ASHRAE Fundamentals of Water System PDF

Fundamentals of
Water System Design

Second Edition
I-P
t
pe
Mark Hegberg
Richard A. Hegberg
Inch-Pound
A Course Book for

Eg Self-Directed or Group Learning
up
ro
G
Includes Skill Development Exercises

for PDH, CEU, or LU Credits
Fundamentals of
Water System Design
e t
Second Edition
gp
pE
Mark Hegberg
Richard A. Hegberg
u
ro
G
Atlanta
Fundamentals of Water System Design (I-P), Second Edition
A Course Book for Self-Directed or Group Learning
ISBN 978-1-936504-66-4 (paperback)
ISBN 978-1-939200-04-4 (PDF)
SDL Number: 98020
© 1996, 2000, 2015 ASHRAE

All rights reserved.
t
ASHRAE is a registered trademark in the U.S. Patent and Trademark Office, owned by the American Society of Heat-
ing, Refrigerating and Air-Conditioning Engineers, Inc.
e
No part of this publication may be reproduced without permission in writing from ASHRAE, except by a reviewer who
may quote brief passages or reproduce illustrations in a review with appropriate credit, nor may any part of this publica-
tion be reproduced, stored in a retrieval system, or transmitted in any way or by any means—electronic, photocopying,
gp
recording, or other—without permission in writing from ASHRAE. Requests for permission should be submitted at
www.ashrae.org/permissions.
ASHRAE has compiled this publication with care, but ASHRAE has not investigated, and ASHRAE expressly disclaims
any duty to investigate, any product, service, process, procedure, design or the like that may be described herein. The
appearance of any technical data or editorial material in this publication does not constitute endorsement, warranty, or
guaranty by ASHRAE of any product, service, process, procedure, design or the like. ASHRAE does not warrant that the
information in this publication is free of errors. The entire risk of the use of any information in this publication is assumed
by the user.
pE
ASHRAE STAFF
ASHRAE Learning Institute Special Publications
Karen Murray Mark Owen
Manager of Professional Editor/Group Manager of
Development Handbook and Special Publications
Martin Kraft Cindy Sheffield Michaels
Managing Editor of Professional Managing Editor
Development Matt Walker
u
Associate Editor
Sarah Boyle
Assistant Editor
Lauren Ramsdell
Assistant Editor
ro
Michshell Phillips
Editorial Coordinator
For course information or to order additional materials, please contact:
ASHRAE Learning Institute Telephone: 404/636-8400

1791 Tullie Circle, NE Fax: 404/321-5478
G
Atlanta, GA 30329 Web: www.ashrae.org/ali

E-mail: edu@ashrae.org
Errors or omissions in the data should be brought to the attention of Special Publications via SDLcorrections@ashrae.org.
Any updates/errata to this publication will be posted on the

ASHRAE Web site at www.ashrae.org/publicationupdates.
Your Source for HVAC&R Professional Development
1791 Tullie Circle, NE • Atlanta, GA 30329-2305 USA • Tel 404.636.8400 • Fax 404.321.5478 • www.ashrae.org
t
Karen M. Murray Email: kmurray@ashrae.org
Manager of Professional Development
e
Dear Student,
gp
Welcome to the ASHRAE Learning Institute (ALI) Fundamentals of HVAC&R Series of self-directed or group
learning courses. We look forward to working with you to help you achieve maximum results from this course.
You may take this course on a self-testing basis (no continuing education credits awarded) or on an ALI-monitored
basis with credits (PDHs, CEUs or LUs) awarded. ALI staff will provide support and you will have access to tech-
nical experts who can answer inquiries about the course material. For questions or technical assistance, contact us
at 404-636-8400 or edu@ashrae.org.
pE
Skill Development Exercises at the end of each chapter will gauge your comprehension of the course material.
If you take this course for credit, please complete the exercises and send copies from each chapter to
edu@ashrae.org (preferred method) or ASHRAE Learning Institute, 1791 Tullie Circle, Atlanta, GA 30329-
2305. Be sure to include your student ID number with each set of exercises. Your student ID can be the last five
digits of your Social Security number or another unique 5-digit number you create. We will return answer
sheets to the Skill Development Exercises and maintain records of your progress. Please keep copies of your
completed exercises for your own records.
When you finish all exercises, please submit the course evaluation, which is located at the back of your course
u
book. Once we receive all chapter exercises and the evaluation, we will send you a Certificate of Completion
indicating 35 PDHs/LUs or 3.5 CEUs of continuing education credit. The ALI does not award partial credit for
SDLs. All exercises must be completed to receive full continuing education credit. You will have two years
from the date of purchase to complete each Self-Directed Learning Course.
ro
We hope your educational experience is satisfying and successful.
Sincerely,
G
Karen M. Murray
Manager of Professional Development
American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

AN INTERNATIONAL ORGANIZATION
Continuing Education Opportunities
from the ASHRAE Learning Institute
Self-Directed or Group Learning
ASHRAE offers texts for self-study or group training with instructor materials. Texts cover
the basics of what a practicing engineer needs, and skill development exercises are included to
evaluate progress.
t
ASHRAE offers the following course books:
e
Fundamentals of Air System Design Fundamentals of Psychrometrics
Fundamentals of Building Operation, Mainte- Fundamentals of Refrigeration
nance, and Management Fundamentals of Steam System Design
Fundamentals of Heating and Cooling Loads Fundamentals of Thermodynamics
gp
Fundamentals of Heating Systems Fundamentals of Water System Design
Each course book includes the following:

• Clear and concise discussion of the technical topic covered
• Examples that show how to apply the lesson’s principles
• Skill development exercises that test students’ ability to apply the newly acquired knowledge
and answer sheets to assess progress in learning the material
pE
Those who complete a course receive a certificate designating continuing education (CE)
credits. Note that individuals are responsible for contacting their relevant governing body to deter-
mine whether an activity qualifies for that body’s continuing education credits.
ASHRAE eLearning
ASHRAE is a continuing education provider of the American Institute of Architects (AIA)
and the Green Building Certification Institute (GBCI), a third-party certification administrator of
u
the United States Green Building Council (USGBC). Continuing Education (CE) hours earned
from ASHRAE courses may be applied toward maintenance of state-licensed professionals
(Architects and Professional Engineers) and Leadership in Energy and Environmental Design®
(LEED®) Green Building Rating System credentials.
ro
ASHRAE offers over 130 eLearning courses focusing on specific topics and 18 course pack-
ages that cover a topical area and include several related courses. ASHRAE offers the following
course packages:
AC Design Electrical System Design
HVAC Systems Standard 62.1
HVAC Control Systems Standard 90.1
G
DDC Controls Standard 90.1 for Architects

Small Office Buildings Standard 189.1 for High-Performance
Small Retail Buildings Green Buildings
Sustainable Buildings Data Center Equipment Load Trends
AC and Refrigeration Principles and Planning
AC and Refrigeration Equipment Data Center Thermal Guidelines
Fundamentals: Electricity Data Center Liquid Cooling
Mark C. Hegberg: Mark Hegberg has spent major portions of his career involved in direct
digital control system application for Tour & Andersson, consulting engineering for A. Epstein
and Sons and others, and hydronic training and product development for ITT Bell & Gossett. In
recent years Mark has been significantly involved in hydronic system testing and balancing and
the issues associated with flow control in the field. Mark has chaired and been involved with sev-
eral ASHRAE committees and is a recipient of the ASHRAE Exceptional Service Award. He too
has been associated with the Illinois Chapter HVAC training class for many years.
t
Richard (Dick) A. Hegberg, PE: Dick Hegberg spent the majority of his career in the area of
temperature control application for Powers Regulator company (now Siemens Building Technol-
e
ogy), where he lead their early application engineering efforts and development of their building
automation systems. In particular, he devoted much time to valves and dampers. Dick was actively
involved in several ASHRAE TCs regarding controls and hydronics, as well as other groups that
used both as part of their processes. Dick is an ASHRAE Life Member, Fellow, and recipient of
gp
the Distinguished Service Award. Locally, he inaugurated and directed for over thirty years the
Illinois Chapter HVAC training class which is taught to this day.
pE
u
ro
G
Table of Contents
Preface . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . viii
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix
Chapter 1: Water System Design Concepts . . . . . . . . . . . . . . . . . . . . 1
Introductory Concepts. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
t
Basic System Components. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
e
Heat Transfer in Hydronic Systems . . . . . . . . . . . . . . . . . . . . . . . . 15
Load Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
Chapter 2: Piping System Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
gp
Basic Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31
Design Philosophy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34
Sizing Piping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
Flow-Rate Measurement . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
Chapter 3: Pipe Materials and Fittings . . . . . . . . . . . . . . . . . . . . . . . . 59
Pipe Materials . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 73
pE
Valves and Fittings . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Backflow Prevention Devices . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
Pipe Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Chapter 4: Centrifugal Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Types of Pumps. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
Operation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 95
Pump Selection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
u
System Design Considerations . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
Variable-Speed Pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127
Chapter 5: Terminal Unit Performance and Control . . . . . . . . . . . 139
ro
Types of Terminals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Performance and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 142
System Control Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . 150
System Control Configurations . . . . . . . . . . . . . . . . . . . . . . . . . . 159
Chapter 6: Expansion Tanks and Air Elimination . . . . . . . . . . . . . . 171
G
Open and Closed Water Systems . . . . . . . . . . . . . . . . . . . . . . . . 171

Hydronic Accessories . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
Sizing Expansion Tanks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 184
Chapter 7: Piping System Development . . . . . . . . . . . . . . . . . . . . . 193
Piping System Design. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 193
Direct-Return Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 197
Reverse-Return Analysis. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Primary/Secondary Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 215
Types of Pumps and Valves . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Primary/Secondary Application Study . . . . . . . . . . . . . . . . . . . . . . 221
Antifreeze Solutions for Low-Temperature Applications . . . . . . . . 226
Pumping Design Factors. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 229
Chapter 8: Matching Pumps to Systems . . . . . . . . . . . . . . . . . . . . . 235
t
Matching the Pump to the System . . . . . . . . . . . . . . . . . . . . . . . . 235
Parallel Pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 247
e
Series Pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
Standby Pumps . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 255
Trimming Pump Impellers and Adjusting Pump Speed . . . . . . . . . 255
gp
Two-Speed Pumping. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 258
Variable-Speed Pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 261
Source Distribution Pumping . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
Chapter 9: Water Chillers and Load Control . . . . . . . . . . . . . . . . . 273
Basic Water Chiller Components . . . . . . . . . . . . . . . . . . . . . . . . . 273
Refrigeration Cycle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 274
pE
Heat Transfer Chiller . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 277
Refrigeration Power. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Chiller Types and Control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
Chiller Piping Arrangements. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
Chiller Energy Performance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 295
Thermal Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
Skill Development Exercises
u
ro
G
Preface
Fundamentals of Water System Design, Second Edition is an entry level

text for the principles and fundamentals of water, or hydronic, HVAC system
design and calculations. It uses the ASHRAE Handbook, and its related chap-
t
ters on hydronics, as well as various research projects, ASHRAE documents,
and articles on hydronics. The text represents the fundamentals of design and is
e
an equally sound footing for work in commissioning, operation, and trouble-
shooting. The manual has been evolving over thirty years under the sponsor-
ship of ASHRAE and is supported by ASHRAE Technical Committee (TC)
6.1, Hydronic and Steam System Design. For the majority of its history, this
gp
writing effort was lead by Richard (Dick) A. Hegberg.
Hydronic systems are a key foundation technology of HVAC system
design. Hydronics provides an important tool in the designer’s toolbox of pro-
viding effective and efficient HVAC systems to buildings around the world.
One can argue exactly when hydronics started to be used for heating buildings;
however, the modern application of hot-water systems may have started in the
1870s in the United States, and the technology has been refined ever since. Of
the many benefits that water plays as a heat transfer medium for heating or
pE
cooling, fundamentally important is the ability to transport a lot of energy in
the small area of a pipe (especially when compared to the equivalent size of an
air duct). That benefit allowed for the development of central systems, elimi-
nating the necessity of a local heating production system for each heated zone.
The advancement allowed us to exercise individuality in our own building zone
comfort control and lead to widespread application of hydronic systems in
HVAC. From the early days of gravity circulation hot-water heating systems to
the most complex of district energy chilled-water systems that are used today,
u
the fundamentals presented here can be employed to successfully design the
modern hydronic system.
ro
G
Acknowledgments
ASHRAE TC 6.1 and its members have played a key role in collecting,
researching, and communicating the details and experiences for the successful
application of hydronic systems over the years. TC 6.1 is an ad-hoc collection
of individuals interested in hydronic system technology who take the time to
t
volunteer and attend meetings and then communicate what they have learned to
the other 53,000 members of ASHRAE.
e
Key in the educational organizing efforts of TC 6.1 in hydronic systems
and ASHRAE, a few individuals stand out when this text and the general con-
cept of the ASHRAE’s professional development series (be they Self Directed
gp
Learning manuals or the face-to-face version of the Professional Development
Seminar) are considered. At the apex of this is William (Bill) J. Coad, former
ASHRAE President and longtime member of several ASHRAE committees.
When the professional development series was first developed, Bill Coad, J.
Barry Graham, and Gerry Williams developed the first two classes on Air Sys-
tem Design and Energy Efficient HVAC. Bill’s significant involvement in lead-
ing many ASHRAE activities left little time to develop the Water Systems
manual, which lead to Dick Hegberg taking up the charge. All of these volun-
pE
teer member contributors are owed a deep debt of gratitude for their devotion
to helping ASHRAE develop the initial active training efforts that have evolved
into the many practical training programs that are available today.
ASHRAE has, from its very inception as ASHVE, represented a unique
cooperative collaboration between all aspects of the HVAC industry. Design-
ers, owners, operators, academicians, and manufacturers have all met and
debated the merits of all things HVAC&R for over 100 years, seeking the
answers to the problems associated with the HVAC system challenges of the
day and providing the fundamental knowledge required by all professionals in
u
a noncommercial manner. In that regard, while there are many professionals
from the manufacturing or “product” side of the business that have contributed
to advancing the area of hydronics, we would be remiss to not mention the sig-
nificant contributions of Gil Carlson of Bell & Gossett. As hydronics shifted
ro
from steam heating system application to hot- and chilled-water systems in the
1950s and 1960s and then rolled into the early days of the first energy crisis of
the 1970s, Gil’s application engineering, research expertise, and ability to com-
municate was an invaluable resource to the industry and ASHRAE TC 6.1. His
significant ASHRAE research and journal articles on hydronic application, as
well as the simple design tools he developed, were the educational precursor to
G
what later became the professional development series. The industry is

indebted to Gil’s contributions and to Bell & Gossett, now a division of Xylem
Corporation, for their support to this day of the fundamentals of application
engineering education in the area of hydronics.
On a personal note, thanks to all those within ASHRAE who, for no other
reason than their commitment to the spirit of mentoring that ASHRAE repre-
sents, helped provide me the practical education that comes from beyond col-
lege corridors. I thank you all and in particular Dick Hegberg, Bill
Coad, and Roy Ahlgren. I have had the great fortune to come to know
many of the ASHRAE members that contributed to this manual over
the years, successfully designing and operating systems even before the
rise of computers and spreadsheets. Amazingly, we can still employ
those fundamentals to the same end today, keeping in mind the guid-
ance of Carlson, “A difference to be a difference must be a big enough
difference to make a difference,” and Coad paraphrasing Einstein, “The
t
secret is to keep it simple… but not too simple.”
e
Mark Hegberg
gp
upE
ro
G
Water System
Design Concepts
e t
Study Objectives
gp
After studying the material in this chapter, you should be able to do the following:
❑ know what determines the load;

❑ understand the difference between closed and open systems;
❑ know the components of a hydronic system;
❑ understand heating versus cooling source devices;
❑ understand how systems meet part-load conditions;
pE
❑ identify temperature and pressure ranges for low-, medium-, and high-
temperature water systems;
❑ know what sensible, latent, and total heat loads are and how they affect
design water flow;
❑ identify examples of heating and cooling load devices; and
❑ know how load diversity suggests a reduction in the total cooling capacity
required.
u
Instructions
Read Chapter 1, and answer all of the questions at the end.
ro
Introductory Concepts
Water system design depends on the designer’s ability to evaluate the space
loads, occupancy patterns, and indoor environment requirements. This chapter
examines the actual process of water system design and provides information
G
on how to evaluate space loads. It also provides strategies and formulas for
mastering the key requirements for water systems.
Water systems that convey heat to or from a conditioned space or process
with hot or chilled water are frequently called hydronic systems. In general,
these systems employ centrifugal pumps to force water flow from a heating
or a cooling source to the conditioned space or load by means of various pip-
ing, pumping, control, and terminal arrangements (ASHRAE 2012).
2 Chapter 1 Water System Design Concepts
Given the design requirements, it is the designer’s task to evaluate the

space loads resulting from building construction, weather distribution, occu-
pancy patterns, indoor environment requirements, and other internal loads to
determine the total load (Figure 1-1) subject to the local building codes. The
loads include transmission, solar radiation, infiltration, ventilation air, peo-
ple, lights, power, appliances, and materials in and out (Sauer and Howell
2013).
The historical weather distribution for the project location is important, and
t
a means to control the systems at part-load conditions to maintain comfort con-
ditions for the occupants must be studied for proper design. The designer must
e
weigh the cost of the source utilities available and also the efficiencies of boil-
ers and chillers or other hydronic sources to determine the most efficient sys-
tem design (Figure 1-2). The system must be able to operate between part-load
gp
and full-load conditions. In many cases, the hydronic system is a support sys-
tem providing the heating or cooling medium for heat transfer equipment in an
air distribution system. This course is intended to acquaint the student with the
various hydronic principles and practices available for consideration in a proj-
ect’s design concept.
Figure 1-3 shows the basic components of a hydronic system that the
designer must define for an HVAC system, namely, a source of heating or cool-
ing, a distribution system, and the load components. There are different classi-
pE
fications of hydronic systems; the most common types are summarized below.
As you familiarize yourself with the various systems, remember that different
factors may come into play for each.
Water systems may be closed or open types. The fundamental difference
between them is the interface of the water with a compressible gas (such as
air) or an elastic surface (such as a diaphragm). A closed water system is
u
ro
G
Figure 1-1 Source/load.

