03 Water Cooled Chillers

COMMERCIAL
HVAC CHILLER
EQUIPMENT
Water-Cooled
Chillers
Technical Development Programs (TDP) are modules of technical training on HVAC theory,
system design, equipment selection and application topics. They are targeted at engineers and
designers who wish to develop their knowledge in this field to effectively design, specify, sell or
apply HVAC equipment in commercial applications.
Although TDP topics have been developed as stand-alone modules, there are logical group-
ings of topics. The modules within each group begin at an introductory level and progress to
advanced levels. The breadth of this offering allows for customization into a complete HVAC
curriculum – from a complete HVAC design course at an introductory-level or to an advanced-
level design course. Advanced-level modules assume prerequisite knowledge and do not review
basic concepts.
Water-cooled chillers range in size from small 20-ton capacity models that can fit in an eleva-
tor to several thousand-ton models that cool the world’s largest facilities such as airports,
shopping centers, skyscrapers, and other facilities. This TDP module will review all sizes of wa-
ter-cooled chillers, but will contain more information on the larger chillers in the range of 200-ton
and upward. Screw and centrifugal compressor water-cooled chillers tend to be the most popular
designs for larger commercial applications, while scroll and reciprocating compressor chillers are
used on the smaller ones. Air-cooled chillers are covered in a companion module, TDP-622.
© 2005 Carrier Corporation. All rights reserved.

The information in this manual is offered as a general guide for the use of industry and consulting engineers in designing systems.
Judgment is required for application of this information to specific installations and design applications. Carrier is not responsible for
any uses made of this information and assumes no responsibility for the performance or desirability of any resulting system design.
The information in this publication is subject to change without notice. No part of this publication may be reproduced or transmitted in
any form or by any means, electronic or mechanical, for any purpose, without the express written permission of Carrier Corporation.
Printed in Syracuse, NY
CARRIER CORPORATION
Carrier Parkway
Syracuse, NY 13221, U.S.A.
Table of Contents
Introduction...................................................................................................................................... 1
The First Centrifugal.................................................................................................................... 1
Water-Cooled versus Air-Cooled Chillers................................................................................... 2
Basic Refrigeration Cycle for Water-Cooled Chillers ..................................................................... 4
Subcooling Cycle..................................................................................................................... 5
Economizer Cycle.................................................................................................................... 6
Water-Cooled Chiller Components.................................................................................................. 7
Evaporator.................................................................................................................................... 7
Brazed Plate ............................................................................................................................. 7
DX Shell-and-Tube.................................................................................................................. 7
Flooded Shell-and-Tube........................................................................................................... 8
Evaporator Pros and Cons...................................................................................................... 10
Parallel and Series Chiller Evaporators ................................................................................. 10
Condenser .................................................................................................................................. 11
Compressors............................................................................................................................... 12
Reciprocating ......................................................................................................................... 12
Scroll...................................................................................................................................... 13
Screw ..................................................................................................................................... 13
Centrifugal ............................................................................................................................. 14
Refrigerant Metering – Expansion Device................................................................................. 16
Waterboxes ................................................................................................................................ 17
Purge .......................................................................................................................................... 18
Storage Tank and Transfer (Pumpout) Unit............................................................................... 19
Relief Valves.............................................................................................................................. 20
Chiller Controls.............................................................................................................................. 20
Compressor Starting Methods.................................................................................................... 21
Across-the-Line ..................................................................................................................... 21
Auto Transformer .................................................................................................................. 22
Primary Reactor ..................................................................................................................... 22
Part-Winding.......................................................................................................................... 22
Wye-Delta.............................................................................................................................. 23
Solid-State.............................................................................................................................. 23
Variable Frequency Drive...................................................................................................... 24
Energy Management .................................................................................................................. 25
Chilled Water Reset ............................................................................................................... 25
Demand Limit and Duty Cycling........................................................................................... 25
Screw Compressor Operational Details ......................................................................................... 26
Design and Off-Design Performance......................................................................................... 27
Centrifugal Compressor Operational Details................................................................................. 27
Head ........................................................................................................................................... 28
Lift ............................................................................................................................................. 28
Compressor Boundaries ............................................................................................................. 29
Compressor Stages..................................................................................................................... 31
Capacity Control Methods for Centrifugals and Screws ............................................................... 31
Inlet Guide Vanes ...................................................................................................................... 31
Screw Unloaders ........................................................................................................................ 32
Hot Gas Bypass.......................................................................................................................... 33
Speed Control ............................................................................................................................ 33
Refrigerant Related Topics ............................................................................................................35
Regulations ................................................................................................................................35
Chiller Construction...................................................................................................................36
Safety .........................................................................................................................................37
Heat Transfer .................................................................................................................................38
Heat Balance of Fluid ................................................................................................................39
Overall Heat Transfer.................................................................................................................40
Heat Transfer Coefficient...........................................................................................................41
Impact of Fouling Factor on (U) ................................................................................................42
Impact of Tube Velocity On (U)................................................................................................43
Impact of Tube Material On (U) ................................................................................................44
Evaporator and Condenser Tubing ............................................................................................44
Freezing of Fluids in Tubes .......................................................................................................45
Pass Arrangement ......................................................................................................................45
Variable Flow Operation............................................................................................................46
Codes and Standards ......................................................................................................................46
ARI Testing Standards ...............................................................................................................46
ASHRAE 90.1............................................................................................................................48
UL/CSA & ETL .........................................................................................................................48
ASHRAE Standard 15 ...............................................................................................................49
Selection Criteria ...........................................................................................................................50
Summary ........................................................................................................................................52
Work Session .................................................................................................................................53
Notes ..............................................................................................................................................56
Appendix........................................................................................................................................57
References..................................................................................................................................57
Work Session Answers ..............................................................................................................58
WATER-COOLED CHILLERS
Introduction
This TDP module on water-cooled chillers starts with a history of the first centrifugal chiller
and describes the first applications for early water-cooled chillers. After a discussion of the rela-
tive merits of water-cooled chillers, the refrigeration cycle for a water-cooled centrifugal chiller is
explained using pressure-enthalpy diagrams.
We will examine the major components used in water-cooled chillers such as evaporators,
condensers, compressors, and metering devices. The types of chiller starters and their applications
are also discussed.
A greater emphasis is placed on the larger screw and centrifugal types of water-cooled chill-
ers in this TDP module. A more detailed discussion of the screw and centrifugal compression
process and its characteristics is included. Current refrigerant issues, phase out dates, and appli-
cable codes and standards for water-cooled chillers are also examined.
Finally, computerized selection software for a centrifugal chiller is used to demonstrate the
required inputs and the selection process for a typical application.
The First Centrifugal

The art of building centrifugal air compressors was already 75 years old in 1916 when Dr.
Willis H. Carrier recognized their potential use in the then infant air-conditioning industry.
By 1922, the Carrier Company had
purchased a German-manufactured
centrifugal air compressor and modi-
fied it for use with dielene refrigerant
(C2 H2 Cl2). After two years of test and
development, this first centrifugal re-
frigeration machine was sold in 1924
to the Onondaga Pottery Company in
Syracuse, New York. The machine ran
for 26 years, providing air condition-
ing until 1950. The compressor of that
first machine was then retired to the
Smithsonian Institute in Washington,
D.C. where it remains today on exhibit
as one of the major technical develop-
ments in the United States.
Figure 1
Dielene Early Centrifugal Chiller
was used in the dry cleaning

industry as a cleaning agent.
Commercial HVAC Chiller Equipment

1
Carrier’s second centrifugal ma-

chine was installed in 1923 in
Cambridge, Massachusetts, at the
candy manufacturing plant of the W.
F. Schraft and Sons Company. The
second one built ended up being in-
stalled prior to the first one.
The first centrifugal machine

was tested for two years at
Carrier. It was sold to the
Onondaga Pottery Company in
Syracuse, New York.
Figure 2
A Carrier chiller is on display at the Smithsonian Institute.
Because of those early efforts in the

1920s, water-cooled chillers have gained
widespread acceptance in both large and
medium systems. Technology has re-
sulted in the evolution of water-cooled
chillers, which are characterized by their
excellent reliability, high efficiency, and
compact, cost-effective construction.
Figure 3
Evolution of Centrifugal Chillers
Water-Cooled versus Air-Cooled Chillers

Two methods are used to con-
dense the refrigerant in chillers. The
condensers can be air-cooled or wa-
ter-cooled. A typical air-cooled chiller
uses propeller fans to draw ambient
air over a finned coil to condense the
refrigerant. It may contain multiple or
single compressors. For a complete
discussion on air-cooled chillers, refer
to TDP-622, Air-Cooled Chillers.
Figure 4
Air-Cooled and Water-Cooled Chiller Benefits
Commercial Chiller Equipment

2
A typical water-cooled chiller uses recirculating condenser water from a cooling tower to
condense the refrigerant. For a complete discussion on cooling towers, refer to TDP-641, Con-
densers and Cooling Towers.
Cost and efficiency are the important factors when
Absorption Chillers considering air or water-cooled chillers. Chilled-water
This TDP will cover water- systems with air-cooled chillers typically have lower in-
cooled chillers using the vapor- stalled and maintenance costs than water-cooled because
compression refrigeration cycle. a condenser water system using a cooling tower is not
Absorption chillers that use required. A condenser water pump and chemical treat-
water as the refrigerant will not ment for the condenser water loop adds to the
be covered in this TDP.
maintenance required with a water-cooled system. How-
ever, water-cooled chillers have higher efficiency and
therefore lower operational costs. Air-cooled chillers are chosen when it is impractical to use a
cooling tower, such as when little water is available or water is highly corrosive.
The refrigerant condensing temperature in an air-cooled chiller is dependent on the ambient
dry-bulb temperature. In a water-cooled chiller, refrigerant condensing temperature is dependent
upon the entering condenser water temperature (and flow rate), which is a function of the ambient
wet-bulb temperature. Since the wet-bulb temperature is always lower than the dry-bulb tempera-
ture, the refrigerant condensing
temperature (and pressure) in a
water-cooled chiller is often
significantly lower than in an
air-cooled chiller. This is why
water-cooled chillers are more
efficient.
In terms of capacity, air-
cooled chillers are available in
packaged sizes ranging up to
approximately 500 tons, while
water-cooled chillers are typi-
cally available up to 3,000 tons, Figure 5
with limited custom designs
Typical Water-Cooled Chiller System
available up to 10,000 tons.
Water-cooled chillers typically last longer than air-cooled chillers. Air-cooled chillers may
last 20 years while water-cooled chillers may last 23 years or more. This may be attributable to
the fact that water-cooled chillers are installed indoors, and most air-cooled chiller configurations
spend their lives outdoors in the elements. Also, some of the larger water-cooled chillers are con-
structed with heavy duty, industrial-grade components.

3
Basic Refrigeration Cycle for Water-Cooled Chillers

In this TDP, we will explain the refrigeration cycle using components from a centrifugal
chiller since that type of chiller is water-cooled. The following temperatures are typical of the
standard refrigeration cycle for
comfort cooling applications. In
the evaporator of a water-cooled
chiller, liquid refrigerant at ap-
proximately 42º F takes on heat
from building return water
(whose entering temperature
may be represented at 54º F)
flowing through the evaporator,
and changes to a vapor. The re-
frigerant vapor is drawn into the
compressor and its temperature
and pressure are elevated. The
compressor provides the work
necessary to compress the gas to
a temperature and pressure re- Figure 6
quired by the condenser, Components of a Centrifugal Chiller Refrigeration Cycle
typically 97º F. The gas is then
discharged into the condenser where it condenses on tubes through which water flows, typically
at 85º F. This is the entering condenser water from the cooling tower. The condensed droplets of
liquid refrigerant then fall to the bottom of the condenser, flow through a pressure reducing de-
vice such as a float valve or an orifice, and return to the bottom of the cooler where the process
repeats itself.
The cycle can be shown on a
pressure-enthalpy (p-h) diagram.
Pressure is the force exerted per unit
area, while enthalpy is the total heat
content expressed in Btu per pound of
the substance. When the compressor
is close-coupled to the evaporator,
there is negligible pressure loss in the
suction line, and gas enters the com-
pressor at approximately the saturated
conditions that exist in the evaporator,
Point 1.
If we follow the steps shown, you
can see that from A to 1 is the refrig-
eration effect. In this step, building Figure 7
heat from the chilled water is ab- Pressure-Enthalpy (p-h) Diagram Showing Lift
sorbed by the refrigerant, and the
refrigerant transitions from a liquid to a saturated vapor, at Step 1. From Step 1 to 2 is the com-
pression stage. This stage raises the temperature and pressure of the saturated refrigerant vapor to
the saturated condensing temperature, so that heat can be rejected to the condensing fluid. This
compression is also called the “lift” of the compressor, which will be discussed later in the TDP.

