system should be used to provide the ASHRAE Standard 62-1999 specified amount of outdoor
ventilation air. To prevent any possibility of condensation on the panels, the air system must maintain
a space dew point temperature and humidity levels lower than the design conditions all the time
when cooling panel is used.
Panel heating and cooling requires greater field experience in design and operation. Refer to
ASHRAE Handbook 1996, HVAC Systems and Equipment, chapter 6 for detailed information.
REFERENCES
Anantapantula, V. S., and H. J. Sauer, Heat Recovery and the Economizer for HVAC Systems, ASHRAE
Journal, no. 11, 1994, pp. 48–53.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
Carrier Corp., Products and Systems 1992 / 1993 Master Catalog, Carrier Corporation, Syracuse, NY.
Clark, S. J., and B. Eversion, Non-HVAC Water Lines Provide Cost Effective Four-Pipe Fan-Coil Systems for
Motel, ASHRAE Journal, no. 3, 1994, pp. 29–31.
Cooper,W. S., Operative Experience with Water Loop Heat Pump Systems, ASHRAE Transactions, 1994, Part I,
pp. 1569–1576.
DOE/ EIA, 1998 Nonresidential Buildings Energy Consumption Survey: Commercial Buildings Consumption
and Expenditures 1995, DOE/ EIA -0318 (95).
Friberg, E. E., Case History—Low-Rise Office Building Using Water-Source Heat Pump, ASHRAE
Transactions, 1988, Part I, pp. 1708–1725.
Gottfried, D. A., E. A. Schoichet, and M. Hart, Green Building Environmental Control: A Case Study, HPAC,
no. 2, 1997, pp. 71–78.
Howell, R. H., and J. H. Zaidi, Analysis of Heat Recovery in Water-Loop Heat Pump Systems, ASHRAE
Transactions, 1990, Part I, pp. 1039–1047.
Hughes, P. J., Survey of Water Source Heat Pump System Configurations in Current Practice, ASHRAE
Transactions, 1990, Part I, pp. 1021–1028.
AIR CONDITIONING SYSTEMS: SPACE CONDITIONING SYSTEMS 28.33
Kilkis, B. I., Radiant Ceiling Cooling with Solar Energy: Fundamentals, Modeling, and a Case Design,
ASHRAE Transactions, 1993, Part II, pp. 521–533.
Kochendorfer, C., Standardized Testing of Cooling Panels and Their Use in System Planing, ASHRAE
Transactions, 1996, Part I, pp. 651–658.
Kush, E. A., Detailed Field Study of a Water-Loop Heat Pump System, ASHRAE Transactions, 1990, Part I, pp.
1048–1063.
Kush, E. A., and C. A. Brunner, Optimizing Water-Loop Heat Pump Design and Performance, ASHRAE
Journal, no. 2, 1992, pp. 14–19.
Lefebvre, R. R., New HVAC System Reduces Operating Costs, ASHRAE Journal, no. 4, 1993, pp. 20–23.
Meckler, M., Integrating Water Source Heat Pumps with Thermal Storage, Heating / Piping / Air Conditioning,
July 1988, pp. 49–64.
Mulroy,W. J., The Effect of Short Cycling and Fan Delay on the Efficiency of a Modified Residential Heat
Pump, ASHRAE Transactions, 1986, Part I B, pp. 813–826.
Pietch, J. A., Optimization of Loop Temperatures in Water-Loop Heat Pump Systems, ASHRAE Transactions,
1991, Part II, pp. 713–726.
Simmonds, P., Practical Applications of Radiant Heating and Cooling to Maintain Comfort Conditions,
ASHRAE Transactions, 1996, Part 1, pp. 659–666.
The Singer Company, Electro Hydronic Systems, The Singer Co., Carteret, NJ, 1977.
The Trane Company, Water Source Heat Pump System Design, The Trane Company, La Crosse, WI, 1981.
The Trane Company, Fan-Coil Units, La Crosse, WI, 1990.
Virgin, D. G., and W. B. Blanchard, Cary School—25 Years of Successful Heat Pump / Heat Reclaim System
Operation, ASHRAE Transactions, 1985, Part I A, pp. 40–45.
Wang, S. K., Air Conditioning, vol. 4, Hong Kong Polytechinc, Hong Kong, 1987.
Weinstein, A., L. D. Eisenhower, and N. S. Jones,Water-Source Heat Pump System for Mount Vernon Unitarian
Church, ASHRAE Transactions, 1984, Part I B, pp. 304–312.
Zaidi, J. H., and R. H. Howell, Energy Use and Heat Recovery in Water-Loop Heat Pumps, Variable-Air-Volume,
Four-Pipe Fan-Coil, and Reheat HVAC Systems, Part 1 and Part 2, ASHRAE Transactions, 1993, Part II,
pp. 13–39.
28.34 CHAPTER TWENTY-EIGHT
CHAPTER 29
AIR CONDITIONING
SYSTEMS: PACKAGED
SYSTEMS AND
DESICCANT-BASED
SYSTEMS
29.1
29.1 PACKAGED SYSTEMS 29.2
Comparison between Packaged and
Central Systems 29.2
Applications 29.3
Types of Packaged Systems 29.4
29.2 SINGLE-ZONE CONSTANT-VOLUME
PACKAGED SYSTEMS 29.4
System Description 29.4
Air and Water Economizers 29.4
Supply Volume Flow Rate, and Cooling
and Heating Coil Loads 29.4
Controls 29.5
Energy Use Intensities 29.5
System Characteristics 29.5
29.3 SINGLE-ZONE VAV PACKAGED
SYSTEMS 29.7
System Description 29.7
System Calculations 29.7
Controls 29.7
System Characteristics 29.8
29.4 COMPLIANCE WITH STANDARD
90.1–1999 THROUGH SIMPLIFIED
APPROACH OPTION FOR SMALL AND
MEDIUM HVAC&R SYSTEMS 29.8
29.5 VAV COOLING PACKAGED
SYSTEMS 29.9
System Description 29.9
Supply Volume Flow Rate and Coil
Load 29.10
Pressure Characteristics and Duct Static
Pressure Control 29.10
System Characteristics 29.12
29.6 VAV REHEAT PACKAGED
SYSTEMS 29.12
System Description 29.12
Supply Volume Flow Rate and Coil
Load 29.12
Night Setback and Morning
Warm-up 29.14
Evenly Distributed Airflow at
DX Coils 29.14
Discharge Air Temperature Control for
Packaged Systems 29.15
Fan Modulation 29.16
Air-Cooled, Water-Cooled, and
Evaporatively-Cooled Condensers 29.17
Controls and Sound Problems for Rooftop
Packaged Units 29.17
Case Study: VAV Reheat Packaged
System for Precision
Manufacturing 29.17
System Characteristics 29.18
29.7 PERIMETER-HEATING VAV
PACKAGED SYSTEMS 29.18
System Characteristics 29.18
29.8 FAN-POWERED VAV PACKAGED
SYSTEMS 29.18
System Description 29.18
Supply Volume Flow Rate and Coil
Load 29.19
Controls 29.20
Case Study: A Fan-Powered VAV
Packaged System with Rooftop
Packaged Unit 29.20
System Characteristics 29.22
29.9 DESICCANT-BASED AIR
CONDITIONING SYSTEMS 29.22
Desiccant-Based Air Conditioning 29.22
Desiccant Dehumidification and Sensible
Cooling 29.22
Desiccant-Based Air Conditioning
Systems 29.24
Desiccants 29.24
Rotary Desiccant Dehumidifiers 29.27
29.10 CASE STUDY: A DESICCANT-BASED
AIR CONDITIONING SYSTEM FOR A
SUPERMARKET 29.27
Loads in Supermarkets 29.27
System Description 29.28
Space Conditioning Line 29.28
Operating Parameters in Rotary
Desiccant Dehumidifier 29.29
Heat-Pipe Heat Exchanger 29.29
Mixing of Process Air and Recirculating
Air 29.30
Indirect Evoporative Cooler or
Refrigeration 29.30
29.1 PACKAGED SYSTEMS
Comparison between Packaged and Central Systems
The differences in construction and operational characteristics between packaged and central systems
include these:
 A packaged system (PS) always uses a DX coil to cool the supply air directly in the PU, whereas
a central system (CS) adopts chilled water as the cooling medium. Refrigerants cool the chilled
water in the chiller, and the chilled water cools the supply air in the AHU. If both packaged and
central systems are using water-cooled condensers, a central system will need a 3 to 7°F (1.7 to
3.9°C) lower evaporating temperature than a packaged system.
 In a packaged system, the size of the air system, refrigeration system, and air conditioning system
that serves a specific area, a typical floor, an area of several floors, or the entire building is the
same. In a central system, the size of the conditioned area of an air and refrigeration system is
usually different.
 In a packaged system gas-fired furnace, electric heaters are often used to heat the air, and DX
coils, often air-cooled, or sometimes evaporative condensers are used. Many packaged units are
rooftop units and installed outdoors. Scroll and reciprocating compressors are usually used in
packaged systems. In a central system, there is often a central plant where boilers and chillers are
installed indoors to supply hot and chilled water to the AHUs and to cool and heat the supply air
there. Centrifugal, screw, and reciprocating chillers and water-cooled condensers are used in central
systems. AHUs are usually installed indoors in the fan rooms.
 In a packaged system, the controls of the heating and cooling system are often a part of the discharge
air temperature control in the packaged unit. In a central system, there are separate water
system controls, heating system controls, and refrigeration system controls in the central plant.
 In general, low- , medium- , and high-efficiency filters are used in packaged systems; and usually,
medium- and high-efficiency filters are used in central systems.
 Packaged units used in packaged systems are factory-fabricated and -assembled, whereas some
components in an AHU may be custom-built in the field.
 The modulation of cooling capacity in a packaged system with scroll and reciprocating compressors
is achieved by cycling of the cylinders or step controls. The control actions of the modulation
29.2 CHAPTER TWENTY-NINE
Gas Heater 29.30
Operating Parameters of the
Desiccant-Based Air Conditioning
Cycle 29.30
Part-Load Operation and Controls 29.31
29.11 CASE STUDY: A DESICCANT-BASED
AIR CONDITIONING SYSTEM FOR
RETAIL STORES 29.31
System Description 29.31
Operating Characteristics 29.31
Performance 29.32
29.12 CASE STUDY: A DESICCANT-BASED
AIR CONDITIONING SYSTEM FOR
OPERATING ROOMS 29.32
Indoor Environment of Operating
Rooms 29.32
System Description 29.33
29.13 APPLICATIONS OF
DESICCANT-BASED AIR CONDITIONING
SYSTEMS 29.34
Comparison between Conventional Vapor
Compression Refrigeration System and
Desiccant-Based Air Conditioning
System 29.34
Conditions to Apply Desiccant-Based Air
Conditioning Systems 29.34
System Characteristics 29.35
REFERENCES 29.35
of capacities in air, water, and refrigeration systems of a central system are often stepless, continuous
modulation controls.
Because of these differences, packaged systems and central systems are compared as follows:
First, consider the IAQ of the conditioned space that the air conditioning systems serve:
 Both the PS and the CS can provide the minimum ventilation air to dilute the air contaminants.
 Both the PS and CS can provide the humidity control and prevent wet surfaces and mold growth
in the ducts and conditioned space.
 Because of the lower fan total pressure provided by the small and medium-size packaged units,
low- and medium-efficiency filters are often used in small and medium-size PUs instead of
medium- and high-efficiency filters in AHUs
Second, the energy efficiency of the centrifugal and screw chillers is higher than that of reciprocating
and scroll compressors in packaged units, and the combined fan, motor, and drive efficiency
is higher in AHUs in CS than in PUs in PS. However, a higher fan total pressure in AHUs, greater
system heat gains, and an additional cooling medium require a greater evaporating-condensing
pressure lift than a direct-expansion (DX) system if PS and CS both are using either water-cooled
or evaporatively cooled condensers. All these cause the energy consumption of packaged systems to
be generally less than that of the central systems.
According to the DOE/EIA 1995 Commercial Buildings Consumption and Expenditures, the
annual energy use intensities of the PS and CS in 1995 in the United States are as shown below:
Third, packaged systems need more maintenance than central systems.
Fourth, rooftop packaged systems may be located directly above the occupied area and create
potential noise problems. Proper design and attenuation are important.
Fifth, chillers and chilled water systems in a central system have greater reliability than more
scattered packaged units in packaged systems. The flexibility of expansion and change of layout are
similar for both packaged and central systems.
Sixth, most of the packaged systems and the central systems need return air to mix with the outdoor
ventilation air, and install mechanical exhaust systems if required. Central systems are more
easily coordinated with smoke-control systems in multistory buildings than packaged systems.
Finally a packaged system has lower initial cost and often requires less space; factory-built units
are easier and faster to install than central systems.
Applications
Most current medium-size and large packaged units are equipped with microprocessor-based DDC
control systems, higher-efficiency scroll compressors, medium- and sometimes high-efficiency filters,
with many energy-efficient features, such as VAV boxes, air economizer, evaporative condenser, fault
detection and diagnostics. All this makes the packaged system the most widely used system in small
and medium-size air conditioning (HVAC&R) systems in both commercial buildings and households
EUI, kBtu/ ft2 yr (kWh /m2  yr)
Packaged systems
Water source heat pumps (HPs),
air source HPs, and other HPs 28.1 (8.2)
VAV systems 37.1 (10.9)
Single-zone, constant-volume 44.0 (12.9)
Central systems 44.7 (13.1)
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.3
in the United States, as discussed in Sec. 1.5. In commercial buildings in 1995, the packaged system
had 48 percent of the floor area among the air conditioned buildings, and in households, 55 percent.
The packaged unit is the primary equipment in a packaged system, as discussed in Sec. 16.5. In
a packaged system, air is often directly heated in the gas furnace or electric heater and is cooled by
DX coils. These are the primary differences between an AHU and a PU, and between a central and
a packaged system. On the other hand, because of the use of the microprocessor-based DDC system
in the packaged system, today the difference in indoor environmental control between a custombuilt
central system and a packaged system is less than ever before.
Types of Packaged Systems
Packaged systems can be subdivided according to their configuration and operating characteristics
into the following air conditioning systems:
 Single-zone constant-volume packaged system (SZCVPS)
 Single-zone VAV packaged system (SZVAVPS)
 VAV cooling packaged system (VAVCPS)
 VAV reheat packaged system (VAVRPS)
 Perimeter-heating VAV packaged system (PHVAVPS)
 Fan-powered VAV packaged system (FPVAVPS)
29.2 SINGLE-ZONE CONSTANT-VOLUME PACKAGED SYSTEMS
System Description
A single-zone, constant-volume packaged system is an air conditioning system that uses a packaged
unit to supply and return a constant-volume flow rate of conditioned air to and from a single-zone
conditioned space. It maintains a predetermined zone parameter at design and part-load conditions
by controlling the capacity of the gas furnace or refrigeration compressors. Single-zone packaged
systems are widely used in residences, indoor stadiums, arenas, and many industrial applications.
A typical single-zone constant-volume packaged system is simlar to that shown in Fig. 20.16a.
It consists of mainly an upflow gas furnace and a DX refrigeration system with the following:
 A supply fan, most often of forward centrifugal type. Centrifugal fans with airfoil blades are usually
used in large packaged units.
 A gas furnace with induced combustion, primary and secondary heat exchangers. An air-source
heat pump or an electric heater may be used to provide winter heating instead of a gas furnace.
 Low- and medium-efficiency filters.
 DX coil connected with an outdoor condensing unit.
 Supply and return ducts, diffusers, and return inlets.
 Space, functional, and safety controls.
 A heating element humidifier and an exhaust / relief / return fan (both are optional).
Supply Volume Flow Rate, and Cooling and Heating Coil Loads
For a single-zone, constant-volume packaged system, the supply volume flow rate , in cfm
[m3 / (60 s)], required to offset the space sensible cooling load at summer design conditions can be
V?
s
29.4 CHAPTER TWENTY-NINE
calculated from Eq. (20.69), as discussed in Sec. 20.17,
(29.1)
The supply volume flow rate required to offset the space heating load at winter design conditions
can be calculated from Eq. (20.70)
(29.2)
Usually, the supply volume flow rate required to offset the summer design sensible cooling load is
greater than the winter heating load. For convenience, the larger is used for both summer and
winter.
The DX coil load required at summer design conditions can be calculated from Eq. (20.48b)
(29.3)
and the heating coil load required to offset the heating load at winter design conditions can be calculated
from Eq. (20.39)
(29.4)
Controls
For a single-zone constant-volume system used in most residences in the United States, the primary
control is the zone temperature control; and a dual thermostat with two set points, for separate
cooling and heating mode operation during year-round zone temperature control, is used, as discussed
in Sec. 20.19.
For single-zone constant-volume systems used in many industrial and commercial buildings
where morning warm-up heating mode should be automatically changed over to cooling mode
when the space is occupied, a single-set-point heating and cooling mode automatic changeover
thermostat should be used. There is often a deadband of 1 to 3°F (0.6 to 1.7°C) between the heating
and cooling modes, or vice versa.
In addition to the zone temperature control, there are specific controls such as ignition control,
capacity conrols, head pressure controls, and safety controls. The burner is always energized first
and then the supply fan. Cooling and heating capacity controls include on /off and two-speed; and
several heating-stage step controls are often used.
Energy Use Intensities
According to DOE/EIA 1998 Commercial Buildings Consumption and Expenditure 1995, the
energy use intensity of the single-zone constant-volume packaged systems (or residential central
systems), in 1995 in the United States was 44 kBtu / ft2 yr (139 kWh/m2 yr). This EUI was
only smaller than that in central air conditioning systems using chilled water as the cooling
medium.
System Characteristics
System characteristics of a single-zone, constant-volume packaged system (SZCVPS) are listed in
Table 29.1.
Qch  60V?s scpa (T1  T2)
Qcc  60V?s s(hae  hcc)
V?
s
V?
s 
Qrh
60 scpa(Ts  Tr)
V?
s 
Qrs
60 scpa(Tr  Ts)
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.5
29.6 CHAPTER TWENTY-NINE
TABLE 29.1 System Characteristics of Packaged Systems
SZCVPS SZVAVPS VAVCPS, VAVRPS, PHVAVPS
Zone thermal and sound control
Control zone Single-zone Single-zone Multizone
Control methods Electric, two-stage DDC, electric DDC
thermostat
Control modes On/off PI, PID, on/ off PI, PID, on/ off
Heating-cooling mode changeover Manual, automatic Automatic Automatic
Sound control 35–45 dBA NC 30–45 NC 30–45
Indoor air quality (IAQ)
Minimum ventilation control Constant-volume DCV, MPC DCV, MPC
Filters Low- or medium- Low-, medium-, high- Low-, medium-,
efficiency efficiency high-efficiency
Humidity control Optional Optional Optional
Types of humidity control Heating element, Heating element, Heating element,
wetted element wetted element wetted element
Air systems
Types Constant-volume VAV, air mixing VAV, air mixing
air mixing
Supply fan Forward-curved Centrifugal, forward, Centrifugal, forward,
centrifugal airfoil airfoil
Fan total pressure 2.5 in. WC 4.0 in. WC 4.5 in. WC
Combined fan, motor, and drive 40% 40% 45%
efficiency
Volume flow control Constant-volume Inlet vanes, variable- Inlet vanes, (VSD)
speed drive (VSD)
Relief / return fan total pressure N/A 0.4–1.0 in. WC 0.4–1.0 in. WC
Combined fan, motor, drive 35% 45%
efficiency
Economizer Air, water Air, water Air, water
Cooling system
Refrigeration compressor Scroll, reciprocating Scroll, reciprocating Scroll, reciprocating
Capacity control On/off cycling On/off cycling, hot- Cycling, hotgas
bypass gas bypass
Refrigerants HCFC-22, HCFC-22, HCFC-22,
HFC-407C, HFC-407C, HFC-407C,
HFC-410A HFC-410A HFC 410A
Evaporator DX coil DX coil DX coil
Condenser Air-cooled Air-cooled, Air-cooled,
evaporatively cooled evaporatively cooled
Refrigerant flow control Expansion TXV, restrictor TXV, electric
valve (TXV), capillary tube expansion valve
Energy performance  5 tons, 9.3 EER 10–12 EER
Heating system
Types of heating system Gas furnace, Gas furnace, VAVCPS N/A;
heat pump heat pump, VAVRPS and PHVAVPS:
electric heating hot water, electric heating
AFUE Gas furnace, Gas furnace, Boiler,
conventional, 78%, conventional 78%, conventional 80%,
condensing 93% condensing 93% condensing 93%
Maintenance
Fault detection and diagnostics Packaged unit Packaged unit Packaged unit
N/A means not available.
29.3 SINGLE-ZONE VAV PACKAGED SYSTEMS
System Description
In a single-zone VAV packaged system, zone temperature is maintained by a DDC controller
through variation of the opening of the inlet guide vanes at the supply fan inlet or the supply fan
speed through a variable-speed drive. Therefore, the supply volume flow rate is varied to match the
zone load variation. A typical single-zone VAV packaged system is similar to that shown in Fig.
21.1 except for the following:
 The AHU is replaced by a packaged unit.
 The water-cooling coil is replaced by a DX coil.
 The heating coil is replaced by a gas furnace with primary and secondary heat exchangers.
An air economizer or a water economizer is often used to save energy. The cooling mode operation
at design conditions and part-load operation and the heating mode operation at design and partload
operation are similar to that discussed in Sec. 21.2.
System Calculations
The supply volume flow rate to offset the zone sensible load at cooling mode design conditions can
be calculated by Eq. (21.1) as
(29.5)
The DX coil load required at cooling mode design conditions can be calculated by Eq. (21.2) as
(29.6)
The heating capacity of the gas furnace required to offset the single-zone heating load during heating
mode design conditions can be calculated by Eq. (21.3) as
(29.7)
Controls
Zone temperature control and sequence of operations are similar to those discussed in Sec. 21.2 except
for the following:
 Cooling capacity is varied by cycling multiple compressors, or cylinder unloaders, or hot gas pass
at low part-load operation. Cooling capacity control is discussed in later sections.
 Heating capacity is varied by controlling the gas valve in on /off or two-stage control.
 Discharge air temperature control is still required for a better control quality and energy saving.
However, for a small packaged unit with only on /off cooling capacity control, zone temperature
control in cooling mode operation is actually the same whether the single-zone VAV packaged
system is employed with or without a discharge air temperature control.
Minimum ventilation control, head pressure control, and safety controls are similar to those discussed
in Chap. 23. Shirey (1995) recommended an improved fan cycling strategy based on the field
operating experience in the Salvador Dali Museum, St. Petersburg, Florida. The museum areas (storage
Qch  60V?s scpa(Thc  Ten)
Qcc  60V?s s(hrn  hcc)
V?
s 
Qrs
60 scpa(Tr  Ts)
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.7
area and lobby) are primarily under thermostat control. Prior to the improvements, all indoor fans operated
continuously, regardless of compressor operation, to maintain ventilation and air circulation for
occupant comfort. Continuous fan operation resulted in large humidity fluctuations and increased indoor
humidity levels at night when space load was reduced. Continuous fan operation also consumed
more energy and needed additional compressor operation to offset fan heat. If there are wetted surfaces
around the coil and condensate drain pan when the compressors are cycling off, moisture will be evaporated
from these wetted surfaces, extracted by the indoor fan, and supplied to the conditioned space.
An alternative improved control strategy “automatic fan mode” was adopted. The indoor fan
energized only when the compressors and heaters were operating. When the thermostat and
humidistat set points are met, the indoor fan stops and the moisture drains from the packaged unit.
The indoor fan that serves the storage room in the Salvador Dali Museum was changed to automatic
fan mode control. The new control strategy provided a better storage indoor environment with significant
energy savings. For the lobby, daytime continuous fan operation for constant air circulation
was used when the lobby was open to the public. Automatic fan mode was used at nighttime.
System Characteristics
System characteristics of a single-zone VAV packaged system are listed in Table 29.1. In Table 29.1,
DCV represents demand-controlled ventilation and MPC indicates mixing plenum pressure control.
29.4 COMPLIANCE WITH STANDARD 90.1–1999 THROUGH SIMPLIFIED
APPROACH OPTION FOR SMALL AND MEDIUM HVAC&R SYSTEMS
Single-zone, constant-volume and single-zone, variable-air-volume packaged systems are the two
widely used air conditioning systems in residential, commercial, and industrial buildings in the
United States. For air conditioning systems in buildings two stories or less in height and with less
than 25,000 ft2 (11,800 m2) gross floor area shall be considered in compliance with the requirements
of HVAC in ASHRAE/IESNA Standard 90.1-1999 by means of a simplified approach if the
following criteria are met:
1. The system serves a single zone.
2. A single or split-packaged unit is used which is either air-cooled or evaporatively cooled and
meets minimum efficiency requirements listed.
3. An air economizer and control shall be installed as required in Sec. 21.2 with either barometric
or powered relief sized to prevent space overpressurization.
4. Heating shall be provided by packaged heat pumps, PTHPs, gas-fired warm air furnaces, electrical
heaters, or hot water heating systems and boilers with all applicable efficiency requirements
met.
5. Outdoor air supplied by the system shall be equal to or less than 3000 cfm (1420 /L/ s) and less
than 70 percent of the supply air quantity at minimum outdoor air design conditions unless an
energy recovery ventilation system is provided.
6. The system is controlled by a manual changeover or dual set point thermostat.
7. Controls shall be provided to prevent auxiliary electric heater operation when the heat pump
alone can meet the required heating load.
8. Controls shall not permit reheat or simultaneous heating and cooling.
9. For spaces other than those requiring continuous operation having a heating or cooling capacity
greater than 65,000 Btu/h (30.7 kW), and having a supply fan motor power greater than 3/4 hp
(0.56 kW), a timeclock shall be provided
a. To start and stop the system under different schedules,
b. That is capable of retaining programming and time setting during a loss of power of at least
10 hours
c. That has an accessible manual override allowing temporary operation up to two hours
29.8 CHAPTER TWENTY-NINE
d. That is capable of being set back down to 55°F (12.8°C) and up to 90°F (32.2°C) during off
hours
10. HVAC&R piping, ducts, and plenums shall be insulated and meet the requirements listed in
Secs. 7.2, 17.2, and 17.4.
11. Duct systems shall be air balanced to within 10 percent of design airflow rates.
12. When separate heating and cooling equipment serve the same control zone, thermostats shall
be interlocked to prevent simultaneous heating and cooling
13. Exhausts with a volume flow rate over 300 cfm (140 L/ s) that do not operate continuously
shall be equipped with gravity or motorized dampers that will automatically shut when the exhaust
systems are not in use.
14. Air systems with a design volume flow rate greater than 10,000 cfm (4720 L/ s) shall have optimum
start controls.
29.5 VAV COOLING PACKAGED SYSTEMS
System Description
A VAV cooling packaged system is a multizone air conditioning system that provides conditioned
air without heating. It uses a packaged unit with DX cooling coils to condition the air and distribute
it to various control zones through VAV boxes, ducts, distributing devices, and controls. The zone
supply volume flow rate is modulated by the damper in the VAV box to match the variation of the
zone sensible load to maintain a preset zone temperature. A typical VAV cooling packaged system
with its system pressure characteristics is shown in Fig. 29.1. It consists of a packaged unit, many
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.9
FIGURE 29.1 A VAV cooling packaged system.
VAV boxes, ducts, diffusers, and controls. The air conditioning cycle of a VAV cooling packaged
system is the same as the air conditioning cycle in cooling mode operation for the interior zone in
Fig. 21.5a. VAV cooling packaged systems are used for interior zones in buildings where the heating
load is negligible.
Supply Volume Flow Rate and Coil Load
The supply volume flow rate for a control zone in the interior zone Vsin, in cfm [m3 / (60 s)], or the
supply volume flow rate of interior zone , in cfm [m3 /(60 s)], can be calculated by Eq. (21.1) as
(29.8)
where Qrin, Qri  sensible cooling load of control zone n and interior zone, Btu /h (W). The DX
coil load Qcc, in Btu / h (W), of the packaged unit that serves the interior zone at cooling mode design
conditions can be calculated by Eq. (21.2) as
(29.9)
Pressure Characteristics and Duct Static Pressure Control
In a VAV cooling packaged system with a system total pressure loss of 4.75 in. WC (1188 Pa), as
shown in Fig. 29.1, the total pressure loss, in in. WC, between various sections of the system at
peak supply volume flow may have values as follows:
The purpose of duct static pressure control is to achieve the following:
 Limit the maximum static pressure in the supply main ducts and maintain a preset static pressure
at the point where the static pressure sensor is located.
 Provide the required volume flow rate of supply air for any branch takeoff at both design load and
part-load conditions, and prevent starving of any VAV box.
 Minimize fan energy use at part load.
A starving VAV box is a VAV box with a supply volume flow rate lower than the amount required
to offset the zone load. This is because the set point of the duct static pressure control is not
properly set, the static pressure sensor is not properly located, or the supply duct and branch takeoffs
are not properly designed and sized. If the static pressure sensor is located at the end of the
main supply duct just before the last branch takeoff at point D, shown in Fig. 29.1, at peak supply
Flow, cfm
Peak flow 0.5 Peak flow
Return system 0.40 0.10
PU 2.25 0.56
Supply main duct 1.15 0.29
VAV box 0.60 0.90
Flexible duct and slot diffuser 0.40 0.10
Qcc  60V?s s(hm  hcc)
V?
si 
Qri
60 scpa(Tr  Ts)
V?
sin 
Qrin
60 scpa(Tr  Ts)
V?
si
29.10 CHAPTER TWENTY-NINE
volume flow rate when the total pressure loss of the VAV box, flexible ducts, and slot diffuser in a
branch takeoff 	p  1.0 in. WC (250 Pa), the static pressure at point D is 1.0 in. WC (250 Pa).
Point C is located upstream of point D. At peak supply volume flow rate, the mean air velocity inside
the main supply duct at point B, denoted by vB, is 1500 fpm (7.5 m / s), at point C the mean velocity
Vc is 1200 fpm (6 m / s), and at point D the mean velocity vD is 1000 fpm (5 m / s). The total
pressure loss between points B and C along the main supply duct is 0.3 in. WC (75 Pa), and between
points C and D it is 0.25 in. WC (63 Pa).
During part-load operation, when the volume flow rate has been reduced to 0.5 of peak supply
volume flow, the static pressure at point D is maintained at 1.0 in. WC (250 Pa). From Eq. (17.11),
pt  ps  pv (29.10)
and from Eq. (17.12)
pt1  pt2 	pf (29.11)
If the duct static pressure sensor is relocated at point C, then the pressure characteristics in the supply
main duct at points B, C, and D at peak and 0.5 of peak supply volume flow rate, in in. WG
(Pa), can be calculated as follows:
From these analyses, the following can be noted:
 At peak (design) supply volume flow rate, points upstream of point C where the duct static pressure
sensor is relocated (such as point B), both total pressure pt,up and static pressure ps,up of the
airstream inside the duct, are higher than duct total pressure and static pressure at point C, denoted
by ptC and psC. Points downstream of point C (such as point D) pt,do are lower than ptC, and
ps,do are lower than psC.
 At part-load operation, such as 0.5 of peak supply volume flow rate, pt,up 	 ptC, ps,up 	 psC, pt,do 
ptC, and ps,do  psC. But their differences are smaller than in design load.
 When duct static pressure sensor is located at point C, compared with the static pressure sensor
located at point D, at part-load operation of 0.5 of peak supply volume flow rate, the duct static
pressure at point D or at points downstream from point D, ps,do  1.0 in. WG. That is, the VAV
box connected to points downstream from where the static pressure sensor is located may starve.
 It is the total pressure of the airstream inside the supply main duct (duct total pressure) that offsets
the total pressure loss in the branch takeoff. However, duct static pressure is steadier than the
duct total pressure, and duct static pressure is usually sensed in a supply main duct for duct static
pressure control.
As discussed in Sec. 23.8, the set point of the duct static pressure control is the duct static
pressure plus the associated velocity pressure, i.e., the total pressure required to overcome the
Point D Point C Point B
Peak supply volume flow rate
Static pressure ps 0.78 1.00 1.25
Velocity pressure pv 0.062 0.09 0.14
Total pressure pt  ps  pv 1.09
pt1  pt2 	pf 0.084 1.39
Total pressure loss 	pf 0.25 0.30
0.5 peak supply volume flow rate
Static pressure ps 0.944 1.00 1.062
Velocity pressure pv 0.015 0.022 0.035
Total pressure pt  ps  pv 1.022
pt1  pt2 	pf 0.959 1.097
Total pressure loss between C and D, 	pf 0.063 0.075
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.11
maximum total pressure loss in any of the branch takeoffs connected to the supply main duct. And
0.1 in. WG (25 Pa) should be added as a safety factor.
For a supply duct system with similar configuration in branch take-offs, the static pressure sensor
of a duct static pressure control should be located near the remote end of the main duct as well
as at a location where steady static pressure can be properly measured, as shown in Fig. 29.1.
If there are two or three supply main ducts, two or three static pressure sensors should be installed.
Each should be located near the remote end of each main duct. A comparator is used so that
the DDC controller can pick the lowest static pressure as the feedback value to modulate the variable-
speed drive to provide a duct static pressure higher than the preset value in all main ducts.
System Characteristics
System characteristics of a VAV cooling packaged system are listed in Table 29.1.
29.6 VAV REHEAT PACKAGED SYSTEMS
System Description
A VAV reheat packaged system is a multizone system that uses a packaged unit with DX coil and
filters to condition the air and that supplies the conditioned air to various control zones in the
perimeter zone of a building through reheating VAV boxes, ducts, and diffusers, and to various control
zones in the interior zones through VAV boxes, ducts, and diffusers. The zone supply volume
flow rate of the cold supply air is modulated to match the variation of the zone loads during partload
operation.
Heating provided by the reheating coil is used to offset the zone heating load in winter in the
perimeter zone as well as to prevent the zone temperature drops below a preset temperature when
the zone supply volume flow rate has been reduced to the minimum setting during cooling mode
part-load operation. A reheating coil is not required in the VAV boxes in the interior zone.
As discussed in Sec. 21.3, for energy saving, the zone supply volume flow rate must be reduced
to a minimum setting, such as 30 percent of the peak supply volume flow rate during cooling mode
part-load operation before the reheating coil is energized. Figure 29.2 shows a VAV reheat packaged
system with a rooftop packaged unit. The air conditioning cycle for cooling mode operation is
the same as that shown in Fig. 21.5b. Usually, the same rooftop packaged unit is used to supply
conditioned air to both perimeter and interior zones. However, in winter, if two packaged units are
used to serve the perimeter and interior zone separately, the mixing air temperature for the perimeter
zone can be raised to a higher value, to save energy.
VAV reheat packaged systems are simple and effective. However, simultaneous cooling and
heating processes should be minimized. VAV reheat packaged systems are used in many commercial
buildings.
Supply Volume Flow Rate and Coil Load
In a VAV reheat packaged system, the zone peak supply volume flow rate , in cfm [m3 /(60 s)],
can be calculated from Eq. (21.4) as
(29.12) V?sn 
Qrsn
60 scpa(Tr  Tsn)
V?
sn
29.12 CHAPTER TWENTY-NINE
From Eq. (21.5), the block supply volume flow rate for perimeter zone , in cfm [m3 / (60 s)], can
be calculated as
(29.13)
From Eq. (21.6), the block supply volume flow rate for interior zone , in cfm [m3 / (60 s)], can be
calculated as
(29.14) V?si 
Qrsi
60 scpa(Tr  Tsi)
V?
si
V?
sx 
Qrsx
60 scpa(Tr  Tsx)
V?
sx
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.13
FIGURE 29.2 A VAV reheat packaged system.
From Eq. (21.8), the DX coil load Qcc, in Btu / h (W), can be calculated as
(29.15)
The reheating coil load depends primarily on the winter zone heating load. From Eq. (21.10), the
reheating coil load for a control zone in the perimeter zone Qchxn, in Btu / h (W), can be calculated as
(29.16)
Night Setback and Morning Warm-up
In multizone VAV reheat rooftop packaged systems, the heating of the zone air during the occupied
period in the perimeter zone in winter is often provided by zone electric reheating coils, or sometimes
by water heating coils. The gas furnace, heat pump, or electric heater in the rooftop packaged
unit is only used to maintain a night setback temperature and for morning warm-up purposes.
The reasons to supply warm air to maintain an indoor night setback temperature (such as 55°F,
or 12.8°C) are
 To prevent freezing of water pipes and water surfaces in areas where the outdoor temperature at
night is below 32°F (0°C)
 To provide an acceptable indoor temperature for emergency access
 To reduce the time required to warm up to a required temperature, say, 68 or 70°F (20.0 or
21.1°C), prior to an occupied period the next morning
During the night setback and morning warm-up periods, the following hold:
1. Outdoor dampers and exhaust dampers should be completely closed.
2. Recirculating dampers, inlet vanes, and inlet cones of the supply fan should be fully open; or the
variable-speed drive should be running at full speed.
3. All VAV boxes or reheating boxes should be fully open.
4. Both the refrigeration compressor and the relief fan should be turned off.
5. When the zone temperature exceeds a certain limit, both the supply fan and the furnace, heat
pump, or heater in the packaged unit can be turned off. The heating device in the packaged unit
will be energized again when the zone temperature drops below a certain limit.
6. The termination of the night setback period is the beginning of the warm-up period. The warm
air supply temperature from the rooftop packaged unit during the warm-up period is generally
between 100 and 120°F (37.8 and 48.9°C).
Evenly Distributed Airflow at DX Coils
In a VAV reheat packaged system, if the air flowing through the DX coil is not evenly distributed
over the entire coil surface, then liquid slugging of the reciprocating compressor, hunting of the
thermostatic expansion valve, and a decrease in the DX coil capacity may all occur at the same
time.
Refrigerant enters the various refrigerant circuits of the DX coil and the evaporator, typically as
a mixture of 75 percent liquid and 25 percent vapor, after passing through the thermostatic expansion
valve and the distributor tubes. If some of the refrigerant circuits have heavy refrigeration
loads and others have only very light loads, the refrigerant in circuits with heavy loads expands
rapidly to vapor, resulting in a greater vapor velocity and greater pressure loss. The refrigerant in
circuits with very low loads remains in a liquid state and flows to the compressor in the form of
 Qrhxn  60V?sxn scpa(Trxn  Tenn)
Qchxn  Qrhxn  Qvenxn
Qcc  60V?s s(hm  hcc)
29.14 CHAPTER TWENTY-NINE
liquid slugging. Liquid slugging may cause reciprocating compressor failure. Therefore, an even
airflow for the DX coil and thorough upstream air mixing must be maintained for both full- and
part-load operations, especially using reciprocating compressors. Liquid refrigerant lowers the temperature
of the sensing bulb of the thermostatic expansion valve and causes hunting.
Discharge Air Temperature Control for Packaged Systems
For a VAV reheat packaged system using reciprocating rooftop unit with a cooling capacity greater
than 20 tons (70 kW), its compressors are often controlled by using four-step capacity controls.
Multiple compressors are usually equipped in medium-size and large packaged units. Each compressor
may have two or more cylinders, so that two can be loaded and two unloaded at the same
time (see Fig. 11.23).
Air-Side Economizer Mode. For a typical DDC controller-activated discharge air temperature
control, as shown in Fig. 5.6, if the set point of the discharge temperature Tdis is set at 53°F
(11.7°C), which gives a supply temperature at the slot diffuser of Ts  55°F (12.8°C), there will be
a control band of 4°F (2.2°C). When a VAV reheat packaged system operates in an air economizer
mode and 51  Tdis  55°F (10.6  Tdis  12.8°C), the system will float within the control band
by mixing outdoor air with recirculating air.
Initiation of Cooling Stages. If Tdis 	 53°F (11.7°C) and due to the air economizer alone can no
longer balance the sensible coil load, then Tdis floats to the upper limit of the control band. Once it
reaches point 1, 55°F (12.8°C) in Fig. 5.6, the first-stage cooling is energized and Tdis drops below
55°F (12.8°C). First-stage cooling is most likely provided by the cooling capacity of two cylinders
in a compressor. It cycles on and off, and Tdis floats within the control band, with proportional-integral
(PI) control mode to maintain Tdis  53°F (11.7°C) with the least deviation. When the firststage
cooling is turned off during cycling, it needs a time delay of at least 4 min before it can be
turned on again, in order to prevent hunting and possible damage to the compressor motor. If, after
the first-stage cooling has been energized, the DX coil capacity still cannot balance the sensible coil
load, then Tdis continues to rise until it reaches point 2, which is 1°F (0.56°C) higher than the upper
limit of the control band.
If the time interval between the time the first stage turns on and the instant when Tdis reaches
point 2 is greater than 4 min, the first-stage cooling is then locked on and the second-stage cooling
is energized to cycle on and off in an attempt to maintain Tdis at 53°F (11.7°C). As in the first
stage, there must be a time delay of 4 min between the time the second-stage cooling turns off and
the monent it can be turned on again. When Tdis floats within the control band, it may drop to point
3 because of the low sensible coil load. If the cooling capacity of having the first-stage cooling
locked on and the second-stage cooling cycling still cannot offset the sensible coil load, Tdis will
rise until it reaches a value 1°F (0.56°C) higher than the upper limit of the control band, such as
point 4. At point 4, the third-stage cooling is cycling, and the first- and second-stage cooling is
locked on.
In this type of packaged unit, the greatest cooling capacity is provided when the fourth-stage
cooling is cycling and the first-, second-, and third-stage cooling is locked on.
As Tdis floats within the control band, any deviation from the set point of 53°F (11.7°C) is integrated
over time as part of the proportional-integral control mode, in an attempt to reduce the deviation
to zero, i.e., to maintain a near-constant temperature at the set point of 53°F (11.7°C), e.g.,
points 5 and 6 in Fig. 5.6.
If Tdis is being controlled in air economizer and cooling mode with refrigeration, the greater the
deviation from the set point, the shorter the time needed for corrective action in order to provide
control stability.
When a decrease in the coil load causes Tdis to drop to a value 1°F (0.56°C) lower than the lower
limit of the control band (e.g., point 7), the currently cycled stage is locked off and the next-lower
cooling stage becomes the cycling stage.
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.15
Successive decreases in Tdis to 1°F (0.56°C) below the lower limit of the control band cause repeated
locking off of cooling stages until the second stage is locked off and the first stage becomes
the cycling stage. When the first stage is cycled off, the liquid line solenoid valve is deenergized, but
the compressor still operates to pump down the refrigerant to the condenser. When the suction pressure
in the DX coil becomes lower than the limit of low-pressure control, it opens the compressor
circuit, and both the compressor and the condenser fans are stopped. A fully intertwined DX coil will
adjust its activating refrigerant circuits accordingly during step control of the compressor capacity.
At design load, the evaporating temperature inside the DX coil is usually between 40 and 45°F
(4.4 and 7.2°C). Hot gas bypass is only used to prevent the DX coil from frosting in case Tev falls
below 32°F (0°C) because of a sudden load decrease.
Reset. For a DDC-controlled VAV packaged system that uses a proportional-integral control
mode, Tdis can be reset based on either the space air temperature floating within the deadband or the
outdoor temperature To. When Tdis is reset based on the space temperature, it is preferable to have
several space temperature sensors connected in series. Their average temperature Trm, in °F (°C), is
used to reset Tdis. As Trm floats within the deadband between 70 and 75°F (21.1 and 23.9°C), each
1°F (0.56°C) drop in Trm corresponds to a 1 to 2°F (0.56 to 1.1°C) increase in Tdis.
During cooling mode, if Tdis is reset based on the outdoor temperature To, then Tdis will not be
reset when To 70°F (21.1°C). After To falls below 70°F (21.1°C), for every 5°F (2.8°C) that To
drops, Tdis will increase 1°F (0.56°C). For instance, at an outdoor temperature To  50°F (10.0°C),
the set point of Tdis will be 57°F (13.9°C). For larger packaged units with multiple scroll compressors,
cooling capacity is often divided into four capacity steps such as when reciprocating compressors
are used.
Fan Modulation
Three types of supply fan modulation are widely used in large VAV packaged systems with rooftop
packaged units to maintain a supply duct static pressure near the most remote branch takeoff in the
supply main duct: (1) a forward-curved centrifugal fan with inlet vanes, (2) an airfoil centrifugal
fan with a variable-speed drive, and (3) an airfoil centrifugal fan with inlet cone modulation. Because
of the smaller fan inlet in an airfoil centrifugal fan, it is not recommended that inlet vanes be
used, in order to prevent an extremely high inlet velocity, such as the actual maximum inlet velocity
which exceeds 5000 fpm (25 m / s).
Both forward-curved centrifugal fans with inlet vane modulation and airfoil centrifugal fans
with inlet cone modulation have a lower cost than airfoil fans that use a variable-speed drive. However,
variable-speed drive fan modulation is more energy-efficient at part-load operations. A cost
analysis is always helpful to determine whether having a variable-speed drive or having inlet vanes
is optimum.
In a VAV packaged system using inlet vanes for its forward-curved centrifugal fan modulation,
consider the following:
 Because all inlet vanes are subject to a certain amount of air leakage at the completely closed
position, the fan volume flow rate will still be about 10 to 30 percent of the design flow when the
inlet vanes are completely closed. A modulation range from 100 to about 30 percent is usually
sufficient for a VAV packaged system.
 A forward-curved centrifugal fan has a smaller fan surge area at a lower fan total pressure and
flow rate than an airfoil centrifugal fan.
 Fan modulation resulting in an external total pressure drop greater than 4 in. WC (100 Pag) in a
VAV packaged unit may cause damage to the unit.
 A discharge damper should not be installed because closing the VAV boxes can provide the same
shutoff function. A discharge damper may cause excessive noise due to overpressurization of the
VAV boxes. It may also overpressurize the ducts.
29.16 CHAPTER TWENTY-NINE
A VAV packaged system is shut off during nighttime unoccupied hours in summer and off seasons.
In locations where there is a cold winter, it operates to provide warm air and maintain a night
setback indoor space temperature during unoccupied hours.
Air-Cooled, Water-Cooled, and Evaporatively Cooled Condensers
In rooftop packaged units using reciprocating compressors, the energy efficiency ratio (EER) for
air-cooled, water-cooled, and evaporatively cooled condensers at an outdoor dry-bulb temperature
of 95°F (35°C) and a wet-bulb temperature of 75°F (23.9°C) is as follows:
Air-cooled condenser 9–12
Water-cooled condenser 11–14
Evaporative-cooled condenser 14–16
The EER for VAV packaged units using scroll compressors is higher than that for those using
reciprocating compressors. Air-cooled and evaporatively cooled condensers often have a higher installation
cost than water-cooled condensers and cooling towers.
Although water-cooled and evaporatively cooled condensers are considerably more energy-effi-
cient at an outdoor condition of 95°F (35°C) dry-bulb and 75°F (23.9°C) wet-bulb temperatures
than air-cooled condensers, in many locations the annual average dry-bulb temperature is rather
low. Shaffer (1987) used a manufacturer’s computer program to compare the annual energy use of
air-cooled and water-cooled reciprocating chillers for multistory buildings using VAV reheat systems
in Boston and San Diego. The annual energy use of a water-cooled condenser and cooling
tower is about 10 percent higher than that of an air-cooled condenser, mainly due to the energy use
of condenser pumps. Comparing the energy use and costs of air-cooled, water-cooled, and evaporatively
cooled condensers is recommended in order to make an energy-efficient and cost-effective selection.
Controls and Sound Problems for Rooftop Packaged Units
Microprocessor-based specific controls, safety controls, and diagnostics for a 1997 manufactured
rooftop packaged unit with a cooling capacity between 20 to 130 tons (70 to 457 kW) are listed in
in Sec. 16.5. Noise considerations for rooftop top packaged units are discussed in Sec. 19.9.
Case Study: VAV Reheat Packaged System for Precision Manufacturing
Desmone and Frank (1992) reported an addition of 10,000 ft2 (930 m2) for sophisticated precision
manufacturing for Oberg Industries, a manufacturer of high-quality carbide and steel stamping dies
in Freeport, Pennsylvania. The required HVAC&R criteria are as follows:
 Temperature is maintained at a constant level of 68 
 1°F (20 
 0.56°C) not only horizontally
but also vertically from the floor to 13 ft (3.96 m) high.
 Humidity should be maintained at 50 percent.
 Airflow noise had to be limited and drafts and turbulence should be tightly constrained.
The selected HVAC&R system is a VAV reheat packaged system with two rooftop packaged units
that feed cold air year-round into a medium-pressure duct loop. Machine oil duct exhaust, grinding
duct exhaust, conditioned air return, medium-pressure supply duct, and the lowest VAV boxes and
reheating hot water piping are arranged in layers to facilitate servicing.
There are 19 pneumatically controlled VAV boxes located throughout the extended plant with a
two-stage modulation in sequence. The modulation of the supply air volume flow by the damper in
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.17
the VAV box provides the first-stage control. If the zone cooling load drops below the minimum setting
of air delivery by the VAV box, the pneumatic controller opens the hot water valve, forcing the
hot water to enter the reheating coil, and modulates the hot water flow to maintain the required exact
zone temperature. The air economizer cycle is adopted to save energy. The HVAC&R system is
designed with a 75 percent backup capacity for reliability.
System Characteristics
System characteristics of VAV reheat packaged systems are listed in Table 29.1.
29.7 PERIMETER-HEATING VAV PACKAGED SYSTEMS
A perimeter-heating VAV packaged system is a VAV multizone system that uses packaged units
with DX coils to condition the air, During cooling mode operation, the conditioned air is supplied
to the control zones in the perimeter and to interior zones through VAV boxes, ducts, and diffusers.
In heating mode operation, heating is provided to the perimeter zone by a perimeter-heating system
using either hot water finned-tube heaters or electric baseboard heaters to offset zone heating loads.
Ventilation is provided by the cold ventilation air supplied from the VAV boxes at 30 percent of the
peak supply volume flow rate minimum settting.
Perimeter VAV packaged systems are suitable for multizone commercial buildings in locations
with long, cold winters.
Similar to that in Sec. 29.5, for VAV reheat packaged systems, the supply volume flow rate for
control zones in the perimeter and in interior zones , in cfm [m3 /(60 s)], can be calculated from
Eqs. (29.12). The supply volume flow rates for perimeter or interior zone Vsx or Vsi , in cfm [m / (60
s)], can be calculated from Eq. (29.13) or (29.14). The heating load for perimeter zone Qchx, in
Btu/ h (W), can be calculated from Eq. (29.16).
System Characteristics
System characteristics of a perimeter-heating VAV packaged system are listed in Table 29.1.
29.8 FAN-POWERED VAV PACKAGED SYSTEMS
System Description
A fan-powered VAV packaged system (FPVAVPS) is a multizone system using a packaged unit to
condition the air and to distribute the conditioned air through fan-powered VAV boxes, ducts, diffusers,
and controls in the perimeter zone for an air system with conventional air distribution as
shown in Fig. 21.13 except that a packaged unit is used to condition the air; or it distributes the conditioned
air through the fan-powered VAV boxes, ducts, diffusers, and controls in either perimeter
or interior zones for an air system with cold air distribution.
After the single-blade volume control damper has been closed to a minimum setting, such as 30
percent, the function of a fan-powered box is, first, to extract recirculating warm plenum air and
mix it with the cold primary air and, second, to energize a reheating coil to maintain a preset zone
temperature during part-load cooling mode operation. In winter heating mode, an electric heater in
several capacity stages is controlled to maintain a preset zone temperature.
There are two types of fan-powered boxes: parallel fan-powered boxes and series fan-powered
boxes. Parallel fan-powered boxes save fan energy more than series fan-powered boxes and are
V?
sn
29.18 CHAPTER TWENTY-NINE
more widely used. Fan-powered VAV packaged systems with parallel fan-powered boxes using an
electric heater or hot water heating coils have been used in many commercial buildings.
Supply Volume Flow Rate and Coil Load
As discussed in Sec. 21.5, to calculate the zone peak supply volume flow rate at summer design conditions
, in cfm [m3 /(60 s)], one must consider whether it is a control zone in the perimeter zone or
in the interior zone, and whether a conventional air distribution is used or a cold air distribution is used.
 For conventional air distribution in the perimeter zone with a cooling supply temperature differential
(Tr  Ts)  15 to 20°F (8.3 to 11.1°C), from Eq. (21.18), the zone peak supply volume flow
rate from the fan-powered box , in cfm [m3 /(60 s)], can be calculated as
(29.17)
For conventional air distribution in interior zones, the zone peak supply volume flow rate from the
VAV boxes , in cfm [m3 /(60 s)], can be calculated as
(29.18)
 For cold air distribution in the perimeter zone with a cooling supply temperature differential typically
Tr  Tsc,dis  32°F (17.8°C), the peak zone supply volume flow rate of the cold primary air
from the fan-powered box , in cfm [m3 / (60 s)], can be calculated as
(29.19)
As discussed in Sec. 21.5, the ratio between the recirculating plenum air from the fan-powered
box and the peak zone supply volume flow rate . Also, .
Therefore,
(29.20)
where , peak zone supply volume flow rate and peak zone supply volume flow rate of
recirculating plenum air of control zone xn in perimeter zone, cfm [m3 /(60 s)]
 For cold air distribution in the perimeter zone during winter heating design conditions, the peak
zone supply volume flow rate should be calculated by Eq. (21.18) so that the supply temperature
differential will be equal to or less than 15°F (8.3°C), to prevent excessive buoyancy and
stratification.
 For cold air distribution in the interior zone, each control zone can be either with or without a fanpowered
box, as discussed in Sec. 18.5. Supply of low-temperature air with high induction nozzle
diffusers is recommended. When fan-powered boxes are used for cold air distribution in the
interior zone, the supply volume flow rate of cold primary air , in cfm [m3 /(60 s)], can be
similarly calculated by Eq. (29.19) except that the zone sensible cooling load in the perimeter
zone Qrsxn should be replaced by zone sensible cooling load in the interior zone Qrsin, both in
Btu/ h (W). As the ratio is equal to 0.67, the zone peak supply volume flow rate of the
recirculating air from the fan-powered box , in cfm [m3 /(60 s)], is therefore determined.
 From Eq. (29.13), the supply volume flow rate for the perimeter zone at summer design conditions
can be calculated. From Eq. (29.14), the supply volume flow rate for the interior zone at
summer design conditions can be similarly calculated, both in cfm [m3 /(60 s)]. V?si
V?
sx
V?
sin,r
V?
sin,r /V?sin,c
V?
sin,c
V?
sxn
V?
sxn,r  V?sxn
V?
sxn,r  V?sxn  V?sxn,c
V?
sxn,c  V?sxn,r  V?sxn V?sxn,r /V?sxn,c  0.67
V?
sxn,c 
Qrsxn
60 scpa(Tr  Tsc,dis)
V?
sxn,c
V?
sin 
Qrsin
60 scpa(Tr  Ts)
V?
sin
V?
sxn,d 
Qrsxn
60 scpa(Tr  Ts)
V?
sxn,d
V?
sn,d
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.19
 From Eq. (29.15), the DX coil load Qcc in the packaged unit of a fan-powered VAV packaged system
can be calculated, and from Eq. (29.16) the reheating coil load Qchxn in each of the fan-powered
boxes in the perimeter zone can be similarly calculated, both in Btu / h (W).
Controls
A fan-powered VAV packaged system has two air mixings: outdoor air and recirculating air mixing
in the mixing plenum of the packaged unit, and the cold primary air and recirculating plenum air
mixing in the fan-powered box. Minimum outdoor ventilation air control is critical. Both demandcontrolled
ventilation using CO2 sensors and mixed plenum pressure control are discussed in Chap.
23 and should be carefully investigated during the peak supply volume flow rate as well as when
the volume flow rates of the cold primary air in control zones have been reduced to minimum
settings.
Zone controls and the sequence of operations of fan-powered boxes are discussed in Sec. 21.5.
Microprocessor-based specific safety controls and diagnostics for a late-1990s manufacturered
packaged unit are similar to those listed in Sec. 16.5.
Case Study: A Fan-Powered VAV Packaged System with
Rooftop Packaged Unit
This case study is a retrofit project. In it, an HVAC&R system is used to serve a 48,000-ft2 (4461-
m2) medical office building in Little Rock, Arkansas. The building was constructed in 1979. It is
owned by a limited partnership of 16 physicians. The renovation of the HVAC&R system of this
project won a 1992 ASHRAE Technology Award, second place, of existing commercial buildings.
As described by Tinsley et al. (1992), prior to the 1990 renovation, the HVAC&R system was a
constant-volume electric terminal reheat rooftop packaged system. There were six rooftop packaged
units with a total capacity of 120 tons (422 kW) and 162 electric duct heaters with an approximate
total capacity of 450 kW.
For the renovation, a fan-powered VAV packaged system with a rooftop packaged unit was designed
and installed. One rooftop packaged unit of 133 tons (468 kW) with 100 electric heated parallel
fan-powered units was equipped to serve this medical office building.
The renovation mainly includes the following:
 The air economizer and the compressors are controlled in sequence by DDC discharge air temperature
control in order to maintain a supply air temperature at 55°F (12.8°C). When the outdoor air
temperature To  60°F (15.6°C), all compressors are locked out.
 An energy-efficient evaporatively cooled condenser is used. The energy use of the compressors,
tower fans, and spray pump at full load and at an outdoor wet-bulb temperature of 80°F (26.7°C)
is 0.79 kW/ ton (4.45 COP).
 An airfoil centrifugal fan with inlet cone modulation is used because it uses less fan energy than a
forward-curved centrifugal fan with inlet vanes.
 A relief fan is used instead of a return fan. This relief fan operates only when the system is in air
economizer mode.
 A lower duct static pressure control set point of 0.75 in. WG (188 Pag) is used.
 The fan-powered VAV packaged system operates only when any one of the 16 suites is occupied.
When any suite is unoccupied, the primary air damper in the VAV box is completely closed. The
fan and the electric heater in the parallel fan-powered box are controlled in sequence to maintain
a zone temperature 10°F (5.6°C) lower than the zone temperature set point during unoccupied
hours.
 All supply ducts have been sealed and insulated.
29.20 CHAPTER TWENTY-NINE
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.21
TABLE 29.2 System Characteristics of Fan-Powered VAV Packaged System and Desiccant-Based Air Conditioning
Systems
FPVAVPS DBACS
Zone thermal and sound control
Control zone Multizone Single-zone or multizone
Control methods DDC or thermostat; Thermostat, humidistat, or DDC
high-mediumlow
fan speed
Control modes On/off, PI, or PID On/off, PI, or PID
Heating-cooling mode changeover Automatic Automatic, manual
Sound control 35–45 dBA NC 30–45
Indoor air quality (IAQ)
Minimum ventilation air control Demand-controlled Demand-controlled
ventilation, or mixing ventilation, or mixing
plenum pressure control plenum pressure control
Filters Medium- or high-efficiency Medium- or high-efficiency
Humidity control Optional Heating element humidifier
and desiccant dehumidifier
Air systems
Types VAV air mixing Constant-volume or VAV
Supply fan Centrifugal, forward- Centrifugal, forwardcurved
or airfoil curved or airfoil
Supply fan total pressure 4.5 in. WC 4.5–6 in. WC
Combined fan, motor, drive efficiency 40–45% 45–55%
Volume flow control Inlet vanes, inlet cone, N/A
variable-speed drive (VSD)
Air economizer Air, temperature, or enthalpy N/A
Fan-powered box fan Centrifugal, forward-curved N/A
Fan total pressure 0.5 in. WC
Combined fan, motor, drive efficiency 25%
Relief / exhaust fan Axial or centrifugal Centrifugal
Fan total pressure 0.4–1.0 in. WC 1–1.5 in. WC
Combined fan, motor, drive efficiency 25–35% 30–40%
Cooling systems
Refrigeration compressor Scroll or reciprocating Centrifugal, screw, scroll,
or reciprocating
Refrigerants HCFC-22, HFC-407C, Centrifugal HCFC-123,
HFC-410A screw HCFC-22, scroll
HFC-407C, HFC-410A
Energy performance 10–12 EER Centrifugal 0.5–0.55 kW/ ton,
screw 0.7–0.75 kW/ ton,
scroll 10–12 EER
Evaporator DX coil DX coil, liquid cooler
Condenser Air-cooled, or evaporatively Air-cooled, water-cooled,
cooled or evaporatively cooled
Capacity control On/off, multiple compressors, Water flow, or multiple
hot-gas bypass compressors
Heating system Gas furnace, hot-water coil, Gas heater, electric heater,
or electric heater or hot-water coil
AFUE Conventional 78%, Conventional 78%,
condensing 93% condensing 93%
Maintenance
Fault detection and diagnostics Packaged unit
 The zone temperature can be maintained between 70 and 80°F (21.1 and 26.7°C) in any control
zone by the DDC system.
The annual energy use for the HVAC&R system in this building before renovation was 169,150
Btu/ ft2 yr (533 kWh/m2  yr), and the annual energy cost was $2.92 / ft2  yr ($31.4 /m2  yr). After
renovation, the energy use intensity had dropped to 55,890 Btu / ft2 yr (176 kWh/m2 yr) and the
energy cost was $1.40 / ft2 yr ($15.1 /m2  yr), drops of 67 and 52 percent, respectively.
System Characteristics
System characteristics of fan-powered VAV packaged systems are listed in Table 29.2.
29.9 DESICCANT-BASED AIR CONDITIONING SYSTEMS
Desiccant-Based Air Conditioning
A desiccant-based air conditioning process is a combination of desiccant dehumidification, evaporative
cooling, supplementary compression refrigeration, and the regeneration or reactivation of the
desiccant by means of waste heat or gas heating. A desiccant-based air conditioning system is a
hybrid system of desiccant dehumidification, evaporative cooling, refrigeration, and regeneration
systems to cool and dehumidify the space air and maintain it at a required temperature and relative
humidity with adequate outdoor ventilation air, at the same time improving the efficiency of energy
use. In 1983, ASHRAE awarded the Willis Carrier Prize to Nancy Banks, a 23-year-old engineer
who proposed a desiccant dehumidification process with evaporative cooling and necessary supplementary
refrigeration in commercial buildings. It was a new idea at that time that considerably
improved the energy use of refrigeration and started a new era of commercial desiccant-based air
conditioning.
Desiccant Dehumidification and Sensible Cooling
When moist air flows over a bed of either solid or liquid desiccant, absorption occurs at the bed of
sorbents. Absorption is a moisture soption process associated with physical or chemical changes,
while adsorption is one without physical or chemical changes. For liquid absorbents, such as liquid
lithium chloride, the absorption of moisture is mainly caused by the vapor pressure difference
between the moist air and the surface of the liquid. Solid adsorbents such as silica gels attract moisture
because of the vapor pressure difference and the electric field at the desiccant’s surface.
During the process of sorption, heat is released. This released heat, often called the heat of sorption,
is the sum of the following:
 Latent heat due to the condensation of the absorbed water vapor into liquid
 Heat of wetting when the surface of the solid sorbent is wetted by the attached water molecules,
or the heat of solution when moisture is absorbed by the liquid sorbents
Heat of wetting varies from a relatively large value when new desiccant first absorbs moisture to a
very low value when it is nearly saturated.
During the desiccant dehumidification process, the humidity ratio of the leaving moist air is decreased.
Although there is a heat loss through the outer casing of the dehumidifier to the ambient
air, it is rather small compared with the release of the heat of sorption and any residual heat remaining
after the regeneration process.
Regeneration uses heat to drive off the accumulated moisture so that the sorbent can be reused.
The temperature of the dehumidified air is considerably increased because of the heat released
29.22 CHAPTER TWENTY-NINE
during sorption.The desiccant dehumidification process line deviates from the thermodynamic wetbulb
temperature line and is represented by line di-do on the psychrometric chart in Fig. 29.3.
The decrease in humidity ratio during the desiccant dehumidification process is given by
Difference in humidity ratio  wdi  wdo (29.21)
where wdi, wdo  humidity ratio at dehumidifier inlet and outlet, lb / lb (kg / kg). The reduction of latent
heat Ql, in Btu / h (W), can be calculated as
(29.22)
where volume flow rate of air entering dehumidifier, cfm [m3 /(60 s)]
o  air density of entering air, lb/ ft3 (kg /m3)
hfg  latent heat of vaporization, Btu/ lb (J / kg)
Desiccant dehumidification is a dehumidification and heating process on the psychrometric chart,
and it is illustrated by a straight line inclined downward, deviating slightly from the thermodynamic
wet-bulb temperature line, as shown in Fig. 29.3. It has a flatter negative slope.
The acute angle between the desiccant humidification process and thermodynamic wet-bulb
temperature depends mainly on the desiccant material and the water-holding capacity of the desiccant.
When the desiccant approaches saturation, angle is smaller than that when a new desiccant
first absorbs moisture from the surrounding air. For a rotary desiccant dehumidifier impregnated
with lithium chloride, if manufacturer’s data are not available, angle can be estimated at 5°.
The regeneration process removes the absorbed moisture from the desiccant. For the regeneration
airstream, it is also a humidifying and cooling process. Molecular sieves require a higher regeneration
temperature than silica gels and lithium chloride.
V?
o 
Ql  60V?o o(wdi  wdo)hfg
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.23
FIGURE 29.3 Desiccant dehumidification and sensible cooling.
After desiccant dehumidifcation, the extremely high-temperature air is usually sensibly cooled
in two stages in order to save energy. This process is represented by line do-sc1 on the psychrometric
chart in Fig. 29.3. The first stage uses an indirect evaporative cooling process, or is cooled by an
outdoor airstream in an air-to-air heat exchanger, or is cooled by the evaporator end of a heat-pipe
heat exchanger. In an indirect evaporative cooling process, the hot air is sensibly cooled by another
evaporatively cooled airstream without direct contact.
The first-stage sensible cooling load Qsc1, in Btu / h (W), can be calculated as
(29.23)
where Tdo, Tsc1  temperature of dehumidified air entering and leaving indirect evaporative cooler,
air-to-air heat exchanger, or heat pipe heat exchanger, °F (°C). The end state of first-stage sensible
cooling sc1 can be determined according to local outdoor dry- and wet-bulb temperatures to use
fully the indirect evaporative cooling to save refrigeration in second-stage sensible cooling.
In the second stage, dehumidified air may be mixed with the recirculated air and then enter a
DX coil for further sensible cooling. The second-stage sensible cooling load Qsc2, in Btu / h (W),
can be calculated as
(29.24)
where Tsc2  temperature of dehumidified air leaving coil, °F (°C).
Desiccant-Based Air Conditioning Systems
In an air conditioning system, the coil load can be divided into sensible load and latent load. For
comfort systems in commercial buildings, the latent load varies from 5 to 35 percent of the total
coil load. In applications such as supermarkets, it may exceed 50 percent of the total coil load.
A desiccant-based air conditioning system is a system in which latent cooling is performed by
desiccant dehumidification and sensible cooling by evaporative cooling or refrigeration, as shown
in Fig. 29.4. Thus, a considerable part of the expensive vapor compression refrigeration is replaced
by inexpensive evaporative cooling. There are two airstreams in a desiccant-based air conditioning
system: a process airstream and a regenerative airstream. Process air can be all outdoor air or a mixture
of outdoor and recirculating air. Process air is also the conditioning air supplied directly to the
conditioned space or enclosed manufacturing process, or to the packaged unit or sometimes to the
air-handling unit, or terminal for further treatment. A regenerative airstream is a high-temperature
airstream used to reactivate the desiccant.
A typical desiccant-based air conditioning system, or simply a desiccant-based system, consists
of mainly the following components: rotary desiccant dehumidifier, heat-pipe heat exchanger,
indirect evaporative cooler, scroll or reciprocating vapor compression unit with DX coil or watercooling
coil, gas-fired heaters, fans, pumps, filters, controls, outer casing, ducts, and piping. Recently,
the heat-pipe heat exchanger has been used instead of a rotary heat exchanger in the earlier
desiccant-based systems because a heat-pipe heat exchanger has no cross-contamination or moving
parts. Any of these components may be replaced by other components of similar function.
In addition to these components, a desiccant-based system may need a return air system, smoke
control systems, and mechanical exhaust systems.
Desiccants
Either solid or liquid desiccant absorbs or releases moisture because of the difference in vapor pressure
between the surface of the desiccants pdes and the surrounding air psur. In a desiccant dehumidi-
fication process, when pdes  psur, the desiccant absorbs moisture from the ambient air. In a regenerative
process, pdes 	 psur, so moisture is released from the desiccant to the surrounding air.
Qsc2  60V?o ocpa(Tsc1  Tsc2)
Qsc1  60V?o ocpa(Tdo  Tsc1)
29.24 CHAPTER TWENTY-NINE
FIGURE 29.4 A desiccant-based air conditioning system for a supermarket: (a) schematic diagram; (b) desiccantbased
air conditioning cycle.
Desiccants can be classified as adsorbents, which absorbs moisture without accompanying physical
and chemical changes, and absorbents, which absorb moisture accompanied by physical or
chemical changes. Three kinds of desiccants are widely used in desiccant-based air conditioning
systems: silica gel, lithium chloride, and molecular sieves.
Silica Gels. These are solid desiccants and adsorbents. Structurally, they contain numerous pores and
capillaries in which water is condensed and contained. Silica gel has a high capacity to absorb moisture
and releases it at a higher temperature. They are low in cost and available in sizes from -in.
(4.8-mm) beads to powderlike grains. Recently, titanium silica gel is used for its stable property.
Lithium Chloride (LiCl). This is an absorbent. It is in dry form when each LiCl molecule holds
two water molecules. If each LiCl molecule holds more than two water molecules, it becomes a liquid
and continues to absorb moisture. LiCl has a high capacity to absorb and to hold moisture.
Lithium chloride is widely used in rotary wheel dehumidifiers.
Molecular Sieves. These are actually synthetic zeolites, a solid desiccant and an adsorbent in the
form of crystalline aluminosilicates produced by a thermal process. Molecular sieves show physical
stability and high moisture-releasing capacity at high regenerating temperatures of 248 to 428°F
(120 to 220°C), and they are recommended in direct gas-fired applications.
316
29.26 CHAPTER TWENTY-NINE
FIGURE 29.5 Sorption isotherms of some desiccants. (Source: ASHRAE Handbook 1989, Fundamentals.
Reprinted with permission.)
As defined in Sec. 3.4, a sorption isotherm is a constant-temperature curve that indicates the relationship
between the moisture content of the desiccant Xdes, or moisture absorbed as a percentage
of its dry mass, and the relative humidity of the surrounding air sur. Figure 29.5 shows the sorption
isotherms of silica gels, lithium chloride, and molecular sieves. LiCl has a much higher water-holding
capacity than silica gels and molecular sieves.
Rotary Desiccant Dehumidifiers
A rotary desiccant dehumidifier is a rotary wheel that removes moisture from the airstream flowing
through it. A rotary desiccant dehumidifier has similar construction and operating characteristics to
the rotary total heat exchanger, as discussed in Sec. 12.5. Both are rotary wheels. The main differences
between a rotary desiccant dehumidfier and a rotary total heat exchanger or rotary heat exchanger
are as follows:
 The purpose of using a rotary desiccant dehumidifier is to remove moisture from an airstream,
while the purpose of using a rotary total heat exchanger is to transfer heat energy from one
airstream to another airstream.
 In a rotary desiccant dehumidifier, there is a process airstream and a regeneration airstream. The
regeneration airstream has a very high temperature, whereas in a rotary total heat exchanger, there
is an outdoor airstream and an exhaust airstream at indoor temperature.
 A rotary desiccant dehumidifier rotates at a speed of 6 and 10 r /h (revolutions per hour), whereas
a rotary heat exchanger rotates at a speed of 10 to 25 r / min.
29.10 CASE STUDY: A DESICCANT-BASED AIR CONDITIONING
SYSTEM FOR A SUPERMARKET
Loads in Supermarkets
A desiccant-based air conditioning system is operated at an open cycle when the process air is entirely
outdoor air. If the process air at the inlet to the rotary desiccant dehumidifier is a mixture of
recirculating air and outdoor air, the desiccant-based system is said to be operated at a closed cycle.
For supermarkets with many frozen food refrigerators, especially single-deck well types and single-
deck island food refrigerators with considerable area exposed to indoor air, there are three types
of loads:
 Refrigeration load. The refrigeration load of the frozen food refrigerator depends mainly on the
temperature of the foods in it, the design ambient conditions (the dew point temperature of the
conditioned space air), and the frozen surface area exposed to the space air. The higher the space
temperature and space relative humidity, and therefore the dew point, the greater the amount of
moisture that may condense on the frozen surface, the larger the accompanying latent heat of condensation,
and the greater the refrigeration load.
 Space cooling load. A considerable portion of the sensible space cooling load is removed by the
cooler surfaces of the frozen food refrigerator and the exfiltrated cold airstreams from the frozen
refrigerators through their openings, which results in a space cooling load with a heavy latent load
that may vary from 50 to 65 percent.
 Coil load. The coil load or refrigeration load of the desiccant-based air conditioning system in a
packaged unit or sometimes an AHU to maintain a required space condition in the sales area is
the sum of the space cooling load, outdoor ventilation air load, and system heat gain. During summer,
the outdoor ventilation load is again mainly latent load, and it is affected by the space temperature
and relative humidity.
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.27
In many supermarkets, 50 percent of the electric energy used annually is consumed by frozen
food refrigerators, and another 15 percent is used for the air conditioning system to maintain the required
space condition.
System Description
Consider a supermarket with an area of 30,000 ft2 (2788 m2) and an outdoor ventilation air requirement
of 3000 cfm (1416 L / s). The condition of outdoor air is dry-bulb temperature of 95°F, relative
humidity of 40 percent, and humidity ratio of 0.0142 lb / lb (100 gr/ lb or 0.0142 kg / kg). The space
cooling load is about 83,000 Btu /h (24,320 W). If a desiccant-based air conditioning system with
an impregnated lithium chloride dehumidifier is used instead of a conventional vapor compression
refrigeration system, it is possible to maintain an indoor space temperature of 75°F (23.9°C) and a
relative humidity of 45 percent (0.0085 lb / lb or kg / kg, 60 gs / lb). Meanwhile, the supply air volume
flow rate can be reduced from the 1 cfm / ft2 (5 L / s m2) of a conventional system to 0.5
cfm/ft2 (2.5 L / s m2) for a desiccant-based system. Thus, there is a significant reduction in refrigeration
load of the frozen food refrigerators because of lower space relative humidity as well as the
fan power consumption in the air conditioning system.
Figure 29.4a is schematic diagram of a desiccant-based air conditioning system for a 30,000 ft2
(2788 m2) supermarket. The process airstream of the desiccant-based air conditioning cycle is indicated
on the psychrometric chart by the solid line in Fig. 29.4b. The state points of the process air at
the exit of various components are as follows:
Outdoor air o
Filter o
Rotary desiccant dehumidifier dl
Heat-pipe heat exchanger pl
Mixing with recirculating air m
Indirect evaporative cooler or refrigeration s
Supply air after supply fan s
Space air r
The state points at the exits of the various components of the regeneration airstream are as follows:
Outdoor air o
Filter o
Heat-pipe heat exchanger xl
Gas heater rg
Rotary desiccant dehumidifier ro
Exhaust fan to atmosphere ro
Space Conditioning Line
For a supermarket of 30,000 ft2 (2788 m2), if 50 percent of the total space cooling load Qrc is latent
load Qrl, then
Qrl  0.5  83,000  41,500 Btu /h (12,160 W)
If the supply air is at a temperature of 72.5°F (22.5°C), a relative humidity of 48 percent, and a humidity
ratio of 0.0079 lb / lb (55 gr / lb or 0.0079 kg / kg), from Eq. (20.72), the supply air volume
29.28 CHAPTER TWENTY-NINE
flow rate of the mixture of the process air and the recirculating air is
It is approximately equal to 0.5  30,000  15,000 cfm (7078 L / s). Similarly, from Eq. (20.69),
for a sensible cooling load of about 0.5  83,000  41,500 Btu /h, the supply volume flow rate to
maintain a space temperature of 75°F (23.9°C) is
The larger of these two calculated values should be the design supply volume flow rate.
Operating Parameters in Rotary Desiccant Dehumidifier
Assuming that the humidity ratio of the process air at the exit of the rotary desiccant dehumidifier
after dehumidification is 0.006 lb / lb (0.006 kg / kg or 42 gr / lb), the difference during the dehumidi-
fication process in the rotary dehumidifier is therefore
wo  wdl  0.0142  0.006  0.0082 lb / lb (kg / kg or 57.4 gr / lb)
For the process airstream, the outdoor air (point o), enters the rotary dehumidifier at 95°F (35°C)
and a humidity ratio of 0.0142 lb / lb (kg / kg or 100 gr / lb). On the psychrometric chart, draw a line
o-dl from point o with an acute angle of 5° between o-dl and the thermodynamic wet-bulb temperature
line. This line intersects the 0.006 lb / lb (kg / kg or 57.4 gr / lb) humidity ratio line at point dl,
which is the state point of process air leaving the rotary desiccant dehumidifier. At point dl, temperature
Tdl  139°F (59.4°C).
For process air at a volume flow rate of 3000 cfm (1416 L / s), the dehumidifying capacity of the
rotary dehumidifier is
If the volume flow rate of the regeneration air is also 3000 cfm (1416 L / s) and the temperature of
the regeneration air entering the rotary desiccant dehumidifier is Trg  185°F (85.0°C) with a humidity
ratio of 0.0142 lb / lb (kg / kg or 100 gr / lb), regeneration air leaves the rotary dehumidifier at
Tro  130°F (54.4°C), and a humidity ratio of 0.023 lb / lb (kg / kg or 159 gr / lb).
Heat-Pipe Heat Exchanger
Assume that the effectiveness of the heat-pipe heat exchanger is 0.65. If the difference between the
specific heat of the process air and regeneration air is ignored, and if the mean density of the
process air flowing through the heat-pipe heat exchanger is hp  0.0685 lb / ft3 (1.096 kg /m3) and
for the regeneration stream mean density rg  0.070 lb / ft3(1.12 kg /m3), then, from Eq. (15.35),
 