Fundamentals of Water System Design I-P 3
e t
gp
Figure 1-2 Source–distribution–load.
u pE
ro
G
Figure 1-3 Source–distribution–part-load.
defined as one with no more than one point of interface with a compressible
gas (air) or surface (Figure 1-4). This definition is fundamental to under-
standing the hydraulic dynamics of these systems (to be discussed in the sec-
tion Expansion Chamber).
An open system has more than one such interface. For example, a cooling
tower has at least two points of interface: the tower basin and the discharge
pipe or nozzles entering the tower.
Comparing Figure 1-4 and Figure 1-5, the differences between the hydrau-
lics of the systems become evident as one analyzes the two systems. However,
one major difference is that certain hydraulic characteristics of open systems
do not occur in closed systems. For example, in a closed system,
t
• flow cannot be motivated by static head difference,
• pumps do not provide static lift, and
e
• the entire piping system is always filled with water.
These factors affect the installation and operating costs of the system over
gp
its service life.
Closed water systems are classified by operating temperature:
• Low-temperature water (LTW) system. This hydronic heating system

operates within the pressure and temperature limits of the ASME Boiler
and Pressure Vessel Code for low-pressure boilers (Figure 1-6). The maxi-
mum allowable working pressure for low-pressure boilers is 160 psig, with
pE
a maximum temperature of 250°F. The usual maximum working pressure
for LTW boiler systems is 30 psig; however, boilers specifically designed,
tested, and stamped for higher pressures are frequently used. Steam-to-
water and water-to-water heat exchangers are also used for heating LTW.
u
ro
G
Figure 1-4 Hydronic system fundamentals (closed system).

• Medium-temperature water (MTW) system. This hydronic heating sys-

tem operates at temperatures between 250°F and 350°F, with pressures not
exceeding 160 psig (Figure 1-7). The design supply water temperature is
approximately 250°F to 325°F, with a pressure rating of 150 psig for boil-
ers and equipment.
• High-temperature water (HTW) system. This hydronic heating system
operates at temperatures over 350°F, with pressures not exceeding 300 psig.
t
The maximum design supply water temperature is about 400°F, with a pres-
sure rating for boilers and equipment of 300 psig (Figure 1-7). The pressure/
e
temperature rating of each component should be checked for compliance
with the system’s design versus the manufacturer’s rating.
gp
u pE
ro
G
Figure 1-5 Cooling tower (open system).

e t
gp
pE
Figure 1-6 Low-temperature water system—direct or reverse return.
u
ro
G
Figure 1-7 Medium- or high-temperature water system.

• Chilled-water (CHW) system. A hydronic cooling system normally oper-

ates with a design supply water temperature of 40°F to 55°F (usually 44°F
or 45°F) with a pressure rating of 120 psig. Figure 1-8 shows a small- to
medium-sized system with constant-speed pumping using three-way valves
to ensure constant flow in the chiller source and balancing valves on each
load for flow measurement and adjustment. Larger systems may employ
two-way control valves and different chiller piping and pumping arrange-
t
ments to reduce pumping power.
• Condenser water (CW) system. Open water systems are typically used in
e
refrigeration CW systems as once-through or cooling tower systems. Fig-
ure 1-9 shows a water-cooled condenser using city, well, or river water. The
return is run higher than the condenser so that the condenser is always full
gp
of water. Water flow through the condenser is modulated by a control valve
in the supply line. This is usually actuated by a condenser head pressure to
maintain a constant condensing temperature with load variations.
Figure 1-10 shows two cooling tower applications to protect against low
outdoor temperature conditions. Water flows to the pump from the tower basin,
and the level should be above the top of the pump casing for positive prime,
pE
and piping pressure drop should be minimized.
u
ro
G
Figure 1-8 Chilled-water system—direct return piping.

e t
Figure 1-9
gp
Condenser open water system (once through).
u pE
ro
G
Figure 1-10 Condenser cooling tower system inside reservoir or heated sump.
Antifreeze or brine solutions may be used for applications (process appli-

cations) that require temperatures below 40°F or for coil freeze protection.
Well water systems can use supply temperatures of 60°F or higher.
In addition, there are other forms of system classification that tend to blur
the lines between system operating temperatures, pipe layout styles, and/or a
reference to the operated system style.
Four-pipe dual temperature water systems are one such example of this
t
blur. This older example of system design combines the heating and cooling
system, circulating hot and/or chilled water through separate supply and
e
return pipes to common terminal coils. Figure 1-11 shows a very simplified
piping diagram for this type of system, which would not normally be piped
this way. This system operates within the pressure and temperature limits of
gp
LTW systems, with usual winter design supply water temperatures of about
100°F to 150°F and summer supply water temperatures of 40°F to 45°F. A
reading of ANSI/ASHRAE/IES Standard 90.1 (ASHRAE 2013) can suggest
that this type of system (especially as shown) is not allowed due to energy
performance issues associated with potential mixing of hot or cold condi-
tioned water. Of course, given enough thought and control logic, this issue
may be overcome.
pE
A more common application is to have a two-pipe dual-temperature sys-
tem where a common piping system serves common loads with both hot and
chilled water supplied seasonally to the system. In either case, some of the
issues a designer must deal with start to show themselves so that the system
operates properly. Think through the components and issues like the
designer!
u
ro
G
Figure 1-11 Distribution orientation.

The design should consider source protection to prevent temperature

shocks to the chiller or boiler on cycle changeover. The designer might con-
sider timing mechanisms, temperature mechanisms, or combinations of both so
that the two sources never operate together and that neither source gets water
that is too hot or too cold.
Think about heat transfer. In general, a hot-water heating coil with its higher
operating-water temperatures and differential temperatures as compared to CHW
requires a much smaller heat transfer coil. The water, on the other hand, does not
t
know or care about this, it just acts as it does because there is heat transfer sur-
face area. Therefore, it performs to whatever capabilities are inherently a part of
e
that seasonal operation. The designer then has to carefully consider the design
temperature differentials and flow rates associated with the coil selection and
whether the sources are capable of handling the developed differentials.
gp
Consider how the coil will be controlled. For example, will flow be propor-
tionally modulated or be positioned for either full or no flow?
Consider whether seasonal changes in operating flow will affect the water flow
velocity, possibly impacting air control and pressure management in the system.
There are other issues to be considered, too. This manual teaches a frame-
work of issues to be dealt with in design that can be adapted to deal with any
type of system.
pE
Basic System Components
Figure 1-12 shows the fundamental components of a closed hydronic system.
Actual systems generally have additional components (such as valves, vents,
etc.), but these are not essential to the basic principles underlying the concept
of the system. These fundamental components are as follows:
u
ro
G
Figure 1-12 Closed hydronic system fundamental components.

• Source system
• Load system
• Pump system
• Distribution system
• Expansion chamber
Source
e t
The source is the point where heat is added in a heating system or removed
from a cooling system. Ideally, the amount of energy entering or leaving the
source equals the amount entering or leaving through the load system. Under
steady-state conditions, the load energy and source energy are equal and oppo-
gp
site. In reality, energy conversion and/or transfer is not perfect and the source
has an efficiency of less than 100%.
Each type of source has its own efficiency characteristics as a function of
load. You must consider this source efficiency in the system design process.
Any device that can be used to heat or cool water under controlled conditions
can be used as a source device. Sources typically function in one of two ways:
pE
• By converting chemical, electrical, or solar energy to heat, which is then
transferred to water in the system
• By transferring heat from one system to another
The most common source devices for heating and cooling systems are the
following:
• Heating-source devices: Hot-water generator or boiler, steam-to-water

u
heat exchanger (Figure 1-13), water-to-water heat exchanger (Figure 1-13),
solar collector or panels, heat recovery or salvage heat device, exhaust gas
heat exchanger, incinerator heat exchanger, heat pump condenser, and air-
to-water heat exchanger
ro
• Cooling-source devices: Electric compression chiller (Figure 1-14), thermal

absorption chiller (Figure 1-14), heat pump evaporator, air-to-water heat
exchanger, and water-to-water heat exchanger
A typical large CHW system with multiple chillers, various load controls, and
compound pumping is shown in Figure 1-14. This system provides variable flow,
G
constant supply temperature CHW, multiple chillers, two-way valve control, and
the advantage of adding CHW storage. One design issue shown is the placement
of the common pipe for the chillers. With the common pipe located at the oppo-
site end of the chiller production section, the chillers will unload from right to
left. With the common pipe in the alternative location (between the CHW produc-
tion and the loads), the chillers will load and unload equally in proportion to their
capacity (see Chapter 8 for further discussion).
e t
Figure 1-13
gp
Steam-to-water or water-to-water heat exchanger.
u pE
ro
Figure 1-14 Multiple chiller variable-flow chilled-water system.

G
The two primary considerations in selecting a source device are the design
capacity and the part-load capability, which, when combined, define the turn-
down ratio. The turndown ratio, expressed in percent of design capacity, is
Minimum Capacity
Turndown Ratio (%) = 100 ---------------------------------------------- (1-1)
Design Capacity
The reciprocal of the turndown ratio is sometimes used. For example, a

turndown ratio of 25% may also be expressed as a turndown ratio of four.
The turndown ratio has a significant effect on the successful performance of
a system, and lack of consideration for this capability of the source system has
been responsible for many systems that do not function properly or that exces-
sively consume energy. The turndown ratio has a significant impact on the ulti-
mate system design selection because operating efficiencies tend to decrease as
the turndown ratio decreases.
t
Generally the larger the boiler or chiller on a single unit basis, the more
e
difficult it is to achieve acceptable operation efficiencies and therefore accept-
able costs for the low-load portion of its seasonal operation. Seasonal opera-
tion at less than 50% load can, in the case of heating, be over 50% of its total
duty hours; in northern climates, a similar ratio may occur during cooling sea-
gp
sons. This presents the designer with a dilemma. You must specify a system
that meets the maximum requirements, but in doing that the system may oper-
ate ineffectively for most of its duty season. One approach to solving this
dilemma is by using multiple sources of lower capacity (Figure 1-14).
Another design consideration is the diversity of the cooling load. Carrier
(1965) states
pE
diversity of cooling load results from the probable non-occurrence of
part of the cooling load on a design day. Diversity factors are applied
to the refrigeration capacity in large air-conditioning systems. These
factors vary with location, type, and size of the application, and are
based entirely on the judgment of the engineer.
The diversity factor, as a ratio of actual load/design load, can be applied

to people and lighting loads in large multistory office, hotel, or apartment
buildings. For example, in an office building the diversity factor for the
u
refrigeration capacity due to people may be from 0.75 to 0.90 and due to the
lighting from 0.70 to 0.85. In addition, the design engineer must consider
the storage load factors for heat gain due to lighting and solar gains by
glass.
ro
Specific methods of load control for CHW plants are discussed in

Chapter 9.
Load
G
The load is the point where heat flows out of or into the system from the
space or process; it is the independent variable to which the remainder of the
system must respond. Outward heat flow characterizes a heating system, and
inward heat flow characterizes a cooling system. The quantity of heating or
cooling is calculated by one of the means discussed in the following section.
Typically, loads are a form of heat exchanger, either air to water or water to
water.
Pump System
If we consider the modern era of hydronic system application to be the last
100 or so years, historically the first hydronic systems were heating systems,
and they did not apply a pump to the system to circulate water. These systems
were gravity fed, relying on the difference in head caused in the system by
water being heated in a boiler and cooled in a radiator and pipes. These were
t
exceptionally simple systems, circulating water with differential pressures (less
than a few inches of pressure). They tended to be small in size and used a large
e
pipe (2 to 3 in.) to circulate flows less than 1 gpm. These systems were slow to
react and operation-sensitive to how well the pipe was sized. There was no real
control short of turning the flow on or off. The first recorded use of a pump in a
heating system appeared in 1926 in the Transactions of ASHVE, forerunner to
gp
ASHRAE (Giesecke 1926). Giesecke, a researcher and professor at Texas
A&M, delivered a paper on testing pipe fittings, and an engineer from Toronto,
H.H. Angus, commented that they had solved flow problems in gravity heating
systems by applying a small pump to boost the water flow. A new era of
hydronic system design dawned and a term that is still used today, booster
pump, came into use. Pumps create flow by converting captured centrifugal
force into a higher pressure of water. In essence, the pump is a differential pres-
sure machine with the flow of water being created as the pressure moves to a
pE
point of lower pressure.
Distribution System
The distribution system in a hydronic system is a series of pipes, transporting
water to the location of the loads where it is required.
u
Expansion Chamber
Recognizing that water is transported in rigid pipes and that water expands and
ro
contracts as it is heated and cooled, water must have a place to expand into to
prevent damage to the distribution system. The expansion chamber is this
place, and there are two types of chambers that are commonly applied. By
using this device, the pressure of the water in the system can be controlled and
managed so that useful work is done with the system without damage.
G
Design Trade-Offs
Remember, improved efficiency comes with an initial installation cost penalty.
As the designer, you must work out the acceptable trade-off in the initial instal-
lation cost that a customer will accept for reduced operational expenses over
the life of the system. This will be easier to accomplish when backup system
requirements are considered in the decision.
System Temperatures
As the designer, you must design temperatures and temperature ranges by con-
sidering the performance and economics of the components. For example, for a
cooling system that must maintain 50% rh at 75°F (Figure 1-15), the dew-point
temperature is 55°F. This sets the maximum return water temperature near
55°F (60°F maximum); the lowest practical temperature for refrigeration, con-
sidering the freezing point and the economics, is about 40°F. This temperature
t
spread then sets constraints for a CHW system.
For a heating system, the maximum hot-water temperature is normally
e
established by the ASME Boiler and Pressure Vessel Code (2015) as 250°F,
and, with space temperature requirements of little over 75°F, the actual operat-
ing supply temperatures and the temperature ranges are set by the design of the
gp
load devices. Most economic considerations relating to the distribution and
pumping systems favor the use of the maximum possible temperature range t.
Heat Transfer in Hydronic Systems

Sensible Heating or Cooling of Air
The quantity of heat entering or leaving the airstream is expressed by
pE
q = 60 a cp Qa t (1-2)
where
cp = specific heat of air, Btu/lb·°F
q = heat transfer rate, Btu/h
u
ro
G
Figure 1-15 Psychrometric chart chilled-water example.

Qa = airflow rate, cfm

t = air temperature change, °F
a = density of air, lb/ft3
For standard air with a density of 0.075 lb/ft3 and a specific heat of
0.24 Btu/lb·°F, this equation becomes
60 min 0.24 Btu 0.075 lb Qa ft3 t°F
q Btu/h =
t
1h lb·°F ft3 min
e
q Btu/h = 1.08 Q a cfm t (1-3)
gp
Example 1-1
In the system shown in Figure 1-16, the air upstream of the heating coil is
60°F and the air temperature leaving the heating coil is 130°F. Given that cp =
0.24 Btu/lb·°F, Qa = 5000 cfm, and the density is 0.075 lb/ft3, calculate the heat
transfer rate of the system.
pE
q = 1.08 Qa t
= 1.08 Btu/h·cfm·°F
q = (1.08)(5000 cfm)(130°F – 60°F)
q = 378,000 Btu/h
Solution The heat exchanger or coil must then transfer this heat to the water. The quan-
tity of sensible heat transferred to the heated or cooled medium in a specific heat
exchanger is a function of the surface area, the mean temperature difference
u
between the water and the medium, and the overall heat transfer coefficient, which
is a function of the fluid velocities, properties of the medium, geometry of the heat
transfer surfaces, and other factors (Figure 1-16). It may be expressed by
ro
G
Figure 1-16 Example system with heating coil.

q = UA LMTD (1-4)
where
A = surface area, ft2
LMTD = logarithmic mean temperature difference, heated medium to
water, °F
Q = heat transfer rate, Btu/h
t
U = overall coefficient of heat transfer, Btu/h·ft2·°F
e
Example 1-2
gp
In Figure 1-17, assume that the coil has a U of 155 Btu/h·ft2·°F per row.
The upstream temperature is 60°F, and the air temperature leaving the coil is
130°F. Water enters the heating coil at 160°F and leaves at 140°F. The duct
size is 48 × 36 in. and the coil has four rows. Find the heat transfer rate (see
also Figure 1-18).
Solution q = UA LMTD
pE
t max – t min t max = 140 – 60 = 80
LMTD = --------------------------------
t max t min = 160 – 30 = 30
Ln -------------
t min
80 – 30- = 51 F
LMTD = -----------------
u
80
Ln ------
30
Next, using LMTD, find q.

ro
G
Figure 1-17 Example system with heating coil.

e t
Figure 1-18 Coil LMTD example.
gp
pE
q = UA LMTD
2
Btu/h·ft · F 3 ft 4 ft 51 F 4 Rows
= 155 ----------------------------
-
Row
q = 379,440 Btu/h
Latent Cooling and Dehumidification of Air

u
The quantity of heat removed from the cooled medium (Figure 1-18) when
both sensible cooling and dehumidification are present is expressed by
ro
qt = W h (1-5)
where
qt = total heat transfer rate, Btu/h
W = mass flow rate of cooled medium, lb/h
G
h = enthalpy difference of entering and leaving conditions of cooled

medium, Btu/lb
Expressed for a cooling coil, this equation becomes
q t = 60Q a a h (1-6)
where
e t
Figure 1-19 Example system with cooling coil.
gp
Qa = airflow rate, cfm
a = density of air, lb/ft3
For standard air, the density is 0.075 lb/ft3 and the formula reduces to
q t = 4.5Q a h (1-7)
Example 1-3
pE
For the system shown in Figure 1-19, determine the heat transfer rate for the
sensible cooling/dehumidification process, assuming entering air is 82°F,
enthalpy is 38.5 Btu/lb, and discharge air is 52°F and saturated at enthalpy of
21.4 Btu/lb.
Solution qt = 4.5Qa h
= (4.5)(5000 cfm)(38.5 – 21.4) Btu/h
u
= 384,750 Btu/h
Heat Transferred to or from Water

ro
The quantity of heat transferred to or from the water is a function of the flow
rate, the specific heat, and the temperature drop or rise of the water as it passes
through the heat exchanger. The heat transferred to or from the water is
expressed by
G
q t = mc p t (1-8)
where
cp = specific heat of water, Btu/lb·°F
m = mass flow rate of water, lb/h
qt = total heat transfer rate, Btu/h
t = water temperature increase or decrease across unit, °F
These equations are also used to express the heat-carrying capacity of the
piping or distribution system of any portion of that system. In this regard, the
temperature differential, sometimes called the temperature range, is established
or identified. For any flow rate through the piping, qw is called the heat-carrying
capacity.
With water systems, it is common to express the flow rate in gallons per
minute (gpm), in which case the equation becomes
t
q t = 8.02 w c p Qw t (1-9)
where
e
Qw = water flow rate, gpm
w = density of water, lb/ft3
For standard conditions in which density is 62.4 lb/ft3 and specific heat is
gp
1 Btu/lb·°F, the equation becomes
q t = 500Q w t (1-10)
Equations 1-9 and 1-10 can be used to express the heat transfer across a
single load or source device or any quantity of such devices connected across a
piping system. In the design or diagnosis of a system, the load side may be bal-
anced with the source side by these equations.
pE
Example 1-4
For a single system similar to that shown in Figure 1-16, assume the heat trans-
fer rate across the coil is 378,000 Btu/h and the t of the water supplying the
coil is 20°F (water and air at standard conditions). Find the water flow rate
required for the system.
Solution q t = 500Q w t
u
qt
Q w = --------------------
-
500 t
ro
378,000 Btu/h
Q w = -------------------------------------------------------------------------
500 Btu·min/°F·h·gal 20 F
Q w = 37.8 gpm
Example 1-5
G
Assume qt (cooling coil) = 384,750 Btu/h and t = 12°F.