4
In step 2 to 3, heat rejection takes place and the refrigerant transitions from superheated vapor
to saturated liquid. Step 3 to A is the refrigerant passing through the expansion device to reduce
pressure and temperature to the necessary conditions in the evaporator.
Subcooling Cycle
To increase the cycle efficiency,

and reduce power input in a water-
cooled chiller, a subcooling circuit,
also called a thermal economizer, may
be built into the condenser. The con-
densed liquid refrigerant is circuited
to a special section in the bottom of
the condenser called the subcooler.
The subcooler cools the refrigerant
below the saturated liquid tempera-
ture, Point 3. As a result of this, the
refrigeration effect will be increased,
as shown as A – 1, thus increasing
cycle efficiency.
Figure 8
RE = refrigeration effect Pressure Enthalpy Diagram Showing Subcooling
Ps = suction pressure
Pc = condensing pressure
ts = suction temperature
tc = condensing temperature
hfc = enthalpy of liquid
hgs = enthalpy of gas
vgs = volume of gas at compressor inlet
In the subcooler, the refrigerant is subjected to the coldest entering condenser water. The liq-
uid refrigerant is subcooled by 10º F to 15º F below the saturated condensing temperature. A
subcooler is an economical way to increase refrigeration cycle efficiency without adding work on
the compressor. It is ideal for use with
positive pressure refrigerants such as
HFC-134a and single-stage compres-
sors, which we will discuss later.
The typical method of subcooling
as described earlier simply cools the
liquid below its saturation temperature
in a tube bundle that cold condenser
water circulates through. This might
also be described as sensible subcool-
ing, because no refrigerant has flashed
or changed state (latent heat transfer).
Figure 9
Condenser Subcooler

5
Another method is called “flash subcooling.” In the flash subcooler a portion of the refriger-
ant is allowed to flash back into vapor further cooling the remaining refrigerant and effectively
lowering the saturated condensing temperature. The flash subcooler is not really a simple sub-
cooler but a lower saturation temperature condenser. Thermodynamically, the net effect of a
properly sized sensible subcooler can be the same as that of a flash subcooler in that the amount
of heat being rejected by the refrigerant to the condenser water is the same.
The “flash subcooler” has two cost advantages over a sensible subcooler:
1. It requires less refrigerant charge, since vapor is less dense than refrigerant and most of the
refrigerant side volume in a flash subcooler is vapor while it is liquid in a sensible sub-
cooler.
2. It has a very good refrigerant-side heat transfer coefficient (two-phase flow has a better
heat transfer coefficient than a pure liquid).
For example, it takes 71.22 Btu/lb to convert each pound of liquid R-134a to a gaseous form.
If the liquid refrigerant did not undergo a phase change, yet underwent 5° F temperature change,
the change in enthalpy is only 1.78 Btu/lb.
Economizer Cycle
Multiple-stage centrifugal and some screw chillers improve the cycle efficiency by using an
economizer, also known as a flash economizer. An economizer is a separate vessel after the con-
denser that improves the refrigeration cycle by allowing a small amount of gas to flash into vapor
after the condensing stage. This phase change decreases the temperature of the remaining liquid
refrigerant. The remaining vapor is then drawn back into the compressor at some secondary stage
of compression to be sent back through the cycle. Shown in Figure 10 is the increase in refrigera-
tion effect produced by the addition of an economizer to the centrifugal cycle.
This increase in refrigeration effect due to an economizer is shown by hfc – hfe.x
Pc = condensing pressure
Pe = pressure at economizer
P2 = pressure at stage 2
Ps = pressure at stage 1
hfe = enthalpy of liquid with
economizer
hfc = enthalpy of liquid without
economizer
h3 = enthalpy of mixture
hge = enthalpy of gas with
economizer
hgs = enthalpy of gas suction
inlet
Figure 10
Refrigeration Cycle with Economizer

6
Water-Cooled Chiller Components

Evaporator
The evaporator, or cooler, as it is often called, is the vessel in which the water is cooled.
There are several designs of evaporator, and their use is typically based on the size of the chillers.
Brazed-Plate
For smaller capacity chillers (15-60 tons), the brazed-plate evaporator is often used. These
evaporators consist of a series of plates brazed together with every second plate turned 180 de-
grees. This design creates two highly turbulent fluid channels that flow in opposite directions that
result in a high heat transfer coeffi-
cient over a small surface area. The
plates are stacked so they form a
multi-layered design for two inde-
pendent paths of fluid to travel. Each
layer or path is linked to an inlet and
outlet via a manifold at either end.
The design of a brazed-plate evapora-
tor maximizes heat transfer at an
attractive first cost for smaller chiller
designs. The waterside of a brazed-
plate evaporator must be kept free of
sediment and debris. Strainers are re-
quired. The brazed-plate evaporator
requires chemical cleaning since the Figure 11
waterside is not mechanically clean- Brazed-Plate Evaporator
able. Photo Courtesy of API Heat Transfer
DX Shell-and-Tube
The direct expansion (DX) evap-

orator is a shell-and-tube heat ex-
changer design in which the
refrigerant flows through the tubes
and the water flows around the tubes
inside the shell section. As heat is
transferred to the colder refrigerant
from the warmer water, the refrigerant
in the tubes boils while the water is
cooled. Baffles within the shell direct
the water flow path over the tubes to
create turbulence, resulting in im-
proved heat transfer. Figure 12
Direct Expansion (DX) Evaporator
Photo Courtesy of Standard Refrigeration

7
However, the shell of a direct expansion Cleaning DX Evaporators

evaporator must be chemically cleaned. This
type of heat exchanger is typically used in the When it is necessary to clean the water
40-60 ton chiller range. Internal baffling circuit on a direct expansion-type
changes are made on DX evaporators to ac- evaporator, a chemical cleaning process is
commodate flow variations and maintain required. This is because water is on the
shell side and mechanical cleaning is not
acceptable pressure drops. possible due to the baffling inside the shell.
Flooded Shell-and-Tube
In larger capacity water-cooled

chillers, the cooler is normally a shell-
and-tube heat exchanger with the re-
frigerant on the shell side and the heat
transfer fluid passing through the
tubes. This fluid can be fresh water,
an antifreeze mixture, or any process
liquid that requires cooling. The ad-
vantage to this type of heat exchanger
is that the tubes can be mechanically
cleaned. The fluid in a chilled water
loop is chemically treated and closed
to the environment to protect the
equipment and piping system from Figure 13
corrosion and fouling. Flooded Shell-and-Tube Evaporator
Waterbox nozzle connections are
used to connect the water distribution
piping to the chiller. A typical evapo-
rator has a suction pipe (also known
as a vapor outlet) and an inlet pipe for
the liquid refrigerant. The inlet pipe
enters at the bottom of the cooler and
empties into a distribution system,
which spreads the flow of refrigerant
evenly over the entire length of the
cooler tube bundle. This provides op-
timal use of the cooler tube surface by Figure 14
producing an even refrigerant level
throughout the cooler. Evaporator Refrigerant Path
Because refrigerant goes through a change of state from liquid to gas within the shell (refrig-
erant) side, some means of separating the vapor from the boiling liquid must be provided. This is
because liquid refrigerant droplets tend to become entrained in the flow of suction gas that is go-
ing to the compressor inlet. These liquid droplets that are “carried over” can be corrosive or
destructive to the centrifugal compressor. Space elimination is the preferred design method. The

8
cooler is designed with enough space between the top row of tubes and the compressor inlet to
ensure that the velocity of the refrigerant vapor is low enough to prevent entrainment of liquid
refrigerant droplets.
Other designs use eliminators to trap the liquid droplets before they reach the inlet of the
compressor. Eliminators are either a series of parallel plates bent in a Z-shape, or a wire mesh.
The disadvantages to eliminators are
that they can eventually fail, they in-
troduce a pressure drop, and they have
an upper limit on face velocity.
Eliminators work by converting mist
into large drops that fall back against
the oncoming vapor flow. If the ve-
locity of the refrigerant vapor is too
high, the liquid droplets that were
eliminated are re-entrained and car-
ried through the eliminator. The
allowable face velocity is a function
of the ratio of liquid to vapor densi-
ties.
Figure 15
Mechanical elimination Space Elimination of Refrigerant Droplets
requires approximately the Tube support sheets at intermediary points are criti-
same area as space elimination. cal to ensure that the tubes do not deflect due to the
boiling action that takes place around the tube bundle.
They also aid in the replacement of tubes, should that be
necessary. The end tube sheets form the ends of the heat exchanger, and tubes are sealed at the
end on these tube sheets to create the refrigerant to water boundary. Welding, chemically sealing,
or rolling the tubes into the grooved tube sheets can accomplish this sealing. Among these,
grooved tube sheets provide the best
seal, and with a double groove, the
seal is even better. To increase the
stability of the tubes, they can also be
swaged (expanded) at the intermediate
tube support sheets, a process that
expands the tube just enough to fill
the machined holes. This process en-
ables the tubes to be replaced by use
of a pulling tool that extracts the tube
from the tube support sheet.
Figure 16
Double Grooved Tube Sheet for Tight Seal

9
Evaporator Pros and Cons
Brazed-Plate Evaporator** Pro Smaller footprint

Pro Lightweight
Pro Cost effective for smaller sizes
Con Sensitive to contamination of the water circuit

Con Not mechanically cleanable
DX Shell-and-Tube Evaporator* Pro Cost effective for small and medium sizes
Pro Familiarity with customers/engineers
Design can accommodate heater cable on the shell to
Pro
prevent freezing of water
Con Larger footprint

Con Not mechanically cleanable
Flooded Shell-and-Tube Evaporator Pro High efficiency

Pro Mechanically cleanable
Con Large footprint

Design cannot accommodate heater cable on the shell to
Con
prevent freezing of water in tubes
*Photo courtesy of Standard Refrigeration

**Photo courtesy of API Heat Transfer
Parallel and Series Chiller Evaporators
Where chiller capacities greater

than can be supplied by a single
chiller are required, or where stand-by
capability is desired, chillers may be
installed in parallel. Units may be of
the same or different sizes. Usually,
equally sized chillers are utilized to
accomplish commonality of parts and
maintain simplicity. If unequal sized
chillers are used, cooler flow rates
must be balanced to ensure proper
flow to each chiller. Software is avail-
able from the chiller manufacturer
that automatically stages multiple
chillers of equal or unequal size. Figure 17
Parallel and Series Chillers

10
Where a large temperature drop (greater than about

18º F) is desired, chillers may have to be installed in Future Expansion Arrangements
series. The 18º F value is a rule of thumb. Actual chiller Parallel chiller arrangement lends
selections should be made to determine when a series itself to future expansion because
chiller configuration is necessary. The evaporator mini- another chiller can simply be
mum entering fluid temperature limitations should be added to the “ladder.” Two
considered for the downstream chiller. Use of a reduced chillers in a series cannot
accommodate future expansion
pass configuration may be required to keep waterside
since pressure drop in the
pressure drop at an acceptable level. This is covered in evaporator circuit would become
detail later in the TDP. Additional piping configurations too great with three chillers.
can be found in TDP-705, Chilled Water Systems.
Condenser
Water-cooled chillers either use brazed-plate or shell-and-tube heat exchangers for the con-
denser. Since the focus of this TDP is on larger chillers, the shell-and-tube design will be
discussed further. As in the evaporator, the refrigerant is on the shell side, and the fluid is on the
tube side. The fluid is generally fresh
water, used with an open cooling
tower.
Other condenser water sources
include lakes, rivers, or cooling
ponds. Use of these water sources
constitutes a once-thru system. Once-
thru condenser applications are very
limited compared to recirculating
cooling tower applications. That is
because they are often considered as a
source of thermal pollution. In coastal
areas, seawater may also be used;
however, appropriate materials, like
titanium, should be used to minimize Figure 18
tube and tube sheet corrosion. Large Chiller Shell-and-Tube Condenser
See the tubing discussion later in

this TDP.
Besides the waterbox connec-
tions, a typical condenser has a
refrigerant hot gas inlet pipe and a
liquid outlet pipe. The inlet pipe en-
ters the top of the condenser where
hot discharge gas from the compres-
sor impacts a refrigerant distribution
baffle.
Figure 19
Large Chiller Condenser Cutaway

11
The baffle redirects the discharge

gas along the length of the condenser,
lowering its velocity and preventing
direct impact of the high velocity dis-
charge flow on the tubes. This
prevents tube failure due to vibration.
The same tube sealing and support
methods as in the evaporator dis-
cussed previously are used in the
condenser.
Figure 20
Refrigerant Path in a Shell-and-Tube Condenser
Compressors
There are several widely used compressors for water-cooled chillers. They are grouped into
two categories: positive displacement or dynamic compression (also called non-positive dis-
placement). Centrifugal compressors are the only type of non-positive displacement compressor.
Reciprocating
Reciprocating compres-
sors, like a reciprocating
engine, have pistons, rods,
discharge, and intake valves.
The valves operate on suc-
tion and discharge pressure.
Compression is achieved by
trapping a fixed amount of
refrigerant gas into a cham-
ber. For this reason,
reciprocating compressors
are positive displacement
type compressors as are
scrolls and screws.
Figure 21
Reciprocating Compressor

12
Scroll
Scroll compressors are a popular alternative to reciprocating compressors. Multiple scroll

compressors are often used in a single chiller design to meet larger capacities. Scroll compressors
are 10-15 percent more efficient than reciprocating and are proven very reliable because they
have approximately 60 percent less moving parts.
Scroll compressors have a unique
compression process. A fixed scroll
coupled with the movement of an or-
biting scroll ingests the suction gas
into several pockets. As the orbiting
scroll moves, the pockets of gas are
compressed to an intermediate pres-
sure. In a final orbit, the pockets reach
discharge pressure and exit through
the discharge port.
Reciprocating and scroll compres-
sors are currently used in the 10 to
400-ton chiller range, and are used in
both water-cooled and air-cooled Figure 22
chillers.
Scroll Compressor
Screw
Screw compressors are another version of positive displacement compression. As the screw
rotors turn, the gas is compressed. Screw compressors have one or more rotors to accomplish the
compression. Due to this characteristic, positive displacement compressors are best suited to han-
dle smaller volumes of refrigerant gas
over high compression ratios. The
HVAC industry is trending to phase
out reciprocating compressors, favor-
ing screw and scroll compressors.
Screw compressors are used in the 70
to 500-ton chiller range in multiples,
and are also used in both water-
cooled and air-cooled chillers. For an
in-depth discussion of screw com-
pressor operation, see the Screw
Compressor Operational Details sec-
tion (page 26). Figure 23
Screw Compressor

13
Centrifugal
Centrifugal compressors are dynamic compression, and they are water-cooled because their
efficiency is optimized at the lower condensing temperatures and pressures. As refrigerant mole-
cules are flung outward by centrifugal
force, new ones are drawn into the
compressor to replace them. The
overall effect is one of continuously
compressing the stream of refrigerant
gas. This process of compression al-
lows large volumes of refrigerant gas
to be compressed resulting in a rela-
tively compact size chiller. Thus,
centrifugal compressors are dominant
in larger capacities.
Note:
The break point from positive
displacement to centrifugal type
compression in air-conditioning Figure 24
chillers generally occurs Centrifugal Compressor
between 300 to 500 tons due to
design and capacity related
reasons.
As an alternative for larger capacity requirements, multiple chillers with positive displace-
ment compressors can be used to match the capacity of a single centrifugal chiller.
There are several possible drivers for refrigeration compressors. These include motors, steam
turbines, gas engines, and gas turbines. As electric motors are the most commonly used method
for driving a compressor, we will focus on these.
Centrifugal compressors fall into two broad categories: hermetic and open. The compressor
driven by these motors may be direct-drive, where the motor rotor and the impeller are on the
same shaft, or gear driven, where the rotor shaft and impeller shaft are coupled by a transmission.
In a direct-drive compressor, the impeller spins at the same speed as the motor (slightly below
3600 rpm for 60 Hz). In a gear-driven compressor, the gearing in the transmission is changed by
the manufacturer to meet different power frequency requirements or compression requirements.