p(Tdl  Tpl)
rg(Tdl  To)
 0.65
m?deh  3000  0.075  0.0082  60  110.7 lb /h (50.2 kg /h)

41,500
60  0.075  (0.243)(75  72.5)
 15,180 cfm (7163 L/s)
V?
s 
Qrs
60 scpa(Tr  Ts)

41,500
60  0.075(0.0085  0.0079)(1061)
 14,487 cfm (6836 L/s)
V?
s 
Qrl
60 s(wr  ws)hfg 32
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.29
The temperature of the process air after the heat-pipe heat exchanger is
The temperature of regeneration air leaving the heat-pipe heat exchanger Txl can then be evaluated as
Mixing of Process Air and Recirculating Air
Process air at a volume flow rate of 3000 cfm (1416 L / s) is mixed with recirculating air with a volume
flow rate of 12,000 cfm (5663 L / s). The condition of the mixture point m can be determined
from the psychrometric chart by drawing a line that connects the space air and the process air after
the heat-pipe heat exchanger, r-pl, so that
From the psychrometric chart, the temperature of the mixture Tm is 81.7°F (27.6°C), and its humidity
ratio wm is 0.0078 lb / lb (kg / kg or 55 gr / lb) with a dew point of 50°F (10°C).
Indirect Evaporative Cooler or Refrigeration
If the desiccant-based air conditioning system is installed in a location in which summer 1 percent
cumulative frequency of occurrence of summer outdoor design wet-bulb temperature is lower than
67°F (19.4°C), then an indirect evaporative cooler is recommended. Otherwise, a scroll or reciprocating
vapor compressor refrigeration system should be used. For a supply volume flow rate of
15,180 cfm (7163 L / s), the sensible cooling capacity can be calculated as
Qcs  60 s scps(Tm  Ts)  60  15,180  0.075  0.243(81.7  72.5)
 169,313 Btu /h (49,609 W)
The refrigeration system is mainly used for sensible cooling. This is only possible when the evaporating
temperature Tev in the DX coil is higher than 50°F (10°C). If Tev is lower than 50°F, a certain
degree of dehumidification exists in sensible cooling process ms. The required refrigeration load is
therefore greater than Qcs.
Gas Heater
The temperature of regeneration air to reactivate the desiccant LiCl is 185°F (85°C). A gas heater is
used to heat the air from Txl  123.5 to 185°F (50.9 to 85°C). The heating capacity of the gas heater is
Qh  60 rg rgcpa(Trg  Txl)
 60  3000  0.243  0.0685(185  123.5)  184,266 Btu /h (53,990 W)
Operating Parameters of the Desiccant-Based Air Conditioning Cycle
From the above calculations, the operating parameters of the desiccant-based air conditioning cycle
are as follows:
V?
V?
rm
r-pl