384,750 Btu/h
Solution Q w = --------------------------------------------------------------------------
500 Btu·min/ F·h·gal 12 F
Q w = 64.1 gpm
Load Systems
Load systems are the devices (terminal units) that convey heat from the water
for heating or to the water for cooling of the space or process. Most load sys-
tems are basically water-to-air finned-coil heat exchangers or water-to-water
heat exchangers. The specific configuration is usually used to describe the
device. Common configurations include the following:
t
• Heating load devices: Preheat coils in central air-handling units, heating
coils in central air-handling units, zone or central reheat coils, finned-tube
e
radiation, baseboard radiation, convectors, unit heaters, fan-coil units, water-
to-water heat exchangers, radiant heating panels, and snow-melting panels
• Cooling load devices: Coils in central units, fan-coil units (Figure 1-20),
gp
induction unit coils, radiant cooling panels, and water-to-water heat
exchangers
It is at this point that further consideration of the coil selection becomes

important. Figure 1-21 shows a representative coil. Typically, as air enters the
coil it passes over tubes carrying the water back to the source to be recondi-
tioned. Those tubes are arranged in a geometrical pattern selected to achieve
the required heat transfer and air and water pressure drop considerations. Tubes
pE
are connected to a supply or return header, which either supplies water to the
coil or directs it back to the source. The designer has a number of criteria that
can be adjusted in achieving the selection, albeit they are bounded by the heat
transfer characteristic and the psychrometric process the HVAC process needs
to achieve. While the application of the psychrometric chart has been previ-
u
ro
G
Figure 1-20 Fan-coil unit.

e t
Figure 1-21 Typical coil.
gp
pE
ously noted, it becomes useful in determining what to select for a heat transfer
device. Load calculation programs are often applied by designers for calcula-
tion of the required heating or cooling, which generates a gross value for the
required heating or cooling. While space and outdoor design conditions are
specified by the designer, these programs do not select the required heat trans-
fer components, because it is just too early in the deign process to do so. The
load calculation program does not have the capability to evaluate how many
air-handling units are required and what their respective sizes should be. The
u
designer’s link then is to take the load calculation data and apply it to the cor-
rect heat transfer device based on the design criteria of applied equipment. This
implies selecting physical equipment to perform the air-conditioning function.
Often getting the equipment to perform like the theoretical equation can be
ro
daunting.
For example, consider the determination of a cooling coil and its character-
istics for a CHW coil per the following criteria:
• Office space with a sensible cooling load of 42,000 Btu/h and a total latent
load of 18,000 Btu/h
G
• Space design conditions are 75°F/50% rh

• Outdoor design conditions are 95°F db/75°F wb
• Ventilation air is 30% of the total
• Supply air is delivered at 55°F
• 65% efficient draw-through blower at 5.0 in. total pressure
• 12°F design t CHW rise through the coil
In this example, we have to incorporate our basic knowledge of psychromet-

rics and load calculation but also account for the energy use as the unit operates,
represented by the fan energy causing a temperature rise. Similarly, we could
also take into account temperature rise for the airflow in the ductwork.
We know the following:
Sensible heating
t
Q
Q = 1.08 cfm t cfm = ----------------------
1.08 t
e
Fan heat energy balance
gp
Q = 1.08 cfm t
0.1175 cfm P IN
P WATTS = --------------------------------------------------
-
F
P BTU = P WATTS 3.412
Energy balance
pE
0.1175 3.412 cfm P IN
1.08 cfm t = ----------------------------------------------------------------------
-
F
(1-11)
0.1175 3.412 cfm P IN 0.371 P IN
t = ----------------------------------------------------------------------
- = -------------------------------
-
1.08 cfm F F
The process could be sketched psychrometrically per Figure 1-22.

u
From the psychrometric chart, the desired space conditions are fixed, as is
the design temperature for the supply air temperature. In order for the process
to work psychrometrically, the supply air condition must be along the sensible
heat ratio (SHR) line. Calculating SHR at 0.7, using either ASHRAE Psychro-
ro
metric Chart No. 1 (1992) compass or other means, a line can be drawn at the
SHR slope and through the space condition and the design supply air tempera-
ture, showing the entering point at about 55°F/80% rh (51.5°F wb).
Blower energy must be accounted for, which is a sensible rise calculated to
be about 3°F. This is also plotted and shows the coil leaving conditions to be
G
52°F/90% rh (about 50°F wb).
Sensible heating
Q = 1.08 cfm t cfm

Q 42,000
= ---------------------
- = --------------------------------------- 1950 cfm
1.08 t 1.08 75 – 55
24
G
ro
u
Chapter 1 Water System Design Concepts
pE
gp
Figure 1-22 Example process diagrammed on ASHRAE Psychrometric Chart.
e t
This allows the linear process line, sometimes called the grand sensible heat
ratio (GSHR), from the coil entering condition at the mixed-air point for 30%
outdoor air at 81°F/48% rh (66.5°F wb) through the leaving-air point. Extend
the graphic plot of the line so that it intersects with the saturation line of the
psychrometric chart; it must intersect for the process to work psychrometri-
cally. If it does not, the process must be modified to one that does. The inter-
section point is known as the apparatus dew point (ADP) and has a specific
function as it relates to coil selection—it helps set the entering CHW tempera-
t
ture for the coil. This is also a relative indicator of how well the process works.
If the entering water temperature is too low, again reevaluate the process.
e
Blower temperature rise
0.371 P IN 0.371 5 in.
t = -------------------------------
- = ------------------------------ = 2.9 F
gp
F 0.65
Blower sensible heat
Q FAN = 1.08 1480 2.9 F = 4600 Btu/h
The ADP is an average temperature representing the coil surface and is

what allows condensation for dehumidification to occur. The ADP in this
example is about 47°F (depending on either how good one’s eyesight is or how
pE
large the psychrometric chart has been printed!) The coil entering water tem-
perature (EWT) can be estimated using the following equation:
Coil EWT = 2 ADP – T WBLVG

(1-12)
Coil EWT = 2 47 F – 50 F = 44 F
It appears plotted on the ASHRAE Psychrometric Chart in Figure 1-22.
Coil selection does not end with the establishment of the selection data;
u
this merely gives the required data to search out the heat transfer device from a
manufacturer. In that regard, most manufacturers will provide the selection ser-
vice or provide complimentary software that will find either a single coil or a
series of coils that provide varying levels of performance. An example of a man-
ro
ufacturer’s coil selection program is shown in Figure 1-23.

Coil selection is an iteration process, and rarely does a real product provide
exactly the desired psychrometric process as specified by the designer. The only
datum not established for selection is that of the required flow rate, and to do that
the designer must set design criteria for the water-side temperature difference for
the coil. The ADP calculation established an entering temperature of 44°F. For the
G
example, we will use a moderate t for design of 12°F, implying a leaving water
temperature of 56°F, just as, previously, design flow rate is calculated as follows:
Q COIL = 4.5 cfm h = 4.5 1950 31.2 – 21 = 89,500 Btu/h
Q COIL 89,500
q gpm = ------------------------
- = --------------------- = 15 gpm
500 tW 500 12
A variety of options are up to the designer in coil selection. Typical tube

sizes are 0.375, 0.5, and 0.625 in. in water coils. The quantity of fins per inch
(FPI) of tube has a large effect on overall heat transfer as well as dehumidifica-
tion. In that regard, the depth of the coil generally referred to as the rows with
the quantity of tubes is equally important.
All air that flows through the coil does not come into physical contact with the
metal surfaces of the coil. The air that does not is called bypass air. Getting a coil to
t
provide the required heat transfer requires playing with all of the dimensions that
affect the surface area exposed to the air as it passes through the coil as well as giv-
e
ing the air enough time to come into contact with that metal. For CHW coils, air
velocities are normally limited to a maximum of 500 fpm to reduce the potential for
condensate carryover in the airstream. Velocity can be lower and often is to achieve
the required latent energy removal. Table 1-1 provides guidance on coil depth and
gp
velocity to achieve dehumidification. In addition, carefully consider the quantity of
FPI of the coil. While packing in the fins increases heat transfer and reduces coil
pE
u
ro
Figure 1-23 Example of manufacturer’s coil selection program.
Table 1-1 Coil Depth and Velocity to Achieve Dehumidification

Coil Load, Face Velocity,
G
Coil Rows
cfm/ton fpm
600 500 3
500 500 4
400 400 4
300 300 6
200 200 8
size, it may degrade latent energy transfer and become a maintenance problem, col-
lecting dirt on the fin surface. (Remember, they get wet!) Generally, try to limit fin
spacing to a maximum of 8 FPI in cooling and dehumidification applications. It
may not always be possible, but it is a good starting point for evaluation.
Note that in this selection, we are very close to the psychrometric process
that we plotted on the chart. We are not exactly at 89,500 Btu/h, but we are
very close. Similarly, we are at a little less than the required 52°F discharge
temperature and similarly at slightly less than the ideal water-side T. All
t
things considered though, this is not a bad first pass at selection. Air-side pres-
sure drop is a reasonable 0.4 in., and water-side pressure drop is a very respect-
e
able 1 ft. Note though that it is 8 tubes deep by 18 tubes high, or 144 copper
tubes. Coil cost is all about the quantity of metal and the labor to connect
everything together, so this might be a pricey coil selection (only your coil pro-
gp
vider knows for sure), and you may only know by selecting a few comparable
coils and asking for estimates.
Less easy to get will be a coil characteristic (as portrayed in Figure 1-24),
which either will require data from the manufacturer for your calculation or may
be given to you by them. This is a helpful graph in determining control strategy
u pE
ro
G
Figure 1-24 Heat transfer versus water flow for a coil.

and part-load performance. It is a normalized graph (percentage heat transfer

versus percentage water flow) so that it is not influenced by the magnitude of
heat transfer (Btu/h versus flow), and generally it is shown in this form holding
the entering air conditions to the coil as fixed.
Generally, the characteristic plots retain this type of shape. It is the sensible
heat curve that is important to the control valve selection and performance,
with the shape typically being the influence in selection of equal percentage
valve characteristics, as is discussed in Chapter 5. Recall that this is generated
t
with a fixed entering condition, and if the same type of evaluation is performed
where specific entering conditions at specific loads are examined, one might
e
find a wide range of shapes and required flow rates. Often, it makes the shape
more linear, but that is not always the case. Similarly, note how dehumidifica-
tion starts at about 30% flow to the coil. That take-off point too can vary based
on the geometry of the selected coil. Typically, with less depth or higher air-
gp
flow velocities, that starting point happens to the right, at larger flow rates. In
this particular coil selection (which is slightly different from the previous
example), a water-side t of 20°F was used, and note that as flow decreases to
the coil, t rises, as would be expected for any throttled condition. This needs
to be checked against the chiller to make sure that entering water to the chiller
does not have a maximum criteria.
Another observation of this coil characteristic is that it is for throttled flow.
pE
Were the coil pumped, the entering water temperature would be controlled by
throttling a percentage of the chilled water being circulated to the coil. That
makes the coil characteristic more linear in performance with respect to the
percentage of chilled water provided to the coil. Why do such a thing? Specifi-
cally, when there may be concern about freezing a coil in situations where that
can arise or in situations where as a designer you always want to provide the
full latent energy (dehumidification) heat transfer effect of the coil. Options
such as pumping the coil and changing the heat transfer characteristic are some
u
of the great tools that chilled-water systems provide in air-conditioning system
design.
The Next Step

ro
In Chapter 2, you will be introduced to piping system design.
Summary
G
Chapter 1 covered the following topics:
• What determines the load

• The difference between closed and open systems
• Components of a hydronic system
• Heating versus cooling source devices
• How systems need to meet part-load conditions
• Temperature and pressure ranges for low-, medium-, and high-temperature

water systems
• Sensible, latent, and total heat loads and how they affect design water flow
• Examples of heating and cooling load devices
• How load diversity suggests a reduction in total cooling capacity required
References
e t
ASHRAE. 1992. ASHRAE Psychrometric Chart No. 1. Atlanta: ASHRAE
ASHRAE. 2012. ASHRAE Handbook—HVAC Systems and Equipment.
Atlanta: ASHRAE.
ASHRAE. 2013. ANSI/ASHRAE/IES Standard 90.1-2013, Energy Standard
gp
for Buildings Except Low-Rise Residential Buildings. Atlanta: ASHRAE.
ASME. 2015. Boiler and Pressure Vessel Code. New York: ASME.
Carrier. 1965. Handbook of Air Conditioning System Design, Chapter 3. New
York, NY: McGraw-Hill.
Giesecke, F.E. 1926. Friction of water in elbows. ASHRAE Transactions
32:303–20. New York: ASHRAE.
Sauer, H., and R. Howell. 2013. Principles of Heating, Ventilating and Air-
Conditioning, Chapter 5. Atlanta: ASHRAE.
u pE
ro
G
Skill Development Exercises for Chapter 1

Complete these questions by writing your answers on the worksheets at the back of this book.
1-1 Water systems that convey heat to or from a conditioned space or process with
hot or chilled water are frequently called _____________________________.
t
1-2 What is the fundamental difference between closed and open types of water
systems?
e
1-3 A cooling tower has at least two points of interface with air. Where are they?
1-4 What is the maximum working pressure for LTW boiler systems?
gp
1-5 What is a CHW system? How is it different from a CW system?
1-6 What are the fundamental components of a closed hydronic system?
1-7 Explain the most common source devices for heating and cooling systems.
1-8 Explain what load means.
1-9 What factors influence the heating and cooling load requirements?
pE
1-10 Define sensible heat transfer.
1-11 Name five heating load devices and describe how each is used in system appli-
cations.
u
ro
G
Piping System
Design
e t
Study Objectives
gp
After completing this chapter, you should be able to do the following:
❑ understand Bernoulli’s principle,

❑ know the three steps in design of a fluid distribution system,
❑ understand the difference between direct-return and reverse-return piping
and if they can be combined,
❑ know methods to allow thermal expansion,
pE
❑ determine pressure drop in piping,
❑ know the difference between laminar and turbulent flow and what index
quantifies it,
❑ understand piping roughness factors, and
❑ know what governs pressure drop in a piping system.
Instructions
u
Read Chapter 2 and answer all of the questions at the end.
Basic Considerations
ro
The piping system is a key component of the distribution system, and good
design practice can significantly affect the performance and energy efficiency of
an HVAC system. This chapter discusses the key aspects of piping system
design, including pipe sizing and system design philosophy. In the design of any
fluid distribution system, you must consider the following steps (Coad 1985):
G
• Establish the piping design philosophy and objectives.

• Size the pipes.
• Calculate or determine the pressure drop in the system as a whole or in var-
ious subparts or branches.
To achieve the best energy efficiency, you may need to repeat these steps
several times to optimize the design.
32 Chapter 2 Piping System Design
Pressure Drop
From an instructional standpoint, it is important to understand the concept of
pressure drop before discussing design philosophy and sizing. In general, to
direct a flow of water through a piping system, a pressure difference must be
created to overcome the friction head due to the piping length, type of fittings,
elevation changes, and pressure requirements at the receiving end.
t
In Hydrodynamica, Bernoulli (1738) analyzed the flow of water through a
piping system and theorized that it must obey the law of the conservation of
e
energy, where the energy can never be created or destroyed but only trans-
formed or directed in its flow (Figure 2-1). Many engineering texts on thermo-
dynamics and fluid mechanics have explained this concept and developed a
general energy equation for analyzing the fluid flow in a process. An energy
gp
balance is made by equating all of the energy entering the process to that leav-
ing plus the heat added or subtracted and the work done by or on the fluid per
unit of time. Bernoulli and Euler developed this concept into the well-known
Bernoulli equation for the flow of an incompressible liquid with the addition
of a term for head loss hL due to flow in the pipe (Euler 1750):
2 2
V P V P
Z 1 + ------1 + -----1- = Z 2 + ------2 + -----2- + h L (2-1)
pE
2g 1 2g 2
where
g = acceleration due to gravity, 32.2 ft/s2
hL = head loss in feet of fluid flowing
u
ro
G
Figure 2-1 Bernoulli’s theorem.

P1, P2 = pressure, lb/ft2

V1, V2 = velocity, ft/s
Z1, Z2 = elevation above or below datum, ft
1, 2 = density, lb/ft3
Figure 2-2 shows an example of a piping system where two gage readings
are taken, the elevation is measured, and the pipe size is the same for the enter-
t
ing and leaving conditions.
e
According to the Bernoulli theorem (Equation 2-1):
2 2
V P V P
Z 1 + ------1 + -----1- = Z 2 + ------2 + -----2- + h L
gp
2g 1 2g 2
Substituting values of pressure (in feet of fluid head) into the equation and
making sure units are consistent yields the following:
2 2 2 2
V1 100 lb/in. 144 in. /ft
0 + ------ + ----------------------------------------------------------------
pE
2g 62.4 lb/ft
3
2 2 2 2
V 30 lb/in. 144 in. /ft
= 100 + ------2 + ------------------------------------------------------------- + h L
2g 62.4 lb/ft
3
h L = 230.8 – 169.2 = 61.6 ft

u
ro
G
Figure 2-2 Bernoulli’s piping example.