14
Figure 25 shows a cut away view of a hermetic transmission-drive centrifugal compressor.

The motor is attached to the transmission and is hermetically sealed from the machine’s ambient
environment. Motor cooling is accomplished by spraying liquid refrigerant on the motor wind-
ings. Hermetic motors reject their heat directly into the refrigerant that is cooling them, which is
then rejected by the condenser. The highly efficient motor cooling results in the use of smaller
motors than could be utilized with an open-drive motor of the same type. Thus, hermetic motors
are extremely reliable, run cooler, and are lighter than comparable open drive, air-cooled motors.
They also are protected from environmental contamination (dirt). Hermetic motors are squirrel
cage induction type, which are well suited for sealed environments.
Figure 25
Hermetic Centrifugal Compressor – Cutaway
Open-drive compressors have the compressor shaft extending through the casing to facilitate
connection to a number of different types of drives as discussed earlier. However, these machines
require shaft seals, which must
be replaced on a regular basis. By
design, shaft seals leak oil and
refrigerant to keep contaminants
out of the chiller. This oil and
refrigerant must be reclaimed,
typically in a separate container
outside the chiller. Additionally,
when routine preventive mainte-
nance is done, the compressor
and motor must be realigned. An
open drive motor also rejects its
heat directly into the
Figure 26
Open-Drive Centrifugal Compressor

15
mechanical room, requiring additional

cooling or ventilation to reject the heat Shaft seals
from the building. This also subjects the in open-drive compressors pass oil to lubricate
motor to the hot, and typically dirty, themselves. The seal chamber area is flooded
atmosphere in a mechanical room. with oil and the oil holds in the refrigerant.
However, a small amount gets by the seal.
Refrigerant Metering – Expansion Device

To maintain the necessary pressure difference between the evaporator and condenser, some
form of metering device must be used. There are several types used on the larger water-cooled
chillers such as screws and centrifugals. They are mainly fixed or variable orifices and expansion
or float valves. Fixed orifices are simply perforated plates that cause a pressure drop in the fluid.
These orifices are designed for the maximum flow of refrigerant, which hinders their performance
at lower flows/loads. Fixed orifices are used in some chiller designs. Variable geometry orifices
can respond better to the changes in loads by changing area, and some require external inputs to
make this change. Expansion valves, for instance, use saturated suction temperature to regulate
the amount of refrigerant that passes
through them, and are typically con-
trolled electronically. Float valves are
used on larger screw and centrifugal
chillers and can automatically vary
the flow of refrigerant in a system by
responding to changes in liquid re-
frigerant levels. These level changes
are the result of changes in load con-
ditions and pressure differences in the
heat exchangers. Float valves are ideal
on larger water-cooled chillers be-
cause they are simple, always
optimize the refrigerant flow, and do
not require any external inputs to vary Figure 27
the flow rates of the refrigerant. Refrigerant Metering – Large Chiller
Another benefit of the variable geometry orifice (float valve) on larger screw and centrifugal
chillers is that it can pass increased mass flow of refrigerant during periods of lower entering
condenser water temperature by increasing the area of the opening into the evaporator. More re-
frigerant flow will result in an approximate 15 percent increase in chiller capacity above the
design full load tons. This increase in capacity is called “maximum capacity.” Fixed orifice de-
signs cannot accomplish this feature.
Why might this be important? On multiple chiller projects, each chiller that can accomplish
maximum capacity will handle a larger portion of the total building load than a fixed orifice de-
sign chiller. This means the second (or third) chiller in a multi-chiller plant can remain inactive as
long as possible saving energy over designs that must run more than one chiller.
For a discussion on refrigerant metering devices used on the smaller chillers with reciprocat-
ing and scroll compressors (and some smaller screw compressors), refer to TDP-622, Air-Cooled
Chillers, and TDP-403, Expansion Devices & Refrigeration Specialties.

16
Waterboxes
Waterboxes are used to connect
the chiller waterside to the chilled or
condenser water loop piping. There
are two types of waterbox construc-
tion in common use on larger screw
and centrifugal chillers: “nozzle-in-
head” and “marine” type.
The most popular and economic
design is the “nozzle-in-head” type.
This design is a result of the industry
trend in air conditioning to make ma-
chines more compact and less costly.
The only drawback to nozzle in head
type waterboxes is that the piping Figure 28
connection must be undone, and the
Nozzle-in-Head Waterboxes
“heads” must be completely removed
to perform any tube maintenance.
The “marine” type waterbox is
larger and more costly. However, it
provides the ability to clean the tubes
without disturbing the piping external
to the machine. For this reason, the
marine waterboxes are generally used
in chillers above 700 tons. Moving the
piping is time consuming and expen-
sive, so making use of a marine type
waterbox offers owners an attractive
feature and reduced maintenance
costs. When available, marine water-
boxes with hinged waterbox covers Figure 29
eliminate rigging of heavy waterbox
Marine Waterboxes
covers.
Waterboxes are typically designed for 150 or 300 psig, the latter used for taller buildings,
where there may be a large water column in the water distribution piping system.
In a high-rise building, the static head imposes pressure on the chiller waterside components
if the chiller is located in the basement. Example: A 50-story building with 12 ft between floors
would be 600 ft high. Since 2.31 ft equals 1 psi, the pressure on the components at the lowest
level would be 600 ft/2.31 ft/psi = 260 psi. The hydronic system components in the basement
MER (mechanical equipment room) must be designed for this pressure. In this example, the
chiller “waterboxes” (area where piping connections are made to the chiller) would have to be
constructed to accommodate 260 psi. That means optional 300 psi waterboxes would have to be
specified.

17
Purge
Some refrigerants used in centrifugal water-cooled chillers, such as HFC-134a, are positive
pressure (operate above atmospheric pressure), and some are negative pressure, such as HCFC-
123, operating below atmospheric
pressure. Negative pressure centrifu-
gal chillers operate in a vacuum and
have gaskets and o-rings that can leak
which allows air to enter the chiller.
When air leaks in, some water vapor
comes with it. This air/water vapor
can displace some of the refrigerant in
the chiller causing a reduction in effi-
ciency.
A device called a purge is re-
quired to collect and expel the air and
the water vapor from the inside of the
chiller. When the air is expelled, some
Figure 30
refrigerant goes with it. A modern
high efficiency purge reduces the on- Purge on Negative Pressure Centrifugal Chiller
going loss of refrigerant to 0.5 lb for
every pound of air removed. Older purges lost from 3 to 5 lbs of refrigerant for every pound of air
removed. Air and non-condensables in the chiller causes the head pressure to rise. As a rule of
thumb, for every 2.5 psig increase in condenser (head) pressure, the chiller’s efficiency falls be-
tween 8-10 percent
The purge is continually sampling gas from the chiller condenser
and condensing the refrigerant and recirculating it back to the evapo- Note:
rator. Air and other non-condensables cannot be condensed. The The purge unit adds
purge will collect these non-condensables and eventually discharge maintenance and
them to the atmosphere. Purging to the atmosphere is in compliance electrical consumption
with EPA regulations. Water vapor from humid mechanical rooms is that is not required
also drawn into the chiller when there are leak areas in the chiller and with a positive
pressure chiller
must be condensed and collected in the purge then drained from the
design.
purge collection chamber.
Water mixed with refrigerant can also produce corrosive acids that break down the oil and at-
tack the internal metal parts of the machine.

18
Storage Tank and Transfer (Pumpout) Unit

In performing certain service procedures on larger screw and centrifugal chillers, it may be
necessary to remove or relocate the operating refrigerant charge from the evaporator or con-
denser. Extreme care must be exercised to prevent releasing the refrigerant to the atmosphere
during this procedure.
Note:
A typical example of when the
operating charge must be
removed from the condenser is
when some of the condenser
tubes are being replaced.
In large positive pressure chillers,

such as those utilizing HFC-134a, the
construction of the evaporator and the
condenser allows either vessel to
serve as a refrigerant storage tank.
Isolation valves are installed on the Figure 31
chiller to isolate the refrigerant charge Positive Pressure Chiller/Transfer Compressor
in either the evaporator or the con-
denser.
This means that the operating charge does not have to be completely removed from the chiller
to perform service work. The entire chiller operating charge can be stored in the evaporator when
working in the condenser, or in the
condenser when working in the
evaporator. Transfer of charge from
one vessel to another is accomplished
with a portable or unit-mounted trans-
fer compressor and isolation valve
accessory.
In contrast, negative pressure
chillers require a separate refrigerant
storage tank and transfer compressor
as well as complete removal of the
refrigerant from the chiller. The rea-
son is a negative pressure chiller, by
virtue of its shell construction, is not Figure 32
rated for in-chiller storage of refriger- Storage Tank and Transfer Unit
ant.
If the refrigerant must be completely removed from the chiller, the chiller has to be leak-
tested and dehydrated before the refrigerant can be pumped back into the chiller. This increases
both the downtime required for maintenance and the potential for refrigerant loss.

19
Relief Valves
Refrigerant relief valves on either the condenser, the evaporator, or both, are provided as
safety devices in the event of overpressurization. Overpressurization might occur during a fire or
if hot water were run through the chiller tubes.
The standard pressure relief de-
vice on a negative pressure centrifugal
chiller is a carbon disk with a mem-
brane that shatters at 15 psig. Over-
pressurization results in total loss of
the refrigerant. Back-up relief valves
are available to minimize this loss in
the event of overpressurization. These
back-up devices contain non-
fragmenting disks with a reseating
plunger that will relieve the pressure
and then reseat. This saves a good
portion of the refrigerant that might Figure 33
otherwise be lost. Relief Valves
Positive pressure chiller designs typically use the reseatable valve design as standard. After a
release, the valve reseats, thus preventing a total loss of refrigerant.
Chiller Controls
All chillers need to be properly controlled for safe and efficient operation. Chiller designs
commonly use microprocessors, electronic sensors, and digital displays that have changed the
look of chiller equipment control sys-
tems. This technology is referred to a
Direct Digital Control (DDC).
DDC systems allow for the use of
digital display systems that give the
owner-operator and service technician
the ability to do a diagnostic evalua-
tion of the operation of the equipment
by way of equipment-mounted or
handheld user interfaces. In DDC sys-
tems, information is digital and can be
shared by other equipment to function
as a system. Most DDC systems will Figure 34
have a communication terminal that Water-Cooled Chiller Control Panel
allows for network communications
or the attachment of a PC for more detailed diagnostic capabilities. DDC systems are capable of
sharing information on a communication bus that can be viewed worldwide.

20
Compressor Starting Methods

There are seven types of starters commonly used with water-cooled chillers. Besides cost, the
selection of a starter depends on many variables, most of which deal with the electrical system
characteristics, voltage, and power company regulations at the installation.
When large horsepower motors on large capacity chillers
(approximately 3000 tons and larger) are connected to a Starters
power distribution system, the motor inrush current can cause are also grouped by voltage
voltage drops that may adversely affect the operation of other ranges. Low voltage is any
equipment (particularly computers) connected to the system. system that utilizes 600 volts
Careful consideration should be given in the selection of the and below, medium voltage
starter type for voltage sensitive applications. Smaller “pack- is considered to be higher
than 600 volts.
aged” chillers typically include an integrated starter as a
standard part of the chiller package.
For larger chillers, starters may be unit-mounted or remote-mounted from the chiller. Unit
mounting saves space and reduces installation costs, and can increase the reliability of the chiller
system. Unit-mounted starters are very popular on centrifugal chillers because the entire chiller’s
electrical requirements can be supplied with power through the starter, thus accomplishing “sin-
gle point” connection. If you do not use a unit-mounted starter, several electrical connections are
required. A separate electrical feed for the compressor, the oil pump, and the unit controls would
be necessary. These separate wiring connections must be field-installed between the remote
starter and the chiller.
Across-the-Line
Across-the-line, also called full voltage, is the simplest and least expensive method. The mo-
tor locked rotor current is drawn directly from the line on starting, and is used when the power
system can withstand high inrush currents such as in a large industrial plant. With an across-the-
line starter, the inrush can be 6 to 8 times the motor rated full-load amp (FLA) value. Therefore,
this method is especially suited to the following motors:
1) low-voltage, single phase motors up to approximately 10 hp
2) low voltage, three phase motors up to approximately 25 hp
3) medium-voltage (over 600 volts) motors of any hp
Because full voltage is applied directly to the
Across-the-line starters motor terminals, this method also provides the
on large, single compressor chillers are quickest acceleration to running speed. This type
mostly used in the voltages (like 4160v) of starter is a common one used on smaller ton-
where the amp draw is low. On smaller nage chillers with multiple compressors. This is
multi compressor chillers, across-the- because the starting amps of the individual com-
line starters can be used for low voltage pressors are relatively low, and the compressors
applications because the inrush per
are staggered on startup.
compressor is manageable.