3000
15,000
 0.2
Txl  95 
0.0685(139  109.8)
0.070
 123.5F (50.9C)
Tpl  139  0.65 
0.070(139  95)
0.0685
 109.8F(43.2C)
29.30 CHAPTER TWENTY-NINE
Part-Load Operation and Controls
When the humidity ratio of the outdoor air drops or the space latent load falls during part-load operation,
there are three methods of maintaining the space humidity ratio wr and the space relative
humidity r if the space temperature remains constant:
 Modulation of the gas heating capacity and the temperature of the regeneration air varies the dehumidifying
capacity of the rotary desiccant dehumidifier.
 In bypass control, a portion of outdoor air bypasses the rotary desiccant dehumidifier, so that the
supply air has a higher humidity ratio.
 The rotational speed of the rotary desiccant dehumidifier can be modulated by using a variablespeed
drive. Lower rotating speed of the dehumidifier means a smaller dehumidifying capacity of
the rotary desiccant dehumidifier.
The space temperature is controlled by the vapor compression refrigeration system. If the sensible
cooling load drops at part-load operation, a cylinder unloader or on/off control in multicompressor
system can be used to maintain the space temperature within predetermined limits.
29.11 CASE STUDY: A DESICCANT-BASED AIR CONDITIONING
SYSTEM FOR RETAIL STORES
System Description
Spears and Judge (1997) reported a field study of a large retail store, a new 188,000 ft2 (17,465 m2)
Wal-Mart supercenter in Norfolk, Nebraska. The outdoor ventilation is provided by an outdoor-air
unit in a gas-fired desiccant-based air conditioning system, as shown in Fig. 29.6. The sales area is
served by two desiccant-based outdoor air units (OAUs) and the grocery area by one OAU. These
OAUs supply the required amount of outdoor air to maintain the preset space relative humidity and
CO2 level. There are also rooftop packaged units (RPUs) to provide most of the sensible cooling
capacity to offset sensible cooling load in the sales area. In the OAU, a DX coil also provides additional
sensible cooling capacity to the space when needed. Each OAU supplies a constant 10,000
cfm (4719 L / s) to the sales area.
Operating Characteristics
During base ventilation operation, 60 percent of the supply air from the OAU, or 6000 cfm (2825 L/ s)
supplied from each of the OAUs is outdoor air. When the CO2 sensor senses that the CO2 level in the
sales area exceeds 1000 ppm, the OAU shifts to high ventilation-air operation; all 10,000 cfm (4719
L/ s) supplied from the OAU is outdoor air until the CO2 level is brought back to an acceptable level.
Point Temperature, °F Humidity ratio, lb / lb and kg / kg (gr / lb)
o Outdoor air 95 0.0142 (100)
dl After dehumidifier 139 0.006 (42)
pl After heat-pipe heat exchanger 109.8 0.006 (42)
m Mixture 81.7 0.0078 (55)
s Supply air 72.5 0.0078 (55)
r Space air 75, 45% RH 0.0083 (57)
Regeneration air
xl After heat-pipe heat exchanger 123.5 0.0142 (100)
rg After gas heater 185 0.0142 (100)
ro After dehumidifier 130 0.023 (161)
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.31
When the relative humidity sensor senses that the relative humidity of a representative zone exceeds
a preset limit and calls for dehumidification, the entire 10,000 cfm (4719 L / s) of supply air in
the OAU is dehumidified in the desiccant dehumidifier. When relative humidity drops to or below a
preset limit and the dehumidification is not required, the entire 10,000 cfm (4719 L / s) bypasses the
rotary desiccant wheel.
Performance
At the Norfolk store, the desiccant-based outdoor ventilation air system provided excellent control
of indoor relative humidity and comfort, generally maintaining relative humidity within 5 percent
of the 45 percent set point. The space temperature was maintained at 78°F (25.6°C). When the two
OAUs were operated in base ventilation mode, the CO2 level in the store never went above
1000 ppm. During high ventilation air mode, the desiccant wheel removes 0.0065 lb / lb (kg / kg
or 42 gr / lb) from the outdoor air. Despite the fact that the ventilation rate had been raised from
0.15 to 0.30 cfm / ft2 (2.73 to 5.46 m3 /hm2) in the Norfolk store, the energy cost increased only 2.6
percent more.
29.12 CASE STUDY: A DESICCANT-BASED AIR CONDITIONING
SYSTEM FOR OPERATING ROOMS
Indoor Environment of Operating Rooms
Because the surgical staff are gowning more heavily to avoid the hazard of infectious diseases such
as AIDS and at the same time there is a higher lighting and electronic equipment load, today surgeons
need 65°F (18.3°C), 45 percent relative humidity, a humidity ratio of 0.0058 lb / lb (kg / kg or
29.32 CHAPTER TWENTY-NINE
RA
To
store
OA EA
EA
OA
EA
RA
Exhaust air OA Outdoor air RA Recirculating air
Dampers
OA
DX-coil
Heat pipe
Rotary desiccant dehumidifier
Gas
Gas heater
FIGURE 29.6 An outdoor air unit of a desiccant-based air conditioning system for a large retail store in Norfolk,
Nebraska.
40 gr / lb), and a dew point of 43°F (6.7°C). Using conventional dew point, there is a 22°F (12.2°C)
reheat and a dedicated refrigeration system.
System Description
Busby (1996) reported a desiccant-based air conditioning system for a six-room operating suite at
Straith Hospital, Southfield, Michigan, as shown in Fig. 29.7. The process airstream is an outdoor
airstream of 7600 cfm (3586 L / s) at an inlet with a temperature of 95°F (35°C) with a humidity ratio
of 0.0142 lb / lb (kg / kg or 100 gr / lb). It flows through a prefilter of medium efficiency and a water
sensible cooling coil; the outdoor air is sensibly cooled to 80°F (26.7°C) and 0.0142 lb / lb
(kg / kg or 100 gr / lb). Before entering the rotary desiccant dehumidifier, the process air is divided
into two airstreams: a dehumidified airstream and a bypass airstream. At summer design conditions,
the bypass air flows through the bypass dampers at 400 cfm (189 L / s). The desiccant used in the
rotary dehumidifier is titanium silica gel. After dehumidification in the rotary dehumidifier, the temperature
of the process air is raised to 138°F (58.9°C) and the humidity ratio reduced to 0.005 lb / lb
(kg / kg or 36 gr / lb). Both the dehumidified and bypass airstreams are extracted by the supply fan,
at the outlet of the supply fan, the process air temperature is increased to 140°F (60°C), and its humidity
ratio is raised to 0.0057 lb / lb (kg / kg or 40 gr / lb).
The process air then flows through the evaporator end of the heat pipe, and its temperature drops
to 95°F (35°C). When the process air flows through a water-cooling coil, its temperature is sensibly
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.33
Gas
heater
Rotary desiccant dehumidifier
Heat-pipe heat exchanger
OA
OA
Supply
fan
Operating room
65 F f  50%, 46 gr/lb
Filter
95%
eff.
140 F
40 gr/lb
95 F
40 gr/lb 138 F
36 gr/lb
80 F
100 gr/lb
95 F
7600 cfm
100 gr/lb
7200 cfm
53 F
40 gr/lb
Filter
Reactivated
fan
EA
Heater DDC
controller Humidifier
Chilled
water
coil
Water
cooling
coil
T H
FIGURE 29.7 A desiccant-based air conditioning system for a six-room operating suite at Straith Hospital,
Southfield, Michigan.
cooled further from 95 to 53°F (35 to 11.7°C) at a volume flow rate of 7200 cfm (3398 L / s). It is
again flowing through a high-efficiency filter of MERV15. At the supply inlet, the supply air temperature
is 53°F (11.7°C) with a humidity ratio of 0.0057 lb / lb (kg / kg or 40 gr / lb). The process air
is then supplied to the operating room through an electric heater and a humidifier, so that the space
temperature can be modulated by a DDC controller to maintain a preset indoor temperature and relative
humidity according to sensed values.
There is an exhaust airstream of 7000 cfm (3303 L / s) to absorb the heat rejected from the condenser
end of the heat pipe. There is also reactivating outdoor airstream of 250°F (121°C) heated by
a gas heater. After it flows through the rotary dehumidifier, this reactivating airstream is exhausted
to outdoors.
29.13 APPLICATIONS OF DESICCANT-BASED AIR
CONDITIONING SYSTEMS
Comparision between Conventional Vapor Compression Refrigeration System
and Desiccant-Based Air Conditioning System
Compared with air conditioning systems using vapor compression refrigeration systems, the desiccant-
based air conditioning systems have the following benefits and savings:
 Thermal energy costs are lower. Desiccant-based systems use lower-cost evaporative cooling to
replace more expensive electric power cost for vapor compression refrigeration, and use lowercost
gas heating to dry the desiccant instead of more expensive electric power cost to remove latent
load by vapor compression refrigeration.
 Desiccant air conditioning system results in drier ductwork to prevent mold and bacterial growth.
 The all outdoor air process airstream provides a beneficial setup for demand-controlled minimum
ventilation control. In a desiccant-based air conditioning system, it is more convenient to have a
built-in heat recovery if the supply and return ducts are located close together.
The disadvantages of a desiccant-based air conditioning system are that it is more expensive and
more complicated and needs more maintenance.
Conditions to Apply Desiccant-Based Air Conditioning Systems
Desiccant-based air conditioning systems are widely used in supermarkets, outdoor air makeup
units, surgical operating rooms, and ice rinks. According to field experience, the following are the
conditions under which a desiccant-based air conditioned system should be used:
 The dew point temperature of the indoor space temperature is equal to 40°F (4.4°C) or lower.
 There is a need of dehumidification of large amount of outdoor air.
 There is a need of low space relative humidity at lower space temperature.
 There is a low sensible heat ratio SHRc in cooling and dehumidification process, such as SHR  0.7.
ASHRAE/IESNA Standard 90.1-1999 specifies that where humidistatic controls, including dehumidification
and humidification, are provided, such controls shall prevent reheating, mixing of
heated and cold airstreams, or other simultaneous heating and cooling of the same airstream. Exceptions
include:
 Systems serving spaces where humidity levels are required to satisfy process requirements, such as
computer rooms, museums, surgical suites, supermarkets, refrigerated warehouses, and ice arenas.
29.34 CHAPTER TWENTY-NINE
 The system can reduce supply air volume to 50 percent or less of the design airflow rate or the
minimal rate specified in ventilation requirements in ASHRAE Standard 62 before simultaneous
heating and cooling takes place.
 The individual fan-coil cooling (refrigeration) unit has a design cooling capacity of 40,000 Btu /h
(11.7 kW) or less, or a design cooling capacity of 80,000 Btu/h (23.4 kW) or less and can unload
to 50 percent capacity before simultaneous heating and cooling takes place.
 75 percent or more of the energy of reheating or for providing warm air in the mixing process is
provided from site recovered heat or from a site solar energy source.
 Where the heat added to the airstream is the result of the use of a desiccant system and 75 percent
of the heat added by the desiccant system is removed by a heat exchanger, either before or after
the desiccant system, with heat recovery.
System Characteristics
System characteristics of a desiccant-based air conditioning system (DBACS) are listed in Table
29.2.
REFERENCES
Acker,W., Industrial Dehumidification:Water Vapor Load Calculations and System Descriptions, HPAC, no. 3,
1999, pp. 49–59.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
Banks, N. J., Desiccant Dehumidifiers in Ice Arena, ASHRAE Transactions, 1990, Part I, pp. 1269–1272.
Burns, P. R., J. W. Mitchell, and W. A. Beckman, Hybrid Desiccant Cooling Systems in Supermarket Applications,
ASHRAE Transactions, 1985, Part I B, pp. 457–468.
Busby, R. L., Relieving the Headache of Humidity Control, Engineered Systems, no. 9, 1996, pp. 45–50.
Carrier Corporation, Products and Systems 1992 / 1993 Master Catalog, Carrier Corporation, Syracuse, NY.
Desmone, C. L., and P. L. Frank, Air Conditioning for Precision Manufacturing, Heating / Piping / Air Conditioning,
no. 2, 1992, pp. 34–36.
Haessig, D. L., A Solution for DX VAV Air Handlers, Heating / Piping / Air Conditioning, no. 5, 1995, pp.
83–86.
Harriman, L. G., The Basics of Commercial Desiccant Systems, Heating / Piping / Air Conditioning, no. 7, 1994,
pp. 77–85.
Jones, R. S., Rooftop HVAC Equipment on Building Roofs, ASHRAE Transactions, 1991, Part I, pp. 442–444.
Jordan, C. H., DX Refrigeration vs. Chilled Water, Heating /Piping /Air Conditioning, no. 10, 1991, pp. 94–98.
Kegel, R. A., Unitary HVAC Enables School-to-Courthouse Conversion, HPAC, no. 3, 1996, pp. 55–60.
Kovak, B., P. R. Heimann, and J. Hammel, The Sanitizing Effects of Desiccant-Based Cooling, ASHRAE Journal,
no. 4, 1997, pp. 60–64.
Manley, D. L., K. L. Bowlen, and B. M. Cohen, Evaluation of Gas-Fired Desiccant-Based Space Conditioning
for Supermarkets, ASHRAE Transactions, 1985, Part I B, pp. 447–456.
Marciniak, T. J., R. N. Koopman, and D. R. Kosar, Gas-Fired Desiccant Dehumidification System in a Quick-
Service Restaurant, ASHRAE Transactions, 1991, Part I, pp. 657–666.
McGahey, K., New Commercial Applications for Desiccant-Based Cooling, ASHRAE Journal, no. 7, 1998,
pp. 41–45.
Meckler, G., Use of Desiccant to Produce Cold Air in Gas-Energized Cold Air HVAC System, ASHRAE
Transactions, 1990, Part I, pp. 1257–1261.
Meckler, G., Comparative Energy Analysis of Gas-Energized Desiccant Cold-Air Unit, ASHRAE Transactions,
1991, Part I, pp. 637–640.
AC SYSTEMS: PACKAGED AND DESICCANT-BASED SYSTEMS 29.35
Parsons, B. K., A. A. Pesaran, D. Bharathan, and B. Shelpuk, Improving Gas-Fired Heat Pump Capacity and
Performance by Adding a Desiccant Dehumidification Subsystem, ASHRAE Transactions, 1989, Part I, pp.
835–844.
Scofield, M., and G. Fields, Joining VAV and Direct Refrigeration, Heating / Piping / Air Conditioning, no. 9,
1989, pp. 137–152.
Shaffer, R., Comparison of Air and Water Cooled Reciprocating Chiller Systems, Heating /Piping /Air Conditioning,
Aug. 1987, pp. 71–87.
Shirey, D. B., Fan Cycling Strategies and Heat Pipe Heat Exchangers Provide Energy Efficient Dehumidification,
ASHRAE Journal, no. 3, 1995, pp. 31–33.
Spears, J. W., and J. Judge, Gas-Fired Desiccant System for Retail Super Center, ASHRAE Journal, no. 10,
1997, pp. 65–69.
Tinsley,W. E., B. Swindler, and D. R. Huggins, Rooftop HVAC System Offers Optimum Energy Efficiency,
ASHRAE Journal, no. 3, 1992, pp. 24–28.
The Trane Company, Packaged Rooftop Air Conditioners, The Trane Company, Clarksville, TN, 1997.
29.36 CHAPTER TWENTY-NINE
CHAPTER 30
AIR CONDITIONING
SYSTEMS: CENTRAL
SYSTEMS AND CLEANROOM
SYSTEMS
30.1
30.1 CENTRAL SYSTEMS 30.2
30.2 FLOOR-BY-FLOOR CENTRAL
SYSTEMS VERSUS CENTRAL SYSTEMS
USING AIR SYSTEMS SERVING MANY
FLOORS 30.2
Size of Air System 30.2
Separate Air Systems 30.2
Floor-by-Floor Air System versus Air
System Serving Many Floors 30.3
30.3 CONTROLS AND OPERATING
CHARACTERISTICS OF CENTRAL
SYSTEMS 30.4
Controls at Part Load for Central
Systems 30.4
Controls of Water, Heating, and
Refrigeration Systems 30.4
Air and Water Temperature
Differentials 30.5
Influence of Inlet Vanes on Small
Centrifugal Fans 30.6
30.4 TYPES OF VAV CENTRAL
SYSTEMS 30.7
30.5 SINGLE-ZONE VAV CENTRAL
SYSTEMS 30.7
System Description 30.7
Supply Volume Flow Rate, Coil Load, and
Zone Temperature Controls 30.7
System Characteristics 30.9
30.6 VAV COOLING CENTRAL SYSTEMS,
VAV REHEAT CENTRAL SYSTEMS, AND
PERIMETER-HEATING VAV CENTRAL
SYSTEMS 30.9
System Description 30.9
Supply Volume Flow Rate and Coil
Load 30.9
Zone Temperature Controls 30.10
System Characteristics 30.10
30.7 DUAL-DUCT VAV CENTRAL
SYSTEMS 30.10
System Description 30.10
System Characteristics 30.11
30.8 FAN-POWERED VAV CENTRAL
SYSTEMS 30.11
System Description 30.11
Zone Supply Volume Flow Rate and Coil
Load 30.11
Case Study: A Fan-Powered VAV Central
System 30.12
System Characteristics 30.13
30.9 CLEAN-ROOM SYSTEMS 30.14
System Description 30.14
Airflow 30.14
Pressurization 30.16
Temperature and Relative
Humidities 30.16
System Characteristics 30.16
30.10 CASE STUDY: CLEAN-ROOM
SYSTEMS FOR SEMICONDUCTOR
INTEGRATED-CIRCUIT
FABRICATION 30.16
Indoor Requirements 30.16
Energy Use of Components 30.17
System Description 30.17
Operating Characteristics 30.18
Summer Mode Operation 30.19
Part-Load Operation and Controls 30.19
Winter Mode Operation and
Controls 30.20
System Pressure 30.21
Effect of Filter Final-Initial Pressure
Drop Difference on System
Performance 30.23
Design Considerations 30.24
REFERENCES 30.24
30.1 CENTRAL SYSTEMS
A central air conditioning system consists of: a central plant in which a boiler and chillers are located,
a water system to transport hot and chilled water from the central plant to the AHUs, and an
air system often with AHUs to condition the mixture of outdoor and recirculating air and distribute
the conditioned supply air to the conditioned space, as discussed in Sec. 1.4. Also, a return air system
for energy saving, smoke control systems for multistory buildings according to fire codes, and
mechanical exhaust systems may be required.
A central air conditioning system, or simply a central system, is either a single-zone or a multizone
air conditioning system. In a central system, air is heated or cooled by the hot or chilled water
in coils in air-handling units (AHUs). Conditioned air is then distributed to various control zones
through ducts, terminals, and diffusers. Hot water and chilled water are heated by the boilers and
cooled by chillers in the central plant.
In a central system, heating and cooling capacities can be accurately modulated by the water
flow; air contaminants can be effectively removed by high-efficiency, even HEPA and ULPA, filters.
Fan rooms and central plants can be located remotely from sensitive areas. Also there is signifi-
cantly less field and space HVAC&R maintenance required; therefore, central systems are always
used in the quietest space, the cleanest space, the most precision-oriented space, and the most demanding
space.
30.2 FLOOR-BY-FLOOR CENTRAL SYSTEMS VERSUS CENTRAL
SYSTEMS USING AIR SYSTEMS SERVING MANY FLOORS
Size of Air System
For a central system, the size of the air systems depends mainly on the following:
 Building occupancy and layout
 The smaller the air system, the less the fan system total pressure with lower energy use, the
smaller the diameter of the main duct, and the greater the number of the AHUs required, resulting
in a higher initial cost
 Advantages to the operation and maintenance of the air system
 Advantages to the compartmentation of fire protection
If the air-handling units are installed indoors, their optimum sizes are usually 15,000 to 25,000
cfm (7079 to 11,798 L / s). Above 25,000 cfm (11,798 L / s), the headroom available in the fan room
to install an AHU and ducts is usually not adequate. AHUs having a supply volume flow rate below
10,000 cfm (4719 L / s) are often more expensive per cfm (L / s) volume flow than larger sizes.
To increase the net rentable floor area, weatherproof rooftop AHUs are often used instead of
indoor units mounted in the fan room for buildings of only a few stories. In such instances, the size
of rooftop AHUs is limited by two things: (1) the products currently available, such as the fact that
the largest rooftop AHU has a volume flow rate of 63,000 cfm (29,730 L / s), and (2) the fact that
too large an air system always results in a higher system total pressure loss and, therefore, a greater
energy cost.
Separate Air Systems
In a control zone with special requirements and a floor area greater than 1000 ft2 (472 m2), it may
be energy-efficient and cost-effective to use a separate air system. Special requirements include the
following:
30.2 CHAPTER THIRTY
 Special process temperature and humidity requirements
 Clean rooms or clean space
 Special health care requirements
 Special operating characteristics, such as after-hours operation
Floor-by-Floor Air System versus Air System Serving Many Floors
In a floor-by-floor air system, AHUs are installed at least one for each floor in fan rooms or mechanical
rooms. For an air system serving several or many floors, conditioned air is supplied or returned
through vertical risers from and to the rooftop units, or from the AHUs installed in the fan
rooms at the middle-level mechanical floor of a high-rise building, as shown in Figs. 1.1 and 1.2, or
from the basement. A high-rise building is a multistory building of four or more floors. A low-rise
building has three floors or less.
Let us compare a floor-by-floor air system and an air system serving many floors, both employing
conventional air distribution.
 Normally, a floor-by-floor air system needs a smaller duct system and less fan energy and is more
decentralized than an air system serving many floors, which is more centralized. Because the peak
loads for all the control zones in a system do not occur simultaneously, a centralized system always
has the benefit of a smaller load diversity factor than a decentralized system. Therefore, its
total capacity is smaller.
 An air system serving many floors needs supply and return risers, which reduces the amount of
available rental space. The air velocity in the return riser is always lower than that in the supply
riser. Linford and Taylor (1989) recommended that rule-of-thumb estimate for the riser’s area
required of 1.3 ft2 (0.12 m2) per 1000 ft2 (472 m2) of conditioned area served.
 For a VAV system, it is far simpler to use a floor-by-floor air system as the supply, and return air
is more easily balanced in each floor when both supply and return volume flow rates are varied at
part load.
 Each floor in a floor-by-floor air system is a separate fire compartment, which avoids supply and
return duct penetration between floors for fire safety. If local codes require smoke control and
purging, the supply and return duct risers for a more centralized system can be easily used as part
of the smoke control system during a building fire.
 The fans or compressors in a more centralized system are often located farther away from the
occupied space than those in a floor-by-floor air system. The duct-borne noise is often the
primary source to be attenuated by sound traps and the longer distances of inner-lined
ductwork.
 A floor-by-floor air system has better redundancy, which confines any malfunction and shutdown
to the individual floor.
 A floor-by-floor air system has greater flexibility than a more centralized system for future development,
after-hours access for overtime workers, easier and more accurate tenant metering, and
staged completion and rental of building space.
 A more centralized system always has larger fans and compressors and, therefore, more efficient
equipment. It is often able to use high-efficiency airfoil fans and adjustable-frequency variablespeed
drives.
 Operating and maintenance of a more centralized system are easier.
 The overall building cost of a more centralized system is often less than that for a floor-by-floor
air system for low-rise buildings of two or three floors. For high-rise buildings, a detailed lifecycle
cost analysis should be undertaken to determine the optimum choice.
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.3
30.3 CONTROLS AND OPERATING CHARACTERISTICS OF
CENTRAL SYSTEMS
Controls at Part Load for Central Systems
For a multizone VAV central system, the zone thermal control actions in the air system, water system,
and refrigeration system during summer cooling mode part-load operation are as follows:
 In any of the control zones, when a drop in zone temperature Trn is detected by the temperature
sensor, the DDC controller adjusts the position of the damper in the VAV box so that the volume
flow rate of zone supply air is modulated to match the reduced sensible cooling load in that control
zone, to maintain a preset zone temperature.
 As the dampers in the VAV boxes served by the same AHU are closed to smaller openings, the
duct static pressure sensor detects the rise in static pressure. The DDC controller then closes
the inlet vanes, or varies the speed of the supply fan, until the reduced supply volume flow rate of
the AHU is matched by the reduced sensible cooling load of that area served by that AHU.
 When the supply volume flow rate of the AHU is reduced at part-load operation, the discharge air
temperature Tdis tends to drop to a lower value. As this signal is sensed by the discharge temperature
sensor, the DDC controller modulates the two-way valve of the water-cooling coil and reduces
the flow rate of the chilled water so as to maintain a nearly constant preset temperature, or a
reset Tdis at part load.
 As the drop in the coil load, represented by the product of the chilled water flow rate and the temperature
difference between the chilled water entering and leaving the coil Twl  Twe, is sensed by
a Btu meter, the DDC controller modulates the inlet vanes of the centrifugal compressor or varies
the compressor speed to reduce the refrigeration capacity to match the fall in the coil load at
the cooling coils, to maintain a nearly constant or reset chilled water temperature leaving the
chiller.
 The capacities of the air, water, and refrigeration systems of a central system should be equal or
nearly equal to each other so that an equilibrium between the space temperature, discharge air
temperature, and the chilled water temperature leaving the chiller can be maintained at a specific
part load.
 Another important characteristic of the capacity control of central systems is that the control actions
in all air, water, and refrigeration systems, except multiple constant-speed pumps used in
building loops, are usually stepless, continuous modulation controls.
Controls in Water, Heating, and Refrigeration Systems
For most VAV central systems discussed in this chapter, the controls for water, heating, and refrigeration
systems are similar.
 Water system. The most widely used water system is the plant-building loop system. There are
four controls for the plant-building loop system: (1) coil discharge air temperature control, (2)
chilled water temperature leaving chiller control, (3) staging control, and (4) pressure differential
control. Water system controls are discussed in Sec. 7.11.
 Heating system. Low-temperature hot water heating system with two-pipe individual loop is
widely used in central systems. Its controls include multiple-boilers staging control, hot water
leaving boiler temperature reset, zone hot water temperature control, and safety controls. These
are discussed in Secs. 8.3 and 8.7.
 Refrigeration system. Centrifugal and screw chillers are widely used in central systems. Leaving
chilled water temperature control and reset, optimizing operation of multiple chillers, and safety
controls are discussed in Secs. 11.11, 13.6, and 13.7.
30.4 CHAPTER THIRTY
Air and Water Temperature Differentials
Both air and water temperature differentials exist in a central air conditioning system. For a central
system in cooling mode operation, the supply air temperature differential indicates the difference
between the space air temperature Tr and the supply air temperature Ts. The discharge air temperature
Tdis from the AHU after the draw-through supply fan is about 3°F (1.7°C) lower than Ts because
of the supply duct heat gain. The temperature of the conditioned air leaving the cooling coil
Tcc is about 5°F (2.8°C) lower than Ts as a result of the supply fan heat gain and duct heat gain.
The implications of a greater supply air temperature differential Tr  Ts, with respect to maintaining
a specified space temperature and relative humidity, include the following:
 A lower supply volume flow rate and, therefore, less investment for ducts, VAV boxes, flexible
ducts, and diffusers
 A lower conditioned air off-coil temperature Tcc
 A greater risk of surface condensation due to a lower Ts
 A lower fan energy use because of the lower
 A higher compressor power input because of the lower Tcc.
The fan energy saving is often greater than the increase in compressor energy input when Tr  Ts is
increased.
In current practice, the supply air temperature differential is divided into two categories:
Conventional Air Distribution. This category has the following operating parameters:
 A supply air temperature differential Tr  Ts of 15 to 24°F (8.3 to 13.3°C)
 A supply air temperature of 52 to 58°F (11.1 to 14.4°C), typically 55°F (12.8°C)
 A lowest temperature of chilled water leaving the evaporator as low as 37°F (2.8°C) because of
the improvements in freeze protection control in evaporators
 A space temperature of 75 to 78°F (23.9 to 25.6°C) and a space relative humidity between 45 and
50 percent
Neither ice storage systems nor glycol is used in this category.
Cold Air Distribution. This category includes the following operating features:
 A supply air temperature differential Tr  Ts of 30 to 36°F (16.7 to 20.0°C)
 A supply air temperature of 42 to 47°F (5.6 to 8.3°C), typically 44°F (6.7°C)
 A supply of water to the water coil at a temperature of 34 to 38°F (1.1 to 3.3°C)
 A space temperature of 75 to 78°F (23.9 to 25.5°C) and a relative humidity between 35 and 45
percent
Cold air distribution is always used with an ice storage system as the brine melts the ice and is
thus cooled to 34°F (1.1°C). Fan-powered units are often used to blend the cold primary air at 44°F
(6.7°C) with the plenum air to produce a supply temperature of 55°F (12.8°C). Ethylene glycol or
propylene glycol mixed with water, or brine, is needed for freeze protection. Adequate insulation
must be provided for ducts, terminals, and diffusers to prevent surface condensation.
As described in Sec. 7.1, under currently accepted procedures, the chilled water temperature differential,
i.e., the difference between the chilled water temperatures entering and leaving the evaporator
Tee  Tel, is between 10 and 18°F (5.6 and 10.0°C) for water systems of the plant-building
loop. A large water temperature differential saves pump power and reduces the pipe size but requires
a lower Tel. For a variable-flow building water loop, although the chilled water temperature
return from the coils Twl may be different from Tee at part load, Tee is nearly equal to Twl at design
V?
s
V?
s
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.5
load. If glycol is not blended with the chilled water, the lowest chilled water temperature should not
be lower than 37°F (2.8°C) to protect against freezing.
Influence of Inlet Vanes on Small Centrifugal Fans
Inlet vanes mounted at the inlet of the centrifugal fan block a certain percentage of the air passage.
For a small centrifugal fan, these vanes have a considerable effect on fan performance compared
with similar types and sizes of fans without inlet vanes, even when the inlet vanes are fully open.
 The ratio of the inlet diameter D1 to the impeller diameter D2, D1 /D2, is different for forwardcurved
and backward-curved centrifugal fans. For forward-curved fans, D1 /D2 varies from 0.8 to
0.9, and for backward fans, it varies from 0.65 to 0.8.
 The percentage of blocked air passage for small centrifugal fans may be between 15 and 25 percent.
 Although the blocked area may be only 15 percent of the total air passage, the eddies and turbulences
after the inlet vanes, even if they are fully opened, result in a greater energy loss.
The following is a comparison of fan performance between backward-curved and forwardcurved
centrifugal fans with inlet vanes when they are fully opened and fans without inlet vanes.
These data are taken from the manufacturer’s catalog of vertical modular AHUs (1990).
If the ratio D1 /D2 for a backward-curved centrifugal fan of impeller diameter D2  20 in. (508 mm)
is taken as 0.75 and the percentage of blocked area at the inlet is 20 percent, the velocity at the fan
inlet when the inlet vanes are fully opened is
 10,590 fpm (53 m / s)
For such an extremely high velocity, the ratio of fan total pressure of this BC centrifugal fan of
20-in. (508-mm) impeller diameter when the inlet vanes are fully opened is pt,i /pt,o  0.63. More
than one-third of the fan energy output is lost as a result of the installation of inlet vanes. Such a decrease
in fan total pressure due to an extremely high velocity vinlet was verified in a field installation
vinlet 
V?
Afree