So, we see a total head loss of 61.6 ft due to the piping and fitting friction and
the elevation head loss (assuming the same size pipe internal diameter at 1 and
2, V1 = V2).
Design Philosophy
In a direct-return system, the length of supply and return piping through the
subcircuits is unequal. This may cause unbalanced flow rates and requires
t
careful balancing to provide each subcircuit with design flow. Ideally, a
reverse-return system provides nearly equal total lengths for all terminal cir-
e
cuits. Will the design provide the most economical geometrical solution
through vertical and/or horizontal distribution?
Figure 2-3 shows a direct-return piping system, as compared to a reverse-
return piping system shown in Figure 2-4. As a designer, you must make sev-
gp
eral important decisions based on design philosophy before starting the design
process. Will this be a basic plan or a combination of direct and reverse sys-
tems (as shown in Figure 2-5)?
Design philosophy and objectives are most often overlooked by designers.
It is in this step that the why and how of the system are addressed, including
other considerations such as the following:
u pE
ro
G
Figure 2-3 Direct-return piping.

e t
gp
pE
Figure 2-4 Reverse-return piping.
u
ro
G
Figure 2-5 Combination of direct and reverse systems.

• Is the system to be constant flow? For example, are three-way valves

inserted to handle part-load conditions by reducing flow through the load
while maintaining nearly constant flow through the source? Or is the flow
through the load heat transfer coils to be constant while allowing the flow
through the source to vary?
• Will the system have intermittent flow, such as on/off control to start/stop a
pump for a zone or a load coil or, if this is a small system, to start/stop the
distribution pumping?
t
• Is variable flow being considered? For example, two-way valves vary the
flow in the load coils that result in variable flow in the source.
e
• Will the pump speeds be varied with the load? How will the variable sys-
tem flow affect the flow through the source? (Variable-speed concepts and
valve arrangements are discussed in Chapters 5 and 8.)
gp
• In the pressure distribution of a direct-return system (Figure 2-3), the avail-
able pressure drop for a load circuit is greatest near the system pump and
decreases the farther away the load is from the source pump. Care must be
taken by the designer to size and select the control valves to ensure ade-
quate flow distribution and proper close-off.
• In the pressure distribution of a reverse-return system (Figure 2-4), the pres-
sure drop for a load circuit is uniform (if the load pressure drops are similar)
pE
even as the distance is increased from the source pump. A key reason for the
reverse-return design is to assist the two-way control valve with a more uni-
form pressure drop. Selection of control valves must ensure adequate flow
and proper close-off, but this is not as critical as in the direct-return design.
• The designer should consider balancing valves because the control valve
may provide a larger flow than the design flow, and the balancing device
will permit field measurement and readjustment.
• Is thermal expansion to be handled through geometrical offset configura-
u
tions (Figure 2-6) or by using mechanical joints (Figure 2-7)? For example,
steel pipe may increase its length by 0.53 in. per 100 ft if its temperature
increases from 0°F to 70°F, or by 1.52 in. if from 0°F to 200°F. The system
must be designed to handle thermal expansion and contraction of the pip-
ro
ing. Failure to properly design for thermal expansion can result in piping
distortion, noise, and possible system failures.
Arrangement of the piping and its suspension from the building structure
must also be given specific attention by the designer if quiet operation is to be
attained. Rigid attachment of the pipe to the structural members of a building,
especially at midspan, provides a direct link that will transmit objectionable
G
vibration and sound. The exception would be at preselected points of the pip-
ing that must serve as anchors to control the amount and direction of move-
ment due to expansion and contraction. Chapter 46 of the 2012 ASHRAE
Handbook—HVAC Systems and Equipment contains recommendations regard-
ing anchor specification and spacing (ASHRAE 2012).
These are examples of concerns and questions that should be answered in
the early steps of the design process.
e t
gp
pE
Figure 2-6 Piping expansion, offset piping.
u
ro
G
Figure 2-7 Piping expansion, mechanical joint.

Sizing Piping
Sizing the pipe is not to be confused with pressure drop calculations, although
the pressure drop is generally used as a primary consideration in the sizing. In
most fluid systems, the size of the piping is established on the basis of the fric-
tion loss per running foot of pipe. The fluid velocity is then used as a limiting
selection parameter.
t
The equation most often used that relates pressure drop, flow rate, and pipe
size is the Darcy-Weisbach equation:
e
2
L V
h f = f ---- ------ (2-2)
D 2g
gp
where
D = pipe diameter, ft
f = friction factor (0.10 to 0.010)
g = gravitational constant, 32.2 ft/s2
hf = energy lost through friction, expressed as fluid head, feet of fluid
flowing
pE
L = pipe length, ft
V = average fluid velocity, ft/s
Figure 2-8 shows an experimental arrangement for determining head loss in
a pipe. Fluid velocity is calculated from the continuity equation (Streeter and
Wylie 1985):
V = Q (2-3)
u
----
A
where
A = cross-sectional area of the pipe, ft2
ro
Q = flow rate, ft3/s

V = fluid velocity, ft/s
Pipe sizing is covered in Chapter 22 of the 2013 ASHRAE Handbook—
Fundamentals (ASHRAE 2013). The general range of pipe friction loss used
for designing hydronic systems is between ~1 and 4 ft/100 ft. For controlling
G
the velocity noise, ASHRAE Handbook—Fundamentals suggests a velocity

limit of 4 ft/s for a 2 in. pipe and smaller. A pressure drop limit of 4 ft/100 ft for
pipe sizes above a 2 in. size is suggested, but this is subject to the designer’s
selection. Maximum water velocity versus operation hours to minimize erosion
may also be considered in the design. After a pipe size has been selected for a
known pipe material, flow rate, and friction factor, the Darcy-Weisbach equa-
tion can be used directly to calculate the head loss in feet of fluid flowing.
e t
Figure 2-8 Experimental arrangement for determining head loss in a pipe.
One approach to simplify the determination of pipe diameter (assuming
gp
turbulent flow, combining some of the formulas, and assuming an approxima-
tion of f = 0.03) reduces the Darcy-Weisbach equation to (Coad 1985):
0.4
D = SQ (2-4)
where
D = pipe diameter, in.
pE
S = pipe sizing constant
= 0.44 assuming 4 ft pressure drop per 100 ft length
= 0.50 assuming 2 ft pressure drop per 100 ft length
This simplified method is useful for quick approximations.
It is important to know some further classic approaches for accuracy and if it
is required to evaluate a pipe sizing program.
u
In an experiment in 1883, Osborne Reynolds showed that fluids can flow
through a pipe under two different conditions: laminar flow and turbulent flow.
He demonstrated that when dye was injected in a glass pipe with low water veloc-
ities (Figure 2-9), the stream of dye stayed in layers (laminar flow) up to Re =
ro
2000. However, as the velocity was increased, the layer of dye wavered and then
broke up, diffusing with the water because of intermingling of the particles or of
the water in what was turbulent flow (Re > 2000).
Reynolds defined the Reynolds number (Re) as
Re = DV --- (2-5)
G
where
D = inside pipe diameter, ft
V = average fluid velocity, ft/s
= dynamic viscosity, lbm/ft·s
= density, lb/ft3
e t
gp
Figure 2-9 Reynolds’s laminar versus turbulent flow demonstration.
The Reynolds number is a nondimensional parameter relating pipe diame-

ter, fluid velocity, and the fluid viscous properties. Relating these fluid flow
pE
parameters enables the development of charts relating flow conditions and pipe
characteristics. These charts present experimental data that can be used in pipe
sizing and pipe system design.
Reynolds’s study showed that the friction factor in the laminar flow range
is equal to the following:
64
f = -----
- (2-6)
Re
u
Johann Nikuradse (1933) demonstrated the effect of pipe surface roughness
on friction for both the laminar and turbulent regions (Figure 2-10) and defined
a roughness factor, /D. Figure 2-10 can be used to determine the friction fac-
tor when the Reynolds number and the pipe roughness factor /D are known.
ro
Lewis Moody (1944) demonstrated that a transition region appears

between Re of 2000 and 10,000, as shown in his diagram (Figure 2-11). The
Moody diagram portrays friction factor f from 0.01 to 0.08 as a function of the
relative roughness /D of the pipe or tubing and the Reynolds number DV / .
The relative roughness /D can be determined from another Moody graph
G
(Figure 2-12) portraying pipe diameter, pipe material, and relative roughness of
pipe.
The kinematic viscosity may also be used in the Reynolds formula since
= / (ft2/s). Substituting the Reynolds number can be written as follows:
V
Re = D ---- (2-7)
e t
Figure 2-10
gp
Relation of Reynolds number, friction flow, and relative roughness for similar
pipes.
pE
Source: Moody (1944).
Figure 2-13 is a nomogram that helps simplify determination of kinematic

viscosity ( ) and the Reynolds number (Re) for different fluids, knowing the
fluid temperature in °F, pipe diameter in in., and velocity in ft/s.
The 2013 ASHRAE Handbook—Fundamentals also refers to the Colebrook
equation for determining the friction factor f in the turbulent flow range
u
(ASHRAE 2013):
1 18.7
-------- = 1.74 – 2log 2 ---- + --------------- (2-8)
f D Re f
ro
The Hazen-Williams equation is also mentioned as an alternative to the

Darcy-Weisbach equation:
1.852 1 -
h f = 3.0221 V
G
---- ----------- (2-9)

C D
1.67
where C is the pipe roughness factor.

Recommended values of the C factor are as follows: 150 for plastic pipe and
copper tubing, 140 for new steel pipe, 100 for steel pipe after 20 years of use,
and down to 80 after 30 years or for badly corroded or very rough pipe (Streeter
and Wylie 1985; Karassik et al. 1986).
42
G
ro
u
Chapter 2 Piping Systems Design
pE
gp
Figure 2-11 Moody chart showing relationship between friction factors and Reynolds number for water flow.
Source: Moody (1944).
e t
G
ro
u pE
gp
Fundamentals of Water Systems Design I-P
Figure 2-12 Friction factors and relative roughness for various pipes.
Source: Engineering Data Book, Figure IIIA-4. (HI 1990).
43
et
44
G
ro
u
Chapter 2 Piping Systems Design
pE
gp
Figure 2-13 Kinematic viscosity and Reynolds number determination nomogram.
Source: 2013 ASHRAE Handbook—Fundamentals, Chapter 3, Figure 13.
et
e t
gp
u pE
Figure 2-14 Friction loss for water in commercial steel pipe (Schedule 40) and friction loss
for water in copper tubing (Types K, L, M).
Source: 2013 ASHRAE Handbook—Fundamentals, Chapter 22, Figures 4 and 5.
ro
The 2013 ASHRAE Handbook—Fundamentals notes that the Darcy-Weis-

bach equation with friction factors from the Moody chart, the Colebrook equa-
tion, or the Hazen-Williams equation are fundamental to calculating pressure
drop in hot and chilled-water piping. Charts calculated from these equations
(such as Figure 2-14) show flow rates and head loss for schedule 40 steel pipe
G
(ASHRAE 2013). The 2013 ASHRAE Handbook—Fundamentals has similar

charts for K, L, and M copper tubing and schedule 80 plastic pipe.
Also, tables and charts can be found in the Crane Company’s Technical Paper
#410 (Crane 1988) and the Hydraulic Institute’s Engineering Data Book (1990).
Reviewing the Engineering Data Book, we find another version of the Colebrook
equation:
1 2.51
-------- = 2log ------------ + --------------- (2-10)
f 3.7D Re f
Be aware that this equation is used for calculation of the friction factor
(f) and, consequently, also the values in the tables of friction loss (hf) in
feet per 100 feet for new pipes from 1/8 to 5 in. nominal schedule 40 steel
pipe and from 6 to 84 in. for schedule 40 steel and cast-iron pipe. These val-
ues will differ slightly from those found in the 2013 ASHRAE Handbook—
t
Fundamentals.
A typical table for 2 in. nominal pipe is shown as Table 2-1. In commercial
e
installations, the tables suggest adding 15% to the friction loss to allow for
aging.
gp
Flow-Rate Measurement
Taken from the 2013 ASHRAE Handbook—Fundamentals, Table 2-2 lists
various means of measuring fluid flow rate. The values for volume or mass
flow-rate measurement are often determined by measuring pressure differ-
ence across an orifice, nozzle, or venturi tube (ASME 1972; Benedict
1984). These types of meters have different advantages and disadvantages.
pE
For example, the orifice plate is more easily changed than the complete
nozzle or venturi tube assembly. However, the nozzle is often preferred to
the orifice because its discharge coefficient is more precise. The venturi
tube is a nozzle followed by an expanding recovery section to reduce net
pressure loss.
Fluid meters use a wide variety of physical techniques to make flow mea-
surements; those more prevalently used are described in the following section
(ASME 1972; Miller 1983; DeCarlo 1984). The search for high-accuracy flow
u
measurement includes the arrangement of appropriate calibration procedures.
While these used to be available only in calibration laboratories, they are now
frequently purchased along with flowmeters so that flow measurements can be
efficiently and effectively ensured and validated at high levels of performance.
ro
To ensure and validate calibration facilities and procedures, realistic traceabil-

ity should be established and maintained for the calibration facilities and pro-
cedures.
Direct and Indirect Flow Measurement Methods

G
Both gas and liquid flow can be measured quite accurately by timing a col-
lected amount of fluid that is determined gravimetrically or volumetrically.
While this method is commonly used for calibrating other metering
devices, it is particularly useful where the flow rate is low or intermittent
and where a high degree of accuracy is required. These systems are gener-
ally large and slow, but in their simplicity they can be considered primary
devices.
Table 2-1 Friction Loss for Water in Feet for 100 ft, 2 in. Nominal Pipe Schedule 40
Schedule 40 Steel Pipe
Inside Diameter = 2.067 in.
2 in. Nominal Discharge
/D = 0.00087
cfs gpm
V, V2/2g, ft per 100 ft of Pipe
ft/s ft
t
0.00446 2 0.191 0.00057 0.0151
0.00669 3 0.287 0.00128 0.0302
e
0.00892 4 0.383 0.00228 0.0497
0.0112 5 0.479 0.00356 0.0733
0.0134 6 0.574 0.00512 0.1008
gp
0.0156 7 0.670 0.00698 0.132
0.0178 8 0.766 0.00911 0.167
0.0201 9 0.861 0.0115 0.206
0.0223 10 0.957 0.0142 0.249
0.0268 12 1.15 0.0205 0.344
pE
0.0312 14 1.34 0.0279 0.454
0.0357 16 1.53 0.0364 0.577
0.0401 18 1.72 0.0461 0.714
0.0446 20 1.91 0.0569 0.865
0.0491 22 2.11 0.0689 1.03
0.0535 24 2.30 0.0820 1.21
0.0580 26 2.49 0.0962 1.40
u
0.0624 28 2.68 0.112 1.60
0.0669 30 2.87 0.128 1.82
ro
0.0781 35 3.35 0.174 2.42

0.0892 40 3.83 0.228 3.09
0.100 45 4.31 0.288 3.85
0.112 50 4.79 0.356 4.69
0.123 55 5.26 0.431 5.61
G
0.134 60 5.74 0.512 6.61

0.145 65 6.22 0.601 7.69
0.156 70 6.70 0.698 8.85
0.167 75 7.18 0.801 10.1
Table 2-1 Friction Loss for Water in Feet for 100 ft, 2 in. Nominal Pipe Schedule 40 (Continued)
Inside Diameter = 2.067 in.
/D = 0.00087
cfs gpm
V, V2/2g, ft per 100 ft of Pipe
ft/s ft
t
0.178 80 7.66 0.911 11.4
0.190 85 8.14 1.03 12.8
e
0.201 90 8.61 1.15 14.3
0.212 95 9.09 1.28 15.8
0.223 100 9.57 1.42 17.5
gp
0.245 110 10.53 1.72 21.0
0.268 120 11.5 2.05 24.8
0.290 130 12.4 2.41 28.9
0.312 140 13.4 2.79 33.4
0.335 150 14.4 3.20 38.1
pE
0.357 160 15.3 3.64 43.2
0.379 170 16.3 4.11 48.6
0.401 180 17.2 4.61 54.3
0.424 190 18.2 5.14 60.3
0.446 200 19.1 5.69 66.6
0.491 220 21.1 6.89 80.2
0.535 240 23.0 8.20 95.1
u
0.580 260 24.9 9.62 111
0.624 280 26.8 11.16 129
ro
0.669 300 28.7 12.81 147

0.714 320 30.6 14.6 167
0.758 340 32.5 16.5 188
0.803 360 34.5 18.4 211
0.847 380 36.4 20.6 234
G
0.892 400 38.3 22.8 259

Table 2-2 Volumetric or Mass Flow-Rate Measurement

Measurement Means Application Range Precision Limitations
Flow through Discharge coefficient
Orifice and
pipes, ducts, and Above Reynolds and accuracy
differential pressure 1%–5%
plenums for all number of 5000 influenced by
measurement system
fluids installation conditions
Nozzle and
t
differential pressure 0.5%–2.0%
measurement system
e
Venturi tube and
differential pressure 0.5%–2.0%
gp
measurement system
Liquids or gases;
Timing given mass or System is bulky and
used to calibrate Any 0.1%–0.5%
volumetric flow slow
other flowmeters
Should be calibrated
Rotameters Liquids or gases Any 0.5%–5.0%
for fluid being metered
Relatively small
pE
As high as 0.1%–2.0% Most types require
volumetric flow
Displacement meter 1000 cfm, depending on calibration with fluid
with high
depending on type type being metered
pressure loss
Short-duration
Total flow limited by
Gasometer or tests; used to
available 0.5%–1.0% —
volume displacement calibrate other
volume of containers
flowmeters
Thomas meter
u
Elaborate setup Uniform velocity;
(temperature rise of
justified by need Any 1% usually used with
stream caused by
for good accuracy gases
electrical heating)
ro
Element of resistance Used for check

Lower limit set by Secondary reading
to flow and where system has
readable pressure 1%–5% depends on accuracy
differential pressure calibrated
drop of calibration
measurement system resistance element
Turbine flowmeters Liquids or gases Any 0.25%–2.0% Uses electronic readout
Accuracy depends on
G
Primarily for
uniformity of flow and
Single or multipoint installed air- Lower limit set by
completeness of
instrument for handling systems accuracy of velocity
2%–10% traverse; may be
measuring velocity at with no special measurement
affected by distur-
specific point in flow provision for flow instrumentation
bances near point of
measurement
measurement
Table 2-2 Volumetric or Mass Flow-Rate Measurement (Continued)