21
Auto Transformer
This method reduces inrush line current and motor starting Auto transformer starters
torque by limiting the voltage applied to the motor windings dur-
ing start-up. It is used typically for medium voltage applications. offer a soft start (relative
The torque is accomplished through a transformer within the to across-the-line) but the
starter with contacts configured to allow stepped acceleration to start is limited to discreet
full current. Changing incoming line voltage taps within the steps.
transformer varies the current drawn from the line, the voltage and current applied to the motor,
and the torque developed by the motor at startup. This tap is usually selected and then fixed at
installation, and is based on the application. For centrifugal chiller applications, the 65 percent tap
is the one that provides the required torque to start the motor compressor.
Primary Reactor
Like an auto transformer, this starter reduces inrush current by limiting the voltage applied to
the motor. A reactor, or resistor, is used to reduce the voltage to the compressor motor instead of
a transformer. Compared to an auto transformer, a primary reactor is the more economical
method for medium voltage. Primary reactor is not the most efficient starter type from an electri-
cal standpoint. A primary reactor starter uses the resistor in the circuit during starting. As the
resistor lowers that amp draw, there is some wasted energy. However, primary reactor starters
provide a smooth acceleration. This is because the starter adds resistance to reduce the voltages
and current at startup. The starter’s incoming line voltage taps may be changed at installation to
set the level of inrush voltage and current.
Part-Winding
Part-winding starters also produce a soft start. An electrical engineer typically determines if
soft starting is required based on an electrical system study that calculates voltage drop in the sys-
tem when a load is started. If the power supply is weak or the system is heavily loaded, the
system is more likely to be sensitive to motor starting, and soft starters are more likely to be re-
quired. It is more likely that 208V systems will require soft starters than 460V systems because of
the higher amp draw.
Part-winding starters are a cost effective means of providing a reduced current start where
smaller horsepower motors are used on water-cooled chiller designs using multiple compressors.
Part-winding starters can only be used with part-winding motors that have two sets of identi-
cal windings that are intended to be operated in parallel. During a part-winding start, only one
winding is energized, reducing the inrush current to 60-70 per-
cent of normal starting values with both windings energized. A part-winding starter
With only one winding in the circuit, a typical part-winding mo- is commonly used on re-
tor will not accelerate to more than half of the motor rated ciprocating compressors.
speed. Because the motor is operated with one winding only Availability varies from
during the initial acceleration period, after the transition from manufacturer to manu-
start to run mode, the current draw of the motor may be close to facturer based on the size
the rated locked rotor amps. A part-winding starter essentially of the compressor motor
provides a two-step start. Part-winding starters are not used on used.
centrifugal chiller compressors.

22
Wye-Delta
Wye-delta, also called Star-delta, starters reduce inrush line current and motor starting torque
by switching the motor winding connections. Three contactors and a timer change the winding
configuration to transition from around 33 percent current in the wye configuration to full current
in the delta configuration. Open transition wye-delta starters stop applying voltage to the motor in
the short transition between windings. A closed transition wye-delta
starter uses shunt resistors. Only closed transition wye-delta starters
Wye-delta starters
are now used in large chillers. Due to its relatively low cost when
compared to the reduced voltage type starters, wye-delta starting is are popular with
the most commonly used starter with hermetic centrifugals under technicians because of
600 volts. their familiarity with
wye-delta trouble
Wye-delta starters are used only on low-voltage applications. shooting procedures.
Solid-State
A solid-state starter’s principle of

operation utilizes a solid-state device
known as a silicon controlled rectifier
(SCR). The SCR is an electronic
“valve,” which allows a volume of
current to flow through it in response
to a controlling electronic signal. For
a solid-state starter, inrush current
decreases in a somewhat linear fash-
ion until the operating current of the
motor is reached. In this manner, the
starter supplies only enough voltage
to overcome the motor starting torque
requirement. The applied voltage can
be controlled to limit inrush at any Figure 35
desired level above that needed to
rotate the compressor motor. Unit-Mounted Solid-State Starter
Solid-state starters are used for both low-voltage and medium-voltage applications although
they are not as common in medium voltage. They are popular for low voltage applications (≤600
volts) because they provide a programmable soft start.

23
Variable Frequency Drive
The newest soft-start technology

in starting equipment for water-cooled
chillers is the variable frequency drive
(VFD), also known as variable speed
drive (VSD). These drives provide the
softest start, because the drive con-
trols start at zero current and ramp up
just until the motor inertia can be
overcome. This results in inrush cur-
rent equaling rated load amps (RLA)
during start up. RLA is a very low
inrush value.
VFDs have become very popular
as a starting method for water-cooled Figure 36
centrifugal chillers. Several reasons VFD Starter for Unit-Mounting
are:
• Low inrush at start-up, low mechanical
stress The advantage of VFDs
• High power factor (the ratio of active power is that they can slow down the
to apparent power) compressor by varying the
frequency applied to the motor. The
A motor that has a high power factor makes speed reduction results in
more efficient use of the incoming energy. reductions in lift. This translates
• Quiet chiller operation at reduced rpm into lower operating costs because
• Energy savings at part load a compressor becomes more
efficient when the speed is reduced
• Best for emergency generator applications to no more than is required to
due to low inrush satisfy the load.
• Utility rebates often available
Paybacks of two
Motor Starting Current as a % of
years or less are not un- Starting Method
common in VFD Locked Rotor Current Full Load Current
applications. This actual Across-the-Line 100 600
power saving is a result Auto Transformer & Primary Reactor

80% 80 480
of the relationship of
65% 65 390
power versus speed, 50% 50 300
which will be discussed
Part Winding 65 390
later in the TDP.
Wye-Delta 33 200
Solid-State 0 - 100 0 - 600
VFD 16.6 100
Figure 37
Motor Starting Current as a Percent of Locked Rotor and Full Load Current

24
Energy Management
Use of energy management practices can significantly reduce operating costs, especially dur-
ing off-peak modes of operation. Demand limiting and chilled water temperature reset are two
techniques for accomplishing efficient energy management.
Chilled Water Reset
Chilled water reset means to change the chilled-water temperature leaving the chiller based
on some parameter. Increasing the leaving chilled-water temperature reduces compressor power
usage by reducing lift at both full and part-load conditions. However, at part-load conditions, de-
sign chilled-water temperature may not be necessary, especially
for comfort cooling applications, so reset is possible. However, Lift and Efficiency
increasing the chilled water temperature may result in greater
humidity levels in the conditioned spaces. Higher coil tempera- Lift refers to the difference
tures resulting from the increased chilled water temperature will between compressor
suction and condensing
reduce the latent heat capacity of the coils and the ability of the temperatures. Reducing lift
air distribution system to remove space humidity. increases efficiency.
Chilled water temperature can be reset as a function of:
• Ambient air temperature (most common) – used when the ambient temperature is the best
indication of load
• Return chilled water temperature – used when return water temperature is the best indica-
tion of load. This method is typically used when it is desired to maintain a fixed ∆t in the
chiller plant.
• Temperature within the building – used when space temperature is the best indication of
load on the chiller. This is used when there are critical areas in a building such as labs
that might require a specific temperature be maintained.
Reset doesn’t always mean an increase in water temperature. For example, space temperature
reset could actually lower the chilled water supply temperature, not raise it, in order to maintain
conditions.
Demand Limit and Duty Cycling
Demand limit is a feature that limits the unit capacity during

periods of peak energy usage. When a utility company’s demand Demand is measured
for electricity exceeds a certain level, users are charged extra in kW; energy charges are
money called demand charges. To avoid these charges, loads in measured in kWh.
the building are limited to keep demand below a prescribed Together, the two form a
maximum level. Demand limiting may be desirable on hot days typical user’s electrical bill.
when air conditioning is most needed. Demand may be limited
on the chiller by increasing the chilled water temperature or by unloading the chiller regardless of
actual load to a predetermined percentage of the current draw which is an indication of load. This
may result in temporary increased building or process temperatures if the actual load is greater
than the demand limited capacity that the chiller can provide.

25
Duty cycling will cycle selected electrical equipment in an installation (building, factory, etc.)
at regular intervals to limit the electrical demand, thereby lowering demand charges. However,
duty cycling is not recommended because constant cycling will cause increased stress and dam-
age to the motor windings, bearing, and controls. If demand must be controlled, the demand limit
sequence available from the manufacturer is recommended.
Screw Compressor Operational Details

The screw compressor is classified as a positive displacement compressor, which means that
a volume of gas is trapped within an enclosed space whose volume is then reduced. Screw com-
pressors are composed of multiple
parallel rotors with external helical
profiles fit into a casing. The drive
rotor can either be coupled directly to
the motor (direct drive) or through a
gear transmission. As it turns, it
moves the other rotor (driven rotor).
It is easy to relate the compression
process to a reciprocating compres-
sor, if you consider the drive rotor as
the piston and the driven rotors as the
cylinder. As the drive rotors and
driven rotors un-mesh, an empty cyl-
inder is created, drawing in suction
gas through the synchronized open-
ing on the rotor suction face. As Figure 38
rotation continues, the suction and
Screw Chiller Technology
discharge rotor faces are sealed off,
trapping the gas in the cylinder.
When this happens, the meshing point moves toward the discharge end of the rotors and drives
the gas ahead of it. As rotation continues further, the drive rotors rotate into the flutes of the
driven rotors and internally reduce the contained volume, resulting in positive displacement com-
pression. The discharge port provided for the gas escape is relatively small, compared to the
suction port, reflecting the smaller volumetric flow rate of compressed vapor.
Volume Ratio
Volume ratio is defined as the ratio of the volume of the com-
pression chamber when it is open to the suction port divided by
the volume of the compression chamber when it is open to the Most screw compressors
discharge port. This is much like in a reciprocating compressor, will be badly damaged by
where the volume ratio is the volume of the compression chamber reverse rotation so there
when the piston is at the bottom of its travel divided by the vol- are built in controls to
ume of the chamber when it is at the top of its travel. Volume stop the compressor if
ratio (vr) is also called volume index (vi). reverse rotation occurs.

26
Design and Off-Design Performance

The design and size of the rotors and the placement of the discharge ports control the amount
of compression for a screw compressor. This design fixes the built-in volume ratio (vi), which
defines the change in volume within the compressor from a larger volume at compressor inlet to a
smaller volume at compressor exit. For a given refrigerant, a fixed volume ratio corresponds
thermodynamically to a certain pressure ratio. The higher the volume ratio (vin/vout) the larger the
corresponding pressure ratio (Pout/Pin). When the required pressure ratio is smaller than the design
pressure ratio, it corresponds to a smaller volume ratio than the built-in volume ratio and over-
compression will occur with subsequent sudden expansion when the overcompressed pocket of
vapor reached the exit port resulting in the so-called over-compression losses. Conversely, when
the required pressure ratio is larger than the design pressure ratio, it corresponds to a larger vol-
ume ratio than the built-in volume ratio and under-compression will occur with subsequent
sudden compression inside the final section of the compressor when the undercompressed pocket
of vapor reached the exit port. The compressor is capable of delivering pressure ratios higher than
the design pressure ratio (as opposed to turbocompressors, which have limited additional pressure
ratio capability above their design pressure ratio because of surge) but these higher-pressure ratio
conditions are suffering from under-compression losses.
Suction volume flow rate defines a screw compressor’s capacity. For a given speed, it is more
or less constant irrespective of the pressure ratio delivered by the compressor. In order to achieve
capacity variation the suction volumetric flow rate has to be changed. There are two ways of do-
ing this in a continuous way. Capacity changes proportional to speed on variable speed machines.
The built-in volume ratio is not affected by the speed of the compressor. Fixed speed screw com-
pressors use a slide valve, a mechanical device that tries to change the inlet volumetric flow rate
of the compressor while maintaining a required built-in volume ratio at lower flow rates. From
this description, it becomes clear that careful consideration must be given by manufacturers in
designing the screw compressor’s operating envelope to efficiently handle the required compres-
sion or work for the HVAC system under all possible operating conditions. It is important to
select the optimum built-in volume ratio for a given HVAC system application in order to assure
that the compressor operates most of the time close to its design pressure ratio, thus limiting the
over-compression and under-compression losses that occur at pressure ratios away from design
conditions.
Manufacturer’s selection software and printed literature incorporates the important perform-
ance criteria described above. When selection data is generated from the software, or is shown in
printed literature, that is your assurance that the application points are operationally sound.
Centrifugal Compressor Operational Details

Imagine a ball attached to the end of a string being whirled by a person holding the other end
of the string. The ball pulls outward on the string, and if the string is released, the ball flies away
on a line perpendicular to its circle of rotation. The heavier the ball, the more it pulls on the
string. The longer the string is, the harder the pull of the ball will be. The faster the ball rotates
the harder the pull of the ball.
This is analogous to the centrifugal compression of gas. Replace the ball with a molecule of
refrigerant gas, replace the string with an impeller, and the effect is the same. The centrifugal
force imparted to the refrigerant molecule will fling it outward, compressing it into the impeller
passageways. The larger the diameter of the centrifugal impeller is, the larger the force on the

27
molecule. The faster the impeller ro-

tates, the larger the force on the
molecule. The above phenomenon de-
scribes how the centrifugal compressor
increases the pressure of the vapor be-
ing compressed in the outer section of
the impeller. An additional pressure
increase results as the high speed im-
peller imparts kinetic energy to the gas.
The diffuser converts the high kinetic
(velocity) energy of the refrigerant gas
to static pressure when the gas is re-
leased into the stationary discharge or
diffuser portion of the compressor. Figure 39
Centrifugal Compressor Theory
Head
Head can be visualized as the height (in feet) of an
imaginary column of refrigerant vapor, which, due to its Head pressure
weight, produces the same pressure as that developed
across the compres- is directly affected by entering
water temperature from the
sor (pressure at
cooling tower. The entering wet
discharge minus bulb temperature to the tower
pressure at suction). affects the tower’s ability to
Head has come to be cool the water.
synonymous with a
measure of the pressure rise developed by a centrifugal
compressor.
Variations in outside conditions result in different en-
tering condenser water temperatures at both full and part
load. This term for this is tower water relief and occurs
when the outside wet bulb temperature is below the cli-
matic design condition, increasing tower heat rejection and
Figure 40 lowering the condenser water temperature entering the
chiller.
Centrifugal Compressor Theory – Head
Lift
Another commonly used term is “lift.” The expression, “55º F of lift” refers to the difference
in saturated refrigerant temperature levels between which the compressor operates. To produce
44º F chilled water, a compressor draws in gas at a saturated suction temperature level of ap-
proximately 42º F and discharges gas at a saturated discharging temperature of approximately
97º F. This is for typical HVAC duty, with 85 to 95º F condenser water. The conditions described
here would produce 55º F of lift (97° F - 42° F). The following pressure enthalpy diagram dem-
onstrates that lift is the difference between saturated condensing and suction temperatures.