10,400
0.8 [(20  0.75)/(12  2)]2
30.6 CHAPTER THIRTY
Without inlet vanes With inlet vanes
Blade BC BC FC BC BC FC
D2, in. 20 22.25 15 20 22.25 15
, cfm 10,400 13,000 6000 10,400 13,000 6000
vf, fpm 2063 2063 2143 2063 2063 2143
ps, in WC 3.75 3.75 3.75 2.25 3.17 3.50
pt, in. WC 4.01 4.01 4.04 2.51 3.43 3.79
bhp, hp 10.93 12.66 5.79 10.38 14.45 5.95
rpm 1841 1645 1296 1839 1646 1291
pt,i /pt,o 0.63 0.86 0.94
BC  backward-inclined or backward-curved centrifugal fan
FC  forward-curved centrifugal fan
 volume flow rate, in cfm (L / s)
vf  air velocity at fan outlet, fpm (m/ s)
ps, pt  fan static pressure and fan total pressure, respectively, in. WC (Pa)
bhp  brake horsepower, hp (kW)
rpm  revolutions per minute
pt,i, pt,o  fan total pressure with and without inlet vanes, respectively, in. WC (Pa)
V?
V?
of a VAV system using inlet vane modulation. A fan total pressure of less than 3 in. WC (750 Pa)
was available. After the inlet vanes had been dismantled, the fan total pressure increased to more
than 4 in. WC (1000 Pa).
For a BC fan with a 22.25-in.- (565-mm-) diameter impeller, the ratio pt,i /pt,o  0.86; and for an
FC fan of 15-in.- (381-mm-) diameter impeller, the ratio pt,i /pt,o  0.94.
During a cost comparison between inlet vanes and adjustable-frequency variable-speed drives,
the total pressure loss due to the inlet vanes when they are fully opened must be taken into consideration.
Therefore, for centrifugal fans in either AHUs or PUs, installation of inlet vanes is not
recommended in backward-curved centrifugal fans with impeller diameters smaller than 25 in.
(635 mm).
30.4 TYPES OF VAV CENTRAL SYSTEMS
Similar to packaged systems, VAV central systems can be subdivided into the following air conditioning
systems according to their configuration and operating characteristics:
 Single-zone VAV central system (SZVAVCS)
 VAV cooling central system (VAVCCS)
 VAV reheat central system (VAVRCS)
 Perimeter-heating VAV central system (PHVAVCS)
 Dual-duct VAV central system (DDVAVCS)
 Fan-powered VAV central system (FPVAVCS)
The dual-duct VAV central system is the only system with a dual-duct VAV air system. Regarding
the operating characteristics of the air system, the difference between a VAV central system and a
VAV packaged system is mainly due to the fact that water heating and cooling coils in an AHU are
used to heat and to cool the supply air in the central system instead of gas or electric heater and DX
coils in a PU to heat and cool the supply air in packaged systems. Their differences are discussed in
detail in Sec. 29.1.
30.5 SINGLE-ZONE VAV CENTRAL SYSTEMS
System Description
A single-zone variable-air-volume central system is an air conditioning system that has a central plant
and water systems to supply hot and chilled water to the water heating and cooling coils in AHUs to
heat and cool the supply air, and distribution to the conditioned space through ducts and diffusers, as
shown in Fig. 21.1. A relief or return fan is employed to extract the recirculating air to the AHU, or it is
exhausted outdoors during the air economizer cycle. The zone air is maintained at a preset temperature
or relative humidity by modulating the water flow rate to the coils by means of a sensor and a DDC
controller. Hot water is heated in the boiler, and the chilled water is cooled in the chiller in a central
plant and transported to the AHU through hot and chilled water systems. Single-zone VAV systems are
widely used in arenas, indoor stadiums, airport terminals, and many industrial applications.
Supply Volume Flow Rate, Coil Load, and Zone Temperature Controls
The supply volume flow rate , in cfm [m / (60s)], of a single-zone VAV central system based on
the zone sensible cooling load at summer design conditions can be calculated from Eq. (29.1); and
V?
s
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.7
the supply volume flow rate based on the zone heating load can be calculated from Eq. (29.2). Usually,
the supply volume flow rate required to offset the zone sensible cooling load is the greater of
the two and is the adopted supply volume flow rate for the single-zone VAV central system.
The cooling coil load Qcc, in Btu / h (W), can be calculated from Eq. (29.3), and the heating coil
load Qch, in Btu / h (W), can be calculated from Eq. (29.4).
Zone temperature controls for a single-zone VAV central system are similar to those discussed in
Sec. 21.2.
30.8 CHAPTER THIRTY
TABLE 30.1 System Characteristics of VAV Central Systems
VAVCCS/VAVRCS,
SZVAVCS PHVAVCS DDVAVCS
Zone thermal and sound control
Control zone Single-zone Multizone Multizone
Control methods DDC, electric DDC DDC
Control modes PI, PID PI, PID PI, PID
Heating-cooling modes changeover Automatic Automatic Automatic
Humidity control Optional Optional Optional
Sound control NC 25–40 NC 25–40 NC 25–40
Indoor air quality
Minimum ventilation control DCV or MPC DCV or MPC DCV or MPC
Filter efficiency Medium or high Medium or high Medium or high
Air system
Types VAV, air mixing VAV, air mixing VAV, air mixing, dual supply
duct
Supply fan (SF) Forward, airfoil, Forward, airfoil, Single- or dual-fan, forward,
centrifugal centrifugal airfoil, centrifugal
Supply fan total pressure 4–5 in. WC 4.5–5.5 in. WC 4.5–6 in. WC
Combined fan-motor-drive efficiency 50% 55% 55%
Relief / return fan Centrifugal or axial Centrifugal or axial Centrifugal or axial
Relief / return fan total pressure Relief 0.6 in. WC, Relief 0.6 in. WC, Relief 0.6 in. WC, return
return 0.5–1 in. WC return 0.5–1 in. WC 0.5–1.5 in. WC
Combined fan-motor-drive efficiency 35% 35–40% 35–40%
Air economizer Fixed or differential dry- Fixed or differential dry- Fixed or differential dry-bulb
bulb or enthalpy bulb or enthalpy or enthalpy
Cooling systems
Refrigeration compressor Centrifugal or screw Centrifugal or screw Centrifugal or screw
Refrigerants HCFC-123, HCFC-22, HCFC-123, HCFC-22, HCFC-123, HCFC-22
HFC-134a HFC-134a HFC-134a
Evaporator Flooded—liquid cooler Flooded—liquid cooler Flooded—liquid cooler
Condenser Water or evaporatively Water or evaporatively Water or evaporatively
cooled cooled cooled
Energy performance, compressor 0.5–0.7 kW/ ton 0.5–0.7 kW/ ton 0.5–0.7 kW/ ton
Heating systems
Type Hot water coil, Hot water coil, Hot water coil,
electric heater electric heater electric heater
PHVAVCS lowtemperature
hot water
baseboard heater
Boiler AFUE Conventional 78%, Conventional 78% Conventional 78%
condensing 93% condensing 93% condensing 93%
Maintenance
Fault detection and diagnostics AHU, chiller AHU, chiller AHU, chiller
System Characterisics
System characteristics of a single-zone VAV central system (SZVAVCS) are listed in Table 30.1.
30.6 VAV COOLING CENTRAL SYSTEMS, VAV REHEAT
CENTRAL SYSTEMS, AND PERIMETER-HEATING VAV
CENTRAL SYSTEMS
System Description
A VAV cooling central system is a multizone system that has a central plant and chilled water systems
to supply chilled water to the AHUs and water-cooling coils to cool and dehumidify the supply
air in the AHUs and distribute it to various control zones through ducts, VAV boxes, diffusers,
and controls in the interior zone of a building. It is similar to that shown in Fig. 29.1 except the
packaged unit should be replaced by an AHU.
A VAV reheat central system is a multizone system that has a central plant and water systems to
supply hot water to the reheating VAV boxes, or an electric heater to heat the air in the reheating
coil, and chilled water to the water-cooling coils to cool the supply air in AHUs; it distributes the
conditioned air to various control zones in the perimeter zone of a building through ducts, reheating
VAV boxes, diffusers, and controls, and to various control zones in the interior zone of a building
through ducts, VAV boxes, diffusers, and controls. A VAV reheat central system is similar to that
shown in Fig. 29.2 except the rooftop packaged unit should be replaced by an AHU.
A perimeter-heating VAV central system is a multizone system that has a central plant and water
systems to supply hot water to the perimeter hot water heating system and chilled water to the cooling
coils in the AHUs to condition the supply air and distribute it to various control zones in the
perimeter and interior zones through ducts, VAV boxes, diffusers, and controls, as shown in Fig.
21.6. In the perimeter zone, winter heating is provided by a low-temperature hot water heating
system, such as basebord finned-tube heaters. During winter heating, the VAV boxes supply cold
primary air at minimum setting to various control zones in the perimeter zone for the required minimum
outdoor ventilation air.
Supply Volume Flow Rate and Coil Load
For VAV cooling central systems, VAV reheat central systems, and perimeter-heating VAV central
systems, the supply volume flow rate , in cfm [m3 /(60 s)], of any control zone in the perimeter or
interior zone at summer design conditions can be calculated from Eq. (29.12) as
(30.1)
The cooling coil load of the AHU that has a supply volume flow rate of can be calculated from
Eq. (29.15) as
(30.2)
In a VAV reheat central system, for any control zone in the perimeter zone, the reheating coil load
of a reheating VAV box can be calculated from Eq. (29.16) as
Qchn  Qrhxn 	 Qvenxn
 Qrhxn 	 sxn scpa(Trxn  Tenn) (30.3) 60V?
Qcc  V?s s(hm  hcc)
V?
s
V?
sn 
Qrsn
60 scpa(Tr  Tsn)
V?
sn
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.9
Zone Temperature Controls
In a VAV cooling central system, as the temperature sensor detects that the zone temperature Trn is
falling below a preset value (set point) in the interior zone during part-load operation, a DDC controller
closes the damper opening in the VAV box until the minimum setting, such as 30 percent, is
reached. When the VAV box damper is at minimum setting, if the zone temperature Trn drops 2 or
3°F (1.1 or 1.7°C) lower than the set point, the DDC controller resets the discharge air temperature
of the AHU Tdis by 2 or 3°F (1.1 to 1.7°C) higher in order to maintain a higher zone temperature Trn.
In a VAV reheat central system, VAV boxes are installed in each of the control zones in the interior
zone.
 Since in the interior zone cold air is often still required to offset the zone cooling load in the winter
season, zone temperature control is the same as that in a VAV cooling central system. For each
control zone in the perimeter zone, a reheating VAV box is installed.
 As discussed in Sec. 21.3, when the zone temperature drops within the range 72.0 	 Trxn 
75.0°F (22.2 	 Trxn 23.9°C), zone temperature control is in deadband mode. The damper in the
reheating VAV box is at 30 percent minimum setting, and the reheating coil is deenergized.
 When the zone temperature Trxn 72°F (22.2°C), zone temperature control is in heating mode. The
DDC controller opens the two-way valve of the reheating coil and modulates its entering water mass
flow rate in a reverse-acting mode to maintain a preset zone temperature. The cold primary air is still
at 30 percent minimum setting to provide outdoor ventilation air to the control zone.
In a perimeter-heating VAV central system, the zone temperature controls of the control zones in
the interior zone during the cooling mode design load and part-load operations are similar to those
in the VAV reheat central system except that a VAV box is installed in each of the control zones in
the perimeter zone.
The operating mode of the zone temperature control in control zones of the perimeter zone is often
automatically changed over from cooling mode to heating mode and vice versa. There is also a
deadband mode between the cooling and heating modes. The width of the deadband mode is often
limited to 1 to 3°F (0.6 to 1.7°C).
When the zone temperature of a control zone in the perimeter zone Trxn 72°F (22.2°C), zone
temperature control of the perimeter-heating VAV central system is in heating mode operation. In
heating mode:
 A DDC controller actuates the reverse action relay.
 The cold primary air supplied from the VAV box in the perimeter zone is still at 30 percent minimum
setting.
 The hot water temperature leaving the boiler is reset according to the outdoor air temperature.
 The controller opens and modulates the mass flow rate of hot water entering the finned-tube baseboard
heater in the perimeter zone to maintain a preset temperature of 72°F (22.2°C).
System Characteristics
System characteristics for VAV cooling central systems, VAV reheat central systems, and perimeterheating
VAV central systems are listed in Table 30.1.
30.7 DUAL-DUCT VAV CENTRAL SYSTEMS
System Description
A dual-duct VAV central system is a multizone system that has a central plant and water systems
to supply hot and chilled water to the water heating and cooling coils in the AHUs to condition
30.10 CHAPTER THIRTY
the supply air; it distributes to various control zones in the perimeter zone by using a warm air
duct and a cold air duct through mixing VAV boxes, diffusers, and controls, and to various
control zones in the interior zone using only a cold duct through VAV boxes, diffusers, and
controls.
Only central systems have field-built dual-duct VAV air systems. The number of supply fans,
operating characteristics of mixing VAV boxes, zone temperature control and sequence of
operations, discharge air temperature control, and zone supply volume flow rate of a dual-duct
VAV central system are the same as those discussed in Sec. 21.4. The dual-fan dual-duct
(cold and warm duct has its own supply fan) VAV central system allows the use of different ratio
of outdoor air to supply air, simplifies its control and is more energy-efficient than a single-supply-
fan dual-duct VAV central system. A dual-fan dual-duct VAV central system is shown in
Fig. 21.9.
The supply volume flow rate, cooling coil load, and the heating coil load can be calculated in the
same way as discussed in Section 21.4.
System Characteristics
System characteristics of a dual-duct VAV central system are listed in Table 30.1.
30.8 FAN-POWERED VAV CENTRAL SYSTEMS
System Description
A fan-powered VAV central system is a multizone system that has a central plant and water
systems to supply hot and chilled water to the AHUs and VAV boxes, water-cooling coils in
the AHUs, and reheating coils in the fan-powered VAV boxes to cool and heat the supply air
and distribute it to various control zones in the perimeter zone through ducts, fan-powered
VAV boxes, diffusers, and controls, and to various control zones in the interior zones through
ducts, VAV boxes, diffusers, and controls. There are two types of fan-powered VAV box: parallel
fan-powered VAV box and series fan-powered VAV box, as discussed in Sec. 21.5. Parallel fanpowered
VAV boxes are more widely used because they are more energy-efficient. A fanpowered
VAV central system with parallel fan-powered VAV boxes is similar to that shown in
Fig. 21.13.
The fan energy use, fan characteristics, zone temperature control, and design considerations of a
fan-powered VAV central system are similar to those discussed in Sec. 21.5.
Zone Supply Volume Flow Rate and Coil Load
Zone supply volume flow rates of a fan-powered VAV central system are the same as those in a fanpowered
VAV packaged system, as discussed in Secs. 21.5 and 29.7.
For conventional air distribution in the perimeter zone, zone peak supply volume flow rate
can be calculated from Eq. (29.17). For conventional air distribution in interior zone, zone peak
supply volume flow rate can be calculated from Eq. (29.18). For cold air distribution in
perimeter zone, zone supply volume flow rate of cold primary air can be calculated from Eq.
(29.19) For cold air distribution, the volume flow rate of the recirculating plenum air can be calculated
from Eq. (29.20) as
The reheating coil load in the fan-powered VAV box Qchxn can be calculated from Eq. (29.16). The
cooling coil load of the AHU at summer design conditions Qcc can be calculated from Eq. (29.15).
V?
sxn,r  V?sxn  V?sxn,c
V?
sxn,c
V?
sin,d
V?
sxn,d
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.11
Case Study: A Fan-Powered VAV Central System
A fan-powered VAV central system was designed and constructed for the 34-story office building of
the Taipei World Trade Center, Taipei, Taiwan. It has a gross floor space of 1.05 million ft2 (97,580
m2) and a total air conditioned floor space of 880,310 ft2 (81,810 m2). The peak refrigeration load of
the building is about 2530 tons (8890 kW). The local utility rate structure strongly favors the use of
the refrigeration system during off-peak hours. This project won first place in the 1991 ASHRAE
Technology Awards for HVAC&R system designs for commercial buildings. The entrants were
Hsing-Chung Yu and Carl E. Claus.
Refrigeration System. The ice storage system was a partial storage one (ice storage systems are
discussed in the next chapter). A separate flooded liquid cooler was connected to the screw compressors
to provide direct cooling during on-peak hours as well as during after-hours operation, to
offset the significant nighttime load. During direct cooling, the compressors were controlled to operate
at a higher suction temperature of 33°F (0.56°C) to conserve energy.
A liquid overfeed system was used instead of direct expansion during the ice-making period
because of its higher heat transfer. About 20 percent of the refrigerant coil surface in the ice
builders could be saved. During ice melting, chilled water was supplied directly through a closedloop
water circuit to the storage tanks, and the tanks were pressurized. Such an arrangement obviates
the use of a heat exchanger and thus avoids the corresponding chilled water temperature
rise.
Condensing heat was efficiently rejected through evaporatively cooled condensers.
Air Systems. Because of the lower chilled water temperature from the storage tank during ice
melting, cold primary air at 45°F (7.2°C) was introduced at the parallel fan-powered unit. It was
mixed with the induced plenum air to produce a supply air of 56°F (13.3°C) at the diffuser. Such a
lower primary air temperature had the following consequences:
 Summer space relative humidity dropped to 40 to 45 percent. The thermal comfort of the occupants
was greatly improved.
 Cold primary air volume flow rate was reduced to 40 percent compared with conventional air distribution.
Two built-up centralized AHUs were installed in the basement. Each was used to supply 200,000
cfm (94,380 L / s) of primary air from the bottom to the top of the entire building. Compared with
the originally proposed conventional system with additional equipment floors at mid-level, this air
distribution system saved considerable rental space.
The centralized air system, having floor-by-floor automatic shutoff dampers, was closely
matched with the roof-mounted smoke control system. In case of a building fire, the return fan
would be stopped and the rooftop exhaust fan would be started to purge the smoke from the fire
floor through the opening and closing of dampers connected to the smoke exhaust system. At the
same time, air would be supplied to the two floors above the fire floor as well as one floor below to
pressurize these floors.
The air system also did the following:
 Provided 15 cfm (7.1 L / s) of outdoor air for each occupant
 Included cartridge-type filters and upstream prefilters with a dust spot efficiency of 85 percent
 Used controllable-pitch axial fans for supply and return air
 Included electric heaters for winter heating
A microprocessor-base DDC system was used for zone temperature; AHU operating parameters;
ice-charging, discharging, and direct cooling; and emergency smoke controls.
30.12 CHAPTER THIRTY
Cost and Electric Demands. The actual bidding price of the HVAC&R system for this project was
approximately $9 million, or $8.57 per ft2 ($92.2 per m2) gross area. Compared with the originally
proposed conventional air conditioning system using centrifugal chillers, the air system power input
was reduced from 1679 to 1136 kW, and the electricity demand dropped from 5960 to 4330 kW. The
estimated total annual energy consumption per ft2 gross area was 33,190 Btu/ ft2
 yr (105
kWh/m2 
 yr).
System Characteristics
System charateristics of fan-powered VAV central systems are listed in Table 30.2.
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.13
TABLE 30.2 System Characteristics for Fan-Powered VAV Central Systems (FPVAVCS) and Clean-Room Systems
FPVAVCS CRS
Zone thermal and sound control
Control zone Multizone Single or multizone
Control methods DDC DDC, or electric
Control modes PI or PID PI or PID
Heating-cooling mode changeover Automatic Cooling mode year-round,
summer-winter mode
changeover, automatic
Sound control 35–45 dBA NC 50–65
Indoor air quality (IAQ)
Minimum ventilation control DCV or MPC Constant-volume
Filters Medium to high efficiency HEPA or ULPA with prefilters
Humidity control Optional Dew point control
Humidifiers Steam, or heating element
Air system
Types VAV air mixing MAU and RAU conditioning and mixing
Supply fan Forward, or airfoil centrifugal Centrifugal or vane-axial
Supply fan total pressure 4.5–5.5 in. WC MAU 4–4.5 in. WC, RAU 2–2.5 in. WC
Combined fan-motor-drive efficiency 55% MAU 55%, RAU 55%
Relief / return fan Centrifugal or axial Centrifugal or axial
Relief / return fan total pressure Relief 0.6 in. WC, return 0.5–1 in. WC Return 0.5–1 in. WC
Combined fan-motor-drive efficiency 35–40% 35–40%
Fan-powered VAV box fan Centrifugal, forward N/A
Box fan total pressure 0.5 in. WC
Combined fan-motor-drive efficiency 25%
Air economizer Fixed or differential dry-bulb or N/A
enthalpy
Cooling systems
Refrigeration compressor Centrifugal or screw Centrifugal or screw
Refrigerants HCFC-123, HCFC-22, HFC-134a HCFC-123, HCFC-22, HFC-134a
Evaporator Flooded liquid cooler Flooded liquid cooler
Condenser Water cooled or evaporatively cooled Water cooled or evaporatively cooled
Energy performance (compressor) 0.5–0.7 kW/ ton 0.5–0.7 kW/ ton
Heating system Hot water coil, or electric heater Space, year-round cooling,
preheating—outdoor air,
reheating—part load; hot water,
electric heating
Boiler AFUE Conventional 78%, condensing 93% Conventional 78%, condensing 93%
Maintenance
Fault detection and diagnostics AHU, chiller AHU, chiller, filters
30.9 CLEAN-ROOM SYSTEMS
System Description
A clean-room or clean space central air conditioning system, or simply a clean-room system (CRS)
or clean space system, has a central plant and water systems to supply hot and chilled water to the
AHUs, uses HEPA and ULPA filters with prefilters and water coils in the AHU to remove air contaminants,
and conditions the air. The conditioned air is supplied to the conditioned space through
ducts, terminals, and air distribution devices in order to provide an indoor environment strictly controlled
with the required cleanliness, temperature, relative humidity, airflow pattern, pressurization,
and noise.
A clean-room system is shown in Fig. 30.1 and has the following characteristics:
 Because of the complexity of the system configuration and the higher requirements in the control
of indoor environments, a clean-room system is a custom-built central system with AHUs and water-
cooling and heating coils to condition the supply air.
 A clean-room system is required to provide airflow of specific velocity to reduce lateral air contamination.
Therefore, a clean-room system is a constant-volume system.
 A clean-room system can be either a single-zone system or a multizone system. In a multizone
clean-room system, the zone reheating coil is used to compensate for the variation in zone sensible
load to maintain a nearly constant preset zone temperature.
 A clean-room system always has a separate makeup air unit (MAU) to condition the outdoor air,
and a recirculating air unit (RAU) recirculates the space air, filters it, cools it, and pressurizes the
mixture of outdoor and recirculating air. Such an arrangement minimizes the cross-contamination
of airstreams as well as consolidates the filtration of the outdoor air.
Clean rooms and clean spaces are widely used in semiconductor, pharmaceutical, aerospace, and
health care industries and facilities.
Airflow
The volume flow rate of the cleaned and conditioned air supplied to the clean room depends on the
desirable air velocity that must be provided in the working area of the clean room. As discussed in
Sec. 20.17, the supply volume flow rate , in cfm (L / s), can be calculated from Eq. (20.73) as
(30.4)
According to ASHRAE Handbook 1999, HVAC Applications, U.S. Federal Standard 209E does not
specify velocity requirements. The 90 fpm (0.45 m / s) figure is still widely used in clean rooms.
Current research suggests lower velocity may be possible if the required cleanliness levels can be
maintained.
Proper airflow pattern is essential to predict the paths of the airstreams as well as to prevent contaminants
from being deposited on critical surfaces in the working area. In clean rooms, there are
two types of airflow pattern: unidirectional airflow, as shown in Fig. 30.1a, and nonunidirectional
airflow. In a unidirectional airflow pattern, airstreams flow through the working area of the clean
room in a single-pass, single direction of parallel airstreams. The unidirectional airflow can be subdivided
into vertical unidirectional airflow and horizontal unidirectional airflow.
When the ceiling of a clean room is fully covered by HEPA or ULPA filters, the downward air-
flow produces a unidirectional flow of ultraclean air that covers the working area of the clean room.
Baylie and Schultz (1994) reported for clean rooms with ceiling partly covered by HEPA or ULPA
filters, a porous membrane that is added beneath the HEPA or ULPA filters forms a small plenum
which equalizes the pressure across the face of the ceiling and minimizes the turbulence created by
the larger grid required to support the HEPA filters and lighting fixtures. When a membrane ceiling
V?
s  Arvr
V?
s
30.14 CHAPTER THIRTY
(b)
FIGURE 30.1 A clean-room system for a class 10 clean room: (a) schematic diagram; (b) air conditioning cycle.
30.15
is added and air is returned from the bottom inlets of the sidewall, the downward airflow underneath
the membrane ceiling is in the form of unidirectional airflow.
For a clean room with a ceiling that is partly covered by HEPA or ULPA filters and without any
porous membrane underneath the filters, the downward airflow in the clean room is in the form of
nonunidirectional airflow.
Pressurization
Clean rooms and clean spaces always maintain a higher pressure than the surrounding less clean
space to minimize the infiltration of air contaminants. The following pressure differentials are often
used:
Pressure control precision is typically between  0.01 and  0.03 in. WC ( 2.5 and  7.5 Pa).
For a door or opening between two clean rooms of different cleanliness requirements, a minimum
air velocity in the range of 15 to 50 fpm (0.07 to 0.25 m / s) should be maintained at the door (when
it is opened) or the other openings.
Temperature and Relative Humidities
Because staff wear heavy gowns in clean rooms with stringent indoor environmental requirements,
a temperature between 66 and 68°F (18.9 and 20°C) is to be maintained. In class 10,000 or class
100,000 clean rooms with less restrictive garments, 70 to 72°F (21.1 to 22.2°C) may be satisfactory.
A tolerance of 2°F (1.1°C) is adequate for comfort purposes.
In clean rooms, space relative humidity is usually controlled at 45  5 percent.
System Characteristics
System characteristics of a clean-room system are listed in Table 30.2.
30.10 CASE STUDY: CLEAN-ROOM SYSTEMS FOR
SEMICONDUCTOR INTEGRATED-CIRCUIT FABRICATION
Indoor Requirements
The fabrication of semiconductor integrated circuits requires a highly sophisticated combination of
advanced technologies. Clean-room central system with vertical unidirectional airflow is one of these
technologies in semiconductor wafer fabrication. It must meet stringent air quality requirements in
air cleanliness, air temperature, humidity, airflow pattern, pressurization, lighting, noise, and vibration
for the sake of successful manufacturing. Contaminated semiconductors result in inferior products.
As discussed in Sec. 4.11, clean-room air cleanliness requirements for semiconductor fabrication
include classes 1, 10, 100, 1000, and 10,000. Vertical unidirectional airflow with average air
velocities from 60 to 90 fpm (0.3 to 0.45 m / s), typically at 90 fpm (0.45 m / s), is widely used.
Pressure diffferential, in. WC (Pa)
Between clean rooms or clean space and
nonclean space 0.05 (12.5)
Between clean rooms and less clean rooms 0.02–0.03 (5–7.5)
30.16 CHAPTER THIRTY
Manufacturing an integrated circuit involves photolithography, etching, and diffusion processes.
These processes may take place over many hours, even days. A stable temperature is extremely important.
Because clean-room production personnel wear smocks that fully cover them, clean-room
temperatures are controlled from 68 to 72°F (20 to 22.2°C) with tolerances of 0.1°F (0.056°C),
0.2°F(0.11°C), 0.5°F (0.28°C), and 1.0°F (0.56°C). A closer tolerance is often maintained
within the manufacturing process itself; e.g., wafer reticle writing by electron beam technology
needs 0.1°F (0.056°C). whereas 1.0°F (0.56°C) is often used for the open-bay area.
In clean rooms, if the relative humidity is too low, static electricity is created and causes defective
products. If the relative humidity is too high, some chemicals may expand and may cause
equipment failure. For etching and diffusing areas, the humidity should be maintained at 40 to 45
percent with a tolerance of 5 percent, whereas in the photolithography area, a relative humidity of
35 to 40 percent with a tolerance of 2 percent is usually maintained.
The manufacture of metal oxide semiconductors requires large quantities of conditioned outdoor
air as makeup air, to replace processing exhaust air and to maintain clean-room pressurization. A
clean room is always maintained at a positive pressure to prevent the infiltration of contaminated air
from surrounding spaces. For some clean rooms, the process exhaust airflow may be as high as 10
cfm/ft2 (182 m3 /h
m2) of floor area. The average process exhaust airflow for semiconductor fabrication
may be between 2 and 3 cfm / ft2 (36 to 54 m3 /h
m2).
Energy Use of Components
In Naughton (1990a, 1990b), the breakdown of energy use in a typical clean-room central system
with vertical unidirectional flow is approximately as follows:
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.17
50 percent
40 percent
6 percent
4 percent
Manufacturing equipment (75 W/ ft2
or 807 W/m2)
Air conditioning (HVAC&R)
Electric lights
Building envelope and others
The chiller and fans each consume 45 percent of the total HVAC&R energy use. Both plant and
building pumps use the remaining 10 percent.
Because of extremely high space sensible cooling load and very small space latent load, the sensible
heat ratio of the space conditioning line SHRs can often be taken as 0.99.
The operating cost of HVAC&R is only 5 to 20 percent of the total cost needed to produce an integrated
circuit (semiconductor wafer). The air cleanliness, temperature, relative humidity, airflow,
and pressurization required for successful fabrication are still extremely important goals of an
HVAC&R system design. Because of high utility rates, however, a reduction in the operating cost of
the clean-room system also becomes a very influential factor in clean-room design and operation.
System Description
Semiconductor clean rooms are of two types: open-bay design or clean tunnel. The clean tunnel
consists of narrow modular clean rooms which may be isolated from one another. Open-bay design
includes large open-construction clean rooms. A typical clean-room air conditioning system, or
simply clean-room system, for a class 10 clean room of the open-bay configuration is shown in Fig.
30.1. This system consists of the following components:
Recirculating Air Unit (RAU). The function of a recirculating air unit is to recirculate the space
air; to filter it, to cool it, to pressurize the mixture of recirculating and conditioned makeup air, and
to force the mixture to flow through the ULPA filters and the clean-room working area. An RAU
comprises the following components:
 A mixing box to mix recirculating and conditioned makeup air
 A prefilter with a dust spot efficiency of 30 percent
 A chilled water sensible cooling coil
 A recirculating fan
 Makeup air and recirculating air dampers
 Two sound attenuators, one located immediately before the fan inlet and the other after the fan outlet
Usually an axial fan is used as the recirculating fan because of its higher fan total efficiency (between
75 and 82 percent) and its better operating characteristics. An unhoused centrifugal fan, often
called a cabinet fan, with a fan total efficiency of 58 to 63 percent is sometimes used because of its
lower sound power level.
The chilled water entering the sensible cooling coil in the RAU is often at a temperature of 50°F
(10°C).
Makeup Air Unit (MAU). The function of an MAU is to supply outdoor air to the clean room for
process exhaust and pressurization, to condition it, and to control the humidity of the clean room by
cooling and dehumidifying, or heating and humidifying, the makeup air. The system components in
an MAU include the following:
 An outdoor air damper
 A prefilter with a MERV 8 (dust spot efficiency of 30 percent)
 A preheating coil
 A chilled water cooling coil
 A makeup air centrifugal fan
 A HEPA filter of 99.97 percent DOP efficiency
 A humidifier, most often a steam humidifier
 MAU shutoff damper
Chilled water entering the cooling coil in the MAU for cooling and dehumidifying is at a temperature
of 40°F (4.4°C). It is more energy-efficient to have a separate chiller to provide chilled water
for an MAU.
ULPA Filters and the Unidirectional Airflow Clean Room. The function of ULPA filters is to
provide ultraclean air for the clean room. The pressurized plenum or ducted ULPA filter modules
are often used for an even distribution of unidirectional downward air flow.
For class 1, 10, and 100 clean rooms, ULPA filters with a DOP efficiency of 99.9997 percent of
0.12-m particles and unidirectional airflow are used. For class 1000 through class 100,000, HEPA
filters with a DOP efficiency of 99.97 percent of 0.3-m particles and nonunidirectional airflow
may provide satisfactory contamination control.
Unidirectional downward airflow produces a uniform air shower of ultraclean air. Internally generated
contaminants will not move laterally against the 90 fpm (0.45 m / s) airflow and will be carried
away by predictable parallel airstreams. Recirculating air enters either the bottom side return inlets directly
or the perforated raised floor panels. It is then returned to the RAU to mix with makeup air again.
Operating Characteristics
The following temperature and relative humidity are to be maintained in a class 10 clean room:
30.18 CHAPTER THIRTY
For a class 10 clean room with an area of 1000 ft2 (472 m2), a supply volume flow rate of 90,000
cfm (42,470 m2) is required to provide an air velocity of 90 fpm (0.45 m / s) in order to produce a
unidirectional flow in this clean room. The outdoor air intake for process exhaust, space pressurization,
and occupants is typically 6000 cfm (2830 L / s).
Summer Mode Operation
In Mandelbaum (1991), the operating modes of clean rooms of classes 1 through 1000 are divided
into summer and winter modes. When the dew-point temperature of the outdoor air To is 46°F
(7.8°C) and above, the clean-room system is in summer mode operation.
Let us consider a hot summer day. Outdoor air at a summer design temperature of 100°F
(37.8°C) and a wet-bulb temperature of 78°F (25.6°C) enters the MAU, as shown in Fig. 30.1a and
b. It is cooled and dehumidified at the cooling coil to a leaving coil condition of air temperature
Tcc  46°F (7.8°C), relative humidity r  99 percent, and dew-point temperature  46°F (7.8°C).
After the conditioned outdoor air absorbs the fan heat of the MAU, the air enters the RAU at a discharge
temperature Tdis of 47.5°F (8.6°C) and a relative humidity of 92 percent.
In the RAU, conditioned air from the MAU is mixed with the recirculating air from the clean
room at a temperature of 69°F (20.6°C), a relative humidity of 42.5 percent, and a dew-point temperature
of 46°F (7.8°C). The ratio of volume flow of recirculating air to makeup air is 12:1. The
mixture m enters the sensible cooling coil at temperature Tm  67°F (19.4°C) and a dew point of
46°F (7.8°C). It is then sensibly cooled to a temperature Tsc.
If the maximum space sensible cooling load is 563,000 Btu / h (265,680 L / s) and if the density
of supply air s  0.078 lb / ft3 (1.248 kg /m3), the temperature of supply air Ts can be calculated as
(30.5)
Usually the temperature rise due to fan heat in the RAU is 1°F (0.56°C). The temperature of the air
leaving the sensible cooling coil is then Tsc  63.5  1  62.5°F (16.9°C). Its dew-point temperature
is still 46°F (7.8°C). Because of the short supply duct, large volume flow, and the fact that the
surrounding space is conditioned, the duct heat gain is usually ignored.
Because the SHRs of the space conditioning line is 0.99, after the air has absorbed the space sensible
cooling load and a very small amount of latent load, the space temperature is then maintained
at 69°F (20.6°C). The space relative humidity is 42.5 percent, and the dew point is 46°F (7.8°C).
To cool and dehumidify the makeup air at the cooling coil to a leaving dew-point temperature of
46°F (7.8°C), the chilled water entering the cooling coil should be provided at a temperature Twe 
40°F (4.4°C) without using glycol for freeze protection. The chilled water leaving the chiller in the
plant loop is usually 1°F (0.56°C) lower, or 39°F (3.8°C).
Part-Load Operation and Controls
Two controls maintain the required space temperature at 69°F (20.6°C) and the relative humidity
within acceptable limits during summer mode part-load operation: the discharge air temperature
control incorporating dew point control of the MAU and the space (zone) air temperature control.
When the dry- and wet-bulb temperatures of the outdoor air To and To drop but the dew-point
 69 
563,000
60  90,000  0.078  0.243
 63.5F (17.5C)
Ts  Tr 
Qrs
60V?s scpa
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.19
69  1.0°F (20.6  0.56°C)
42.5  2.5 percent
Room temperature, year-round
Space humidity, year-round
temperature is still at 46°F (7.8°C) or above, the discharge air temperature sensor T2 detects the fall
in Tdis and the DDC controller modulates the two-way valve of the chilled water cooling coil in the
MAU to reduce its water flow until a 47.5°F (8.6°C) discharge air temperature and a discharge air
dew point of 46°F (7.8°C) are maintained.
If the space sensible cooling load falls below the design load, the space temperature drops accordingly.
As the space temperature sensor T1 detects such a fall in temperature, the DDC controller
modulates the two-way valve of the sensible cooling coil in the RAU, reduces its water flow,
and tends to maintain a constant space temperature of Tr  69°F (20.6°C).
In addition to the MAU discharge air temperature and space temperature controls, there are controls
for space humidity, space pressurization, filter pressure drop monitoring, and smoke detection.
Because the discharge air dew point from the MAU is 46°F (7.8°C), the space dew point is also
46°F (7.8°C), and the mixture of recirculating and makeup air still has a dew point of 46°F (7.8°C).
After sensible cooling in the RAU, the dew point of supply air, point s, remains at 46°F (7.8°C). Because
the sensible heat ratio of the space conditioning line SHRs  0.99, the space relative humidity
will be always around 42.5 percent if the space temperature Tr is maintained at 69°F (20.6°C).
For a space like a clean room having only a negligible latent load, supply air dew-point control
can always maintain both space temperature and relative humidity within required limits by means
of sensible cooling and reheating when the space sensible cooling load is varied.
In Fig. 30.1a, there is a pressure sensor P1 in the clean room. This sensor measures the pressure
differential pr between the clean room and the surrounding area. This pressure differential is generally
maintained at 0.05 in. WC (12.5 Pa) to prevent infiltration of contaminated air.
According to the signal from P1, the DDC controller modulates the opening of the interlocked
makeup air and recirculating dampers. When there is an increase in makeup air and a decrease in
recirculating air in the mixing box of RAU, the space pressure differential tends to rise. A decrease
in makeup air and an increase in recirculating air cause a drop in space pressure differential. The
supply of makeup air is balanced by the process exhaust air and the exfiltrated air from door gaps
due to the pressure differential.
Pressure sensors are also used for MERV 8 prefilters, HEPA filters, and ULPA filters. If the
pressure drop exceeds a predetermined limit, an indicating lamp flashes or an alarm is energized.
Both signals call for replacement.
When the smoke detector is energized, a signal is sent to the building fire control system. The
fire alarm and the smoke control system are energized accordingly.
Winter Mode Operation and Controls
When the dew point of the outdoor air To is 44 °F (6.7°C) or below, the clean-room system is in
winter mode operation. If To 44°F (6.7°C) and To 	 53.5°F (11.9°C), the outdoor air is preheated
at the preheating coil to a temperature Tph  53.5°F (11.9°C). After absorbing the fan heat
of the MAU, it enters the RAU at a discharge temperature Tdis  55°F (12.8°C). If To 44°F
(6.7°C) and To  53.5°F (11.9°C), the outdoor air is sensibly cooled to 53.5°F (11.9°C). Again, it is
discharged to the RAU at Tdis  55°F (12.8°C) after absorbing fan heat.
If makeup air at a temperature of 55°F (12.8°C) and a dew point of 44°F (6.7°C) is mixed with
winter mode recirculating air of temperature 69°F (20.6°C) and 40 percent relative humidity (a dew
point of 44°F, or 6.7°C), the resulting mixture will be at a temperature of 67.5°F (19.7°C). The dew
point will still be 44°F (6.7°C). If the mixture is cooled at the sensible cooling coil to 62.5°F
(16.9°C) according to the signal of the temperature sensor T1, and if it is supplied to the clean room
at a temperature of 63.5°F (17.5°C) and a dew point of 44°F (6.7°C) after absorbing the fan heat of
RAU, then it will maintain a space temperature Tr  69°F (20.6°C), r  40 percent, and a dew
point of 44°F (6.7°C) at a maximum space sensible cooling load of 563,000 Btu / h (164,960 W). A
lower space sensible cooling load results in a decrease in the space temperature Tr. As the sensible
cooling capacity at the cooling coil is modulated and reduced by the DDC controller, the space temperature
will still be maintained at 69°F (20.6°C).
If To 	 44°F (6.7°C), space relative humidity r may be less than 40 percent. When the space
30.20 CHAPTER THIRTY
humidity sensor detects such a shortage, the DDC controller energizes the steam humidifier in the
MAU to humidify the makeup air until r  40 percent. In winter mode operation, r will always
be maintained around 40 percent.
During winter mode operation, the chilled water temperature entering the cooling coil in the
MAU is reset from 40 to 45°F (4.4 to 7.2°C) for a more energy-efficient operation. Also, the preheating
coil and the cooling coil in the MAU are sequentially controlled; they are not energized simultaneously.
In winter operation, outdoor air is preheated in the MAU, and the mixture of outdoor and recirculating
air is again sensibly cooled in the RAU. Such simultaneous heating and cooling are mainly
a result of the requirement for steam humidification. It is not appropriate to humidify the air to near
saturation. Uneven distribution may cause surface condensation.
During winter mode operation, if the dew point of the outdoor air To is between 44 and 46°F
(6.7 and 7.8°C), the operation mode remains in winter mode. When To  46°F (7.8°C), it will
change over to summer mode operation. Similarly, if 44 	 To 	 46°F (6.7 	 To 	 7.8°C), the
operating mode remains in summer mode until To 44°F (6.7°C). Then it changes over to winter
mode operation. Such a control strategy prevents hunting between summer and winter mode operation
during intermediate seasons.
System Pressure
According to Naughton (1990a, 1990b) and Hunt et al. (1990), the system pressure loss of an MAU
is about 3.8 to 4 in. WC (950 to 1000 Pa). For an RAU, the system pressure loss is usually between
2.0 and 2.50 in. WC (500 and 625 Pa). A typical breakdown of the fan total pressure of the recirculating
air fan in an RAU is as follows:
HEPA filters of lower pressure drop, such as a final loaded pressure drop of 0.50 in. WC (125
Pa), are also available. However, they are expensive and require more space. This fact should be
carefully analyzed and considered by the production engineer and architect.
To reduce the pressure drop across coils, the most effective method is to reduce their face velocities
to 300 to 400 fpm (1.5 to 2 m / s). The result is a larger MAU and RAU. Again, the compromise
is between initial and operating costs.
All duct fittings along the airflow must be carefully designed to reduce pressure losses. Square
elbows should always be installed with splitter vanes. High-efficiency fans and pumps should be
used, along with high-COP and -EER chillers. According to Naughton (1990a, 1990b), the energy
use of fans, chillers, and pumps in a clean-room system for a class 1 clean room may be around 50
W/ft2 (538 W/m2) of floor area.
System components Pressure drop, in. WC (Pa)
RAU
Prefilter 0.25 (62)
Cooling coil 0.35 (87)
Inlet sound attenuator 0.10 (25)
Discharge sound attenuator 0.15 (38)
Discharge elbow 0.10 (25)
External pressure drop
HEPA filters, final 0.80 (200)
Distribution duct work 0.20 (50)
Perforated raised floor panel 0.10 (25)
Return path 0.15 (38)
Total 2.20 (550)
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.21
30.22 CHAPTER THIRTY
FIGURE 30.2 Difference in final and initial pressure drops of filters affects system
performance. (a) Airfoil blade centrifugal fan; (b) forward-curved centrifugal
fan; (c) vane-axial fan.
Effect of Filter Final-Initial Pressure Drop Difference on
System Performance
During the calculation of the system pressure loss of an air system for the purpose of selecting fans
and AHUs, one should use the final pressure drop of the filter pf.f, in in. WC (Pa), to provide the
required volume flow rate whenever filters are either clean or loaded with specified capacity. For
purposes of energy estimation, the average pressure drop of the filters during the working period
pf,m, in in. WG, should be employed instead of final or initial pressure drops.
For most prefilter and ultrahigh-efficiency air filter assemblies used in AHUs, the difference between
the final pressure drop and the initial pressure drop of the filter assembly is 0.4 to 0.8 in. WC
(100 to 200 Pa). Let us consider an air system whose design volume flow rate is 30,000 cfm
(14,160 L / s), and whose system pressure loss when filters are loaded psy is 2.2 in. WC (550 Pa).
At cfm (14,157 L / s), the difference between the final and initial pressure drops is 0.5
in. WC (125 Pa).
The effect of the difference in final and initial pressure drops of a filter on system performance
is illustrated in Fig. 30.2. In Fig. 30.2a, an airfoil-blade centrifugal fan is used. The fan and AHU
are selected according to these criteria: cfm (14,160 L / s) and psy  2.2 in. WC (550
Pa). If the filters are loaded, the operating point is P. If the filters are clean, the system pressure
drops to 1.7 in. WG (425 Pa). The operating point moves along the selected fan curve to point Q,
which has cfm (15,573 L / s) at p  1.7 in. WC (425 Pa).
In Fig. 30.2b, a forward-curved centrifugal fan is used. A forward-curved fan has a slightly flatter
fan curve at operating point Q, and the system volume flow rate increases to 33,500 cfm (15,810
L/ s). For Fig. 30.2c, a vane-axial fan is used. A vane-axial fan has a very steep fan curve. Therefore,
at point Q the system volume flow rate is increased to only about 31,000 cfm (14,630 L / s).
The air system in a clean-room system is always a constant-volume system. The fan power inputs
at point Q for airfoil blade and vane-axial fans are approximately the same as at point P,
whereas for forward-curved fans, their fan power input is comparatively greater.
V?
33,000
V?
 30,000
V?
 30,000
V?
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.23
FIGURE 30.2 (Continued)
During the selection of AHUs, if the fan performance of the AHU is expressed in volume flow
versus fan total pressure pt, the designer should ascertain that the pressure drops of the loaded filters
are included in pt, not those of the clean ones.
Design Considerations
The successful construction of a class 1, 10, or 100 clean room depends on the combined effort of
architects, mechanical engineers, production engineers, owners, contractors, HVAC&R equipment
manufacturers, and many others. To maintain a temperature tolerance of  0.1°F (0.056°C) for a
manufacturing process, it is important to stabilize all internal loads during the processing period.
The process should be surrounded by conditioned space for which the tolerance is  0.5°F
(0.3°C). Even movement of production personnel toward the processing area may cause radiant
heat turbulence and temperature fluctuations.
Energy use per ft2 (m2) floor area of a clean-room system for class 1 through class 100 clean
rooms is 5 to 10 times higher than that of an air conditioning system for a commercial building.
Careful analyses and improvements based on experience with similar projects will provide a satisfactory
indoor environment for the manufacturing process and also will reduce the energy use of the
clean-room system. Semiconductor products that are being produced now may become obsolete
within a few years. As with many other HVAC&R systems, air conditioning system design for clean
rooms must incorporate flexibility for change and future development.
REFERENCES
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, Atlanta, GA, 1999.
ASHRAE, ASHRAE/IESNA Standard 90.1-1999, Energy Standard for Buildings Except New Low-Rise
Residential Buildings, Atlanta, GA, 1999.
Atkinson, G. V., and G. R. Martino, Control of Semiconductor Manufacturing Cleanrooms, ASHRAE
Transactions, 1989, Part I, pp. 477–482.
Austin, S. B., HVAC System Trend Analysis, ASHRAE Journal, no. 2, 1997, pp. 44–50.
Baylie, C. L., and S. H. Schultz, Manage Change: Planning for the Validation of HVAC Systems for a Clinical
Trials Production Facility, ASHRAE Transactions, 1994, Part I, pp. 1660–1668.
Garr, H. B., Clean Room Humidity Control, Heating / Piping / Air Conditioning, no. 3, 1992, pp. 67–70.
Hunt, E., D. E. Benson, and L. G. Hopkins, Fan Efficiency vs. Unit Efficiency for Cleanroom Application,
ASHRAE Transactions, 1990, Part II, pp. 616–619.
Linford, R. G., and S. T. Taylor, HVAC Systems: Central vs. Floor-by-Floor, Heating / Piping / Air Conditioning,
no. 7, 1989, pp. 43–58.
Mandelbaum, I., HVAC Modifications for Semiconductor Fabrication, Heating / Piping / Air Conditioning, no. 3,
1991, pp. 29–32.
Naughton, P., HVAC Systems for Semiconductor Cleanrooms—Part 1: System Components, ASHRAE
Transactions, 1990a, Part II, pp. 620–625.
Naughton, P., HVAC Systems for Semiconductor Cleanrooms—Part 2: Total System Dynamics, ASHRAE
Transactions, 1990b, Part II, pp. 626–633.
Ottmer, J. H., Central Plant Cools and Heats World’s Largest Airport, Heating / Piping / Air Conditioning, no. 5,
1994, pp. 43–47.
Rose, T. H., Noise and Vibration in Semiconductor Clean Rooms, ASHRAE Transactions, 1986, Part I B,