Measurement Means Application Range Precision Limitations
Heat input and
Check value in
temperature changes
heater or cooler Any 1%–3% —
with steam and water
tests
coil
Measure liquid or
t
gas volumetric Fluid must be free of
Laminar flow
flow rate; nearly dirt, oil, and other
element and
e
linear relationship 0.0001–2000 cfm 1% impurities that could
differential pressure
with pressure plug meter or affect its
measurement system
drop; simple and calibration
easy to use
gp
Measures
electrically
At present state of the
Magnetohydrody- conductive fluids,
art, conductivity of
namic flowmeter slurries; meter 0.1–10,000 gpm 1%
fluid must be greater
(electromagnetic) does not obstruct
than 5 mho/cm
flow; no moving
parts
pE
Measure liquid or
Swirl flowmeter and Above Reynolds
gas flow in pipe; 1% —
vortex shedding meter number of 104
no moving parts
Source: 2013 ASHRAE Handbook—Fundamentals, Table 5, Chapter 36.
The variable-area meter or rotameter is a convenient direct-reading flowme-

ter for liquids and gases. This is a vertical, tapered tube in which the flow rate is
indicated by the position of a float suspended in the upward flow. The position of
u
the float is determined by its buoyancy and the upwardly directed fluid drag.
A velocity traverse (made using a pitot tube or other velocity-measuring
instrument) measures airflow rates in the field or calibrates large nozzles. This
method can be imprecise at low velocities and impractical where many test
ro
runs are in progress.
Venturi, Nozzle, and Orifice Flowmeters

Flow in a pipeline can be measured by a venturi meter (Figure 2-15), flow noz-
zle (Figure 2-16), or orifice plate (Figure 2-17). The ASME publication MFC-
G
3M (ASME 2004) describes measurement of fluid flow in pipes using the ori-
fice, nozzle, and venturi; ASME Performance Test Code 19.5-72 specifies their
construction (ASME 1972).
Assuming an incompressible fluid (liquid or slow-moving gas), uniform
velocity profile, frictionless flow, and no gravitational effects, the principle of
conservation of mass and energy can be applied to the venturi and nozzle
geometries to give the following:
e t
gp
pE
Figure 2-15 Typical Herschel-type Venturi meter.
u
ro
G
Figure 2-16 Dimensions of ASME long-radius flow nozzles.

e t
gp
pE
Figure 2-17 Sharp-edge orifice with pressure tap locations.
2g p 1 – p 2
u
w = V 1 A1 = V 2 A 2 = A 2 --------------------------------
- (2-11)
4
1–
where
A1 = area 1
ro
A2 = area 2
p1, p2 = absolute pressure, lb/ft2
V1, V2 = velocity, ft/s
G
w = flow rate, lb/s

= D2/D1 for venturi and sharp-edged orifice and d/D for flow nozzle
= density, lb/ft3
Because the flow through the meter is not frictionless, a correction factor C
is deemed to account for friction losses. If the fluid is at a high temperature, an
additional correction factor Fa should be included to account for thermal
expansion of the primary element. Because this amounts to less than 1% at

500°F, it can usually be omitted. Equation 2-11 then becomes
2g p 1 – p 2
w = C A 2 --------------------------------
- (2-12)
4
1–
The factor C is a function of geometry and Reynolds number. Values of C are
given in ASME Performance Test Code 19.5-72 (ASME 1972). The approach
t
factor can be combined with the discharge coefficient, as described later.
e
The jet passing through an orifice plate contracts to a minimum area at the
vena contracta, which is located a short distance downstream from the orifice
plate. The contraction coefficient, energy loss coefficient, and approach factor
can be combined into a single constant K, which is a function of geometry and
gp
Reynolds number. The orifice flow rate equations then become
Q = K A 2 2g p1 – p2 (2-13)
where
A2 = orifice area, ft2
pE
K = values are shown in ASME Performance Test Code 19.5-72 (ASME
1972)
p1, p2 = pressure drop, lb/ft2, as obtained by pressure taps
= density, lb/ft3
Valves, bends, and fittings upstream from the flowmeter can cause errors.
Long, straight pipes should be installed upstream and downstream of the flow
devices to ensure fully developed flow for proper measurement. ASHRAE
u
Standard 41.8 specifies upstream and downstream pipe lengths for measuring
flow of liquids with an orifice plate (ASHRAE 1989). ASME Performance Test
Code 19.5-72 gives the piping requirements between various fittings and
valves and the venturi, nozzle, and orifice (ASME 1972).
ro
The 2015 ASHRAE Handbook—HVAC Applications recommends straight

pipe a minimum of 15 pipe diameters upstream and 5 diameters downstream
for any flow-measuring device (ASHRAE 2015).
Variable-Area Flowmeters (Rotameters)

G
In permanent installations where high precision, ruggedness, and operational

ease are important, the variable-area flowmeter is satisfactory. It is frequently
used to measure liquids or gases in small-diameter pipes. However, for ducts or
pipes over 6 in. diameter, the expense of this meter may not be warranted. In
larger systems, the meter can be placed in a bypass line and used with an orifice.
The variable-area meter (Figure 2-18) commonly consists of a float that is free to
move vertically in a transparent tapered tube. The fluid to be metered enters at
the narrow bottom end of the tube and moves upward, passing at some point
through the annulus formed between the float and the inside wall of the tube. At
any particular flow rate, the float assumes a definite position in the tube; a cali-
brated scale on the tube shows the float’s location and the fluid flow rate.
Turbine Flowmeters
t
Turbine flowmeters are volumetric flow-rate-sensing meters with a magnetic
stainless steel turbine rotor suspended in the flow stream of a nonmagnetic
e
meter body. The fluid stream exerts a force on the blades of the turbine rotor,
setting it in motion and converting the fluid’s linear velocity to an angular
velocity. Design motivation for turbine meters is to have the rotational speed of
the turbine proportional to the average fluid velocity and thus to the volume
gp
rate of fluid flow (Miller 1983; DeCarlo 1984; Mattingly 1992).
The rotational speed of the rotor is monitored by an externally mounted pickoff
assembly. Magnetic and radio frequency are the most commonly used pickoffs.
The magnetic pickoff contains a permanent magnet and coil. As the turbine rotor
u pE
ro
G
Figure 2-18 Variable-area flowmeter.

blades pass through the field produced by the permanent magnet, a shunting action
induces alternating-current voltage in the winding of the coil wrapped around the
magnet. A sine wave with a frequency proportional to the flow rate develops.
With the radio frequency pickoff, an oscillator applies a high-frequency
carrier signal to a coil in the pickoff assembly. The rotor blades pass through
the field generated by the coil and modulate the carrier signal by shunting
action on the field shape. The carrier signal is modulated at a rate correspond-
ing to the rotor speed, which is proportional to the flow rate.
t
With both pickoffs, the frequency of the pulses generated becomes a mea-
e
sure of flow rate, and the total number of pulses measures total volume (Mat-
tingly 1992; Woodring 1969; Shafer 1961).
The lubricity of the process fluid and the type and quality of rotor bearings
determine whether the meter is satisfactory for the particular application.
gp
When choosing turbine flowmeters for use with fluorocarbon refrigerants,
attention must be paid to the type of bearings used in the meter and to the oil
content of the refrigerant. For these applications, sleeve-type rather than stan-
dard ball bearings are recommended. The amount of oil in the refrigerant can
severely affect calibration and bearing life.
In metering liquid fluorocarbon refrigerants, the liquid must not flash to a
vapor (cavitate). This would cause a tremendous increase in flow volume.
pE
Flashing results in erroneous measurements and rotor speeds that can damage
the bearings or cause a failure. Flashing can be avoided by maintaining an ade-
quate backpressure on the downstream side of the meter (Liptak 1972).
The Next Step

In Chapter 3, you will learn about pipe materials and fittings.
u
Summary
Chapter 2 covered the following topics:
ro
• Bernoulli’s principle
• The three steps in designing a fluid distribution system
• The difference between direct and return piping and if they can be combined
• Methods to allow for thermal expansion
G
• How to determine pressure drop and sizing of piping

• The difference between laminar and turbulent flow, and what index quantifies it
• Piping roughness factors
• The factors that govern pressure drop in a piping system
• Flow-rate measurement
• Flow-measuring device location
References
ASHRAE. 1989. Standard 41.8-1989, Standard Methods of Measurement
of Flow of Liquids in Pipes Using Orifice Flowmeters. Atlanta:
ASHRAE.
Atlanta: ASHRAE.
ASHRAE. 2013. ASHRAE Handbook—Fundamentals. Atlanta: ASHRAE.
t
ASHRAE. 2015. ASHRAE Handbook—HVAC Applications. Atlanta:
e
ASHRAE.
ASME. 1972. Application of fluid meters. ASME Performance Test Code PTC
19.5-72. New York, NY: ASME.
ASME. 2004. MFC-3M-2004, Measurement of Fluid Flow in Pipes Using Ori-
gp
fice, Nozzle and Venturi. New York: ASME.
Benedict, R. 1984. Fundamentals of Temperature, Pressure and Flow Mea-
surements. New York: John Wiley and Sons.
Bernoulli, D. 1738. Hydrodynamica. New York: Dover Publications (1968).
Coad, W. 1985. Sizing of pipes & ducts. Heating, Piping, Air-Conditioning: July.
Crane. 1988. Flow of fluids through valves, fittings and pipe. Technical Paper
#410. Joliet, IL: Crane Co.
DeCarlo, J. 1984. Fundamentals of Flow Measurement. Research Triangle
pE
Park, NC: Instrumentation Society of America.
Euler, L. 1750.
HI. 1990. Engineering Data Book, 2nd ed. Cleveland, OH: Hydraulic Institute.
Karassik, I., J. Messina, P. Cooper, and C. Heald. 1986. Pump Handbook. New
York, NY: McGraw-Hill.
Liptak, B., ed. 1972. Instrument Engineers Handbook. Philadelphia: Chiton
Book Co.
Mattingly, G. 1992. The characterization of a piston displacement-type flow-
u
meter calibration facility and the calibration and use of pulsed output type
flowmeters. Journal of Research of the National Institute of Standards and
Technology. Gaithersburg, MD: NIST.
Miller, R. 1983. Measurement Engineering Handbook. New York: McGraw-
ro
Hill.
Moody, L. 1944. Friction factors for pipe flow. ASME Transactions 66(8).
Nikuradse, J. 1933. Strommung Gesetze in rouhen Rohren. Forsch. Arb. Ing.-
Wes. No.361.
Reynolds, O. 1883. An experimental investigation of the circumstances which
determine whether the motion of water shall be direct or sinuous, and of the
G
law of resistance in parallel channels. Philosophical Transactions of the

Royal Society. London.
Shafer, M. 1961. Performance characteristics of turbine flowmeters. Proceed-
ings of the ASME Winter Annual Meeting, New York, NY.
Streeter, V., and E. Wylie. 1985. Fluid Mechanics. New York, NY: McGraw-Hill.
Woodring, E. 1969. Magnetic turbine flowmeters. Instruments and Control
Systems 6:133.

2-1 What causes unbalanced flow rates in direct-return piping arrangements?

2-2 What is the most often used equation that relates to pressure drop?
t
2-3 Fluids can flow through a pipe under two different conditions. Name them.
e
Explain the differences between these flow conditions.
2-4 Explain friction factor.
2-5 In commercial installations, it is suggested that _____% should be added to the
gp
friction loss to allow for aging.
2-6 What is the Bernoulli principle?
2-7 What factors determine pressure drop in piping?
2-8 What methods allow thermal expansion?
2-9 What is the minimum distance upstream and downstream for a water flow-
pE
measuring device (in pipe diameters)?
u
ro
G
G
ro
upE
gp
et
Pipe Materials
and Fittings
e t
Study Objectives
gp
After completing this chapter, you should be able to
❑ name the U.S. organizations that issue codes and standards for piping sys-
tems and components;
❑ list the different types of pipe used in HVAC water system applications and
describe the characteristics of each;
❑ list the pipe-joining methods commonly encountered in HVAC water sys-
pE
tems and describe the characteristics of each;
❑ list common factors that support or promote corrosion, the five methods of
corrosion control, and the two corrosion environments of particular concern
to the HVAC piping system designer;
❑ describe the function, selection, and installation of backflow prevention
devices; and
❑ solve a basic pipe selection problem.
u
Instructions
Pipe Materials
ro
This section covers pipe materials commonly used for heating, air-conditioning,
and refrigerating systems. When selecting and applying pipe, applicable local
codes, state or provincial codes, and voluntary industry standards (some of
which have been adopted by code jurisdictions) must be followed.
The following U.S. organizations issue codes and standards for piping sys-
G
tems and components:

• ASME
• ASTM International
• National Fire Protection Association (NFPA)
• International Code Council (ICC)
• Manufacturers Standardization Society, Valve and Fitting Industry (MSS)
60 Chapter 3 Pipe Materials and Fittings
• American National Standards Institute (ANSI)

• American Water Works Association (AWWA)
Parallel federal specifications also have been developed by government
agencies for many public works projects. ANSI/ASME Standard B31.9, Build-
ing Services Piping, lists applicable American codes and standards for HVAC
piping (ASME 1988). In addition, it gives the requirements for safe design and
construction of piping systems for building heating and air conditioning.
t
Metal Pipe
e
Steel Pipe
Steel pipe is manufactured by several processes. Seamless pipe made by piercing
or extruding has no longitudinal seam. Other manufacturing methods roll skelp
gp
into a cylinder and weld a longitudinal seam. A continuous-weld (CW) furnace-
butt-welding process forces and joins the edges together at a high temperature.
An electric current welds the seam of electric resistance welded (ERW) pipe.
ASTM Standards A53 and A106 specify steel pipe (2012, 2014a). Both specify
A and B grades. The A grade has a lower tensile strength and is not widely used.
The ASME (2004, 2014) pressure piping codes require that a longitudi-
nal joint efficiency factor E be applied to each type of seam when calculat-
ing the allowable stress, as listed in Table 3-1. ASME Standard B36.10
pE
(2004) specifies the dimensional standard for steel pipe. Up to 12 in. diame-
ter, nominal pipe sizes (NPSs) are used, which do not match the internal or
external diameters. For 14 in. and larger pipe, the size corresponds to the
outside diameter. Steel pipe is manufactured with wall thicknesses identified
by schedule and weight. Although schedule numbers and weight designations
are related, they are not constant for all pipe sizes. Standard-weight (ST) and
schedule 40 pipe have the same wall thickness through 10 in. NPS. For 12 in.
and larger standard-weight pipe, the wall thickness remains constant at
u
Table 3-1 Allowable Stressesa for Pipe and Tube

Minimum Basic Allowable
ro
Available Joint Allowable

ASTM Manufacturing Tensile Allowable Stress
Grade Type Sizes, Efficiency Stressb SE,
Specification Process Strength, Stress Rangec
in. Factor E psi
psi S, psi SA, psi
A53 Steel — F CW 1/2 to 4 45,000 11,250 0.6 6,800 16,900
A53 Steel B S Seamless 1/2 to 26 60,000 15,000 1.0 15,000 22,500
G
A53 Steel B E ERW 2 to 20 60,000 15,000 0.85 12,800 22,500

A106 Steel B S Seamless 1/2 to 26 60,000 15,000 1.0 15,000 22,500
B88 Copper — — Hard Drawn 1/4 to 12 36,000 9,000 1.0 9,000 13,500
a Listed stresses are for temperatures to 650°F for steel pipe (to 400°F for Type F) and to 250°F for copper tubing.
b To be used for internal pressure stress calculations in Equations 1 and 2 of ASHRAE 2012, Chapter 46.
c To be used only for piping flexibility calculations; see Equations 3 and 4 of ASHRAE 2012, Chapter 46.
Source: 2012 ASHRAE Handbook—HVAC Systems and Equipment, Table 1, Chapter 46.
0.375 in., while schedule 40 wall thicknesses increase with each size. A simi-
lar equality exists between extra strong (XS) and schedule 80 pipe up through
8 in.; afterward, XS pipe has a 0.500 in. wall, while schedule 80 increases in
wall thickness. Table 3-2 lists properties of representative steel pipe.
Joints in steel pipe are made by welding or using threaded, flanged, grooved,
or welded-outlet fittings. Unreinforced welded-in-branch connections weaken a
main pipeline, and added reinforcement is necessary unless the excess wall thick-
nesses of both mains and branches are sufficient to sustain the pressure. ANSI/
t
ASME Standard B31.1, Power Piping, gives formulas for determining when rein-
forcement is required (ASME 2014). Such calculations are seldom needed in
e
HVAC applications because standard-weight pipe through 20 in. NPS at 300 psig
requires no reinforcement, full-size branch connections are not recommended, and
fittings such as tees and reinforced outlet fittings provide inherent reinforcement.
gp
Copper Tube
Because of their inherent corrosion resistance and ease of installation, copper and
copper alloys are often used in heating, air conditioning, refrigeration, and water
supply installations. There are two principal classes of copper tube. ASTM Stan-
dard B88 (ASTM 1995a) includes types K, L, M, and drain, waste, and vent
(DWV) for water and drain service. ASTM Standard B280 (ASTM 1995b) spec-
pE
ifies air conditioning and refrigeration (ACR) tube for refrigeration service.
Types K, L, M, and DWV designate descending wall thicknesses for cop-
per tube. All types have the same outside diameter for corresponding sizes.
Table 3-3 lists properties of ASTM Standard B88 copper tube (ASTM 1995a).
In the plumbing industry, a tube of nominal size approximates the inside diam-
eter (ID). The heating and refrigeration trades specify copper tube by the out-
side diameter (OD). ACR tubing has a different set of wall thicknesses.
Types K, L, and M tube may be hard-drawn or annealed (soft) temper.
u
Copper tubing is joined with soldered or brazed wrought or cast copper capil-
lary socket-end fittings. Table 3-4 lists the pressure temperature ratings of sol-
dered and brazed joints. Small copper tubes are also joined by flare or
compression fittings.
ro
Hard-drawn tubing has a higher allowable stress value than annealed, but if
hard tubing is joined by soldering or brazing, the annealed allowable stress
value should be used. Brass pipe and copper pipe are also made in steel pipe
thicknesses for threading. High cost has eliminated these materials from the
market, except for special applications.
The heating and air-conditioning industry generally uses types L and M
G
tubing, which have higher internal working pressure ratings than the solder
joints used at fittings. Type K may be used with brazed joints for higher
pressure and temperature requirements or for direct burial. Type M should be
used with care where exposed to potential external damage.
Copper and brass should not be used in ammonia refrigerating systems. The
Special Systems section in this chapter covers other limitations on refrigerant
piping.
Table 3-2 Steel Pipe Data