28
Under many circumstances, an increase

in lift requires an increase in compressor
and/or heat exchanger sizes. Some factors
that increase lift are lower leaving chilled
water temperatures, lower condenser flow
rate, higher condensing temperature, fouling
on the waterside of the tubes, and the thermal
properties of fluids being used in the heat
exchangers.
Any reduction in lift means a reduction
in the power needed for the compressor
and savings in operating costs, as well as
increased capacity for chillers equipped Figure 41
with variable metering devices. Colder
Lift
condenser water temperatures lower the lift
and have the greatest positive impact on
the refrigerant saturated condensing tem-
perature.
Head and Lift

Represent the same thing. Head Is
the pressure difference; lift is the
temperature difference between
compressor suction and discharge.
Changing weather conditions produce

a reduction in outdoor wet bulb tempera-
ture that in turn allow the cooling tower to
produce colder than design entering con- Figure 42
denser water temperature to the water- Lift Reduction
cooled chiller. Chillers take advantage of
this natural occurrence called cooling tower relief in two areas: stable operation down to ap-
proximately 55º to 60º F entering condenser water temperature, and an increase in capacity from
design tonnage. This is often referred to as “maximum capacity.”
Compressor Boundaries
There are two important factors when discussing the operational envelope of a centrifugal
compressor. The first is stonewall, which describes the condition of maximum compression. As
the volume of gas increases, its velocity through the compressor also increases. When the velocity
of the gas exceeds the local sonic or acoustic velocity at the existing pressure, shock waves de-
velop. When this happens, high head losses occur and the compressor head output drops off
sharply. This is the “flow limitation,” and the resultant steep part of the compressor curve show-
ing this effect is referred to as “stonewall.” Centrifugal chiller software automatically keeps
selections the proper distance from this line.

29
The other limitation for the centrifugal compressor is surge. Surge occurs when the refriger-
ant pressure in the condenser is high enough that the impeller can no longer maintain the required
discharge gas flow. Stated another way, the required lift is too high for the compressor. This
higher condenser pressure results in
flow breakdown in the impeller pas-
sageways, and a partial or complete
flow reversal through the impeller
occurs. A lowering of the system
pressure should follow, and this al-
lows the impeller to function normally
again, and gas to flow in the correct
direction. The system pressure then
builds up again until it exceeds the
impeller capability, and the reversal of
flow repeats itself. This pulsing of
pressures is nearly instantaneous, and
continues until chiller controls take
corrective action to lower the con-
denser pressure, allowing proper flow Figure 43
of the refrigerant through the impeller, Centrifugal Chiller Compressor Map
diffuser, and volute. Surging is char-
acterized by an increase in the
operating noise level, and by wide fluctuations in discharge pressure and motor current. No det-
rimental mechanical effects are caused by periodic, short durations of surging. However,
prolonged periods of surging may damage the chiller.
Several conditions can cause this pressure imbalance.

An example would be a reduction in load, during normal Most centrifugal chillers
operation, without a reduction in condensing temperature can be selected to run to a
(low load, high lift). The refrigerant flow rate is too low in certain level of low load and
this case, and so the velocity of the gas, and therefore the high lift, but these
pressure generated by the compressor is too low to over- conditions must be planned
come the pressure in the condenser. A second example for at time of selection. The
would be the operation of a centrifugal chiller at higher key is to verify the
entering water temperatures than were used to make the requirements for each
project.
initial selection. This condition may create a lift too great
for the compressor.
These situations can be avoided by properly selecting the chiller for the actual operating con-
ditions for the application.

30
Compressor Stages
In a centrifugal compressor, each impeller used is referred to as a stage. We have seen from
the ball and string example that a single-stage compressor with a large diameter impeller running
at a slow speed, or a single-stage
compressor with a small diameter
impeller running at a fast speed, can
produced the same amount of lift. A
two-stage compressor with even
smaller diameter wheels running at a
faster speed can also produce the
same amount of lift. Thus, there is a
design trade off between speed, stag-
ing, and impeller diameter to achieve
a desired result. For air-conditioning
duty (approximately 55º F of lift) ei-
ther one or two-stage compressors are Figure 44
used.
Centrifugal Compressor Stages
Capacity Control Methods for Centrifugals and Screws

In most water-cooled chiller applications, the machine must respond to changes in two basic
variables. These are: entering condenser water temperature and refrigeration load. Both parame-
ters affect the lift requirements on the system at any time, and any reduction in either is
considered a part load condition. Centrifugal and screw compressors can accommodate these
changes, even though they have different means to control refrigerant flow (or capacity).
Inlet Guide Vanes

One method of capacity control
involves keeping the speed of the
compressor constant and varying the
flow of the refrigerant to the compres-
sor. To maintain stable operation in a
centrifugal, the most common method
of flow control is by using an inlet
guide vane assembly (IGV). These
vanes are usually located just ahead of
the inlet to the impeller and are con-
trolled to the leaving chiller water set
point. As the IGVs close, they reduce
the mass flow of refrigerant through
the compressor. Figure 45
Centrifugal Chiller Capacity Control - IGV

31
At part load, closing the vanes

imparts prewhirl to the gas, minimiz-
ing impeller-entry pressure loss.
Inlet guide vanes change the
head-volume characteristic of the
compressor by changing the angle at
which the suction gas enters the com-
pressor. At full load, the gas is
directed against the compressor rota-
tion, and the compressor actually
"bites into" the suction gas, picking
up more gas than if the gas entered
radially. At part load, the angle
changes such that the gas enters with
rotation minimizing any pressure loss Figure 46
through the vanes by imparting a pre- Compressor Efficiency Characteristics with Inlet Guide Vanes
whirl to the gas.
This method offers very good efficiency over a wide range of capacity. At half load condition
for example, the input power required is approximately 45 percent of full load input power.
Shown here are the compressor head-volume characteristics at part load. The compressor can bal-
ance head and volume changes efficiently down to 10 or 15 percent load without the use of hot
gas bypass as long as the operating point falls within the surge boundary.
Screw Unloaders
For a screw chiller, either slide valves or solenoid actuated port valves are used to change the
volume ratio of the compression chamber.
A common means of capacity
control is changing the volume ratio
with slide valves. The slide valve
opens a passageway along the side of
the compression chamber, allowing a
portion of the gas to bypass the rotors.
This bypassed gas is recycled to the
suction cavity and is then compressed.
Essentially, a slide valve shortens the
length of the compression path, reduc-
ing the capacity of the screw
compressor. The slide valve offers
infinite volume ratio, and therefore Figure 47
infinite capacity control for screw Capacity Control – Slide and Port Valves
compressors.
Another means of reducing the volume by shortening the compression path is port valves.
Port valves are staged along the compression path, and the number of opened valves determines
the volume of gas compressed. These valves are less expensive, easier to control, and simpler to
maintain than slide valves, but they do not offer infinite control as they open in discrete steps, and
therefore reduce the volume of gas compressed in the same discrete steps.

32
Hot Gas Bypass

Bypassing gas around the condenser has the effect of falsely loading the compressor. Hot gas
bypass is used to accomplish stable operation if a load exists beneath the minimum step of normal
unloading capability of the chiller compressor. On large screw and centrifugal chillers, hot gas
bypass is available either factory-installed or field-installed. Hot gas bypass may also be used as a
surge prevention method in centrifugal chillers.
Speed Control
Variable speed control provides the most efficient method to vary capacity, but in the past, it
has been expensive. Recent improvements in technology and the widespread acceptance of vari-
able frequency drives (VFD) have lowered the cost such that today’s VFD technology is often in
the 1-2 year payback period. Any time the lift requirements of a chilled water system are reduced,
a VFD applied to the compressor can realize power consumption savings that come from the
lower speeds required at the reduced lift. For a centrifugal compressor, a reduction in speed is
represented by the reduction in gas flow rate, lift capability, and input power as shown.
Figure 48
Laws for Centrifugal Loads
1. Flow is proportional to speed

2. Lift is related to speed squared
3. Power is related to speed cubed VFD retrofit
All of these centrifugal laws are based on how fast the im- is an increasingly popular
energy saving measure on
peller is spinning, and anytime there is a reduction in lift and
existing large chillers.
flow, the compressor speed can be reduced by a VFD, resulting
in energy savings.

33
VFD control provides a much

more efficient means to control capac-
ity for a centrifugal compressor,
because as inlet guide vanes close, the
efficiency of the compressor rapidly
declines.
The example graph shows a plot
of the kW per ton of a partly loaded
centrifugal chiller at reduced lift con-
ditions. The reduced lift conditions
are the direct result of a reduction in
entering condenser water temperature
at part load. Besides the energy sav-
ings, there are lower sound levels that Figure 49
result from running the chiller at re- VFD Versus Constant Speed Centrifugal
duced speeds.
The effect of VFD control is to continuously “shrink” the compressor, keeping the operation
in the maximum efficiency region over a much broader range of operation. Essentially, the VFD
adjusts the speed of the compressor to keep the IGVs as open as possible to meet the system lift
requirements, with the lowest power consumption. Combined with condenser water temperature
reduction that occurs naturally in an
air-conditioning system, the variable
speed centrifugal compressor more
efficiently meets the flow and lift con-
dition or state point required by the
system.
A centrifugal compressor’s dy-
namic design needs sufficient
rotational speed to overcome system
lift requirements. As a result, centrifu-
gal chillers cannot be controlled by
speed variation alone. Centrifugal
chillers must be controlled by a com-
bination of speed variation and inlet
guide vane actuation in variable speed
applications. Thus, centrifugal com- Figure 50
pressors are limited in the amount that
they reduce compressor speed under Variable Speed Centrifugal Operating Characteristics
high lift conditions.

34
Refrigerant Related Topics

Design pressures, cycle efficiencies, toxicity, chemical stability, and environmental impact
have limited the refrigerants in common use to a group of fluorinated hydrocarbons. Members of
this group include Refrigerants 22, 123 and 134a. There are also several refrigerants that are
blends of other refrigerants. These blends are classified as either azeotropic or zeotropic.
The former means that the blend behaves as if it were a
New refrigerant choices… single refrigerant; the latter describes a blend where the dif-
Include blends, like R-410a ferent refrigerants in the blend evaporate at different
(PuronTM) and R-407c that temperatures. The disadvantage of a zeotropic refrigerant is
are being used to replace that if there is a leak, a change in the composition of the blend
R-22 in applications other may reduce the system capacity and/or efficiency. In some
than in centrifugal chillers. cases, the entire charge might be unusable if a leak were to
occur.
Regulations
Recently, refrigerant selection has been driven by two international environmental agree-
ments, the Montreal Protocol on Substances that Deplete the Ozone Layer and the Kyoto
Protocol, which is focused in mini-
mizing the impact of climate change.
The Montreal Protocol mandates the
phase out of ozone depleting sub-
stances. The CFC’s, R-11, R-12, R-
113, R-114 and R-115, were phased
out on January 1, 1996. The HCFCs
(e.g. R-22 and R-123) will be phased
out on January 1, 2030 with interim
reductions taking place on January 1,
2004, January 1, 2010, January 1,
2015, and January 1, 2030. From
January 1, 2020 to January 1, 2030, Figure 51
consumption will be reduced to 0.5 Montreal Protocol Showing EPA Equipment Phaseout Dates
percent of the baseline and new
HCFCs may only be used to service existing equipment. The US and Europe, as well as a major-
ity of developed countries, have adopted the Montreal Protocol and in some cases have passed
regulations that accelerate the phase
Note: out of some or all HCFCs. In
Europe, most uses of HCFCs have
The Montreal Protocol treats HCFC-22 and HCFC-
already been banned.
123 the same, however, the U.S. has adopted
chemical-specific phase-out dates. HCFCs with
higher ODP (R-22)will be phased out first followed
by HCFCs with lower ODP. Therefore, in the U.S.,
HCFC-22 Is being phased out ahead of HCFC-123.
For example, as of 2010, no new R-22 equipment
can be produced, and as of 2020 no new R-123
equipment can be produced.

35
The Kyoto Protocol mandates the reduction of emissions of greenhouse gases (i.e. carbon di-
oxide, methane, nitrous oxide, HFCs). It does not call for the phase out of the chemicals. The
Kyoto Protocol has affected chiller design by influencing manufacturers to find more efficient
refrigerants, and therefore designing more efficient chillers to reduce the electrical consumption
of HVAC systems.
It is also important to note that any refrigerant for use in chillers must be on the EPA’s Sig-
nificant New Alternative Policy approved list, and in ASHRAE Standard 34 to ensure that it has
an “R” designation and is assigned a safety classification based on toxicity and flammability data.
If it is not, using it is a violation of the Clean Air Act, and is potentially hazardous. Several re-
frigerants that are described as drop-in replacements for HCFC chillers are considered dangerous
and are not approved by the EPA. In addition to this, the equipment that the substitute refrigerant
will go into must undergo design changes to ensure safe operation of the approved refrigerant
alternative. One significant reason for this is the different operating pressures of refrigerants. As
an example, R-22 replacement R-410A operates at about 30 percent higher pressure, and if the
original equipment is not designed to operate at that pressure, it would not be safe to apply R-
410A into the system without modifications. For complete discussion on this topic, refer to TDP-
402, Refrigerants.
Chiller Construction
The choice of refrigerant affects the
construction of the chiller as well. Dif-
ferent refrigerants have different
operating pressures, so various codes
dictate their construction. The cooler and
condenser may need to be constructed to
ASME pressure vessel standards as de-
fined in Section VIII, Division I of the
current ASME code. ASME stands for
the American Society of Mechanical
Engineers. A data report for each posi-
tive pressure centrifugal chiller is on file Figure 52
with the National Board of Boiler and Chiller Construction ASME “U” Stamp
Pressure Vessel Inspectors and is avail- Photo Courtesy of Honeywell Genetron Refrigerants
able for public review. In this report, an
authorized independent inspector will have approved the welds and verified the pressure tests.
These inspectors, working on behalf of ASME, not the chiller manufacturer, ensure that the posi-
tive pressure chillers are manufactured and tested in accordance with the strict standards of
ASME and will bear the “U” stamp indicating compliance with ASME. They provide third party
certification during the manufacturing process.
For large screw and centrifugal chillers, the quality of the construction enables hermetic posi-
tive pressure chillers to have approximately 0.1 percent annual refrigerant leak rate. ASME
currently requires that the refrigerant side (shells) of the R-134a chiller be pneumatically tested to
1.1 times the design pressure, a value typically around 200 psig. Negative pressure chillers are not
inspected by an ASME authority, and are tested to 30 psig as required by ANSI the American
National Standards Institute. Waterside testing of the tubes is typically 1.3 times the design pres-
sure of the water. This design pressure is usually 150 or 300 psig.