pp. 289–298.
Schneider, R. K., How to Design a Cleanroom, Engineered Systems, no. 3, 1996, pp. 58–62.
V?
30.24 CHAPTER THIRTY
Schneider, R. K., Healthy Cleanroom Design, Engineered Systems, no. 7, 1997, pp. 46–54.
Schuler, M., Dual Fan, Dual-Duct System Meets Air Quality, Energy-Efficiency Needs, ASHRAE Journal, no. 3,
1996, pp. 39–41.
Stokes, R., The System Choice, Heating / Piping / Air Conditioning, July 1983, pp. 87–89.
Tao,W., and R. R. Janis, Modern Cooling Plant Design, Heating / Piping / Air Conditioning, May 1985,
pp. 57–81.
Yu, H. C., and C. E. Claus, Thermal Storage System Cools Office Building, ASHRAE Journal, March 1991,
pp. 15–17.
AC SYSTEMS: CENTRAL SYSTEMS AND CLEAN-ROOM SYSTEMS 30.25
CHAPTER 31
AIR CONDITIONING
SYSTEMS: THERMAL
STORAGE SYSTEMS
31.1
31.1 THERMAL STORAGE SYSTEMS 31.1
System Description 31.1
Building Energy Consumption and
Thermal Storage Systems 31.2
Electric Deregulation and the Impact on
Thermal Storage Systems 31.2
Benefits and Drawbacks of Thermal
Storage Systems 31.2
Full Storage and Partial Storage 31.3
Ice Storage and Chilled Water
Storage 31.5
31.2 ICE-ON-COIL, INTERNAL-MELT ICE
STORAGE SYSTEMS 31.6
System Description 31.6
Brine and Glycol Solution 31.6
Ice Storage Tank 31.7
Case Study: Operating Modes of
Ice-on-Coil Ice Storage System 31.7
System Characteristics 31.9
31.3 ICE-ON-COIL, EXTERNAL-MELT ICE
STORAGE SYSTEMS 31.10
System Description 31.10
Ice Builders 31.11
Refrigerant Feed 31.11
Ice-Charging Control 31.11
System Characteristics 31.11
Case-Study: An Ice-on-Coil External-Melt
Ice Storage System 31.13
31.4 ENCAPSULATED ICE STORAGE
SYSTEMS 31.13
System Description 31.13
Location of Chiller and Storage
Tank 31.14
Controls 31.14
Charging and Discharging 31.15
System Characteristics 31.15
31.5 ICE-HARVESTING ICE STORAGE
SYSTEMS 31.15
System Description 31.15
Ice Making or Charging 31.16
Chiller Operation 31.17
System Characteristics 31.17
31.6 COMPARISON OF VARIOUS ICE
STORAGE SYSTEMS 31.17
31.7 STRATIFIED CHILLED WATER
STORAGE SYSTEMS 31.18
System Description 31.18
Basics for Chilled Water Storage 31.18
Storage Tanks 31.19
Stratified Tanks 31.19
Temperature Gradient and
Thermocline 31.20
Diffusers 31.20
Charging and Discharging Temperature
versus Tank Volume 31.22
System Characteristics 31.23
31.8 CASE STUDY: A STRATIFIED
CHILLED WATER STORAGE
SYSTEM 31.23
Chilled Water Storage System 31.23
Concentric Double-Octagon
Diffusers 31.24
Charging Process 31.26
Discharging Process 31.26
Part-Load Operation 31.27
System Performance 31.28
REFERENCES 31.28
31.1 THERMAL STORAGE SYSTEMS
System Description
A thermal energy storage air conditioning system, or simply a thermal storage system, consists of a
central plant, a chilled water or brine system incorporated with a thermal storage system, a hot water
system, an air system including AHUs , terminals, return air system, smoke control systems, and
mechanical exhaust systems. In addition, the electric-driven refrigeration compressors in the central
plant are operated at off-peak or at off-peak and on-peak hours. Stored ice or chilled water in tanks
is used to provide cooling in buildings during on-peak hours when there is a higher electric rate.
Hot water from the boilers in the central plant is used to provide heating in winter. Air is conditioned
in the AHUs and terminals. The conditioned air is then distributed to various control zones
through ducts, terminals, diffusers, and controls.
A thermal storage system is
 Always a central system that uses stored chilled water or brine from the central plant to cool the air
 Often a multizone VAV system
 Often a cooling storage system
Building Energy Consumption and Thermal Storage Systems
Lorsch (1993) analyzed the energy consumption for a 264,000 ft2 (24,536 m2) building using the
Ontario Hydro electric rate structure. The chiller equipment needed 31 percent of the on-peak demand
but consumed only 8 percent of the annual kWh. The annual load factor for this building is
only 37 percent. Commercial buildings have poor load profiles of electric power demand. Electric
utilities are a capital-intensive industry. To improve their load factor and to sell more kWh, electric
utilities tend to shift daytime, more energy-inefficient diesel and gas-turbine plant operation to
nighttime base-load highly efficient coal and nuclear plant operation. Electric utilities offer a higher
energy rate in daytime on-peak hours and a lower energy rate in off-peak hours. A thermal storage
system shifts a part of the electric power demand of the HVAC&R system from the daytime onpeak
hours, usually from noon to 8 p.m., to nighttime off-peak hours.
A thermal storage system often consumes approximately the same amount of electricity as a
conventional air conditioning system; i.e., a thermal storage system does not necessarily save energy.
However, a thermal storage system significantly reduces energy cost.
Electric Deregulation and the Impact on Thermal Storage Systems
Because of the electric deregulation and the use of real-time pricing (RTP) and other electricity rate
structures instead of the time-of-use (TOU) rate structure, as discussed in Sec. 25.7, open-market
competition creates pressure on utilities to cut costs. According to Silvetti and MacCracken (1998)
and the analyses from EIA, the following trend may dominate:
 There is a conflict between the interests of the generation-owning utilities in higher-generation
prices and the effects of some demand side management (DSM) programs to reduce demand and
possibly to help hold down competitive prices for generation. DSM programs are causing problems.
 Regarding energy-efficient equipment-related incentive plans, something is wrong when a power
producer gives money to a customer to purchase less power, particularly if the financial incentive
might eventually benefit a competitor.
 Thermal storage systems benefit the power provider, the customer, the power user, and the industry
setting the price for that energy.
Benefits and Drawbacks of Thermal Storage Systems
Silvetti and MacCracken (1998) noted why a thermal storage system is attractive:
 Thermal storage systems are one of the few legitimate tools which shift the higher electric
demand/ electric rate for HVAC&R fully or partially from on-peak hours to a lower electric demand/
electric rate in off-peak hours, and therefore lower operating costs.
31.2 CHAPTER THIRTY-ONE
 Thermal storage systems reduce the equipment size and save initial cost.
 Thermal storage systems do not negatively impact the building’s indoor environmental control
operations, as the load shedding or some load control programs have done.
 As a thermal storage system is a central system, it uses chilled water from the central plant to cool
the air in the AHUs. Thermal storage systems are easily adaptable to the water system of a central
air conditioning system. As discussed in Sec. 1.5, central systems served 24 percent of the floor
space of commercial buildings in 1995.
 Real-time pricing and other rate structures are available as the power providers maneuver to offer
the most competitive structure in the electric deregulation environment.
 Thermal storage systems have longer operating hours of compressors and pumps, and chillers and
cooling tower at full-load operation, at lower outdoor temperatures at nighttime, and a backup
source for cooling during emergencies.
Drawbacks of thermal storage systems include high initial cost and complicated operation, maintenance,
and control.
Full Storage and Partial Storage
The ton-hour, or ton h (kWh), is the unit of stored refrigeration. One ton-hour is the refrigeration
or heat absorption of 12,000 Btu (3.516 kWh) performed by a refrigeration system during a 1-h
period. The aim of thermal storage strategy is to incur the lowest possible energy cost and initial
investment so that life-cycle costs are minimized. The economic benefit of a thermal storage system
can also be assessed by calculating the simple payback period of its cost. A simple payback or lifecycle
cost analysis of the building load profile, utility electric rate structure, and system characteristics
is always necessary.
Determination of the optimum size of a thermal storage system is based on the utility’s electric
rate structure (difference between on-peak and off-peak unit charges) and the building refrigeration
load profile.
Direct cooling denotes the process by which compressors produce chilled water to cool the
building directly. When the cost of the direct cooling by a refrigeration system is lower than the
cost of stored energy, the operation of the thermal storage system is said to be at chiller priority. On
the other hand, if the cost of the direct cooling is higher than the cost of stored energy, the operation
is said to be at storage priority. Construction cost and the utility’s incentive payments should be
considered. If the energy cost difference between on-peak and off-peak hours is great, the full use
of stored energy during on-peak hours may be most economical.
There are two kinds of thermal storage: full storage and partial storage. Figure 31.1 shows the
load-time diagrams for full storage and partial storage. For a full-storage, or load shift thermal storage,
system, all refrigeration compressors cease to operate during on-peak hours, and the building
refrigeration load during that period is entirely offset by the chilled water supplied from the thermal
storage tank, as shown in Fig. 31.1a.
Partial storage, or load leveling, can be either in load-leveling mode, in which refrigeration compressors
are operated at full capacity during on-peak hours, as shown in Fig. 31.1b, or in demandlimited
mode, in which building electric demand limits only part of the refrigeration compressors
operated, as shown in Fig. 31.1c.
A utility’s demand charge is the building total demand charge, which is the sum of the demand
charges for HVAC&R systems and other uses including the electricity for lighting, escalators, computers,
and electric appliances.
The following is a comparison of different storage strategies of a thermal storage system with
heat recovery and cold air distribution for an office building in Dallas, Texas, as described by
Tackett (1989).
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.3
31.4 CHAPTER THIRTY-ONE
24 24
Midnight
Refrigeration load Qrl, ton
12 6
1200
800
400
18
(a)
Noon Midnight
Off-peak
Ice burning
Off-peak On-peak
Refrigeration
load
24 24
Midnight
Refrigeration load Qrl, ton
12 6
1200
800
400
18
(b)
Noon Midnight
Off-peak Off-peak On-peak
24 24
Midnight
Refrigeration load Qrl, ton
12 6
1200
800
400
18
(c)
Noon Midnight
Off-peak Off-peak On-peak
Direct cooling Ice making
FIGURE 31.1 Full and partial storage: (a) full storage; (b) partial storage, all compressors
operating; (c) partial storage, 50 percent of compressors operating.
The larger the thermal storage system and the capacity of refrigeration compressors, the greater the
savings. Actually, selection of the size of the storage system is better determined according to lifecycle
cost, and system reliability should also be considered.
Ice Storage and Chilled Water Storage
In the late 1970s, engineers began experimental thermal storage applications in buildings. Throughout
the late 1990s, there are several thousand thermal storage systems operated in various commercial
buildings including offices, shopping centers, schools, and hospitals and industrial applications
in the United States.
Two thermal storage media are widely used for air conditioning systems: ice storage and chilled
water storage. At a temperature difference of 18°F (10°C), 1 lb (2.2 kg) of chilled water can store
18  1  18 Btu (19 kJ) of thermal energy, whereas 1 lb of ice can store 1  144  60  35 
169 Btu (178 kJ). If the density of water is 62.3 lb / ft3 (997 kg /m3) and the density of ice is 57.5
lb / ft3 (920 kg /m3), for the same stored cooling capacity, the storage volume for ice is only
about 0.12 that of the chilled water. In addition, ice storage systems generally provide chilled water
at a temperature of 34 to 35°F (1.1 to 1.7°C) to produce cold supply air between 42 and 49°F
(5.6 and 9.4°C). Ice storage systems incorporating cold air distribution significantly reduce the
volume flow rate of supply air, so air-side fan energy consumption and initial investment drop
accordingly.
Generally, ice storage systems have a rather low incremental capital cost compared with
conventional air conditioning systems without thermal storage. Ice storage systems can be easily
incorporated with cold air distribution. A lower incremental capital cost than that of chilled water
storage systems and the incorporation with cold air distribution are two main benefits of the ice
storage system.
In addition to the ice storage and chilled water storage systems, phase-change material storage
systems are sometimes used. The most common phase-change material used for cool thermal storage
is a mixture of inorganic salts, water, and nucleating and stabilizing agents which melts and
freezes at 47°F (8.3°C). Phase-change materials have high discharge temperatures and a high storage
tank volume of 6.0 ft3 /ton h (0.048 m3 /kWh) instead of 2.4 to 3.3 ft / ton h (0.019 to 0.027
m3 /kWh) for other ice storage systems; therefore, they have limited applications in commercial
buildings.
Currently, ice storage and chilled water storage systems can be classified into the following categories:
 Ice-on-coil, internal-melt ice storage system (IMISS)
 Ice-on-coil, external-melt ice storage system (EMISS)
 Encapsulated ice storage system (EISS)
 Ice-harvesting ice storage system (IHISS)
 Stratified chilled water storage system (SCWSS)
Chilled water storage systems are often large-capacity storage systems. In newly installed chilled
water storage systems, the stratified chilled water storage system is the most widely used.
Full storage Partial storage Demand limited
Utility incentive, $ 288,750 132,500 215,000
Net incremental cost, $ 291,750 21,750 124,250
Demand limit saving, kW 1,200 400 800
Total saving, $ 124,250 64,000 92,750
Simple payback, years 2.35 0.34 1.34
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.5
31.2 ICE-ON-COIL, INTERNAL-MELT ICE STORAGE SYSTEMS
System Description
An ice-on-coil, internal-melt ice storage system uses brine flowing inside coils to make ice and to melt
ice in the water that surrounds the coil. The central plant (cooling) and a chilled water or brine-incorporated
ice storage system of an ice-on-coil, internal-melt ice storage system consists of the following
components: chillers, ice storage tanks, chiller pumps, building pumps, controls, piping, and fittings as
well as AHUs, terminals, return air system, smoke control systems, and mechanical exhaust systems.
Centrifugal, screw, and reciprocating chillers are usually used in ice-on-coil internal-melt ice
storage systems depending on the size of the plant and types of condenser (water-cooled, aircooled,
or evaporatively cooled) used. In locations where the outdoor air temperature during nighttime
off-peak hours drops 20°F (11.3°C) lower than the daytime maximum temperature, air-cooled
chillers may sometimes be more efficient than water-cooled chillers.
Figure 31.2 is a schematic diagram of a typical ice-on-coil, internal-melt ice storage system for
an office building near Dallas, Texas.
Brine and Glycol Solution
Brine is a salt solution or an aqueous glycol solution used as a heat-transfer medium. Its freezing
point is lower than that of water, and depends on the concentration of salt or glycol in solution.
31.6 CHAPTER THIRTY-ONE
FIGURE 31.2 Schematic diagram of ice-on-coil, internal-melt ice storage system.
Brine is also used as a liquid coolant to absorb or to store heat energy in refrigeration and thermal
storage systems.
Ethylene glycol and propylene glycol are brines. They are colorless, nearly odorless liquids.
They are often mixed with water at various concentrations and used as freezing point depressants to
lower the freezing point of water. Inhibitors must be added to ethylene and propylene glycols to
prevent metal corrosion.
The freezing point of an aqueous ethylene glycol solution with a concentration of 25 percent by
mass drops to 10°F (12.2°C), and its rate of heat transfer is about 5 percent less than that of water.
The freezing point of a propylene glycol solution with a concentration of 25 percent by mass drops
to 15°F (9.4°C).
The physical properties of aqueous ethylene glycol solution are more appropriate for thermal
storage systems than those of aqueous propylene glycol solution. In certain applications, toxicity
considerations may be dictated by federal EPA requirements and local codes and regulations.
Ice Storage Tank
In an ice storage system, ice making or charging is the process in which compressors are used to
produce ice. Ice burning (ice melting or discharging) means that ice in the storage system is melted
in order to produce chilled water to offset the required refrigeration load.
In an ice-on-coil, internal-melt ice storage system, ice is produced, or charged, in multiple
storage tanks where closely spaced multicircuited polyethylene or plastic tubes are surrounded by
water, as shown in Fig. 31.3. Brine, an aqueous ethylene glycol solution with 25 to 30 percent ethylene
glycol and 70 to 75 percent water, circulates inside the tubes at about 24°F ( 4.4°C). The
water surrounding the tubes freezes into ice up to a thickness of about 0.5 in. (12.7 mm). Tubes
containing glycol solution entering and leaving the tank are arranged side by side alternately to provide
more uniform heat transfer.
Brine typically leaves the storage tank at 30°F ( 1.1°C). Plastic tubes occupy about one-tenth
of the tank volume, and another one-tenth is left empty to accommodate the expansion of ice during
ice making. Multiple ice storage tanks are always connected in parallel. During ice burning or ice
melting, brine returns from the cooling coils in the air-handling units at a temperature of 46°F
(7.8°C) or higher. It melts the ice on the outer surface of the tubes and is thus cooled to 34 to 36°F
(1.1 to 2.2°C). The brine is then pumped to the air-handling units to cool the air again.
In the storage tank, ice is stored and the high-pressure brine inside the tubes is separated from
the water, usually at atmospheric pressure, surrounding the tubes in the storage tank.
Case Study: Operating Modes of Ice-on-Coil Ice Storage System
In a typical ice storage system using ice-on-coil, internal-melt storage tanks in a 550,000 ft2 (51,115
m2) office building near Dallas, Texas, as described by Tackett (1989), there are two centrifugal
chillers. Ethylene glycol is used as the coolant. Each chiller has a refrigeration capacity of 568 tons
(1997 kW) when it produces 34°F (1.1°C) brine at a power consumption of 0.77 kW /ton (COP
4.56). If the brine leaves the chiller at 24°F ( 4.4°C), the refrigeration capacity then drops to 425
tons (1494 kW) with a power consumption of 0.83 kW / ton (COP 4.24).
A demand-limited partial-storage strategy is used; i.e., one chiller is operated during on-peak
hours, as shown in Fig. 31.1c. Meanwhile, ice is also burned during on-peak hours to reduce the demand
charge. Ice is charged during off-peak hours to reduce energy costs. The system uses 90 iceon-
coil, internal-melt storage tanks.
For summer cooling, the daily 24-h operating cycle can be divided into three periods: off-peak,
direct cooling, and on-peak.
Off-Peak. This is the period from 8 p.m. until the air-handling units start the next morning. During
this period, the primary operating mode is ice making, at a maximum capacity of 650 tons
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.7
(2285 kW). The chillers also provide direct cooling at a capacity less than 200 tons for refrigeration
loads that operate 24 h.
In this operating mode, the ice-on-coil, internal-melt storage tanks are charged. At the same
time, a 34°F (1.1°C) ethylene glycol solution is supplied to the air-handling units for nighttime
cooling. The DDC controller controls the ice storage system in the following operating sequence, as
shown in Fig. 31.2:
1. Open control valves CV-1, CV-2, CV-5, CV-6, CV-8, and CV-11; and close control valves CV-3,
CV-4, CV-6, CV-7, CV-9, and CV-10.
2. Reset the temperature of the glycol solution leaving the chiller Tel to 22°F ( 5.6°C).
3. Reset the load limit of both chillers to 100 percent.
4. Start the chiller and condenser pumps. Chiller pumps operate at high speeds during ice making
to provide a higher flow rate and a high rate of heat transfer in the storage tanks as well as a
greater head to overcome the pressure drop for both the evaporator and the coils in the ice storage
tanks. Chiller pumps operate at low speeds in direct cooling mode.
5. Start chillers 1 and 2 following the lead / lag sequence.
6. After chillers are started, open control valves CV-3 and CV-7.
7. Start the building chilled water circulating pumps in sequence.
31.8 CHAPTER THIRTY-ONE
FIGURE 31.3 Ice-on-coil, internal-melt ice storage tank.
8. Modulate control valves 4 and 5, and maintain a 34°F (1.1°C) glycol solution supply temperature
to the air-handling units.
If the ice-on-coil, internal-melt ice storage tanks are all 100 percent charged, the ice storage capacity
is 7500 ton h (26,370 kWh).
When the sensors detect that the ice storage tanks are 100 percent charged, the ice-making mode is
terminated. If nighttime after-hours cooling is not needed, the ice storage system shuts down. If the ice
inventory (the amount of stored ice in the tanks) falls below 90 percent, ice making starts again.
There are two additional operating modes during this period: ice making without direct cooling
for after-hours use and ice burning for after-hours use with all chillers shut off.
Direct Cooling. Direct-cooling operation lasts from the start of the air-handling units until noon
on weekdays. This period has two operating modes:
 Direct cooling mode. In this operating mode, chillers are operating and are reset to 34°F (1.1°C).
 Direct cooling with ice-burning or ice-melting mode. In this mode, both chillers are turned on.
When the required refrigeration load exceeds both chillers’ capacity, some ice storage will be discharged
to supplement the chillers.
On-Peak. On-peak hours are from noon until 8 p.m. weekdays. Ice-burning mode, with or without
chiller operation, is used in this period. In ice-burning or ice-melting mode, one chiller is operated
at the demand limit. This is the primary operating mode during summer cooling. The operating
sequence is as follows:
1. Open control valves CV-3, CV-6, CV-7, and CV-10; and close CV-8, CV-9, and CV-11.
2. Open control valve CV-1 and close CV-2 if chiller 1 is required to operate. Open CV-2 and
close CV-1 if chiller 2 is required to operate.
3. Modulate control valves 4 and 5 at normal open positions.
4. Reset chilled water temperature leaving the chiller to 32°F (0°C).
5. Set the load limit of the operating chiller to 400 kW.
6. Start one condenser pump.
7. Start chiller pumps 1 and 2 at low speed. Both pumps will operate during ice burning.
8. Start one chiller according to the lead / lag sequence.
9. Modulate control valves 4, 5, 6, and 7 to maintain a 34°F (1.1°C) chilled water supply temperature
to the air-handling units.
10. Start the brine circulating pumps in sequence. During on-peak hours, the brine circulating
pump needs a greater head to overcome the pressure drop of the coil in the AHU as well as the
pressure drop of the coil in the ice storage tanks.
System Characteristics
An ice-on-coil, internal-melt ice storage system is a modular system consisting of many closely
packed storage tanks connected in parallel. Such an ice storage system is more flexible during the
installation of storage tanks, especially for retrofit projects.
In ice-on-coil, internal-melt ice storage systems, off-peak cooling of the building can be provided
by direct cooling from the chillers. During ice burning, melted water separates the tube and
ice. Water has a much lower thermal conductivity (0.35 Btu /h ft °F, or 0.61 W/m°C) than that of
ice (1.3 Btu /h ft °F, or 2.25 W/m°C), so the capacity of the ice-on-coil, internal-melt ice storage
system is dominated by the rate of ice burning or melting.
Other system characteristics of ice-on-coil, internal-melt ice storage systems are listed in
Table 31.1.
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.9
31.3 ICE-ON-COIL, EXTERNAL-MELT ICE STORAGE SYSTEMS
System Description
In an ice-on-coil, external-melt ice storage system, ice builds up on the outer surface of coils or
tube banks, which are submerged in water in a storage tank. The refrigerant flows and evaporates
inside the tubes. When the ice melts, it cools the water at a temperature between 34 and 38°F (1.1
and 3.3°C) for cooling in the AHUs. The central plant (cooling) and chilled water incorporated ice
storage system of an ice-on-coil, external-melt system consists of chillers, evaporating coils, stor-
31.10 CHAPTER THIRTY-ONE
TABLE 31.1 System Characteristics of Ice-on-Coil, Internal-Melt Ice Storage Systems
IMISS/EMISS, EISS IHISS SCWSS
Zone thermal and sound control
Control zone Single or multizone Single or multizone
Conrol methods DDC or electric DDC or electric
Control modes PI, PID, or on /off PI, PID, or on /off
Sound control NC 25–40 NC 25–40
Indoor air quality
Minimum ventilation control DCV or MPC DCV or MPC
Filter efficiency Medium- to high- Medium- to high-
Humidity control Optional Optional
Air system
Types VAV air mixing or CV VAV air mixing or CV
Supply fan Centrifugal Centrifugal
Supply fan total pressure 4.5–5.5 in. WC 4.5–5.5 in. WC
Combined fan-motor-drive efficiency 55% 55%
Air economizer Fixed or differential dry bulb Fixed or differential dry bulb
or enthalpy or enthalpy
Return / relief fan Centrifugal or axial Centrifugal or axial
Relief / return fan total pressure Relief 0.6 in. WC, return Relief 0.6 in. WC, return
0.5–1 in. WC 0.5–1 in. WC
Combined fan-motor-drive efficiency 35–40% 35–40%
Cooling systems
Refrigeration compressor Centrifugal or screw Centrifugal or screw
Refrigerants HCFC-123, HCFC-22, HCFC-123, HCFC-22,
HFC-134a HFC-134a
Evaporator Flooded liquid cooler IHISS: harvester; SCWISS:
flooded liquid cooler
Condenser Water or evaporatively Water or evaporatively cooled
cooled
Energy performance IMISS: 0.85–1.2 kW/ ton, IHISS: 0.95–1.3 kW/ ton,
EMISS: 0.85–1.4 kW/ ton, SCWSS: FOM 0.85 to 0.92
EISS: 0.85–1.2 kW/ ton
Water pumps Centrifugal Centrifugal
Combined pump-motor efficiency 65% 65%
Heating systems Hot water coil or electric Hot water coil or electric heater
heater
Boiler AFUE Conventional 78%, Conventional 78%,
condensing 93% condensing 93%
Maintenance
Fault detection and diagnostics AHU, chiller AHU, chiller
FOM indicates figure of merit. Refer to Sec. 31.7 for details.
age tanks, condenser, heat exchanger, refrigerant pumps, chilled water pumps, air system controls,
piping, and fittings. A typical ice-on-coil, external-melt ice storage system is shown in Fig. 31.4.
Screw and centrifugal compressors are often used because of their higher efficiency. For an iceon-
coil, external-melt ice storage system with a capacity less than 2400 ton h (8438 kWh), a reciprocating
compressor may also be used. Evaporatively cooled condensers have a higher system
energy efficiency ratio (EER) and are often used in many new projects.
Ice Builders
Ice builders are large, well-insulated steel tanks containing many serpentine coils, usually made of
steel pipes of 1- to -in. (25- to 31-mm) diameter. HCFC-22 is often used as the refrigerant. The
refrigerant-filled serpentine coils are submerged in water in the ice builder and function as evaporators.
The ice build up on the coil is between 1 and 2.5 in. (25 and 64 mm) thick. When ice builds up
on the coil, the suction temperature of the compressor falls to 22 to 24°F ( 5.6 to  4.5°C).
Ice is melted by the water circulating over it. The steel tubes of the serpentine coil should be
spaced so that the built-up ice cylinders do not bridge each other. If the cylinders are bridged, the
paths of water circulation are blocked. Baffle plates are sometimes added to guide the water flow
and provide a secondary heat-transfer surface between the refrigerant and water.
The storage tanks containing refrigerant coils are usually located at a lower level or on a grade
because of their weight. Because the chilled water system in a multistory building is always under a
static head at lower levels, a heat exchanger, usually plate-and-frame type, is used to isolate the storage
tank brine system from the chilled water system connected to the AHUs. An alternative is to supply
chilled water directly to the storage tanks and pressurize the tanks. This arrangement obviates the
use of a heat exchanger and a corresponding increase in brine temperature of about 3°F (1.7°C).
Refrigerant Feed
Two kinds of refrigerant feed are widely used in ice-on-coil, external-melt ice storage systems: direct
expansion and liquid overfeed. Direct expansion (DX) uses the pressure difference between the
receiver at the high-pressure side and the suction pressure to force the refrigerant to flow through
the ice builder. A suction-liquid heat exchanger can be used to cool the liquid refrigerant from the
condenser for better efficiency. Direct expansion is simple, and no refrigeration pump is required.
Its main drawback is that 15 to 20 percent of the coil surface is used for superheat and is not available
for ice buildup.
As described in Secs. 10.1 and 10.4, liquid overfeed uses a refrigerant pump to feed ice-builder
coils about 3 times the evaporation rate they need. Because the liquid refrigerant wets the inner surface
of the ice-builder coils, it has a higher heat-transfer coefficient than direct expansion.
Ice-Charging Control
The thicker the ice built up on the coils, the greater the amount of ice stored in the tank. The thickness
of ice on the coil should be measured to meet ice-burning requirements during on-peak operating
hours or in the direct cooling period. Because ice has a higher volume than water, as the ice
builder is charged (i.e., as ice builds up on the coil), the water level rises. An electric probe can
sense the water level in the tank and thereby determine the amount of ice stored in the tank.
System Characteristics
The ice-on-coil, external-melt ice storage system is the oldest type of ice storage system. It may be
costly and complex. In the storage tank, stored ice occupies only about one-half the volume of the
tank, so the ice builder must be larger and heavier.
114
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.11
31.12 CHAPTER THIRTY-ONE
24
Refrigeration load, tons
12 16 20 0 4
137.5
(b)
Noon Hours
100
200
300
400
460
8
Off-peak Off-peak On-peak
Ice burning
and direct
cooling
Ice burning
Ice making
(a)
Compressor
Plate-and-frame
heat exchanger
Storage
tank
Ice
builder
Water
level
Water
Refrigerant
Condenser Suction-liquid
heat exchanger
CWS
CWR
FIGURE 31.4 A typical ice-on-coil, external-melt ice storage system: (a) schematic diagram; (b) operating diagram.
Case Study: An Ice-on-Coil, External-Melt Ice Storage System
Gilbertson and Jandu (1984) presented an ice-on-coil, external-melt ice storage system for a 24-
story office building with an area of 265,000 ft2 (24,628 m2) in San Francisco, California. This project
was completed in the early 1980s. A schematic diagram of this system is shown in Fig. 31.4a.
This office tower had a peak refrigeration load of 460 tons (1617 kW), and an ice storage capacity
of 3300 ton h (11,603 kWh) of refrigeration over 24 h. It required a 137.5-ton (4843-kW) refrigeration
capacity for the ice storage system for a hot summer day at summer outdoor design conditions.
For redundancy, two identical refrigeration systems using HCFC-22 as refrigerant were installed.
Each had a 70-nominal-ton (246-kW) compressor, an ice builder of 960 ton h (3375 kWh), a watercooled
condenser, and other accessories. (Nominal ton is the refrigeration capacity, in tons or kW,
at rating conditions.) Direct-expansion refrigerant coils with a suction-liquid heat exchanger were
used. When the thickness of built-up ice was about 2 in. (50 mm), the ice storage capacity reached
960 tonh (3375 kWh).
A plate-and-frame heat exchanger was used to isolate the storage tank from the tower chilled
water system. The 34°F (1.1°C) water from the ice builder entering the heat exchanger with an 8°F
(4.4°C) temperature increase cooled the chilled water from 54 to 38°F (12.2 to 3.3°C). This was a
partial-storage system. The operating modes during a 24-h cycle for this 24-story office tower,
shown in Fig. 31.4b, are as follows:
The HVAC&R and plumbing system, including ice storage, for this office building cost $2.4 million,
which was slightly less expensive than a conventional central system using centrifugal chillers.
System characteristics of ice-on-coil, external-melt ice storage systems are listed in Table 31.1.
31.4 ENCAPSULATED ICE STORAGE SYSTEMS
System Description
In an encapsulated ice storage system, plastic containers, filled with deionized water and ice-nucleating
agent, are immersed in a secondary coolant ethylene glycol solution in a steel or concrete
tank. Ice is charged and stored when the secondary coolant is at a temperature between 22 and 26°F
( 5.6 and  3.3°C) circulated through the tank. Ice is melted when the warm coolant returned
from the AHUs is circulated through the tank. Chillers can also provide direct cooling at a coolant
temperature from 36 to 42°F (2.2 to 5.6°C). The central plant (cooling) and the chilled water incorporated
ice storage system of an encapsulated ice storage system consist of the following components:
chillers, steel tank, encapsulated containers, pumps, air system controls, piping, and accessories.
Two types of encapsulated ice containers are currently available in the United States: dimpled
spheres of 4-in. (100-mm) diameter and rectangular containers approximately by 12 by 30 in.
(35 by 300 by 750 mm). The containers are made of high-density polyethylene and are designed to
withstand the pressure due to the expansion during freezing.
When containers are put or stacked inside the storage tank, they allow free circulation of fluid
and do not provide unwanted short-circuit fluid flow which causes degradation of performance. The
storage tank can be an open, nonpressurized type or pressurized type. An open storage tank needs a
barrier to keep the frozen containers submerged into the coolant.
138
Time period Operating mode
6.00 p.m. to 8.00 a.m. Off-peak, ice making
8.00 a.m. to 9.00 a.m. On-peak, ice making and direct cooling
9.00 a.m. to 6.00 p.m. On-peak, ice melting and direct cooling
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.13
Location of Chiller and Storage Tank
In encapsulated ice storage systems and ice-on-coil, internal-melt ice storage systems, the chillers
and storage tanks are usually connected in series, as shown in Fig. 31.5. When partial storage is
used, their relative location can be either in chiller upstream or chiller downstream arrangements. In
a chiller upstream arrangement, as shown in Fig. 31.5a, the chilled water returned from AHUs at
46°F (7.8°C) is often first cooled in the chiller to 40°F (4.4°C), and then it enters the storage tank
and is cooled down to 34°F (1.1°C). In the chiller upstream arrangement, since the chilled water
cooled at the chiller is at a higher temperature, this results in a higher COP at the chiller. However,
the usable portion of the total storage capacity will be reduced because of the lower storage tank
discharge temperature.
In a chiller downstream arrangement, as shown in Fig. 31.5b, the chilled water returned from the
AHUs at 46°F (7.8°C) is often first cooled in the storage tank to 40°F (4.4°C), and then it enters the
chiller and is cooled down to 34°F (1.1°C). In a chiller downstream arrangement, the COP of the
chiller is lower, and the usable portion of the total storage capacity of the ice storage tanks is increased.
Usually, during partial storage, the chiller upstream arrangement is often used for higher effi-
ciency in the chiller. When ice storage capacity becomes a problem, an analysis should be undertaken
to make an optimum choice.
Controls
According to Dorgan and Elleson (1993), ice-charging inventory in the storage tank is measured
based on the displacement of water in the tank when the ice is formed inside the encapsulated
31.14 CHAPTER THIRTY-ONE
Tank
Chiller
AHU
Tank
Tank
Tank
Tank
Tank
Chiller
AHU
(a) (b)
FIGURE 31.5 Relative location of chiller and storage tank: (a) chiller upstream; (b) chiller downstream.
containers. For open tanks, a static pressure transducer is often used to detect the water level in the
storage tank. In pressurized tanks, the expansion of the frozen containers forces the secondary
coolant overflowing into a separate inventory tank, and its water level is measured. A pump is used
to pump the overflowing fluid in the inventory tank back into the storage tank after discharging.
With nondimpled spherical containers that expand very little as the encapsulated ice freezes, storage
inventory can be monitored based on the integrated flow and temperature measurements.
Encapsulated ice storage systems use a storage tank bypass three-way modulating valve to control
the chilled water leaving temperature.
Encapsulated ice storage systems with a chiller upstream arrangement are well suited to chiller
priority control. When the system refrigeration load is less than the chiller capacity, the chilled water
bypasses the storage tanks completely. As soon as the system refrigeration load exceeds the
chiller capacity and the chiller leaving temperature increases above the leaving set point, the control
system diverts part of the chilled water flow through the storage tanks to maintain the required supply
temperature to the AHUs.
Storage priority is more complicated to achieve. A required refrigeration load prediction algorithm
to forecast the chiller cooling is needed each day. Chiller capacity is then limited by increasing the
chilled water leaving setpoint, and most of or all the refrigeration load is then met by the ice storage.
Generally, chillers should be controlled at full load during charging to prevent the reduction of
system efficiency and incomplete charging of ice storage. The chiller leaving temperature setpoint
should be set at or below the minimum required charging temperature so that the chiller is fully
loaded throughout the charging cycle.
Charging and Discharging
For encapsulated ice storage systems, the charging temperature decreases during the charge cycle as
the thickness of ice through which heat is transferred increases. Arnold (1991) reported that encapsulated
containers are subject to supercooling, i.e., cooling of the liquid water inside the container
below its freezing point prior to the ice formation. Supercooling occurs only in fully discharged
containers and results in a reduced rate of heat transfer at the beginning of the charging process. Supercooling
can be significantly reduced by the addition of nucleating agents.
For an entering chilled water temperature at the beginning of charging of 32°F (0°C) and a chilled
water temperature at the end of charging of 20 to 26°F ( 6.7 to  3.3°C), a typical range of charging
temperatures is between 4 and 12°F (2.2 and 6.7°C) corresponding to a charging period of 8 to 16 h.
Encapsulated ice storage systems have a steadily falling discharge rate when the discharge temperature
is kept constant, or a steadily rising discharge temperature when the discharge rate is constant.
This is due to the decreasing area of ice in contact with the container as the ice melts. The encapsulated
ice storage discharge temperature typically begins at 32°F (0°C) and ends at a discharge temperature
of 38 to 45°F (3.3 to 7.2°C) with a discharge temperature range of 6 to 13°F (3.3 to 7.2°C).
System Characteristics
System characteristics of encapsulated ice storage systems are listed in Table 31.1.
31.5 ICE-HARVESTING ICE STORAGE SYSTEMS
System Description
In an ice-harvesting ice storage system, ice is produced in a harvester, which is separate from the
storage tank where ice is stored. The evaporator of the chiller is a vertical plate heat exchanger
mounted above a water / ice storage tank. Low-pressure liquid refrigerant is forced through the
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.15
hollow inner part of the plate heat exchanger, in which liquid refrigerant is vaporized, and produces
a refrigeration effect.
The central plant (cooling) and brine incorporated ice storage system in an ice-harvesting ice
storage system consists of the following equipment and main components: chillers, an ice harvester,
storage tank, air system controls, piping, and accessories, as shown in Fig. 31.6.
Ice Making or Charging
During ice-making or -charging mode, a chilled aqueous ethylene glycol solution with a concentration
of 25 to 30 percent is pumped from the storage tank and distributed over the outer surface of
the evaporator plates at a temperature equal to or slightly above 32°F (0°C). It then flows downward
along the outer surface of the plate in a thin film. Water is cooled and then frozen into ice sheets approximately
0.2 to 0.3 in. (5 to 7.5 mm) thick. Periodically, hot gas is introduced into one-fourth of
the evaporator plates by reversing the refrigerant flow. Ice is harvested, or released from the outer
surface of the plates, in the form of flakes or chunks and falls into the storage tank below. Ice is
formed in 20 to 30 min and is harvested within 20 to 40 s. During harvesting, this section of plate
evaporator acts as a condenser.
Ice accumulates in the storage tank to occupy slightly less than 60 percent of the volume of the
tank. Because the ice flakes are usually smaller than 6 in. by 6 in. by 0.25 in. (1500 mm by 1500
mm by 63 mm), there is a large contact area between the return brine from the cooling coils and the
ice. The time required to melt the ice in the storage tank is less than one-tenth of the time needed in
ice making or charging.
For a reciprocating compressor using an evaporative condenser, the power consumption of the
chiller during ice making is about 0.95 to 1.1 kW / ton (COP 3.7 to 3.22). Because the evaporator
31.16 CHAPTER THIRTY-ONE
FIGURE 31.6 Schematic diagram of a typical ice-harvesting ice storage system.
plates must be located above the storage tank, ice-harvesting systems need more headroom than
other ice storage systems.
Chiller Operation
During off-peak hours, an ice-harvesting system can also be used to lower the temperature of brine
returning from the AHUs. Brine at a temperature of 34°F (1.1°C) is supplied to the air-handling
units to cool the air to a supply temperature of 42 to 45°F (5.6 to 7.2°C) during direct cooling. It is
then returned to the ice harvester at 50 to 60°F (10 to 15.6°C) and distributed over the evaporator
plates directly. After falling from the evaporator plates, brine is again cooled to a temperature of
34°F (1.1°C) before it leaves the storage tank.
Because of the higher temperature of return brine distributed over the evaporator plates, the capacity
of the ice harvester increases, and its power consumption decreases during chiller operation.
These changes in capacity and power consumption depend mainly on the temperature of return
brine distributed over the evaporator plates. During chiller operation, power consumption usually
varies between 0.75 and 0.85 kW / ton (COP 4.69 and 4.14).
Ice-harvesting ice storage systems are effective in ice-making and ice-burning operations. The
temperature of brine from the storage tank of the ice harvester can be lowered to 34°F (1.1°C),
which is 2°F (1.1°C) lower than in the ice-on-coil, internal-melt ice storage system. Ice-harvesting
systems have been successfully used in load shifting and load leveling to reduce electric demand
and energy cost. However, melting of the ice during the harvesting process not only decreases the
amount of ice harvested, but also adds an incremental refrigeration load to the system, which increases
the power consumption during the ice-making process.
An ice-harvesting ice storage system with brine system is an open system. More water treatment
is required than in an ice-on-coil, internal-melt ice storage system whose brine system is a closed
system.
System Characteristics
System characteristics of an ice-harvesting ice storage systems are listed in Table 31.1.
31.6 COMPARISON OF VARIOUS ICE STORAGE SYSTEMS
According to Dorgan and Elleson (1993), the discharge temperature, the energy consumption of the
chillers, the roughly estimated installed cost of storage tanks and accessories, and the chiller cost
without installation are as shown below:
According to Potter et al. (1995), a survey of 196 thermal storage systems was undertaken. The
users of the thermal storage systems were asked to assess the ice storage systems according to an
index of 1 to 10. One indicated complete dissatisfaction, whereas 10 meant complete satisfaction.
The ratings for the ice storage systems are as follows:
IMISS EMISS EISS IHISS
Discharge temperature, °F 34–38 34–36 34–38 34–36
Chiller energy use, kW/ ton 0.85–1.2 0.85–1.4 0.85–1.2 0.95–1.3
Storage tank volume, ft3 /ton h 2.4–2.8 2.8 2.4–2.8 3.0–3.3
Storage installed cost, $/ tonh 50–70 50–70 50–70 20–30
Chiller cost, $/ ton 200–500 200–500 200–500 1000–1500
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.17
In addition to these systems, there is another ice storage system called an ice slurry ice storage system
in which suspended ice crystals are formed. Ice slurry systems have limited commercial application
because of their higher costs.
31.7 STRATIFIED CHILLED WATER STORAGE SYSTEMS
System Description
A stratified chilled water storage system often uses a large storage tank to store chilled water at a
temperature between 40 and 45°F (4.4 and 7.2°C). The stored chilled water offsets the building
refrigeration load during on-peak hours to shift the load to the off-peak hours and reduces the
energy cost. Chilled water in the storage tank is stratified into three regions because of its gravity:
top warmer return water from the AHUs, middle region of steep temperature gradient, and bottom
colder chilled water from the chillers. The central plant (cooling) and the chilled water incorporated
storage system consist of chillers, a cylindrical storage tank, pumps, piping, air system controls,
and accessories.
Basics for Chilled Water Storage
The stored cooling capacity of a chilled water storage system depends on the temperature difference
between the warm water return from the AHUs and the chilled water stored in the tank, and
the amount of water stored. The larger the storage tank, the lower the capital cost per unit stored
volume. According to Dorgan and Elleson (1993), a chilled water storage system is economical
when its storage capacity exceeds 2000 ton h or 200,000 gal (7000 kWh or 760 m3). Currently,
chilled water systems achieve thermal separation between cold charged water and warm return water
by stratification, multiple tanks, membrane, diaphragm, and baffles. The stratified tank is the
simplest and most efficient method. Chilled water storage systems need a storage tank volume of 11
to 21 ft3 / tonh (0.089 to 0.169 m3 /kWh) compared to 2.4 to 3.3 ft3 / tonh (0.019 to 0.027
m3 /kWh) for ice storage systems.
Charging and Discharging. Charging is the process of filling the storage tank with chilled water
from the chiller, usually at a temperature between 40 and 45°F (4.4 and 7.2°C). Meanwhile, the
warmer return chilled water from the air-handling units or terminals, usually at a temperature between
55 and 60°F (11.1 and 15.6°C), is extracted from the storage tank and pumped to the chiller
to be cooled.
Discharging is the process of discharging the chilled water, at a temperature between 41
and 45°F (5.0 and 7.2°C), from the storage tank to the air-handling units and terminals. At the
same time, the warmer return chilled water from the coils fills the tank by means of storage water
pumps.
Loss of Cooling Capacity during Storage. During the storage of chilled water, the following
processes result in losses in cooling capacity:
 Stored chilled water is warmed by direct mixing of warmer return chilled water and stored colder
chilled water.
Ice-on-coil, internal-melt ice storage systems 7.82
Ice-on-coil, external-melt ice storage systems 6.78
Encapsulated ice storage systems 7.25
Ice-harvesting ice storage systems 5.02
31.18 CHAPTER THIRTY-ONE
 Heat from prviously stored warmer return chilled water is transferred from the warmer tank wall
to the stored chilled water.
 Heat is transferred through the tank wall from the warmer ambient air.
Figure of Merit. A more easily measured, enthalpy-based figure of merit (FOM) is often used to
indicate the loss of cooling capacity of the stored chilled water during the charging and discharging
processes in a complete storage cycle. The FOM is defined as
(31.1)
where Qdis  cooling capacity available during discharge process, Btu/ h (W)
Qch  theoretical cooling capacity available during charging process, Btu/ h (W)
, mass flow rate and summation of mass flow rate of water, lb/ h (kg / s)
cpw  specific heat of water, Btu / lb°F (J /kg°C)
Trc  warmer return chilled water temperature filling storage tank during discharge process,
°F (°C)
To  outlet temperature of stored chilled water, °F (°C)
Trc,m  mass-weighted average temperature of return chilled water at inlet during discharge
process, °F (°C)
Ti  inlet temperature of stored chilled water during charging process, °F (°C)
The smaller the losses of cooling capacity during chilled water storage, the greater the value of FOM.
Storage Tanks
Chilled water storage tanks are usually flat-bottomed vertical cylinders. A cylindrical tank has a
lower surface-to-volume ratio than a rectangular tank. Large cylindrical tanks typically have a
height-to-diameter ratio of 0.25 to 0.35. Steel is the commonly used material for above-grade tanks,
and concrete is widely used for underground tanks. In certain projects, precast, prestressed, cylindrical
concrete tanks with enclaved watertight steel diaphragms are used for large chilled water
storage facilities with a volume over 2.5 million gal (9463 m3). All outdoor above-grade structures
should have a 2-in.- (50-mm-) thick external insulation layer spray-on polyurethane foam, a vapor
barrier, and a highly reflective top coating.
Stratified Tanks
Stratified tanks rely on the buoyancy of warmer return chilled water, which is lighter than colder
chilled water, to separate these two chilled waters during charging and discharging. Diffusers are
used to lower entering and leaving water velocity to prevent mixing. In a stratified tank, colder
stored chilled water is always charged from the bottom diffusers arranged concentrically, as shown
in Fig. 31.7a. It is also discharged from the same bottom diffusers.
The warmer return chilled water is introduced to and withdrawn from the tank through the top
lateral diffusers.
According to field measurements, stratified tanks have a figure of merit between 0.85 and 0.92.
Tran et al. (1989) showed that there is no significant difference in FOM between stratified tanks and
membrane tanks or empty tanks. A membrane tank is a storage tank in which a membrane separates
the colder stored chilled water and warmer return water. An empty tank is a storage tank in which
walls are used to separate the colder and warmer chilled water.
Compared with membrane tanks and empty tanks, stratified tanks have the advantages of simpler
construction and control, greater storage capacity, and lower cost. Stratified tanks are widely
used in chilled water storage installations.
	m?w  m?w
FOM 
Qdis
Qch