Working Pressurec
Cross
Schedule Wall Inside Surface Area Weight ASTM A53
Nominal Pipe Section
Number Thickness Diameter to 400°F
Size, OD,
or t, d, Metal Flow
in. in. Outside, Inside, Pipe, Water, Mfr. Joint
Weighta in. in. Area, Area, psig
ft2/ft ft2/ft 2 2 lb/ft lb/ft Process Typeb
in in
t
1/4 0.54 40 ST 0.088 0.364 0.141 0.095 0.125 0.104 0.424 0.045 CW T 188
80 XS 0.119 0.302 0.141 0.079 0.157 0.072 0.535 0.031 CW T 871
e
3/8 0.675 40 ST 0.091 0.493 0.177 0.129 0.167 0.191 0.567 0.083 CW T 203
80 XS 0.126 0.423 0.177 0.111 0.217 0.141 0.738 0.061 CW T 820
gp
1/2 0.84 40 ST 0.109 0.622 0.22 0.163 0.25 0.304 0.85 0.131 CW T 214
80 XS 0.147 0.546 0.22 0.143 0.32 0.234 1.087 0.101 CW T 753
3/4 1.05 40 ST 0.113 0.824 0.275 0.216 0.333 0.533 1.13 0.231 CW T 217
80 XS 0.154 0.742 0.275 0.194 0.433 0.432 1.47 0.187 CW T 681
1 1.315 40 ST 0.133 1.049 0.344 0.275 0.494 0.864 1.68 0.374 CW T 226
80 XS 0.179 0.957 0.344 0.251 0.639 0.719 2.17 0.311 CW T 642
pE
1-1/4 1.66 40 ST 0.14 1.38 0.435 0.361 0.669 1.5 2.27 0.647 CW T 229
80 XS 0.191 1.278 0.435 0.335 0.881 1.28 2.99 0.555 CW T 594
1-1/2 1.9 40 ST 0.145 1.61 0.497 0.421 0.799 2.04 2.72 0.881 CW T 231
80 XS 0.2 1.5 0.497 0.393 1.068 1.77 3.63 0.765 CW T 576
2 2.375 40 ST 0.154 2.067 0.622 0.541 1.07 3.36 3.65 1.45 CW T 230
80 XS 0.218 1.939 0.622 0.508 1.48 2.95 5.02 1.28 CW T 551
2-1/2 2.875 40 ST 0.203 2.469 0.753 0.646 1.7 4.79 5.79 2.07 CW W 533
u
80 XS 0.276 2.323 0.753 0.608 2.25 4.24 7.66 1.83 CW W 835
3 3.5 40 ST 0.216 3.068 0.916 0.803 2.23 7.39 7.57 3.2 CW W 482
ro
80 XS 0.3 2.9 0.916 0.759 3.02 6.6 10.25 2.86 CW W 767

4 4.5 40 ST 0.237 4.026 1.178 1.054 3.17 12.73 10.78 5.51 CW W 430
80 XS 0.337 3.826 1.178 1.002 4.41 11.5 14.97 4.98 CW W 695
6 6.625 40 ST 0.28 6.065 1.734 1.588 5.58 28.89 18.96 12.5 ERW W 696
80 XS 0.432 5.761 1.734 1.508 8.4 26.07 28.55 11.28 ERW W 1209
G
8 8.625 30 0.277 8.071 2.258 2.113 7.26 51.16 24.68 22.14 ERW W 526
40 ST 0.322 7.981 2.258 2.089 8.4 50.03 28.53 21.65 ERW W 643
80 XS 0.5 7.625 2.258 1.996 12.76 45.66 43.35 19.76 ERW W 1106
Table 3-2 Steel Pipe Data (Continued)

Working Pressurec
Sche- Cross
Inside Surface Area Weight ASTM A53
Pipe dule Wall Section
Nominal Dia- to 400°F
OD, Number Thick-
Size, in. meter d, Metal Flow
in. or ness t, in. Outside, Inside, Pipe, Water, Mfr. Joint
a in. Area, Area, psig
Weight ft2/ft ft2/ft 2 2 lb/ft lb/ft Process Typeb
in in
t
10 10.75 30 0.307 10.136 2.814 2.654 10.07 80.69 34.21 34.92 ERW W 485
40 ST 0.365 10.02 2.814 2.623 11.91 78.85 40.45 34.12 ERW W 606
e
XS 0.5 9.75 2.814 2.552 16.1 74.66 54.69 32.31 ERW W 887
80 0.593 9.564 2.814 2.504 18.92 71.84 64.28 31.09 ERW W 1081
gp
12 12.75 30 0.33 12.09 3.338 3.165 12.88 114.8 43.74 49.68 ERW W 449
ST 0.375 12 3.338 3.141 14.58 113.1 49.52 48.94 ERW W 528
40 0.406 11.938 3.338 3.125 15.74 111.9 53.48 48.44 ERW W 583
XS 0.5 11.75 3.338 3.076 19.24 108.4 65.37 46.92 ERW W 748
80 0.687 11.376 3.338 2.978 26.03 101.6 88.44 43.98 ERW W 1076
14 14 30 ST 0.375 13.25 3.665 3.469 16.05 137.9 54.53 59.67 ERW W 481
pE
40 0.437 13.126 3.665 3.436 18.62 135.3 63.25 58.56 ERW W 580
XS 0.5 13 3.665 3.403 21.21 132.7 72.04 57.44 ERW W 681
80 0.75 12.5 3.665 3.272 31.22 122.7 106.05 53.11 ERW W 1081
16 16 30 ST 0.375 15.25 4.189 3.992 18.41 182.6 62.53 79.04 ERW W 421
40 XS 0.5 15 4.189 3.927 24.35 176.7 82.71 76.47 ERW W 596
18 18 ST 0.375 17.25 4.712 4.516 20.76 233.7 70.54 101.13 ERW W 374
30 0.437 17.126 4.712 4.483 24.11 230.3 81.91 99.68 ERW W 451
u
XS 0.5 17 4.712 4.45 27.49 227 93.38 98.22 ERW W 530
40 0.562 16.876 4.712 4.418 30.79 223.7 104.59 96.8 ERW W 607
ro
20 20 20 ST 0.375 19.25 5.236 5.039 23.12 291 78.54 125.94 ERW W 337
30 XS 0.5 19 5.236 4.974 30.63 283.5 104.05 122.69 ERW W 477
40 0.593 18.814 5.236 4.925 36.15 278 122.82 120.3 ERW W 581
a
Numbers are schedule numbers per ASME Standard B36.10M; ST = standard weight, XS = extra strong.
b T = thread, W = weld.
c
Working pressures were calculated per ANSI/ASME B31.9 using furnace butt-weld (CW) pipe through 4 in. and electric resistance welded (ERW)
G
thereafter (ASME 1988). The allowance A has been taken as (1) 12.5% of t for mill tolerance on pipe wall thickness, plus (2) an arbitrary corrosion allow-
ance of 0.025 in. for pipe sizes through NPS 2 and 0.065 in. for NPS 2.5 in. through 20, plus (3) a thread cutting allowance for sizes through NPS 2.
Because the pipe wall thickness of threaded standard pipe is so small after deducting the allowance A, the mechanical strength of the pipe is impaired. It
is good practice to limit standard-weight threaded pipe pressure to 90 psig for steam and 125 psig for water.
Source: 2012 ASHRAE Handbook—HVAC Systems and Equipment, Table 2, Chapter 40.
Table 3-3 Copper Tube Data

Working
Pressurea,b,c
Wall Diameter Surface Area Cross Section Weight
Nominal ASTM B88
Thickness (1995a) to 250°F
Diameter, Type
t,
in. Outside Inside Metal Flow
in. Outside, Inside, Tube, Water, Annealed, Drawn,
D, d, Area, Area,
ft2/ft ft2/ft lb/ft lb/ft psig psig
in. in. in2 in2
t
1/4 K 0.035 0.375 0.305 0.098 0.08 0.037 0.073 0.145 0.032 851 1596
e
L 0.03 0.375 0.315 0.098 0.082 0.033 0.078 0.126 0.034 730 1368
3/8 K 0.049 0.5 0.402 0.131 0.105 0.069 0.127 0.269 0.055 894 1676
L 0.035 0.5 0.43 0.131 0.113 0.051 0.145 0.198 0.063 638 1197
M 0.025 0.5 0.45 0.131 0.118 0.037 0.159 0.145 0.069 456 855
gp
1/2 K 0.049 0.625 0.527 0.164 0.138 0.089 0.218 0.344 0.094 715 1341
L 0.04 0.625 0.545 0.164 0.143 0.074 0.233 0.285 0.101 584 1094
M 0.028 0.625 0.569 0.164 0.149 0.053 0.254 0.203 0.11 409 766
5/8 K 0.049 0.75 0.652 0.196 0.171 0.108 0.334 0.418 0.144 596 1117
L 0.042 0.75 0.666 0.196 0.174 0.093 0.348 0.362 0.151 511 958
3/4 K 0.065 0.875 0.745 0.229 0.195 0.165 0.436 0.641 0.189 677 1270
pE
L 0.045 0.875 0.785 0.229 0.206 0.117 0.484 0.455 0.209 469 879
M 0.032 0.875 0.811 0.229 0.212 0.085 0.517 0.328 0.224 334 625
1 K 0.065 1.125 0.995 0.295 0.26 0.216 0.778 0.839 0.336 527 988
L 0.05 1.125 1.025 0.295 0.268 0.169 0.825 0.654 0.357 405 760
M 0.035 1.125 1.055 0.295 0.276 0.12 0.874 0.464 0.378 284 532
1-1/4 K 0.065 1.375 1.245 0.36 0.326 0.268 1.217 1.037 0.527 431 808
L 0.055 1.375 1.265 0.36 0.331 0.228 1.257 0.884 0.544 365 684
u
M 0.042 1.375 1.291 0.36 0.338 0.176 1.309 0.682 0.566 279 522
DWV 0.04 1.375 1.295 0.36 0.339 0.168 1.317 0.65 0.57 265 497
1-1/2 K 0.072 1.625 1.481 0.425 0.388 0.351 1.723 1.361 0.745 404 758
L 0.06 1.625 1.505 0.425 0.394 0.295 1.779 1.143 0.77 337 631
ro
M 0.049 1.625 1.527 0.425 0.4 0.243 1.831 0.94 0.792 275 516
DWV 0.042 1.625 1.541 0.425 0.403 0.209 1.865 0.809 0.807 236 442
2 K 0.083 2.125 1.959 0.556 0.513 0.532 3.014 2.063 1.304 356 668
L 0.07 2.125 1.985 0.556 0.52 0.452 3.095 1.751 1.339 300 573
M 0.058 2.125 2.009 0.556 0.526 0.377 3.17 1.459 1.372 249 467
G
DWV 0.042 2.125 2.041 0.556 0.534 0.275 3.272 1.065 1.416 180 338
2-1/2 K 0.095 2.625 2.435 0.687 0.637 0.755 4.657 2.926 2.015 330 619
L 0.08 2.625 2.465 0.687 0.645 0.64 4.772 2.479 2.065 278 521
M 0.065 2.625 2.495 0.687 0.653 0.523 4.889 2.026 2.116 226 423
Table 3-3 Copper Tube Data (Continued)

Working
Pressurea,b,c
Nominal Wall Diameter Surface Area Cross Section Weight
ASTM B88
Diameter, Type Thickness (1995a) to 250°F
in. t, in.
Outside Inside Outside, Inside, Metal Flow Tube, Water, Annealed, Drawn,
D, in. d, in. ft2/ft ft2/ft Area, in2 Area, in2 lb/ft lb/ft psig psig
t
3 K 0.109 3.125 2.907 0.818 0.761 1.033 6.637 4.002 2.872 318 596
L 0.09 3.125 2.945 0.818 0.771 0.858 6.812 3.325 2.947 263 492
e
M 0.072 3.125 2.981 0.818 0.78 0.691 6.979 2.676 3.02 210 394
DWV 0.045 3.125 3.035 0.818 0.795 0.435 7.234 1.687 3.13 131 246
3-1/2 K 0.12 3.625 3.385 0.949 0.886 1.321 8.999 5.12 3.894 302 566
gp
L 0.1 3.625 3.425 0.949 0.897 1.107 9.213 4.291 3.987 252 472
M 0.083 3.625 3.459 0.949 0.906 0.924 9.397 3.579 4.066 209 392
4 K 0.134 4.125 3.857 1.08 1.01 1.68 11.684 6.51 5.056 296 555
L 0.11 4.125 3.905 1.08 1.022 1.387 11.977 5.377 5.182 243 456
M 0.095 4.125 3.935 1.08 1.03 1.203 12.161 4.661 5.262 210 394
DWV 0.058 4.125 4.009 1.08 1.05 0.741 12.623 2.872 5.462 128 240
5 K 0.16 5.125 4.805 1.342 1.258 2.496 18.133 9.671 7.846 285 534
pE
L 0.125 5.125 4.875 1.342 1.276 1.963 18.665 7.609 8.077 222 417
M 0.109 5.125 4.907 1.342 1.285 1.718 18.911 6.656 8.183 194 364
DWV 0.072 5.125 4.981 1.342 1.304 1.143 19.486 4.429 8.432 128 240
6 K 0.192 6.125 5.741 1.603 1.503 3.579 25.886 13.867 11.201 286 536
L 0.14 6.125 5.845 1.603 1.53 2.632 26.832 10.2 11.61 208 391
M 0.122 6.125 5.881 1.603 1.54 2.301 27.164 8.916 11.754 182 341
DWV 0.083 6.125 5.959 1.603 1.56 1.575 27.889 6.105 12.068 124 232
u
8 K 0.271 8.125 7.583 2.127 1.985 6.687 45.162 25.911 19.542 304 570
L 0.2 8.125 7.725 2.127 2.022 4.979 46.869 19.295 20.28 224 421
M 0.17 8.125 7.785 2.127 2.038 4.249 47.6 16.463 20.597 191 358
ro
DWV 0.109 8.125 7.907 2.127 2.07 2.745 49.104 10.637 21.247 122 229
10 K 0.338 10.125 9.449 2.651 2.474 10.392 70.123 40.271 30.342 304 571
L 0.25 10.125 9.625 2.651 2.52 7.756 72.76 30.054 31.483 225 422
M 0.212 10.125 9.701 2.651 2.54 6.602 73.913 25.584 31.982 191 358
12 K 0.405 12.125 11.315 3.174 2.962 14.912 100.554 57.784 43.51 305 571
L 0.28 12.125 11.565 3.174 3.028 10.419 105.046 40.375 45.454 211 395
G
M 0.254 12.125 11.617 3.174 3.041 9.473 105.993 36.706 45.863 191 358
a When using soldered or brazed fittings, the joint determines the limiting pressure.
b Working pressures were calculated using ANSI/ASME Standard B31.9 allowable stresses (ASME 1988). A 5% mill tolerance has been used on the wall
thickness. Higher tube ratings can be calculated using the allowable stress for lower temperatures.
c If soldered or brazed fittings are used on hard-drawn tubing, use the annealed ratings. Full-tube allowable pressures can be used with suitably rated flare
or compression-type fittings.
Source: ASHRAE Handbook—HVAC Systems and Equipment, Table 3, Chapter 46, 2012.
Table 3-4 Internal Working Pressure for Copper Tube Joints

Internal Working Pressure, psi
Sat. Steam
Service Water and Noncorrosive Liquids and Gasesa and
Alloy Used for Joints Temperature, Condensate
°F
Nominal Tube Size (Types K, L, M), in.
1/4 to 1 1 1/4 to 2 2 1/2 to 4 5 to 8a 10 to 12a 1/4 to 8
t
100 200 175 150 130 100 —
e
50-50 Tin/leadb solder 150 150 125 100 90 70 —
(ASTM B32 Gr 50A
[2014b]) 200 100 90 75 70 50 —
250 85 75 50 45 40 15
gp
100 500 400 300 270 150 —
95-5 Tin/antimonyc solder 150 400 350 275 250 150 —
(ASTM B32 Gr 50TA
[2014b]) 200 300 250 200 180 140 —
250 200 175 150 135 110 15
100 to 200 d d d d d —
Brazing alloys melting
250 300 210 170 150 150 —
at or above 1000°F
pE
350 270 190 150 150 150 120
a Solder joints are not to be used for (1) flammable or toxic gases or liquids or (2) gas, vapor, or compressed air in tubing over 4
in., unless maximum pressure is limited to 20 psig.
b Lead solders must not be used in potable-water systems.
c Tin/antimony solder is allowed for potable-water supplies in some jurisdictions.
d Rated pressure for up to 200°F applies to the tube being joined.
Source: ASHRAE Handbook—HVAC Systems and Equipment, Table 4, Chapter 46, 2012 and based on ANSI/ASME Standard
B31.9, Building Services Piping (ASME 1988).
u
Ductile Iron and Cast Iron Pipe
Ductile iron pipe is used for city water mains and waste drainage piping per
ro
ANSI/AWWA C150/A21.50 (AWWA 1996). These pipes use bell and spigot
joints, or mechanical or flanged joints. Cast iron is not used for pressure piping
and has been replaced by ductile iron pipe.
Joining Methods for Metal Pipe

G
Threading
Threading is the most commonly used method for joining small-diameter steel
or brass pipe, as shown in ANSI/ASME Standard Bl.20.1 (ASME 1983). Pipe
with a wall thickness less than standard weight should not be threaded. ANSI/
ASME Standard B31.5, Refrigeration Piping, limits the threading for various
refrigerants and pipe sizes (ASME 1992).
Soldering and Brazing

Copper tube is usually joined by soldering or brazing socket end fittings. Braz-
ing materials melt at temperatures over 1000°F and produce a stronger joint
than solder. Health concerns have caused many jurisdictions to ban solders
containing lead or antimony for joining pipe in potable-water systems. In par-
ticular, lead-based solders must not be used for potable-water systems.
Flared and Compression Joints
t
Flared and compression fittings can be used to join copper, steel, stainless
e
steel, and aluminum tubing. Properly rated fittings can keep the joints as strong
as the tube.
Flanges
gp
Flanges can be used for large pipes and all piping materials. They are com-
monly used to connect to equipment, valves, and wherever it may be necessary
to open the joint to permit service or replacement of components. For steel
pipe, flanges are available in pressure ratings to 2500 psig. For welded pipe,
weld neck, slip-on, or socket weld connections are available. Thread-on
flanges are available for threaded pipe.
pE
Flanges are generally flat faced or raised face. Flat-faced flanges with full-
faced gaskets are most often used with cast iron and materials that cannot take
high bending loads. Raised-face flanges with ring gaskets are preferred with
steel pipe because they facilitate increasing the sealing pressure on the gasket
to help prevent leaks. Other facings (such as O-rings and ring joints) are avail-
able for special applications.
All flat-faced, raised-face, and lap-joint flanges require a gasket between
the mating flange surfaces. Gaskets are made from rubber, synthetic elasto-
u
mers, cork, fiber, plastic, polytetrafluorethylene, metal, or a combination of
these materials. The gasket must be compatible with the flowing media and the
temperatures at which the system is operating.
Welding
ro
Welding steel pipe joints over 2 in. in diameter offers the following advantages:
• Welded joints do not age, dry out, or deteriorate as do gasketed joints.