36
On larger chiller designs like centrifugals and

ASME stamp requirements screws using positive pressure refrigerants such as
dependent on the physical size and the R-134a, the evaporator and condensers meet or
refrigerant pressure rating on the exceed the minimum size requirement and they
chiller. For example, cooler and must bear the symbol signifying compliance with
condenser vessels less than 6 inches ASME. On negative pressure centrifugal chillers,
inside diameter may be constructed to the refrigerant pressures involve exclude the
comply with another nationally chiller construction from consideration by ASME.
recognized testing laboratory other than
ASME. This may vary from
manufacturer to manufacturer.
Chiller construction is also af-

fected by the physical size of the
refrigerant molecules. The chart
shows the relative molecular size of
the refrigerants currently in use. This
means that for the same capacity, an
HFC-134a chiller can often be ap-
proximately 30 percent smaller than a
negative pressure chiller of the same
cooling capacity. The smaller foot-
print results in savings in installed
costs and mechanical room size. Figure 53
Refrigerant Impact on Chiller Size
Safety
ASHRAE Standard 34 divides re-
frigerants into two groups, lower and
higher toxicity, designated by the let-
ters A and B, respectively.
Refrigerants with allowable exposure
limits (AEL) of 400 ppm or higher are
classified as group A (lower toxicity)
refrigerants, and those with AELs of
less than 400 ppm are group B (higher
toxicity). R-134a with an exposure
limit of 1000 ppm is classified as
Group A and R-123 with an exposure Figure 54
limit of 50 ppm is classified as
Allowable Exposure Limits
Group B.
Refrigerants are further classified by their degree of flammability: (1) no flame propagation,
(2) lower flammability, and (3) higher flammability. HFC-134a is a safe choice, because it has a
low toxicity and no flame propagation properties

37
HFC-134a has become the lead-

ing choice for a long-term refrigerant
solution because it is chlorine-free
and therefore has zero ozone deple-
tion potential and is not subject to
phaseout under the Montreal Protocol.
Today, HFC-134a is being applied in
foam blowing, asthma inhalers, auto
air conditioning, residential air condi-
tioning, water fountains, and domestic
refrigerators.
Figure 55
Refrigerant Safety Groups
Heat Transfer
This section will look at the basic
heat transfer relationships of the heat
exchangers for a water-cooled chiller.
The information provided is general
in nature and should provide a basic
understanding of the heat transfer
principles associated with shell-and-
tube heat exchangers.
Shown is the longitudinal cross
section of a typical condenser and
evaporator together with their respec-
tive temperature profiles. The heat
transfer relationships of the evapora-
tor and condenser are similar. Figure 56
Heat Transfer – Condenser, Fluid Heated t1 –t2
In looking at the figure, it can be
observed that there are two basic heat
balances occurring in each heat ex-
changer. They are:
1. Heat given off or gained by the
fluid as it passed through the
tubes.
2. The overall transfer of heat be-
tween the fluid in the tubes and
the refrigerant in the shell.
.
Figure 57
Heat Transfer – Evaporator, Fluid Cooled t1 –t2

38
Heat transfer technology has improved to the point where the difference between the refriger-
ant temperature in a heat exchanger (evaporator or condenser) and the leaving fluid temperature
can be very low…in the range of 2-3 ºF. That means, for instance, if the saturated condensing
temperature is 97º F, the water returning to the cooling tower may be as high as 95º F. If the satu-
rated suction temperature is 42º F, the leaving chilled water may be as low as 44º F. Later in this
TDP module we will examine the components affecting heat transfer between the refrigerant and
water in the tubes as well as tubing technology that promotes increased heat transfer.
Heat Balance of Fluid

The fluid in the cooler tubes is cooled from t1 to t2 as it gives off heat to the colder refrigerant.
Similarly, the fluid in the condenser tubes is heated from t1 to t2 as it takes on heat from the hotter
refrigerant. The total amount of heat exchanged in either vessel can be expressed as follows:
q = W ∗ CP ∗ ∆tF
Where:
q = quantity of heat exchanged (Btuh)
W = flow rate of fluid in tube (lb/h)
CP = specific heat of fluid [Btu/lb(ºF)]
∆tF = temperature rise or drop of fluid in tubes
sp gr = specific gravity
The above equation can be rearranged in the following form, and it applies to both evaporator
and condenser.
gpm ∗ ∆t F ∗ sp gr ∗ CP
tons =
24
For the evaporator, the tons equals cooling capacity. For the condenser, the tons equal cooling
capacity plus the heat equivalent (in tons) of the work of compression.
The above equation for tons is commonly used to determine flow rate, temperature rise, or
load when any two of the factors are known. When the fluid is fresh water, the specific gravity
(sp gr) and specific heat (CP) have values of approximately 1.0, and the equation reduces to:
gpm ∗ ∆t F
tons = for the evaporator and the condenser.
24 In converting
Where: the tons equation into a
gpm = flow rate for evaporator or condenser more recognizable form
involving tons, some
∆tF = difference in fluid entering and leaving standard units of
24 = conversion to tons for fresh water conversion were used.
Heat balance for a water-cooled chiller means that the evaporator capacity in Btuh or tons
when added to the compressor work, numerically equals the total heat rejected by the condenser
water flow. A few simple equations allow checking the heat balance when hermetic motor type
chillers are being evaluated. There is a small degree of latitude in the heat balance permitted by
the industry to account for measurement and manufacturing tolerances.

39
A heat balance is very useful in checking chiller selections for careless mistakes or verifying
chiller performance of an existing chiller. In using a heat balance on an existing chiller, it may be
difficult to predict the exact water flow.
Another equation that is useful in converting water flows into heat transfer is:
Btuh = 500 * (gpm) * ∆t, where 500 is a constant for fresh water.
Example: Heat balance for a centrifugal chiller selection.
Shown in the capacity printout:
Capacity 640 tons
Compressor kW 402
Condenser gpm 1393
Condenser rise °F 13
Let’s calculate the heat balance for the chiller:
Heat In = Heat Out
Chiller capacity + compressor kW = condenser heat rejection
640 tons ∗ 12,000 Btuh/ton + (402 kW ∗ 3413Btuh/kW) = 500 ∗ 1393 gpm ∗ 13° F
3413 converts Btuh to kW
9052 MBtuh is the evaporator and compressor heat
9054 MBtuh is the condenser total heat of rejection
This heat balance is basically equal at less than 1 percent difference. General industry prac-
tice is a balance within 5 percent is considered acceptable.
Overall Heat Transfer

Heat added to or subtracted from the refrigerant at a constant pressure results in a change of
state of the refrigerant. The heat required to change the refrigerant state in the evaporator from a
liquid to a vapor is known as the “latent heat of vaporization.” In the condenser, the heat required
to change the refrigerant state from a vapor to a liquid is known as “the latent heat of condensa-
tion.”
In either the evaporator or the condenser, this process takes place at a constant refrigerant
saturation temperature corresponding to the saturation pressure in the vessel. The rate at which
heat transfer takes place is directly related to the entering temperature difference, and the leaving
temperature difference.
The overall heat exchanged between the fluid and the refrigerant for the cooler or the con-
denser can be expressed as:
q = U ∗ A ∗ LMTD
Where:
q = overall heat exchange rate (Btuh)
U = overall heat transfer coefficient
A = area of heat exchanger tubes (ft2)
LMTD = log mean temperature difference between the fluid and the refrigerant (ºF)

40
The LMTD of either heat exchanger can be calculated from an equation that takes into ac-
count the entering and leaving differences between fluid and refrigerant temperatures.
∆t F
LMTD =
Loge ∆tent
∆tlvg
Where:
∆tF = rise or drop in temperature of the fluid flowing through the tubes
∆tent = temperature difference (ºF) between the entering condenser fluid temperature and
the saturated condensing temperature, or the entering evaporator fluid temperature
and the evaporator suction temperature
∆tlvg = temperature difference (ºF) between the leaving condenser fluid temperature and the
saturated condensing temperature, or the leaving evaporator fluid temperature and
the evaporator suction temperature
This equation for overall heat transfer looks
familiar to us. The heat transfer through a wall Note
assembly when figuring heating and cooling loads This section on overall heat transfer is
is quite similar ( U * A * ∆t ). The heat transfer in included to establish a foundation on
a chiller involves both a sensible heat transfer the factors involved in heat transfer so
along with a phase change of the refrigerant; the that the reader may appreciate the
∆t is not a simple linear function, it is a logarith- impact of LMTD and U on chiller
efficiency and performance. These
mic function.
terms will be utilized in the next section.
Heat Transfer Coefficient

The mean temperature difference (LMTD) between the refrigerant and the fluid in the tubes
is the driving force that overcomes the resistance to heat transfer. The resistance to heat transfer is
made up of four components, which are HEAT TRANSFER RESISTANCES BETWEEN
related to the overall heat transfer coeffi- FLUID IN THE TUBES AND THE REFRIGERANT
cient (U) by the following equation:
RW 0.000469 46% Fluid Film
1
= RW + RF + RM + RR
U RF 0.000250 24% Fouling
Where: RM 0.000029 3% Tube Material

RW = fluid film resistance
RF = fouling deposits resistance RR 0.000277 27% Refrigerant Film
RM = tube wall metal resistance RT 0.001025 100% Total

RR = refrigerant film resistance
Figure 58
Shown are the typical magnitudes for
each component of the total resistance. Typical Resistances to Heat Transfer- Baseline
The refrigerant film resistance RR comes
from manufacturer test data and is a function of the heat exchanger design, the properties of the
refrigerant used, and the geometry of the outside tube surface.

41
The other three, RF, RW, and RM, are all variables (representing 75 to 80 percent of the total
resistance) and are within the control of those applying or selecting the heat exchangers. The im-
pact of these variables on the resultant refrigerant temperature (tR) is discussed below.
Impact of Fouling Factor on (U)

Fouling is the build up of scale deposits on
the inside surface of the heat exchanger tubes RW 0.000469 26% Fluid Film
as minerals or suspended materials in the fluid RF 0.001000 56% Fouling

precipitate out. Fouling begins as soon as the
heat exchanger is placed in operation. RM 0.000029 2% Tube Material
Shown in Figure 59 is the resultant impact RR 0.000277 16% Refrigerant Film

on the overall resistance when fouling is al-
lowed to build up in the tubes of a typical RT 0.001775 100% Total
condenser. The impact of the RF factor has in-
Figure 59
creased from 24 percent at 0.00025 to 56
percent of the total resistance at 0.001 – a dra- Resistances with Increase in Fouling
matic change making RF the dominant factor in
the resistance to heat flow.
With a rise in RF the overall heat transfer coefficient U will decrease. Thus, the temperature
driving force LMTD must increase in order for the heat exchanger to transfer the same amount of
heat q. The result is an
increase in the saturated
condensing temperature.
The degree of foul-
ing from buildup of
scale deposits depends
upon the quality of the
fluid circulating though
the tubes, and the
amount and nature of
any material it may have
in suspension.
The application of
the proper fouling factor Figure 60
in the selection of heat
exchangers is important Fouling
and must be evaluated
for every job. Larger than standard heat exchangers
may be required to accommodate higher fouling factors, ARI
which in effect is increasing the area of heat transfer.
(Air Conditioning and Refrigeration
This will increase the first cost of the machine. If equip- Institute) Standard 550/590 details
ment is selected on the basis of small fouling factors, methods for rating chillers and
and actual conditions dictate otherwise, it will result in provides guidelines for comfort
capacity deficiencies and might also affect surge char- cooling applications.
acteristics of a centrifugal chiller.

42
Impact of Tube Velocity On (U)

Fluid Film
Fluid flowing through a tube forms a static film or boundary layer that has a zero velocity at
the tube wall. This film acts as an insulator and hinders the transfer of heat from the tube wall. At
lower fluid velocities, this film can become thicker, and a higher resistance to heat transfer will
occur.
Comparing the initial parameters where Fluid Film
RW 0.000985 64% (4 fps velocity)
the tube velocity was 10 fps, we now show the
results when the tube velocity is 4 fps. The RF 0.000250 16%
water resistance RW increases from 46 percent
to 64 percent of the overall resistance to heat RM 0.000029 2%
transfer. RR 0.000277 18%
In addition, because the overall resistance
RT is now higher, both the LMTD and satu- RT 0.001541 100%
rated condensing temperature will have to Figure 61
increase in order to transfer the same amount
Resistances with Lower Water Velocity in Tubes
of heat at the lower 4 fps velocity.
Pass arrangements can be increased or decreased to raise or lower the tube velocity and to
circulate more water. The velocity increases with higher passes because the area through which
the water flows decreases. In selecting a heat exchanger, the tube velocity should generally be
kept between 3 and 12 fps. The higher limit is arbitrary and is based on operation with reasonable
pressure drops while simultaneously minimizing any possible tube erosion. Velocities less than
3 fps may result in laminar flow (thick boundary layer), which causes water resistance RW to in-
crease. Typical HVAC duty requires 1, 2, or 3 passes, based on application conditions such as
pressure drop and required capacity. See the pass arrangement section in this TDP.
Tube Erosion
In order for erosion to occur, an agent
must penetrate the fluid layer. Agents that Fluid velocity
cause the tube damage can be chemical, me- itself is not the sole cause of tube damage,
chanical, or a combination of both. Chemical however, if the fluid does contain attacking
agents diffuse through the fluid film to attack agents, increased fluid velocity will increase
the tube. Mechanical agents cause damage by the rate of attack. A good water treatment
the impingement of entrained gas bubbles or program can help prevent premature tube
erosion. For a discussion on water
suspended material against the tube wall. Ex-
treatment, refer to TDP-641, Condensers
cess fluid velocity will accelerate erosion. and Cooling Towers.
Pressure Drop
Pressure drop increases as the square of the velocity. The higher the velocities, the higher the
fluid pumping costs. A practical application approach is to specify limits and allow heat exchang-
ers to be selected up to the 12 fps limit, as long as the pressure drop limits are met. This approach
would assure reasonable operating costs, minimal erosion problems, and still allow low first cost
heat exchangers to be selected.