	m?wcpw(Trc  To)
m?wcpw(Trc,m  Ti)
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.19
Temperature Gradient and Thermocline
Vertical temperature profiles are formed during charging or discharging in stratified tanks at various
time intervals. Temperature profiles may be illustrated on a height-temperature (H-T) diagram at the
beginning, the middle, and near the end of the charging process, as shown in Fig. 31.8. In the middle
of the charging process along the vertical height of the storage tank, chilled water is divided into
three regions: bottom colder-and-heavier stored chilled water, thermocline, and top warmer-andlighter
return chilled water.
A thermocline is a stratified region in which there is a steep temperature gradient. The water temperature
often varies from 42 to 60°F (5.6 to 15.6°C). The thermocline separates the colder stored
chilled water from the warmer return chilled water. The thinner the thermocline, the smaller the mixing
loss.
Diffusers
The layout and configuration of diffusers in a stratified tank have a significant effect upon the mixing
of the colder and warmer chilled water, as well as on the formation of the thermocline. The purpose
of the diffusers and their connecting piping is to distribute the incoming chilled water evenly,
so that it flows through the inlet openings with sufficiently low velocity (usually lower than 0.9 ft / s,
or 0.27 m / s) to minimize mixing of colder and warmer chilled water. Wildin (1990) recommended
that diffuser design and installation take into account the following:
 Warmer return chilled water should be introduced as closely as possible to the top water level in
the stratified tank. Colder stored chilled water should be introduced just above the bottom floor of
the tank.
 The inlet temperature of chilled water should be controlled within a narrow band ( 2°F or 
1.1°C) during charging to avoid additional mixing.
31.20 CHAPTER THIRTY-ONE
FIGURE 31.7 Double-octagon diffuser for a cylindrical stratified tank: (a) plan view of bottom diffusers; (b) 12-in.
diameter PVC pipe diffuser.
 Obstructions in the flow crossing the tank, other than diffusers and the connecting piping, should
be minimized.
 The primary function of diffusers is to reduce mixing. Mixing can occur at two points: at the start
of the charging and discharging processes during the formation and reformation of the thermocline,
and at the inlet side of the thermocline after the thermocline has been formed.
 Mixing near the inlet diffuser can be minimized if the incoming chilled water initially forms a
thin layer of gravity current that travels across the tank because of the gravity difference,
rather than inertia. Gravity current slowly pushes the chilled water originally in the tank out
of the way so that mixing only occurs at the front of the gravity current when it first crosses
the tank.
 There are two kinds of diffusers: linear diffusers and radial disk diffusers. Large stratified tanks
usually incorporate linear diffusers.
 Inlet flow from the top diffusers should be upward or horizontal. Bottom diffusers should flow
downward and have slots spreading at 120°. The cross-sectional inlet area of the branch pipe leading
to the diffuser should be at least equal the total area of the diffuser openings in that branch.
 Mixing on the inlet side of thermocline depends on the inlet Reynolds number Rei and Froude
number Fri. The inlet Reynolds number is closely related to the inlet velocity and is defined as
(31.2)
where volume flow rate of chilled water, ft3 /s V?w 