• Welded joints can accommodate greater vibration and water hammer and
higher temperatures and pressures than other joints.
G
• For critical service, pipe joints can be tested by any of several nondestructive
examination (NDE) methods (such as by radiography or ultrasound).
• Welded joints provide maximum long-term reliability.
The applicable section of the ASME Boiler and Pressure Vessel Code
(2015) provides rules for welding. The code requires that all welders and weld-
ing procedure specifications (WPSs) be qualified. Separate WPSs are needed
for different welding methods and materials. The qualifying tests and the vari-
ables requiring separate procedure specifications are set forth in Section IX of
the code. The manufacturer, fabricator, or contractor is responsible for the weld-
ing procedure and welders. ASME Standard B31.9, Building Services Piping
(1988), requires visual examination of welds and outlines limits of acceptability.
The following welding processes are often used in the HVAC industry:
• Shielded-metal arc welding (SMAW), also called stick welding. The mol-
t
ten weld metal is shielded by the vaporization of the electrode coating.
• Gas metal arc welding (GMAW), also called metal inert gas (MIG). The
e
electrode is a continuously fed wire, which is shielded by argon or carbon
dioxide gas from the welding gun nozzle.
• Gas tungsten arc welding (GTAW), also called tungsten inert gas or heliarc.
This process uses a nonconsumable tungsten electrode surrounded by a shield-
gp
ing gas. The weld material may be provided from a separate noncoated rod.
Reinforced Outlet Fittings

Reinforced outlet fittings are used to make branch and takeoff connections and
are designed to permit welding directly to pipe without supplemental reinforc-
ing. Fittings are available with threaded, socket, or butt-weld outlets.
pE
Other Joints
Grooved joint systems require that a shallow groove be cut or rolled into the pipe
end. These joints can be used with steel, cast iron, ductile iron, and plastic pipes.
A segmented clamp engages the grooves, and the seal is provided by a special gas-
ket designed so that internal pressure tightens the seal. Some clamps are designed
with clearance between tongue and groove to accommodate misalignment and
thermal movements, while others are designed to limit movement and provide a
u
rigid system. Manufacturers’ data give temperature and pressure limitations.
Another form of mechanical joint consists of a sleeve slightly larger than
the OD of the pipe. The pipe ends are inserted into the sleeve, and gaskets are
packed into the annular space between the pipe and coupling and held in place
ro
by retainer rings. This type of joint can accept some axial misalignment, but it
must be anchored or otherwise restrained to prevent axial pullout or lateral
movement. Manufacturers provide pressure and temperature data.
Ductile iron pipe may be furnished with a bell-spigot end adapted for
caulk, gasket, and retainer ring, and mechanical or flanged joints. This joint is
also not restrained.
G
Threaded Unions
Unions allow disassembly of threaded pipe systems. Unions are three-part fit-
tings with a mating machined seat on the two parts that thread onto the pipe
ends. A threaded locking ring holds the two ends tightly together. A union also
allows threaded pipe to be turned at the last joint connecting two pieces of
equipment. Companion flanges (a pair) for small pipe serve the same purpose.
Special Systems
Certain piping systems are governed by separate codes or standards, which are
summarized below. Generally, any failure of the piping in these systems is dan-
gerous to the public, so local areas have adopted laws enforcing the codes.
• Boiler piping: ANSI/ASME Standard B31.1 (ASME 2014) and the ASME
Boiler and Pressure Vessel Code (2015) specify the piping inside the code-
required stop valves on boilers that operate up to 15 psig with steam or
t
160 psig with water and are limited to 250°F. Above these conditions,
codes require fabricators to be certified for such work. The field or shop
e
work must also be inspected while it is in progress by inspectors commis-
sioned by the National Board of Boiler and Pressure Vessel Inspectors.
• Refrigeration piping: ASME Standard B31.5, Refrigeration Piping and
Heat Transfer Components (1992), and ANSI/ASHRAE Standard 15, Safety
gp
Standard for Refrigeration Systems, cover the requirements for refrigerant
piping (ASHRAE 2013).
• Plumbing systems: Local codes cover piping for plumbing systems.
• Sprinkler systems: NFPA Standard 13, Standard for the Installation of
Sprinkler Systems, covers this field (2016).
• Fuel gas: ANSI Z223.1/NFPA 54, National Fuel Gas Code, prescribes fuel
gas piping in buildings (ANSI 2015).
pE
Plastic Pipe
Plastic pipe is gaining wider usage in HVAC and plumbing systems where
local building codes permit. Plastic is usually lighter in weight than metal, gen-
erally inexpensive, and corrosion-resistant. It also has a higher C factor (see
Chapter 2, Equation 2-9), requiring lower pumping power and allowing smaller
pipe sizes. The disadvantages of plastic pipe include the rapid loss of strength
at temperatures above ambient and the high coefficient of linear expansion.
u
The modulus of elasticity of plastics is low, resulting in short support span dis-
tances. Some jurisdictions do not allow certain plastics in buildings because of
toxic products emitted under fire conditions.
ro
Plastic piping materials fall into two main categories: thermoplastic and
thermoset. Thermoplastics melt and are formed by extruding or molding. They
are usually used without reinforcing filaments. Thermosets are cured and can-
not be reformed. They are normally used with glass fiber reinforcing filaments.
Plastic piping materials include the following:
• Polyvinyl chloride (PVC)
G
• Chlorinated polyvinyl chloride (CPVC)

• Polybutylene (PB)
• Polyethylene (PE)
• Cross-linked polyethylene (PEX)
• Polypropylene (PP)
• Acrylonitrile butadiene styrene (ABS)
• Polyvinylidene fluoride (PVDF)
Thermosetting piping systems used in the HVAC industry are referred to as

reinforced-thermosetting resin (RTR) and fiberglass-reinforced plastic (FRP).
RTR and FRP are interchangeable and refer to pipe and fittings commonly
made of fiberglass-reinforced epoxy resin, fiberglass-reinforced vinyl ester,
and fiberglass-reinforced polyester. Because pipe and fittings made from epoxy
resin are generally stronger and operate at higher temperatures than those made
from polyester or vinylester resins, they are often used in HVAC applications.
PEX is generally made from high-density polyethylene (HDPE). Cross-
t
linked polymer bonds are formed in the manufacturing process, changing the
characteristics to a thermoset. Bonds may be formed by irradiating the tube
e
with an electron beam or using other chemical means. PEX is often associated
with radiant heating or cooling systems; however, it is also applied in domestic
water distribution. Upper-end HVAC design criteria limit application to sys-
gp
tems under 180°F operating temperature. Combined with application of energy
design standards that might require high system design differential tempera-
tures, implying the use of condensing hot-water heating systems, PEX then
may also be considered as a viable piping alternative where allowed by code.
What makes this consideration interesting, aside from beneficial HVAC system
operating characteristics of the material, is consideration of material econom-
ics. While steel piping is often applied for larger piping runs, it has not been
uncommon to see piping predominantly installed with copper piping. However,
pE
copper prices have been quite volatile. During the period from 2004 to 2011,
copper went through two significant increases in price at the metals market
exchange level, rising from a low of about $1.50 per pound to just under $5.00
per pound. Prices for finished product, such as pipe, have risen accordingly.
Compared to rigid metal piping systems, alternative methods of pipe support
are required; the significant difference in material costs may make this attrac-
tive.
u
Allowable Stresses
Both thermoplastics and thermosets have allowable stresses derived from a
ro
hydrostatic design basis stress (HDBS). The HDBS is determined by a statistical

analysis of both static and cyclic stress-rupture test data as set forth in ASTM
Standard D2837 (1992) for thermoplastics and ASTM Standard D2992 (1991)
for glass fiber RTRs.
The allowable stress, which is called the hydrostatic design stress (HDS), is
obtained by multiplying the HDBS by a service factor. The HDS values recom-
G
mended by some manufacturers and those allowed by ANSI/ASME B31.9

(1988) are listed in Table 3-5.
The pressure design thickness for plastic pipe can be calculated using the
code stress values and the following formula:
pD
t = --------------------- (3-1)
2 S+p
Table 3-5 Properties of Plastic Pipe Materialsa
Hydrostaticb Upper
Design Stress, psi Temperature
Designation
G Material
Type and
Grade
Cell
No.
Tensile
Strength,
psi
(at 73°F)
(at 73°F)
Mfr.
ASME
B31
Limit, °F
Mfr.
ASME
B31
HDSb
Upper
Limit,
psi
Specific
Gravityc
Impact
Strength,
ft·lb/in.
(at 73°F)
Elasticity, Expansion,
psi (at 73°F) per °F/106 in.
Thermal
Modulus of Coefficient of Conductivity,
Btu·in/
h·ft2 ·°F
Relative
Pipe
Costd
Thermoplastics
PVC 1120 T I,G1 12454-B 7,500 2,000 2,000 140 150 440 1.40 0.8 420,000 30.0 1.1 1.0
PVC 1200 T I,G2 12454-C 2,000 150 410,000 35.0
PVC 2120 T II,G1 14333-D 2,000 150 30.0
CPVC 4120 T IV,G1 23447-B 8,000 2,000 2,000 210 210 320 1.55 1.5 423,000 35.0 0.95 2.9
ro
PB 2110 T II,G1 4,800 1,000 1,000 180 210 <500 0.93 38,000 72.0 1.5 2.9
PE 2306 Gr. P23 630 140 90,000 80.0
PE 3306 Gr. P34 630 160 130,000 70.0
PE 3406 Gr. P33 630 180 150,000 60.0
HDPE 3408
PP
ABS
Gr. P34
Acrylonitrile
u
355434-C
6-3-3
5,000
5,000
5,500
1,600
705
800 140
212
176
180
210
800 0.96
0.91
1.06
12
1.3
8.5
110,000
120,000
240,000
120.0
60.0
56.0
2.7
1.3
1.7
1.1
2.9
3.4
copolymer
ABS 1210 T I,G2 5-2-2 1,000 180 640 250,000 55.0
ABS 1316 T I,G3 3-5-5 1,600 180 1,000 340,000 40.0
ABS 2112 T II,G1 4-4-5 1,250 180 800 40.0
PVDF 7,000 1,275 280 275 306 1.78 3.8 125,000 79.0 0.8 28.0
pE
Thermosetting
Epoxy-Glass RTRP- 44,000 8,000 300 7,000 1,000,000 9 to 13 2.9
11AF
Polyester- RTRP-12EF 44,000 9,000 200 200 5,000 1,000,000 9 to 11 1.3
Glass
For Comparison
Steel A 53 B ERW 60,000 12,800 800 9,200 7.80 30.0 27,500,000 6.31 344 1.3
Copper Type L Drawn 36,000 9,000 400 8,200 8.90 17,000,000 9.5 3.5
gp
a Properties listed are for specific materials listed as each plastic has other formulations. Consult the manufacturer of the system chosen. These values are for comparative purposes.
b The hydrostatic design stress (HDS) is equivalent to the allowable design stress.
c Relative to water at 62.4 lb/ft3.

d
Based on cost of pipe only, without factoring in fittings, joints, hangers, and labor.
Source: 2014 ASHRAE Handbook—Refrigeration.
71
e t
where
D = pipe OD, in.
t = pressure design thickness, in.
p = internal design pressure, psig
S = HDS, psi
The minimum required wall thickness can be found by adding allowances
for mechanical strength, threading, grooving, erosion, and corrosion to the cal-
t
culated pressure design thickness. As there are many formulations of the poly-
mers used for piping materials and different joining methods for each system,
e
manufacturers’ recommendations should be observed. Most catalogs give the
pressure ratings for pipe and fittings at various temperatures up to the maxi-
mum the material will withstand.
gp
Plastic Material Selection
The selection of a plastic for a specific purpose requires careful attention. All
are suitable for cold water. However, plastic pipe should not be used for com-
pressed gases or compressed air if the pipe is made of a material subject to brit-
tle failure. For other liquids and chemicals, refer to charts provided by plastic
pipe manufacturers and distributors. Table 3-6 gives a synopsis of some appli-
pE
cations that are pertinent to the HVAC industry. A brief description follows:
Table 3-6 Manufacturers’ Recommendations for Plastic Materialsa,b

Application PVC CPVC PB HDPE PP ABS PVDF RTRP
Cold-water service R R R R R R R R
Hot (140°F) water N R R R R R R R
u
Potable-water service R R R R R R R R
Drain, waste, and vent R R N — R R — —
Demineralized water R R — — R R R —
ro
Deionized water R R — — R R R R
Salt water R R R R R R — R
Heating (200°F) hot water N N N N N N — R
Natural gas N N N R N N — —
Compressed air N N — R N R — —
G
Sunlight and weather resistance N N N R — R R R

Underground service R R R R R R — R
Food handling R R — — R R R R
R = Recommended, N = Not recommended, — = Insufficient information
a Before selecting a material, check the availability of a suitable range of sizes and fittings and of a satisfactory joining method.
Also have the manufacturer verify the best material for the purpose intended.
b Local building codes should be consulted for compliance of the materials listed.
• PVC: PVC has the best overall range of properties at the lowest cost; it is
the most widely used plastic. It is joined by solvent cementing, threading,
or flanging. Gasketed push-on joints are also used for larger sizes.
• CPVC: CPVC has the same properties as PVC but can withstand higher tem-
peratures before losing strength. It is joined by the same methods as PVC.
• PB: A lightweight, flexible material, PB can be used up to 210°F. Check
local building codes for using PB. Generally, PB is not used in building
construction due to several well-publicized product failure incidents.
t
• LDPE: LDPE is a flexible, lightweight tubing with good low-temperature
e
properties. It is used in the food and beverage industry and for instrument
tubing. It is joined by mechanical means such as compression fittings or
push-on connectors and clamps.
• HDPE: A tough weather-resistant material, HDPE is used for large pipe-
gp
lines in the gas industry. Fabricated fittings are available. It is joined by
heat fusion for large sizes, and flare, compression, or insert fittings can be
used on small sizes.
• PP: A lightweight plastic, PP is used for chemical waste lines and also for
pressure applications, as it is inert to a wide range of chemicals. A wide
variety of drainage fittings are available. For pressure uses, regular fittings
are made. It is joined by heat fusion.
pE
• ABS: ABS is a high-strength, impact- and weather-resistant material. Cer-
tain formulations can be used for compressed air, and ABS is also used in
the food and beverage industry. A wide range of fittings are available. It is
joined by solvent cementing, threading, or flanging.
• PVDF: Widely used for ultrapure water systems and in the pharmaceutical
industry, PVDF has a wide temperature range. This material is over
20 times more expensive than PVC. It is joined by heat fusion, and fittings
are made for this purpose. For smaller sizes, mechanical joints can be used.
u
Local building codes should always be consulted for material compliance
for any HVAC, water supply, gas supply, or compressed-air applications.
Corrosion
ro
For the purposes of this section, corrosion may be deemed as the destruction of
a metal or alloy by chemical or electrochemical (galvanic) reaction with its
environment. Usually this is an electrochemical reaction similar to the one that
occurs in a dry-cell battery. While a complete discussion of the theory and
G
mechanism of corrosion is far outside the scope of this course, some basic prin-
ciples should be kept in mind when designing a piping system.
Factors that support or promote corrosion include the following:
• Oxygen
• Solutes (soluble chemical salts that combine with moisture to form elec-
trolytes)
• Moisture (combines with oxygen and solutes to form electrolytes)
• Dissimilar metals (such as brass and steel) in electrical contact

• Stresses in metals
• Temperature
• Pressure
• Velocity (for example, velocity of water flowing in a pipe)
Corrosion control methods include the following:
t
• Materials selection: Constructing HVAC systems of highly corrosion-
resistant material is often not possible because of economic and physical
e
limitations.
• Cathodic protection: Cathodic protection causes an electrochemical reac-
tion to occur somewhere other than on the protected metal. Two types of
cathodic protection are encountered: sacrificial and impressed current.
gp
With sacrificial cathodic protection, the metal to be protected (the cathode)
is electrically connected to a sacrificial anode, which corrodes instead of
the protected metal. Impressed current cathodic protection uses an external
voltage source to effect the required protection.
• Protective coatings: Protective coatings work either as a barrier, prevent-
ing the metal to be protected from coming into contact with an electrolyte,
sacrificially (as in galvanized pipe where the zinc coating protects the steel
pE
pipe), or as a combination of both barrier and sacrificial coating.
• Environmental treatment: Corrosion inhibitors that retard the corrosion
reaction may be added to water systems or the water may be mechanically
deaerated to remove oxygen.
• Design: Equipment design modifications, such as eliminating crevices
where moisture can accumulate and providing weep holes to allow mois-
ture to drain, may be used to reduce the likelihood of corrosion.
u
Corrosion environments of particular concern to the piping system designer
include the following:
• Underground corrosion: Corrosion on buried pipes must always be antic-
ro
ipated. A corrosion survey to determine the specific conditions in which the

piping will be placed should be conducted so that informed decisions about
corrosion control measures may be made. Underground factors to consider
include types of soils, bacterial activity, and thermal insulation.
• Water-side corrosion and deposits: The most common water problems
include corrosion, scale formation, biological growths, and suspended solid
G
matter.
Control of water-side corrosion and deposits is effected using water treat-

ment with corrosion-inhibiting chemicals and mechanical treatment (filtering
and mechanical deaeration).
The 2015 ASHRAE Handbook—HVAC Applications contains a more com-
plete discussion of corrosion, corrosion control, protective measures, and water
treatment (ASHRAE 2015).
Valves and Fittings