43
Impact of Tube Material On (U)

Compared to Figure 59, which was based RW 0.000469 35%
on copper tubes, Figure 62 shows the impact
on the overall resistance to heat flow caused RF 0.000250 18%
by the use of 70-30 cupro-nickel tubes. Add-
ing CuNi tubes increases the impact of the
RM 0.000350 26% 70 – 30 CuNi
Tube Material
metal resistance RM from 3 percent to 26 per- RR 0.000277 21%
cent of the overall resistance.
R
T 0.001346 100%
The overall increase in resistance causes
an increase in condensing temperature or a Figure 62
decrease in suction temperature, affecting Resistances with Another Tube Material
unit capacity. Manufacturers can offer larger
water-cooled screw and centrifugal chillers with a selection of tubing materials ranging from the
commonly used copper to an exotic type like titanium. Metal resistances of tube materials vary
greatly and can significantly
Approximate Cost Multiplier
Application Tube Material
vs. Copper Tubes impact machine size and
power input requirements.
Fresh Water Copper 1.0
Glycols Copper 1.0 The conditions of the
fluid used in the application
Corrosive Water Cupro-Nickel 1.3
should determine the type of
Special Process Stainless Steel 2 to 3 metals used for the tubing.
Sea Water Titanium or Cupro-Nickel 3 to 4
Figure 63
Tubing Materials Chart
Evaporator and Condenser Tubing

Having concen-
trated on the impact of
RW, RF, and RM on the
overall heat transfer
coefficient (U), we will
now consider the ef-
fects of internally and
externally enhanced
tubes on RR and RW.
The heat transfer
coefficient of boiling or Figure 64
condensing refrigerant Standard tubing – Internally and externally enhanced
is several times less
than that of water flowing through tubes. Therefore, the outer surfaces of the tubes are usually
finned by rolling grooves in the tubes. The net result is to increase the heat transfer surface area.
This extra area reduces RR, and increases heat transfer rate. For that reason, it is standard practice
in current large chiller technology to provide both internal and external enhancement to the

44
evaporator and condenser tubes. En- Fouling tests indicate

hancements are made to the waterside to that fouling rates of enhanced tubes is no
increase surface area and to provide a tur- greater than that of smooth tubes. In fact, test
bulence that reduces the chance of fouling data shows that the RF (fouling resistance) in
be maintaining a scrubbing action. the enhanced tube is less than that in the
smooth tube.
Freezing of Fluids in Tubes

Whenever the saturated refrigerant temperature is below the freezing temperature of the fluid
flowing through the tubes, fluid freeze-up is possible. Freeze-up is most likely to occur at the end
of the last pass, because that is where the temperature of the fluid is at its coldest point. There-
fore, for safety, the inside tube surface temperature at the end of the last pass should be no lower
than 32.4º F when chilling water, and no lower than 1º F above the brine freezing temperature
when chilling a brine.
Pass Arrangement
Pass arrangements are normally
related to maximum allowable tube
velocity or maximum allowable pres-
sure drop requirements. A general rule
of thumb is to use as high a pass ar-
rangement as possible, since leaving
temperature differences decrease with
larger pass arrangements. This has a
beneficial effect on compressor input
power. For example, based on a 10º F Figure 65
rise in the evaporator, the LTD for a
Pass Arrangements for Heat Exchangers with Water in the Tubes
one-pass and two-pass arrangement
would be 5º F and 2º F respectively. Using a one-pass arrangement would therefore require the
compressor to operate at a saturated suction temperature approximately 3º F lower than would be
possible with a two-pass.
Although higher pass arrangements tend to de- The most common method
crease compressor input power, this trend also
increases pumping horsepower for the chilled and of selecting of heat exchanger passes
condensing water circuits. This is a result of increas- is to specify reasonable pressure
ing tube velocities caused by increasing the number drop limits and allow the
manufacturer to recommend the
of passes. Thus, an economic balance must be struck
optimal pass arrangement for the load
between higher pumping horsepower and lower and application under consideration.
compressor input power.
Heat Exchanger Performance
It should be noted that centrifugal compressor performance and horsepower requirements will
be directly impacted by the choices made in applying the heat exchangers. The compressor oper-
ates between the pressure levels determined by the saturated suction and saturated condensing
temperatures. These temperatures are in turn directly related to the LTDs established for the
cooler and condenser. The higher the LTD, the greater will be the compressor head and horse-
power, and vice versa. High efficiency tubes can provide LTD approaches that are 1.5º F to 2º F.

45
Loading Rate
Each heat exchanger is designed by the manufacturer to transfer a nominal amount of heat.
For a given heat exchanger, the LTD will decrease as the load reduces below the nominal design
point. Thus, the LTD will decrease on a machine operating at part load causing the suction tem-
perature to rise or drop and the condensing temperature to drop. Or, for a given load, using larger
and more expensive heat exchangers initially, will require smaller LTD's and subsequently lower
compressor horsepower. This technique is commonly used today to select higher performance
(low kW/Ton) machines using standard heat exchangers and compressors.
Variable Flow Operation

Variable flow operation refers to varying the chilled water flow rate through the chiller
evaporator based on the cooling load in order to minimize water-pumping costs. In effect, the
chilled water flow rate is reduced when cooling loads are lower and increased when the loads are
greater. The maximum gpm would tend to correspond to the gpm at design load.
Modern electronic chiller controls have allowed for variable evaporator flow applications be-
cause of the ability of the controls to respond to changing conditions. However, there are both
minimum and maximum flow rate restrictions for water flow through the evaporator that apply. If
the flow rate through the evaporator becomes too low, the temperature differential between the
entering and leaving water becomes too large and may result in leaving water temperatures ap-
proaching freezing. Additionally, too low of a flow rate may result in a loss of turbulence that
would significantly reduce the heat transfer. Conversely, flow rates that are too high result in
higher pressure losses within the evaporator. Addi-
tionally, the rate of change of the chilled water flow
In flooded shell-and-tube evaporators
is very important in variable flow applications. The
recommendation is to use a maximum flow rate a typical minimum flow is that which
change of 10 percent of full flow per minute. corresponds to 3 fps velocity in the
tubes. The minimum evaporator flow
Consult the manufacturer’s selection software is readily available from the chiller
for actual minimum gpm for variable flow systems. selection software.
Codes and Standards

ARI Testing Standards
The Air Conditioning and Refrigeration Institute (ARI) is a trade association for the industry
that has established a chiller certification program with defined testing procedures and tolerances.
They also certify and label a manufacturer’s selection programs and equipment though random
testing at calibrated test facilities. The ratings are based on standardized HVAC conditions, and
provide an accurate measure of chiller performance.
The ARI testing standard 550/590-98, “Water-Chilling Packages Using the Vapor Compres-
sion Cycle,” measures two power consumption ratings. The first is full load kW/ton, which
measures a chiller’s efficiency at full load based upon the rating point. However, ARI recognizes
that on average a chiller operates at full load about 1 percent of its run time. Therefore, the second
measures of performance are called Integrated Part Load Value (IPLV) and Nonstandard Part
Load Value (NPLV) which are weighted kW/ton efficiency indicators that more closely indicate a
chiller’s overall efficiency than the full load value alone.

46
IPLV is commonly expressed as either an IkW/ton value or in unites of EER. There is a fixed
relationship between IkW/ton and EER.
EER = 12(IkW/ton)
This relationship shows that EER increases as IkW/ton decreases, and vice versa. Therefore, a
“better” IPLV is shown as a lower value when the units are IkW/ton, and, due to the relationship
just described, a “better” IPLV is a higher value when the units are expressed in terms of EER.
The weightings derived by ARI are at
100, 75, 50, and 25 percent of full load. ARI
weights the part load values more, recogniz-
ing that chillers operate more at these
points. This weighting of part loads is also
evaluated with tower relief water, which
more accurately reflects the way chilled wa-
ter systems are operated. IPLV is only
appropriate for standard ARI conditions,
and NPLV is used when the conditions are
not identical to standard ARI conditions.
Figure 66
The standard ARI conditions for IPLV ARI Weighting Factors
for water-cooled chillers are:
• Leaving chilled water: 44º F
• Chilled water flow rate: 2.4
gpm/ton
• Cooler fouling factor: 0.0001 (ft2 ⋅ h ⋅ ºF)/Btu
• Entering condenser water temp: 85º F
• Condenser water flow rate: 3.0 gpm per ton
• Condenser fouling factor: 0.00025 (ft2 ⋅ h ⋅ ºF)/Btu
• Entering condenser water temperature: 85º F at 100 percent load, 75º F at 75 percent
load, 65º F at 50 percent and 25 percent load
IPLV is then calculated using the following formula:
1 Relief
IPLV =
⎛ 0.01 ⎞ ⎛ 0.42 ⎞ ⎛ 0.45 ⎞ ⎛ 0.12 ⎞ is defined as the natural
⎜ A ⎟+⎜ B ⎟+⎜ C ⎟+⎜ D ⎟ reduction in entering
⎝ ⎠ ⎝ ⎠ ⎝ ⎠ ⎝ ⎠
condenser water temperature
Where: made possible by a reduction
in ambient wet bulb
A = kW/ton at 100% load
temperature. For a complete
B = kW/ton at 75% load discussion on this topic, refer
C = kW/ton at 50% load to TDP-641, Condensers and
Cooling Towers.
D = kW/ton at 25% load
ARI also allows a tolerance to be applied to the run testing of chillers. These tolerances are
applied to power consumption and capacity, and are based on the full load temperature difference
across the evaporator. Tolerances increase as evaporator load decreases. These tolerances are
used every time a chiller is tested in a manufacturer’s ARI-certified test facility. Tolerances are
based on aggregate tolerances on instrumentation, as well as engineering and manufacturing
processes.

47
ASHRAE 90.1
The American Society of Heating, Refrigeration and Air Conditioning Engineers developed a
standard, ASHRAE 90.1, to require a minimum level of efficiency in water-cooled chillers. This
efficiency is based on the ARI 550/590 Standard, which uses 44º F leaving chilled water and 2.4
gpm/ton flow rate, and 85º F entering condenser water and 3 gpm/ton flow rate. As the design
conditions move away from this accepted, baseline rating point, the efficiency levels change to
account for the corresponding changes in lift conditions for the compressor. If the leaving chilled
water temperature is reduced, the lift therefore increases, and the efficiency level decreases.
When a lower entering condenser water temperature is used to rate the machine, the lift decreases
and the efficiency level is set higher. If the condenser flow rate is lowered, lift increases and the
efficiency level decreases.
WATER CHILLING PACKAGES-MINIMUM EFFICIENCY REQUIREMENTS
Size Min Effic

Minimum Test
Category in
Efficiency Procedure
Equipment Type kW/TON***
Water-Cooled, Electrically Operated, 0.837 4.20 COP
All Capacities
Positive Displacement (Reciprocating) 0.696 IPLV 5.05 IPLV
0.790 4.45 COP
< 150 tons
0.676 IPLV 5.20 IPLV
Water-Cooled, Electrically Operated,
≥150 tons and 0.718 4.90 COP
Positive Displacement
< 300 tons 0.628 IPLV 5.60 IPLV
(Rotary Screw and Scroll)
0.639 5.50 COP
≥ 300 tons ARI 550/590
0.703 5.00 COP
< 150 tons
Water-Cooled, Electrically Operated ≥ 150 tons and 0.634 5.55 COP
Centrifugal < 300 tons 0.596 IPLV 5.90 IPLV
0.576 6.10 COP
≥ 300 tons
***kW/TON = 3.516/COP
Figure 67
ASHRAE 90.1 Standards, Table 6.2.1C
UL/CSA & ETL

Several safety standards apply to chillers. There are also several widely accepted agencies
such as Underwriters Laboratories, Inc., Canadian Standards Association, and ETL Testing Labo-
ratories, which test for compliance to those standards. When a chiller is approved as meeting the
required testing, it will bear a mark or label from the certifying agency.
UL (Underwriters’ Laboratories, Inc.) is an independent nonprofit organization that tests
products for safety and certifies them. The Canadian Standards Association (CSA) is a non-profit
association serving business, industry, government, and consumers in Canada. Among many
other activities, CSA develops standards that enhance public safety. For heating, ventilation, and
air conditioning, UL/CSA Standard UL 1995 / CSA C22, Heating and Cooling Equipment ap-
plies.