0.00223V?gal
ldif 	w
Rei 
V?
w
ldif 	w
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.21
FIGURE 31.8 Temperature profile and thermocline in a stratified tank.
volume flow rate of chilled water, gpm
ldif  linear length of diffuser, ft
	w  kinematic viscosity of water, ft2 /s
According to Wildin (1990), when Rei 850, loss due to mixing and loss of cooling capacity during
discharge can be significantly reduced. If Rei is determined, the length of the linear diffuser can
be calculated from Eq. (31.2). The inlet Froude number Fri is defined as
(31.3)
where g  acceleration of gravity, ft / s2 (m/s2)

i  density of inlet water, lb/ ft3 (kg /m3)

a  density of ambient water stored in tank at diffuser level, lb / ft3 (kg /m3)
In Eq. (31.3), hi indicates the inlet opening height, ft (m). It is the vertical distance occupied by the
incoming flow when it leaves the diffuser and forms the gravity current. For the bottom diffusers,
inlet opening height hi indicates the vertical distance between the tank floor and the top of the opening
of the diffuser.
Self-balancing. Self-balancing means that the water flow introduced to or extracted from the tank
should be self-balanced according to requirements, i.e., evenly distributed at all flow conditions.
This includes the following requirements:
 The piping design should be symmetric.
 Branch pipes should be equal in length.
 Flow splitters should be added at the appropriate points.
 Pipe diameter reduction should be combined with the flow splitter.
 Long-radius elbows should be used.
Charging and Discharging Temperature versus Tank Volume
The FOM of a stratified tank is closely related to the temperature difference of the outlet temperature
of stored chilled water during discharging To and the inlet temperature of stored chilled water
during charging Ti. For a complete charging and discharging cycle, the average To during discharging
is always greater than the average Ti because of the mixing loss and heat gains, provided that
the water flow rate is constant. The smaller To  Ti, the higher the FOM.
Figure 31.9 shows curves of chilled water temperature versus tank volume during the charging
and discharging processes of a complete chilled water storage cycle in a large stratified tank. Inlet
and outlet temperatures are measured at the openings of the top and bottom diffusers. During the
charging process, return chilled water is extracted from the top diffusers of the stratified tank,
cooled in the chiller, and charged into the stratified tank again through the bottom diffusers. The inlet
temperature of the stored chilled water Ti gradually decreases as the stored volume increases.
This is due to a comparatively lower rate of heat transfer to the inlet water from the warmer ambient
water, piping, and tank wall after the beginning of the charging process.
During the discharging process, stored chilled water is extracted from the stratified tank and
supplied to the cooling coils in the air-handling units and terminals. The return chilled water is introduced
to the stratified tank through the top diffusers. The outlet temperature of the stored chilled
water To gradually increases as the stored volume decreases.
Both the outlet temperature of return chilled water during charging Trc,o and inlet temperature of
return chilled water during discharging Trc.i should be controlled between 55 and 60°F (12.8 and
15.6°C) so that stratification can be maintained in the storage tank.
Fri 
V?
w
ldif[ghi
3( 
i 