48
ETL Testing Laboratories, like UL, conducts electrical performance and reliability testing.
OSHA (Occupation Safety and Health Administration) recognizes ETL as a nationally Recog-
nized Testing Laboratory as is Underwriters Laboratories. The ETL Listed Mark and Canadian-
ETL Listed Mark are accepted throughout the United States and Canada compliance with nation-
ally recognized standards such as ANSI (American National Standards Institute), UL, and CSA.
This certification mark indicates that the product has been tested to and has met the minimum
requirements of a widely recognized U.S. product safety standard, that the manufacturing site has
been audited, and that the applicant has agreed to a program of periodic factory follow-up inspec-
tions to verify continued conformance. If the mark includes a small “US” and/or “C,” it follows
product safety standards of United States and/or Canada respectively.
ASHRAE Standard 15
ASHRAE Standard 15, “Safety Standard for Refrigeration Systems,” has the primary goal of
mitigating safety risks to the environment, to mechanical room operators, and ultimately to the
public by incorporating specific design requirements for the safe installation and operation of me-
chanical refrigeration systems. Some of the design requirements of ASHRAE Standard 15 are:
• Location of refrigerant relief piping discharge
• Sizing of refrigerant pressure relief devices and piping per ASHRAE recommendation
• Restricting access to the mechanical room
• Providing sensors and a refrigerant monitor capable of detecting refrigerant leakage
• Installing a mechanical room ventilation system
• If the type of refrigerant in a chiller is changed, or if the chiller is replaced, then ASHRAE
15 will be applied.
Consulting engineers’ specifications and drawings incorporate ASHRAE 15 requirements
since it is has been adopted by building code authorities.
Figure 68
ASHRAE Standard 15 Requirements for a Typical Mechanical Room

49
Selection Criteria
Now that we have discussed all the items that go into chiller design, construction, and test-
ing/certification, let us look at the parameters that are used to select a centrifugal chiller using a
manufacturer’s ARI-certified selection program.
The first two pieces of information needed are the required capacity and efficiency. This
could be full load, IPLV, or ASHRAE 90.1 minimum efficiency requirements. Next, the stability
requirements of the chiller are selected. For humid regions, typical stability is chosen at approxi-
mately 50 percent of full load at design entering condenser water, to safeguard against surge
conditions.
Stability is a term that is used in evaluating the part load operating condition for a centrifugal
chiller. If the head pressure during part load operation is higher than the chiller was selected for,
the impeller may not be able to overcome the lift, and the chiller may enter an unstable opera-
tional condition causing the compressor to surge.
Centrifugal chillers are typically selected for full load and/or part load kW/ton targets. Then
they are checked for part load stability using software provided by the chiller manufacturer. A
typical part load stability check may involve running the chiller at part load points at entering
condenser water temperatures that follow a relief profile representative of the project geography.
For more information on the subject of cooling tower relief, refer to TDP-641, Condensers and
Cooling Towers.
Figure 69
Centrifugal Chiller Input Selection Screen 1

50
After that, the chilled and condenser fluid conditions are entered. These can be based on the
designer’s requirements, or at typical ARI conditions. For both the chilled and condenser water,
two of three inputs are required. For the chilled water, leaving chilled water temperature is typi-
cally set, and then either the flow rate or entering chilled water is entered. In the condenser, the
entering condenser water temperature at design is entered, and then either the leaving condenser
water temperature or the flow rate is entered.
Other parameters, such as the highest pressure drop allowed for the fluid, and the fouling fac-
tors, are also used here to help eventually determine the size of the heat exchangers and the size
of the compressor.
The next set of selection criteria is the type of starting equipment and the electrical require-
ments. These are typically determined by the site and designer or owner preference. The same is
true for the number of chiller passes, but this may also be dictated by the requirements of the
chiller capacity and water conditions.
Options that affect performance, and therefore the final chiller selection, are entered here.
Waterbox-type and waterside working pressure can be entered. Tubing material, wall thickness
and tube enhancements can be chosen as well, based on the application and design requirements.
Figure 70
Centrifugal Chiller Input Selection Screen 2

51
It is best to allow the program to select the cooler, condenser, compressor, and motor for ini-
tial selection. The program will pick the best options based on all the other inputs entered thus
far. However, these inputs may be entered manually at any time in the selection process. Once the
selection program has chosen an initial design, the user can go back and modify these inputs to
maximize the performance of the chiller. As an example, the compressor may be increased one
impeller size, or the condenser may be increased to improve performance.
These are the main inputs that determine the selected components of a water-cooled chiller
from a typical selection program. The format will vary among manufacturers, but the basic inputs
remain the same.
Summary
The intent of this TDP was to describe the water-cooled refrigeration cycle, as well as the de-
tails of the components, operation, and regulations that determine the chiller’s construction. From
reading this TDP, the student should be able to understand the main components in the centrifugal
refrigeration cycle, and be able to follow the path of water and refrigerant through the chiller. The
different types of compressors were described, so the reader should be familiar with the compres-
sion theories and have the ability to describe the differences in the compressor operating ranges
and chiller construction. The TDP then detailed the construction and uses of the components in a
water-cooled refrigeration cycle, and the student should be able to detail the differences and ad-
vantages to the various options.
Various starter types and were covered, and the TDP demonstrated the differences in starting
current and application for water-cooled chillers. The issue of refrigerants and the regulations
governing its usage, safety and chiller construction impacts was covered to allow the reader to
make informed decisions on what refrigerant should be used in a water-cooled chiller. The TDP
defined different capacity control methods, and the affects on efficiency should be recognized.
Heat transfer theory for heat exchangers provided the basic concepts for shell-and-tube heat trans-
fer and the process used to determine the heat exchanger construction and performance. The TDP
briefly outlined the different standards that relate to water-cooled chiller performance, construc-
tion, and installation. From this section, the student should now understand the agencies and
organizations that rate and certify water-cooled chillers. Finally, the TDP covered the necessary
selection inputs for a centrifugal chiller.

52
Work Session
1. List five major components of a large water-cooled chiller.
____________________________ ___________________________
____________________________ ___________________________
____________________________
2. What are two cycles used to improve efficiency in water-cooled chillers?

________________________________________
________________________________________
3. List four types of compressors used in water-cooled chillers.

________________________________________
________________________________________
________________________________________
________________________________________
Which type is non-positive displacement? ____________________
4. List three types of evaporators used in water-cooled chillers.

________________________________________
________________________________________
________________________________________
5. Describe the two types of compressor motor designs used in centrifugal chillers.
_________________________________________________________________________
_________________________________________________________________________
6. List some attributes of each motor type described in Question 5.
_________________________________________________________________________
_________________________________________________________________________

53
7. List seven types of motor starters
____________________________ __________________________
____________________________ __________________________
____________________________ __________________________
____________________________
8. Define head and lift.
_________________________________________________________________________
_________________________________________________________________________
9. What are two operation envelope limits for the centrifugal compressor?
_________________________________________________________________________
_________________________________________________________________________
10. What are three methods of capacity control?

________________________________________
________________________________________
________________________________________
11. Name three important effects refrigerant has on the construction of a chiller?
________________________________________
________________________________________
________________________________________
12. List four components of the heat transfer coefficient.

________________________________________
________________________________________
________________________________________
________________________________________

54
13. List three testing agencies that certify water-cooled chiller electrical code compliance.
________________________________________
________________________________________
________________________________________
14. Name 4 specific requirements of ASHRAE 15
___________________________________________________________________________
___________________________________________________________________________
15. List four reasons for the attraction of VFDs on centrifugal chillers.
________________________________________
________________________________________
________________________________________
________________________________________

55
Notes

56
Appendix
References
API Heat Transfer, Buffalo, NY. http://www.apiheattransfer.com/
ARI Standard 550/590, “Water Chilling Packages Using the Vapor Compression Guide.”
http://www.ari.org/std
ASHRAE Standard 15, “Safety Standard for Refrigeration Systems”
ASHRAE Standard 34, “Designation and Safety Classification of Refrigerants”
ASHRAE Standard 90.1, “Energy Efficient Design of New Buildings Except Low-Rise
Residential Buildings
Honeywell International, LLC, Morristown, NJ. http://www.honeywell.com/
Standard Refrigeration, Melrose Park, IL, http://www.stanref.com

57
Work Session Answers

1. evaporator, compressor-motor, starter, condenser, and pressure reducing device
2. subcooling, economizer
3. a. reciprocating, scroll, screw, centrifugal

b. centrifugal
4. brazed-plate, D-X shell-and-tube, flooded shell-and-tube
5. hermetic motors, which are sealed to the atmosphere and utilize refrigerant cooling
open-drive motors, which are coupled to the compressor through a shaft seal and exposed to
the mechanical room atmosphere.
6. Hermetic motors are smaller, run cooler, and have lower inrush current. Open drive motors
offer flexibility of motor size and types of drivers that can be applied to the compressor.
7. across the line, auto-transformer, primary reactor, part winding, wye-delta, solid state, vari-
able frequency drives
8. Head is the measure of the pressure rise developed in the compressor, and lift is the differ-
ence between the saturated suction and condensing temperatures.
9. stonewall, the maximum compression or refrigeration flow

capacity, and surge, a condition created when condensing pressure is greater than the pressure
generated by the compressor
10. refrigerant flow control, hot gas bypass, and speed control combined with flow control
11. design working pressure and construction certification, physical size of chiller, and ability to
store refrigerant in chiller
12. refrigerant film resistance, tube wall metal resistance, fouling deposits resistance, fluid film
resistance
13. UL, CSA, ETL
14. ASHRAE 15 will help you determine if you need a separate mechanical room. If you do, then
you should provide a tight fitting access door, a refrigerant monitor, an alarm system to notify
operators if there is a refrigerant leak, a mechanical room ventilation fan, and properly sized
relief lines. There are other requirements in addition to these listed in the standard.
15. high power factor, low inrush current, increased part load efficiency, low sound levels at part
load

58
Prerequisites:
This module assumes the participant has an understanding of industry terminology, basic con-
cepts of the air conditioning, and the mechanical refrigeration process. The following TDPs are
good reference for this material:
Book Instructor Presentation

Form No. Cat. No. Cat. No. Title
TDP-102 796-026 797-026 ABCs of Comfort
TDP-103 796-027 797-027 Concepts of Air Conditioning
TDP-400 796-037 797-037 Principles of Mechanical Refrigeration, Level 1: Introduction
TDP-401 796-084 797-084 Principles of Mechanical Refrigeration, Level 2: Analysis
TDP-301 796-034 797-034 Load Estimating, Level 2: Fundamentals
Learning Objectives:
After reading this module, participants will be able to:
• Compare the advantages of water-cooled versus air-cooled chillers.
• Identify and diagram the different components of a basic refrigeration cycle as it applies to
as water-cooled chiller.
• Compare and describe the differences among scroll, reciprocating, centrifugal, and screw
compressors.
• Discuss the differences in construction of water-cooled chillers of various sizes.
• Identify the standards that relate to water-cooled chillers.
• Understand the typical inputs required to select a water-cooled chiller.
Supplemental Material:
Book Instructor Presentation
Form No. Cat. No. Cat. No. Title
TDP-622 796-054 797-054 Air-Cooled Chillers
TDP-705 796-070 797-070 Chilled Water Systems
Instructor Information
Each TDP topic is supported with a number of different items to meet the specific needs of the
user. Instructor materials consist of a CD-ROM disk that includes a PowerPoint™ presentation
with convenient links to all required support materials required for the topic. This always includes:
slides, presenter notes, text file including work sessions and work session solutions, quiz and
quiz answers. Depending upon the topic, the instructor CD may also include sound, video,
spreadsheets, forms, or other material required to present a complete class. Self-study or student
material consists of a text including work sessions and work session answers, and may also
include forms, worksheets, calculators, etc.
Carrier Corporation
Technical Training
800 644-5544
www.training.carrier.com
Form No. TDP-623 Cat. No. 796-055

Supersedes Form No. T200-29 Supersedes Cat. No. 791-330

03 Water Cooled Chillers

Uploaded by

Copyright:

Available Formats

03 Water Cooled Chillers

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

03 Water Cooled Chillers

Uploaded by

Copyright:

Available Formats

COMMERCIAL

© 2005 Carrier Corporation. All rights reserved.

The First Centrifugal

was used in the dry cleaning

Commercial HVAC Chiller Equipment

Carrier’s second centrifugal ma-

The first centrifugal machine

Because of those early efforts in the

Water-Cooled versus Air-Cooled Chillers

Commercial Chiller Equipment

Commercial HVAC Chiller Equipment

Basic Refrigeration Cycle for Water-Cooled Chillers

Commercial Chiller Equipment

To increase the cycle efficiency,

Commercial HVAC Chiller Equipment

Commercial Chiller Equipment

Water-Cooled Chiller Components

The direct expansion (DX) evap-

Commercial HVAC Chiller Equipment

However, the shell of a direct expansion Cleaning DX Evaporators

In larger capacity water-cooled

Commercial Chiller Equipment

Commercial HVAC Chiller Equipment

Evaporator Pros and Cons

Brazed-Plate Evaporator** Pro Smaller footprint

Con Sensitive to contamination of the water circuit

Con Larger footprint

Flooded Shell-and-Tube Evaporator Pro High efficiency

Con Large footprint

*Photo courtesy of Standard Refrigeration

Parallel and Series Chiller Evaporators

Where chiller capacities greater

Commercial Chiller Equipment

Where a large temperature drop (greater than about

See the tubing discussion later in

Commercial HVAC Chiller Equipment

The baffle redirects the discharge

Commercial Chiller Equipment

Scroll compressors are a popular alternative to reciprocating compressors. Multiple scroll

Commercial HVAC Chiller Equipment

Commercial Chiller Equipment

Figure 25 shows a cut away view of a hermetic transmission-drive centrifugal compressor.

Commercial HVAC Chiller Equipment

mechanical room, requiring additional

Refrigerant Metering – Expansion Device

Commercial Chiller Equipment

Commercial HVAC Chiller Equipment

Commercial Chiller Equipment

Storage Tank and Transfer (Pumpout) Unit

In large positive pressure chillers,

Commercial HVAC Chiller Equipment

Commercial Chiller Equipment

Compressor Starting Methods

Commercial HVAC Chiller Equipment

Commercial Chiller Equipment

A solid-state starter’s principle of

Commercial HVAC Chiller Equipment

Variable Frequency Drive

The newest soft-start technology

applications. This actual Across-the-Line 100 600