a) / 
i]0.5
V?
gal 
31.22 CHAPTER THIRTY-ONE
System Characteristics
System characteristics of stratified chilled water storage systems are listed in Table 31.1.
31.8 CASE STUDY: A STRATIFIED CHILLED WATER STORAGE
SYSTEM
A full-storage stratified chilled water storage system was completed in August 1990 to serve a
1.142 million ft2 (106,134 m2) electronics manufacturing facility in Dallas, Texas. The following
are the details of this project as described by Fiorino (1991).
Chilled Water Storage System
The 2.68-million-gal (10,144-m3) stratified tank is a precast, prestressed, cylindrical concrete water
tank. The design parameters of this stratified chilled-water storage system are as follows:
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.23
FIGURE 31.9 Chilled water temperature versus tank volume curves during charging and discharging
processes.
There are two 1200-ton (4220-kW) and two 900-ton (3164-kW) centrifugal chillers to serve zone 1
and zone 2, as shown in Fig. 31.10. There are also five chiller pumps, one of which is a standby.
Two variable-speed building pumps are installed for each zone, one of which is a standby pump.
Three storage pumps are used for chilled water storage, one of which is standby.
Concentric Double-Octagon Diffusers
Concentric double-octagon diffusers are used at both the top and bottom of the stratified tank, as
shown in Fig. 31.7a. Each octagon introduces 50 percent of the total volume flow during the charging
process. Over time, a double-octagon arrangement provides nearly twice the effective linear
length for diffusers of a single-octagon arrangement.
The total effective length of the eight diffusers in the outer octagon is 559 ft (170 m). For a total
flow of 0.5  5120  2560 gpm (162 L / s) through the outer octagon, the volume flow rate per linear
ft (0.30 m) of diffuser is
The kinematic viscosity of water at 42°F is 1.66  10  5 ft2 /s, so the inlet Reynolds number is calculated
as
Similarly, the inlet Reynolds number for inner octagon is 1068. Both are close to the upper limit of 805.
The acceleration of gravity g  32.2 ft / s2, and
For a common inlet opening height hi  5.64 in, or 0.47 ft (0.14 m), the inlet Froude number for
diffusers in the outer octagon is calculated as
Fri 
0.0102
[(32.2)(0.47)3(0.00064)]0.5
 0.22

i 

a

i

62.4263  62.3864
62.3864
 0.00064
Rei 
V?
dif
ldif 	w

0.0102
1.66  105
 615
V?
dif
ldif

2560  0.1337
559  60
 0.0102 ft2/s
31.24 CHAPTER THIRTY-ONE
24,500 ton h (861,420 kWh)
3200 tons (11,251 kW)
16 h
8 h
40°F (4.4°C)
42°F (5.6°C)
56°F (13.3°C)
5120 gpm (323 L / s)
105.5 ft (32.2 m)
41 ft (12.5 m)
2.68 million gal (10,144 m3)
90%
Storage cooling capacity
Maximum refrigeration load
Charge process duration
Discharge process duration
Inlet temperature during charging Ti
Limiting outlet temperature during
discharging To
Inlet temperature during discharging Tre,i
Maximum volume flow rate
Tank diameter
Tank height
Tank volume
Usable tank volume
31.25
FIGURE 31.10 Schematic diagram of a stratified chilled water storage system for an electronic facility in Dallas,
Texas.
Similarly, the inlet Froude number for diffusers in the inner octagon is 0.38.
At a lateral slot spacing of 0.5 ft (0.15 m) in the inner octagon diffusers and 0.87 ft (0.26 m) in
the outer octagon diffusers, each linear diffuser has 32 lateral slots. If the maximum inlet velocity is
0.9 ft / s (0.27 m / s), the opening area for each lateral slot is
Because the lateral slot is spread at an angle of 120° downward, if the length of the lateral slot is
about 1 ft (0.30 m), the width of the slot is 0.025 /1  0.025 ft, or about 0.3 in. (7.6 mm). If the
cross-sectional area of each linear diffuser is equal to the slot openings, the diameter of the diffuser
is therefore
To provide even distribution of incoming water flow, flow splitters are used to divide the water flow
evenly to the split mains and branches.
Charging Process
During full-load operation in hot weather, charging is performed from 8 p.m. until noon the next
day. All four chillers and all water pumps except the standby pumps are operated simultaneously to
provide direct cooling during off-peak hours as well as the required stored cooling capacity during
on-peak hours the next day.
The direct-cooling refrigeration load during off-peak hours varied from 1980 to 2600 tons (6962
to 9142 kW), with a total of 34,800 ton h (122,357 kWh). The required stored cooling capacity to
meet the refrigeration load during on-peak hours was 21,300 ton h (74,891 kWh) on July 17, 1989.
For a charging process of 16 h, the four chillers are operated at an average refrigeration load of
about 3600 tons (12,658 kW). The set point of the chilled water temperature leaving the chiller is
39.5°F (4.2°C), and the outlet return chilled water temperature is around 56°F (13.3°C).
Before charging, the chillers and water pumps are started. During the charging process, the control
valves CV-1 and CV-2 and solenoid valves SV-3, SV-5, SV-8, and SV-9 are opened, as shown in
Fig. 31.10; and solenoid valves SV-4, SV-6, and SV-7 are closed. This provides both the charging of
the storage tank and the supply of chilled water to building pumps during off-peak hours.
Return chilled water at around 56°F (13.3°C) is extracted from the stratified tank through the top
diffusers and the control valve CV-1 and solenoid valve SV-3 by storage pumps SP-1 and SP-2. It
flows through solenoid valves SV-5 and SV-9 and the chiller pumps, cools in the chiller, and leaves
the chiller at 39.5°F (4.2°C). After that, the chilled water is divided into two streams. One of the
streams flows through solenoid valve SV-8 and control valve CV-2 and is charged to the stratified
tank through the bottom diffusers at 40°F (4.4°C). The other stream is extracted by the building
pump and supplied to the air-handling units and terminals for direct cooling.
Discharging Process
Before the shutoff of the chillers and chiller pumps, the control valves and solenoid valves should be
switched over to the followings: control valves CV-1 and CV-2 and solenoid valves SV-1, SV-2, SV-4,
SV-6, and SV-7 should be open, and solenoid valves SV-3, SV-5, SV-8, and SV-9 should be closed.
Stored chilled water at 41°F(5°C) is extracted by storage pumps SP-2 and SP-3 via bottom diffusers
and control valve CV-2. It then flows through SV-6, is extracted again by the building pumps,
and is supplied to the cooling coils in the air-handling units and terminals. Return chilled water at a
D   4A

0.5

4  0.025  32

(0.5)  1.01 ft (305 mm)
5120  0.1337
60(0.9)(32)(8  8)
 0.025 ft2 (0.023m2)
31.26 CHAPTER THIRTY-ONE
temperature Trc  56°F (13.3°C) is then forced through solenoid valve SV-7 and control valve CV-1
and introduced to the stratified tank through the top diffusers.
Temperature Trc is maintained by changing the chilled water temperature supplied to the air-handling
units from 45 to 52°F (7.2 to 11.1°C) instead of 41°F (5°C). As soon as temperature sensor T2
senses a return chilled water temperature Trc drop below 56°F (13.3°C), the DDC controller closes
the control valve CV-z slightly. Less stored chilled water is extracted by the building pumps. The
required amount of return chilled water at 56°F (13.3°C) bypasses the crossover and mixes with the
41°F (5°C) stored chilled water, which results in a higher supply temperature to the cooling coils.
During on-peak hours, the discharging process requires a refrigeration load from 2500 to 2800
tons (8790 to 9845 kW), with a total of 21,250 ton h (74,715 kWh) for the 8-h on-peak period.
Part-Load Operation
This electronic manufacturing facility includes clean rooms, computer rooms, compressed air aftercoolers,
and manufacturing equipment, all of which need 24-h continuous cooling. Its daytime
refrigeration load averages about 1240 tons (4360 kW).
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.27
Refrigeration load, tons Electric demand, MW
Off-peak Off-peak On-peak
12
Noon Midnight
Refrigeration
load
Direct cooling Direct cooling
Midnight
2000
2400
2800
3200
3600
5
6
7
8
9
10 8 6 4 2 0 14 16 18 20 22 24
Charging
Discharging
Charging
Estimated
refrigeration
load curve
Electric
demand with
storage
Electric
demand without
storage
Drop in electric
demand in on-peak
hours
FIGURE 31.11 Electric demand curves and the storage cycle of the stratified chilled water storage system for an
electronic manufacturing facility in Dallas, Texas.
When the refrigeration load is reduced or the entering temperature of condenser water drops
because of a lower wet-bulb temperature, the following adjustments are made during part-load
operation:
 The temperature of inlet water entering the stratified tank during the charging process Ti is raised
from 39.5° to 42.5°F (4.2 to 5.8°C), which increases the chillers’ capacity and lowers their power
consumption. All four chillers and their auxiliary equipment do not have to operate simultaneously.
 Instead of 8 p.m. daily, the start of the charging process can be delayed until nearly all the stored
chilled water in the stratified tank is discharged. The discharging process might last 10 to 14 h
during part-load operation instead of 8 h in full-load operation.
System Performance
According to the operating cycle extended from August 24 to August 26, 1990, the electric demand
dropped about 2.5 MW during on-peak hours as intended because four chillers and their corresponding
auxiliary equipment were shut down during on-peak hours, as shown in Fig. 31.11. The
maximum storage capacity of the stratified tanks was 27,643 ton h (97,193 kWh). The difference
between the average outlet temperature during discharging and the average inlet temperature during
charging To  Ti was 1.1°F (0.61°C), and the figure of merit was 92.2 percent.
REFERENCES
Arnold, D., Laboratory Performance of an Encapsulated Ice-Storage, ASHRAE Transactions, 1991, Part II,
pp. 1170–1178.
ASHRAE, ASHRAE Handbook 1996, HVAC Systems and Equipment, ASHRAE Inc., Atlanta, GA, 1996.
ASHRAE, ASHRAE Handbook 1997, Fundamentals, Atlanta, GA, 1997.
Denkmann, J. L., Performance Analysis of a Brine-Based Ice Storage System, ASHRAE Transactions, 1985,
Part I B, pp. 876–891.
Dorgan, C. E., and J. S. Elleson, Design of Cold Air Distribution Systems with Ice Storage, ASHRAE
Transactions, 1989, Part I, pp. 1317–1322.
Dorgan, C. E., and Elleson, J. S., Design Guide for Cool Thermal Storage, ASHRAE Inc., Atlanta, 1993.
Fields,W. G., and D. E. Knebel, Cost Effective Thermal Energy Storage, Heating / Piping / Air Conditioning,
July 1991, pp. 59–72.
Fiorino, D. P., Case Study of a Large, Naturally Stratified, Chilled-Water Thermal Storage System, ASHRAE
Transactions, 1991, Part II, pp. 1161–1169.
Gatley, D. P., Successful Thermal Storage, ASHRAE Transactions, 1985, Part I B, pp. 843–855.
Gilberston, T. A., Ice Cools Office-Hotel Complex, Heating / Piping / Air Conditioning, August 1989, pp.
47–52.
Gilbertson, T. A., and R. S. Jandu, 24-Story Office Tower Air Conditioning System Employing Ice Storage—A
Case History, ASHRAE Transactions, 1984, Part I B, pp. 387–398.
Harmon, J. J., and H. C. Yu, Design Consideration for Low-Temperature Air Distribution Systems, ASHRAE
Transactions, 1989, Part I, pp. 1295–1299.
Hittle, D. C., and T. R. Smith, Control Strategies and Energy Consumption for Ice Storage Systems Using Heat
Recovery and Cold Air Distribution, ASHRAE Transactions, 1994, Part II, pp. 1221–1229.
Lorsch, H. G., Air-Conditioning System Design Manual, ASHRAE Inc., Atlanta, GA, 1993.
Lumpkin, R. M., Thermal Storage: A Reversible Process, HPAC, no. 1, 1998, pp. 136–142.
MacCracken, C. D., Off-Peak Air Conditioning: A Major Energy Saver, ASHRAE Journal, December 1991,
pp. 12–22.
Pearson, F. J., Ice Storage Can Reduce the Construction Cost of Office Buildings, ASHRAE Transactions, 1989,
Part I, pp. 1308–1316.
31.28 CHAPTER THIRTY-ONE
Potter, R. A., D. P. Weitzer, D. J. King, and D. D. Boettner, ASHRAE RP-766: Study of Operational Experience
with Thermal Storage Systems, ASHRAE Transactions, 1995, Part II, pp. 549–557.
Schiess, K., RTP  TES  ?, Engineered Systems, no. 10, 1998, pp. 102–110.
Silvetti, B., and M. MacCracken, Thermal Storage and Deregulation, ASHRAE Journal, no. 4, 1998, pp. 55–59.
Sohn, C. W., and J. J. Tomlinson, Diurnal Ice Storage Cooling Systems, ASHRAE Transactions, 1989, Part I,
pp. 1079–1085.
Spethmann, D. H., Optimal Control for Cool Storage, ASHRAE Transactions, 1989, Part I, pp. 1189–1193.
Stamm, R. H., Thermal Storage Systems, Heating / Piping / Air Conditioning, January 1985, pp. 133–151.
Tackett, R. K., Case Study: Office Building Uses Ice Storage, Heat Recovery, and Cold Air Distribution,
ASHRAE Transactions, 1989, Part I, pp. 1113–1121.
Townsend, S. B., and J. G. Asbury, Cooling with Off-Peak Energy: Design Implications of Different Rate
Schedules, ASHRAE Transactions, 1984, Part I B, pp. 360–373.
Tran, N., J. E. Kreider, and P. Brothers, Field Measurement of Chilled Water Storage Thermal Performance,
ASHRAE Transactions, 1989, Part I, pp. 1106–1112.
The Trane Company, Ice Storage Systems 1987, Applications Engineering Manual, American Standard, Inc., La
Crosse, WI, 1987.
Trueman, C. S., Operating Experience with a Large Thermally Stratified Chilled-Water Storage Tank, ASHRAE
Transactions, 1987, Part I, pp. 697–707.
Wildin, M. W., Diffuser Design for Naturally Stratified Thermal Storage, ASHRAE Transactions, 1990, Part I,
pp. 1094–1101.
Williams, C. D., Optimizing TES Chiller Management, ASHRAE Journal, no. 4, 1996, pp. 43–48.
AIR CONDITIONING SYSTEMS: THERMAL STORAGE SYSTEMS 31.29
CHAPTER 32
COMMISSIONING AND
MAINTENANCE
32.1
32.1 HVAC&R COMMISSIONING 32.1
Commissioning and Design Intent 32.1
Necessity of HVAC&R
Commissioning 32.1
Scope of HVAC&R Commissioning 32.2
Testing, Adjusting, and Balancing
(TAB) 32.3
HVAC&R Commissioning Team 32.4
When to Perform HVAC&R
Commissioning 32.4
Cost of HVAC&R Commissioning 32.5
32.2 HVAC&R MAINTENANCE 32.5
Basics 32.5
Maintenance Contractors and
Maintenance Personnel 32.5
Monitoring and Fault Detection and
Diagnostics Assisting Predictive
Maintenance 32.6
REFERENCES 32.6
32.1 HVAC&R COMMISSIONING
Commissioning and Design Intent
According to Wilkinson (1999a, 1999b), commissioning is defined as a systematic, documentated,
and collaborative process that includes inspection, testing, and training conducted to confirm that a
building and its associated serving systems are capable of being operated and maintained in conformance
with the design intent. Design intent is the occupants’ assumed intention of the design as well
as their assumed operation of the building and the servicing systems. Design intent sets the requirements
for occupant satisfaction and forms the foundation and basis of technical criteria and documents—
design drawings and specifications. HVAC&R is one of the building servicing systems.
Wilkinson (1999a, 1999b) recommended that the design intent document include the following:
 General description of the building type and occupancy category
 Particular needs such as air cleanliness, and outdoor air volume flow rate, as well as processes
that require special indoor environments
 Outdoor and indoor design conditions
 Space pressurization and relative pressurization of adjacent spaces
 Emergency operation during utility outages
 Applicable codes and fire / life safety requirements
 HVAC&R system selection if it is mandated by the owner
 Operations and maintenance manuals and staff training
Necessity of HVAC&R Commissioning
Air conditioning or HVAC&R systems in buildings are commissioned based on a sequence of planning,
design, bidding, and construction. Many installed air conditioned systems do not work as
expected by the owners. Because of the higher demands of indoor environmental control, indoor air
quality, and energy-efficient operation as well as the advances in technology in recent years, especially
the microprocessor-based controls, an air conditioning system becomes more and more complicated
and is required to provide considerably more functions than several decades ago.
Tseng (1998) reported on a survey of 60 commercial buildings; the results were presented in the
1994 National Conference on Building Commissioning sponsored by Portland Energy Conservation,
Inc. This survey found the following:
 More than one-half of the buildings suffered control problems.
 Of the buildings 40 percent had HVAC&R equipment problems.
 Of the buildings 25 percent had energy management control systems, economizers, and variablespeed
drives that did not run properly.
There are many of building owners and facility managers who claim that the functional performance
of their HVAC&R systems does not meet their expectations.
ASHRAE/IESNA Standard 90.1-1999 mandates that HVAC control systems shall be tested to
ensure that control elements are calibrated, adjusted, and in proper working condition. For projects
larger than 50,000 ft2 (4650 m2) conditioned space area, except warehouses and semiheated spaces,
detailed instructions for commissioning HVAC systems shall be provided by the designer in plans
and specifications.
Portland Energy Conservation noted top deficiencies discovered by HVAC&R commissioning:
 Incorrect scheduling of HVAC&R
 Incorrect cooling and heating sequence of operation
 Incorrect calibration of sensors and instrumentation
 Lack of control strategies for optimum comfort and energy-efficient operations
 Improper air and water economizer operations
 Microprocessor-based DDC systems not fully utilized
 Short cycling of HVAC&R equipment
 Lack of design intent
 Lack of training of HVAC&R operators or service contractor for complex systems
Tseng (1998) cites the benefits of HVAC&R commissioning:
 HVAC&R commissioning is the quality control tool for building owners, and they know that the
savings from fewer costly change orders and lower operating and maintenance costs will accrue.
 For those jobs with commissioning requirements, contractors and subcontractors know that it is
not advisable for them to cut corners, and they benefit from lowering their cash set-asides for warranty
reserves and callbacks.
 The architects can expect a building with far fewer postconstruction headaches to handle after
commissioning.
 The engineers know that the HVAC&R systems are virtually assured of operating as intended,
thus eliminating postoccupancy troubleshooting visits.
 The building occupants enjoy a higher-performing building with a smoothly functioning
HVAC&R system which provides a healthy and comfortable indoor environment with the benefits
of increased productivity.
Scope of HVAC&R Commissioning
Tseng (1998) and Ellis (1998) described the scope of HVAC&R commissioning as follows:
32.2 CHAPTER THIRTY-TWO
 Clarifly owner priorities and design intent. The commissioning outcomes include a healthy and
comfortable indoor environment, acceptable indoor air quality, energy efficiency, optimized functional
performance, maintainability, and constructibility. Constructibility is the ability of a project
to be built as designed.
 Document and monitor all facets of construction. Documentation includes preparation of the
commissioning specification, preparation of the verification test procedures, and review of key
equipment and instrumentation submittals. Monitoring includes preparation of system readiness
checklists such as electricity to the fans, pumps, and compressors; monitoring and verifying
equipment / system start-up and operation as well as the performance of temperature and EMCS
controls; and testing, adjusting, and balancing (TAB) work.
 Verify TAB work. In HVAC&R commissioning, TAB is a tool and is a part of the commissioning.
It is often provided by a TAB subcontractor.
 Extensively test all subsystems and their components and controls (acceptance commissioning).
Verify and document functional performance testing of all systems including control systems, so
that all systems comply with the contract documents. Establish an as-delivered performance
record, and verify the as-built record for all systems. The documentation should also include the
corrective action reports for every deficiency found, the follow correction and retest.
For acceptance, the final commissioning report should be completed and submitted, and thus,
the recommendation of acceptance determined. All documents should be turned over to the owner.
 Provide detailed training of operating personnel for each major system and equipment. Training
plans and attendance records should be supervised. It is important to conduct system training as
well as its operation so that all the individual components are functioning together as a system.
System operators cannot be expected to operate the system properly unless they know how it is
intended to work and why. If possible, operation and maintenance staff are encouraged to participate
in the verification tests as an excellent training.
 Continue post-acceptance commissioning and ongoing monitoring. Post-acceptance commissioning
and ongoing monitoring are a continuation of adjustment, optimization, and improvement of
HVAC&R systems to meet a specified goal by monitoring specific system operating parameters.
Data trend and diagnostic capabilities of DDC control systems are valuable during HVAC&R
commissioning and post-acceptance commissioning.
Testing, Adjusting, and Balancing (TAB)
Testing means to determine quantitative performance of equipment. Adjusting denotes that fluid
flow rate and air flow pattern are adjusted as specified in design. Balancing indicates that flows are
adjusted in proportion within the distribution system based on design requirements.
ASHRAE/ IENSA Standard 90.1-1999 mandates that construction documents shall require that
all HVAC systems be balanced in accordance with generally accepted engineering standards. Ducted
air and water flow rates shall be measured and adjusted within 10 percent of design flow rates. Variable-
speed, variable-volume flow distribution systems need not be balanced upstream of a pressureindependent
control device. A written balance report shall be provided to the owner for HVAC&R
systems serving zones with a total conditioned area exceeding 5000 ft2 (465 m2). Standard 90.1-1999
also mandates that air systems shall be balanced first to minimize throttling losses. Then, in systems
with fans of power greater than 1 hp (0.75 kW), fan speed shall be adjusted to meet design flow conditions,
except variable-flow distribution systems need not be balanced upstream of the controlling
device (such as a calibrated VAV box).
Standard 90.1-1999 also mandates that hydronic (water) systems shall be proportionally balanced
first to minimize throttling losses, and then the pump impeller shall be trimmed or pump
speed shall be adjusted to meet design volume flow requirements. Each water system shall have
facility either to measure pressure increase across the pump or have test ports on each side of each
COMMISSIONING AND MAINTENANCE 32.3
pump. Exceptions include pumps with motors of 10 hp (7.5 kW) or less or when throttling losses
are no greater than 5 percent of the power draw or 3 hp (2.3 kW), whichever is greater, above that
required after the impeller is trimmed.
HVAC&R Commissioning Team
An HVAC&R commissioning team should be organized to perform commissioning. According to
Tseng (1998), this team should include the following members and perform the duties listed below:
Owner. The owner should be actively involved in commissioning, such as by defining requirements
and assigning commissioning responsibilities, defining the scope and tasks for the commissioning
authority, and developing enforceable contractual provisions to ensure compliance by the
contractor.
Commissioning Authority or Comission Consultant (CC). The commissioning authority or commission
consultant should have specific duties including reviewing the plans and specifications that
related to the commissioning processes, scheduling and conducting all pertinent commissioning activities,
such as precommissioning activities, training, walk-through inspections, tests and balances,
and documentation. The commission consultant plays a key role in an HVAC&R commissioning.
For a large HVAC&R project, an outside expert can be hired as an independent CC. For a small
HVAC&R project, the owner, the engineer, the general contractor, or the principal mechanical subcontractor
may serve as the CC.
Architect. The architect ensures that the commissioning authority reviews the shop drawing
submittal.
Engineer. The engineer must be actively included in the commissioning activities: design intent
documentation, verification testing, and training.
General Contractor. The general contractor must include the cost for commissioning requirements
in the bid price and ensure that such requirements in the mechanical and electrical subcontracts
are complied with.
Mechanical /Electrical Subcontractor. The mechanical/ electrical subcontractor must also
include commissioning requirements in the contract price, ensure participation of subcontractors,
coordinate all testing with pertinent specialty subcontractors, conduct walk-through inspections and
hands-on trainings, provide certification of system performance and functional performance of major
equipment to the CC, and turn over a complete set of as-built drawings to the design engineer.
Test, Adjusting, and Balancing Contractor. The TAB contractor conducts TAB work, demonstrating
to the CC the tested performance and participating in training sessions.
Energy Management and Control System (EMCS) Contractor. A properly installed and debugged
EMCS system is essential for HVAC&R commissioning to succeed.
When to Perform HVAC&R Commissioning
Ellis (1998) recommended three kinds of HVAC&R commissioning processes:
First, design-through-occupancy commissioning is ideal for the HVAC&R commissioning
process. The CC serves as the owner’s technical liaison throughout the design, construction, and
start-up of the project. The CC reviews the design engineer’s plans for compliance with the design
32.4 CHAPTER THIRTY-TWO
intent. The CC also makes sure that during verification testing, the temperature, pressure, fluid flow,
control system signals, and actions are commissionable and included in the design.
Second, postconstruction commissioning is most prevalent recently. Postconstruction commissioning
often requires at least 1 year after the completion of the construction so that both summer
cooling and winter heating mode operations can be included. In this case, the commission consultant
is most often an outsider who was not involved in the problems and politics of the design and
construction. Therefore, conciliatory and professional interpersonal skills are extremely important
for a CC. The CC also will learn about, evaluate, test, and document the installed HVAC&R system
and become involved in the operators’ training.
Third, existing building recommissioning helps to reduce HVAC&R energy consumption, improves
IAQ, and improves indoor environmental control.
Cost of HVAC&R Commissioning
Because the HVAC&R commissioning process is still being developed and standardized, a rule of
thumb for HVAC&R commissioning cost is between 2 and 5 percent of the construction cost of the
systems to be commissioned. The cost of HVAC&R commissioning can be set aside as an independent
budget item or added to the architectural / engineering or contractor fees. The commissioning
cost will be recovered by reductions in change orders and claims.
32.2 HVAC&R MAINTENANCE
Basics
HVAC&R maintenance is the work required to maintain or restore HVAC&R systems, including
equipment, instrumentation, components, and materials, to condition such that they can be effectively
operated to meet specified requirements.
ASHRAE Handbook 1999, HVAC Applications, defines repair as to make good or restore to
good or sound conditions, and defines service as what is necessary to effect a maintenance program
short of repair. In the ASHRAE handbook, maintenance also is classified into the following categories:
 Run-to-failure is a kind of arrangement such that no money is spent on maintenance prior to
equipment or system breakdown.
 In planned maintenance, all functions and resources in this category must be planned, budgeted,
and scheduled. Planned maintenance can be subdivided into preventive and corrective maintenance.
Preventive maintenance is a kind of scheduled maintenance for an HVAC&R system,
equipment, or components, in order to maintain durability, reliability, efficiency, and safety. In
corrective maintenance, corrective action is often the remedial action performed before failure occurs.
Corrective action taken during a shutdown in response to a failure is called a repair.
 Predictive maintenance is based on equipment and system monitoring, the operating conditions,
and performance to discover faults and degradations and thus the remedy performed.
Most manufacturers give detailed instructions in how to operate and maintain the specific equipment
and instruments. These instructions must be followed.
Maintenance Contractors and Maintenance Personnel
In most small facilities with comparatively simple HVAC&R systems, outside maintenance contractors
often provide maintenance service based on the specified maintenance program. The time inter-
COMMISSIONING AND MAINTENANCE 32.5
val between two maintenance programs depends on the operating hours of the HVAC&R system;
dirt accumulation in the filter, coil, and distribution devices; and recalibration requirements of the
instruments. For small and complicated HVAC&R systems, whenever the operator cannot repair
and service the HVAC&R system, the owner or the facility manager should ensure that qualified
contractors are hired for maintenance.
For more complicated medium-size multizone HVAC&R systems, operating personnel are often
responsible for HVAC&R system maintenance. The maintenance programs should be detailed in
the operation and maintenance manual and tailored to each specific building. For highly technical
equipment and systems, an outside maintanance contractor with the specific expertise required
should be called for service.
For large HVAC&R systems with central plants, a management organization with one operations
person and one maintenance person is required. Computerized maintenance programs should be used
to provide detailed timing of system maintenance procedures. Logged information and proper data
management associated with predictive maintenance will reduce failure response requirements. An
outside maintenance service contractor with specific expertise may need to be called.
Monitoring and Fault Detection and Diagnostics Assisting
Predictive Maintenance
As discussed in Sec. 5.15, in EMCS and fault detection and diagnostics, many operating parameters
of the HVAC&R system are monitored and measured. Addition of necessary monitoring parameters
for predictive maintenance may be considered.
Faults occur when the actually measured parameters deviate from the normal operating values.
Performance degradation is an evolving fault accumulated during a certain period of time.
By using rule-based expert systems, autoregressive with exogenous inputs (ARX) models, and
artificial neural network (ANN) models, performance degradation and faults can be determined,
and the corrective maintenance action recommended.
HVAC&R maintenance is already an important part of optimizing an energy-efficient HVAC&R
system operation, to improve IAQ, and to maintain a safe, healthy, and comfortable indoor environment.
More and more computer software will be developed to combine optimizing, energy-efficient
HVAC&R system operation and predictive maintenance in the future.
REFERENCES
ASHRAE, ASHRAE Handbook 1999, HVAC Applications, ASHRAE Inc., Atlanta, GA, 1999.
Austin, S. B., HVAC System Trend Analysis, ASHRAE Journal, no. 2, 1997, pp. 44–50.
Bearg, D. W., The Use of Multipoint Monitoring as a Tool for Commissioning Buildings for IAQ, ASHRAE
Transactions, 1999, Part I, pp. 1101–1108.
Chamberlain, C. S., Statistical Validation Testing vs. Commissioning, Heating / Piping / Air Conditioning, no. 8,
1995, pp. 59–62.
Ellis, R. T., Building Systems Commissioning, Engineered Systems, no. 1, 1998, pp. 108–116.
Seem, J. E., J. M. House, and R. H. Monroe, ASHRAE Journal, no. 7, 1999, pp. 21–26.
Taber, G., Preventive Maintenance, Engineered Systems, no. 6, 1994, pp. 68–73.
Tseng, P. C., Building Commissioning: Benefits and Costs, HPAC, no. 4, 1998, pp. 51–59.
Tseng, P. C., D. R. Stanton-Hoyle, and W. M. Withers, Commissioning through Digital Controls and an
Advanced Monitoring System–A Project Perspective, ASHRAE Transactions, 1994, Part I, pp. 1382–1392.
Wilkinson, R. J., Integrated Commissioning Avoids IAQ Pitfalls through Quality Construction, HPAC, no. 3,
1999a, pp. 79–83.
Wilkinson, R. J., The Commissioning Design Intent Narrative, ASHRAE Journal, no. 4, 1999b, pp. 31–35.
32.6 CHAPTER THIRTY-TWO