The 2013 ASHRAE Handbook—Fundamentals points out that valves and fit-
tings cause pressure losses greater than those caused by the pipe alone
(ASHRAE 2013b). The pressure drop h, in feet of fluid, is shown as
V2
h = K ------ (3-2)
2g
t
where
e
g = 32.2 ft/s2, gravitational constant
K = resistance coefficient for valve or fitting (see Tables 3-7 and 3-8)
2
V /2g = velocity head, ft (Table 2-1), based on flow gallons per minute in
gp
connect pipe
V = fluid velocity, ft/s
h = pressure drop in feet of fluid
To use Equation 3-2, values for K are tabulated for a range of typical valves
and fittings. Tables 3-7 and 3-8 show K values for typical fittings and valves.
Table 3-9 shows the approximate range of variations for the K values deter-
mined by the Hydraulic Institute. The Hydraulic Institute Engineering Data
pE
Book shows reference tables for typical K values for valves, fittings, increasers,
diffusers, and reducers (Figures 3-1 through 3-4) (HI 1990).
Table 3-7 K Factors—Threaded Pipe Fittings

Nom
90° 90°
-inal Ret- Swing Bell Pro-
Stan- Long- 45° Tee- Tee- Globe Gate Angle Square
Pipe urn Check Mouth jected
u
dard Radius Elbow Line Branch Valve Valve Valve Inlet
Dia., Bend Valve Inlet Inlet
Elbow Elbow
in.
3/8 2.5 — 0.38 2.5 0.9 2.7 20 0.4 — 8 0.05 0.5 1
ro
1/2 2.1 — 0.37 2.1 0.9 2.4 14 0.33 — 5.5 0.05 0.5 1
3/4 1.7 0.92 0.35 1.7 0.9 2.1 10 0.28 6.1 3.7 0.05 0.5 1
1 1.5 0.78 0.34 1.5 0.9 1.8 9 0.24 4.6 3 0.05 0.5 1
1 1/4 1.3 0.65 0.33 1.3 0.9 1.7 8.5 0.22 3.6 2.7 0.05 0.5 1
G
1 1/2 1.2 0.54 0.32 1.2 0.9 1.6 8 0.19 2.9 2.5 0.05 0.5 1
2 1 0.42 0.31 1 0.9 1.4 7 0.17 2.1 2.3 0.05 0.5 1
2 1/2 0.85 0.35 0.3 0.85 0.9 1.3 6.5 0.16 1.6 2.2 0.05 0.5 1
3 0.8 0.31 0.29 0.8 0.9 1.2 6 0.14 1.3 2.1 0.05 0.5 1
4 0.7 0.24 0.28 0.7 0.9 1.1 5.7 0.12 1 2 0.05 0.5 1
Source: Engineering Data Book (HI 1990).
Table 3-8 K Factors—Flanged Welded Pipe Fittings

Nom- 90° 90° 45° Return Return
Swing
inal Stan- Long- Long- Bend Bend Tee- Tee- Globe Gate Angle
Check
Pipe dard Radius Radius Stan- Long- Line Branch Valve Valve Valve
Valve
Dia., in. Elbow Elbow Elbow dard Radius
1 0.43 0.41 0.22 0.43 0.43 0.26 1 13 — 4.8 2
1 1/4 0.41 0.37 0.22 0.41 0.38 0.25 0.95 12 — 3.7 2
t
1 1/2 0.4 0.35 0.21 0.4 0.35 0.23 0.9 10 — 3 2
e
2 0.38 0.3 0.2 0.38 0.3 0.2 0.84 9 0.34 2.5 2
2 1/2 0.35 0.28 0.19 0.35 0.27 0.18 0.79 8 0.27 2.3 2
3 0.34 0.25 0.18 0.34 0.25 0.17 0.76 7 0.22 2.2 2
4 0.31 0.22 0.18 0.31 0.22 0.15 0.7 6.5 0.16 2.1 2
gp
6 0.29 0.18 0.17 0.29 0.18 0.12 0.62 6 0.1 2.1 2
8 0.27 0.16 0.17 0.27 0.15 0.1 0.58 5.7 0.08 2.1 2
10 0.25 0.14 0.16 0.25 0.14 0.09 0.53 5.7 0.06 2.1 2
12 0.24 0.13 0.16 0.24 0.13 0.08 0.5 5.7 0.05 2.1 2
pE
Table 3-9 Approximate Range of Variation for K Factors
Threaded,
±20% above 2 in. ±25%
Regular line or branch
Tee
threaded Flanged,
±40% below 2 in. ±35%
line or branch
u
90° Elbow Long-radius
±25% Threaded ±25%
threaded Globe valve
Regular flanged ±35% Flanged ±25%
Long-radius
ro

flanged
Gate valve
Regular
±10% Flanged ±50%
threaded
45° Elbow
Long-radius
flanged
Angle valve
G
Regular
±25% Flanged ±50%
threaded
Return bend Regular flanged ±35% Threaded ±50%
(180°)
Long-radius Check valve +200%
±30% Flanged
flanged –80%
e t
gp
u pE
ro
G
Figure 3-1 Resistance coefficients like Figure 3-2 for valves and fittings.
Source: HI (1990).
e t
gp
u pE
ro
G
Figure 3-2 Resistance coefficients for valves and fittings.

Source: HI (1990).
e t
gp
pE
Figure 3-3 Resistance coefficients for increasers and diffusers.
Source: HI (1990).
u
ro
G
Figure 3-4 Resistance coefficients for reducers.

Source: HI (1990).
Backflow Prevention Devices

As discussed in the 2012 ASHRAE Handbook—HVAC Systems and Equipment,
backflow prevention devices prevent reverse flow between the city water sup-
ply and the building connection (ASHRAE 2012). The American Backflow
Prevention Association (ABPA 2015) states that
Water distribution systems are designed with the intention of the
t
water flowing in a certain direction—from the distribution system to
the consumer. However, hydraulic conditions within the system may
e
deviate from the “normal” conditions, causing water to flow in the
opposite direction. Therefore, it is possible (and common) for the
water to flow in the opposite direction in unprotected systems. This is
called backflow.
gp
ABPA continues their discussion by defining cross-connections:
A cross-connection is an unprotected actual or potential connection
between a potable water system used to supply water for drinking
purposes and any source or system containing unapproved water or a
substance that is not or cannot be approved as safe, wholesome, and
potable. Bypass arrangements, jumper connections, removable sec-
tions, swivel or changeover devices, or other devices through which
pE
backflow could occur, shall be considered to be cross-connections
(ABPA 2015).
ABPA correlates this with backsiphonage being caused by mainline pip-
ing failures or drafting due to high demands.
Vacuum breakers prevent backsiphonage in nonpressurized water sys-
tems, while backflow preventers prevent backflow in pressurized systems
(Figure 3-5).
u
ro
G
Figure 3-5 Backflow prevention device.

Source: ASHRAE Handbook—HVAC Systems and Equipment (2012), Chapter 47, Figure 29.
Selection of Devices
Vacuum breakers and backflow preventers should be selected on the basis of the
local plumbing codes, the water supply impurities involved, and the type of
cross-connection. Impurities are classified as contaminants, which are substances
that could create a health hazard if introduced into potable water, and pollutants,
which could create an objectionable condition but not a health hazard.
Cross-connections are classified as nonpressurized or pressurized connec-
t
tions. In a nonpressurized cross-connection, a potable-water pipe connects or
extends below the overflow or rim of a receptacle at atmospheric pressure. When
e
this type of connection is not protected by a minimum air gap, it should be pro-
tected by an appropriate vacuum breaker or an appropriate backflow preventer.
In a pressurized cross-connection, a potable-water pipe is connected to a
gp
closed vessel or a piping system that is above atmospheric pressure and con-
tains a nonpotable fluid. This connection should be protected by an appropriate
backflow preventer only. Note that a pressure vacuum breaker should not be
used alone with a pressurized cross-connection.
Vacuum breakers should be corrosion resistant. Backflow preventers
(including accessories, components, and fittings 2 in. and smaller) should be
made of bronze with threaded connections. Sizes larger than 2 in. should be
made of bronze, galvanized iron, or fused epoxy-coated iron inside and out,
pE
with flanged connections. All backflow prevention devices should meet appli-
cable standards of ANSI, the CSA Group, and/or the required local authorities.
Installation of Devices
Vacuum breakers and backflow preventers equipped with atmospheric vents or
with relief openings should be installed and located to prevent any vent or relief
opening from being submerged. They should be installed in the position recom-
u
mended by the manufacturer. Backflow preventers may be double-check-valve
(DCV) or reduced-pressure-zone (RPZ) types. Refer to manufacturers’ informa-
tion for specific application recommendations and code compliance.
ro
Pipe Selection
The following example demonstrates how sizing and selection are accom-
plished using the engineering data information in Chapters 2 and 3.
G
Example 3-1
Note: In this exercise, feet of pressure drop is indicated as ft, and feet of distance is
indicated as ft.
Determine the pipe sizing and total pressure drop (PD) for the piping zone
A-B of the system shown in Figure 3-6. You are given that the system consists
of the following:
e t
gp
Figure 3-6 Pipe loop sizing example.
• 1 heating coil rated at 38 gpm and 3 ft drop

pE
• 1 gate valve: 2 in. screwed pipe
• 1 control valve: 1.5 in. rated at 38 gpm at 8.2 ft PD
• 1 balance valve: 2 in. rated at 40 gpm at 2.0 ft PD
• 2 tees (branch flow)
• 4 elbows
• Total pipe length = 300 ft (screwed pipe)
u
Solution 1. Recalling that pipe selection guidelines suggest keeping friction loss less than
4 ft/100 ft and design velocity greater than 2 fps, Table 2-1 allows the designer
to plot out a preliminary pipe selection. For 40 gpm, limited by a head loss of
ro
4 ft/100 ft, a 2 in. pipe has much less PD than a 1 1/2 in. pipe as an alternate
size. At 40 gpm, a 2 in. pipe has a PD of 3.l ft/100 ft. Checking 1 1/2 in. pipe,
the PD is 8.5 ft/100 ft; 2 1/2 in. pipe has a PD of 1.28 ft/100 ft, but the velocity
is starting to approach 2 fps. The pipe size of 2 in. is selected because it meets
most closely the friction loss and water velocity requirements (Figure 3-7).
Pipe that is 2 in. will also be less expensive than pipe that is 2 1/2 in.
G
2. Determine the pressure loss due to friction for the 300 ft of piping:
hf = 300 ft × 2.88 ft/100 ft = 8.4 ft
Adding in the effect of aging:
hf = 1.15 × 8.4 ft = 9.66 ft
From Table 2-1, the value of V2/2g = 0.227 ft
e t
gp
Figure 3-7 Pipe sizing using friction loss chart.
pE
3. Calculate the PD for the gate valve. Looking at Tables 3-7 and 3-8, K for a
2 in. gate valve is 0.17.
hf = kV2/2g
= (0.17)(0.227)
= 0.04 ft
4. Determine the PD due to the 90° regular pattern elbows PD we know:
K = 1 (from Tables 3-7 and 3-8)
hf = (4 elbows) × k(V2/2g)
u
= (4) × (1.0)(0.227)
= 0.91 ft
5. Similarly, the PD for the tee joints is as follows:
ro
K = 1.4 (from Tables 3-7 and 3-8, 2 in. tee branch)

hf = (2 tee joints) × k(V2/2g)
= (2) × (1.4)(0.227)
= 0.64 ft
6. The total PD is the sum of all component contributions.
G
Total hf = coil + pipe + balance valve + control valve + gate valve

+ tee branches + elbows
= (3.0 ft) + (9.66 ft) + (2.31 ft) + (8.34 ft) + (0.04 ft) + (0.64 ft) + (0.91 ft)
= 24.89 ft
System Characteristic Curve

The system characteristic curve, or system curve for short, is an engineering
convenience for visualizing the relationship of flow and PD (head loss) through
the system at various operating points.
The relationship follows the essential relationship of the Darcy-Weisbach
equation comparing what the friction loss is in a fixed system at one flow rate
t
and then comparing the friction loss developed at a different flow rate in the
same flow path. As pipe diameter and length do not change, and as long as
e
gravity remains constant, then knowing friction loss at a given flow rate allows
one to calculate a new friction loss (pressure drop) for a given flow rate per the
following:
gp
Q2 2 h
-------- = ----2-
Q1 h1
where
h1 = known pressure drop in feet of fluid
h2 = unknown pressure drop in feet of fluid
pE
Q1 = known flow rate, gpm
Q2 = new flow rate, gpm
For example, in Equation 3.1 we determined that for the flow of 40 gpm,
there was 24.89 ft of head loss. If we wanted to double the flow rate, we could
predict that if the pipe sizes did not change, the following would be true:
80 2 h2
u
--------- = ------------
-
40 24.89
4 24.89 = h 2 = 99.56 ft
ro
Plotting a collection of points from zero flow to whatever is desired devel-

ops an exponential curve (Figure 3-8). How precise is the calculation? Rhetori-
cally, one can already see an issue if one ascribes to the Hazen-Williams
equation because it uses an exponent of 1.85. However, as this is a basic entry
text, further and deeper study will eventually lead one to understand that fitting
G
loss factors do not always stay constant at every velocity, and there can be vari-
ation also impacting accuracy, especially when one considers the relationship
of fitting quantity to straight pipe. Yet, examined closely, there is not a whole
heck of a lot of graphical difference between an exponent of 1.85 and 2, and 2
is much easier to work with, so a square (2) relationship is always used; as the
industry is fond of noting, it is not a big enough difference to make a differ-
ence.
e t
Figure 3-8
gp
Exponential curve of flow versus pressure drop.
pE
We will see as we progress that this relationship helps us to understand the
proper pump selection and also allows us to calculate and simulate different sys-
tem operation points for system analysis. It is a simple and powerful calculation.
If a system is employing two-way valves and the load requires part-load
conditions, then the room thermostats will partly close the control valves. This
reduces the total flow through the system by increasing the pressure loss on the
control valve. The result is that a new system curve exists, and the curve rotates
u
counterclockwise as the valve closes (Figure 3-9).
What if the same 38 gpm at the head loss of 24.89 existed in an open system
such as that diagrammed in Figure 3-10? If there were a 15 ft elevation difference
between where water enters the fill of the tower and the top of the water level in
ro
the sump, how would the system curve look?

As previously discussed, we use the concepts of Bernoulli, recalling that
we have to allow for the difference in elevation between points one and two:
2 2
V P V P
z 1 + ------1 + -----1- = z 2 + ------2 + -----2- + H L
2g 2g
G
1 2
As the only difference between the two examples is the elevation difference,
all other terms cancel out and we have only to compensate for the independent
head loss of the difference. That yields a curve as shown in Figure 3-11, plotted
from the data attributed to Figure 3-12.
In the next chapter, we will see this relationship’s effect on the centrifugal
pump curve.
e t
gp
pE
Figure 3-9 Curve rotates counterclockwise as the valve closes.
u
ro
G
Figure 3-10 Open-system diagram.

e t
gp
pE
Figure 3-11 Change in curve because of elevation difference.
u
ro
G
Figure 3-12 Data for plotting Figure 3-11.

The Next Step

In the next chapter you will learn about centrifugal pumps.
Summary
In this chapter we covered the following:
t
• The U.S. organizations that issue codes and standards for piping systems
and components
e
• The different types of pipe used in HVAC water system applications and
the characteristics of each
• The pipe-joining methods commonly encountered in HVAC water systems
and the characteristics of each
gp
• Common factors that support or promote corrosion, the five methods of
corrosion control, and the two corrosion environments of particular con-
cern to the HVAC piping system designer
• The function, selection, and installation of backflow prevention devices
• How to solve a basic pipe selection problem
Bibliography
pE
ABPA. 2015. About the ABPA. http://www.abpa.org/?page=ABout.
ANSI. 1992. Z223.1-1992, National Fuel Gas Code. New York: American
National Standards Institute.
ANSI. 2015. ANSI Z223.1/NFPA 54, National Fuel Gas Code. New York:
American National Standards Institute.
ASHRAE. 2012. Handbook—HVAC Systems and Equipment. Atlanta: ASHRAE.
ASHRAE. 2013a. ANSI/ASHRAE Standard 15-2013, Safety Standard for
u
Refrigeration Systems. Atlanta: ASHRAE.
ASHRAE. 2013b. ASHRAE Handbook—Fundamentals. Atlanta: ASHRAE.
ASHRAE. 2014. ASHRAE Handbook—Refrigeration. Atlanta: ASHRAE.
ASHRAE. 2015. ASHRAE Handbook—HVAC Applications. Atlanta: ASHRAE.
ro
ASME. 1983. ANSI/AMSE Standard Bl.20.1-1983, Pipe Threads, General

Purpose. New York: ASME.
ASME. 1988. ANSI/ASME Standard B31.9-1988, Building Services Piping.
New York: ASME.
ASME. 1992. ANSI/ASME Standard B31.5-1992, Refrigeration Piping and
Heat Transfer Components. New York: ASME.
G
ASME. 2004. ASME Standard B36.10M-2004, Welded and Seamless Wrought

Steel Pipe. New York: ASME.
ASME. 2014. ASME B31.1-2014, Power Piping. New York: ASME.
ASTM. 1991. ANSI/ASTM Standard D2992-91, Practice for Obtaining
Hydrostatic or Pressure Design Basis for Fiberglass Pipe and Fittings.
West Conshohocken, PA: ASTM International.
ASTM. 1992. ASTM Standard D2837-92, Obtaining Hydrostatic Design Basis

for Thermoplastic Pipe Materials. West Conshohocken, PA: ASTM Inter-
national.
ASTM. 1995a. ASTM Standard B88-95a, Specification for Seamless Copper
Water Tube. West Conshohocken, PA: ASTM International.
ASTM. 1995b. ASTM Standard B280-95a, Seamless Copper Tube for Air
Conditioning and Refrigeration Field Service. West Conshohocken, PA:
ASTM International.
t
ASTM. 2012. ASTM A53/A53M-12, Standard Specification for Pipe, Steel,
Black and Hot-Dipped, Zinc-Coated, Welded and Seamless. West Con-
e
shohocken, PA: ASTM International.
ASTM. 2014a. ASTM A106/A106M-14, Standard Specification for Seamless
Carbon Steel Pipe for High-Temperature Service. West Conshohocken, PA:
ASTM International.
gp
ASTM. 2014b. ASTM B32-08(2014), Standard Specification for Solder Metal.
West Conshohocken, PA: ASTM International.
AWWA. 1996. ANSI/AWWA C150/A21.50, American National Standard for
Thickness Design of Ductile-Iron Pipe. Denver: American Water Works
Association.
Bonds, Richard W. 2007. Ductile iron pipe joints and their uses. http://
c.ymcdn.com/sites/www.isawwa.org/resource/resmgr/handouts_for_seminars
pE
/joints.pdf.
HI. 1990. Engineering Data Book, 2d ed. Cleveland, OH: Hydraulic Institute.
NFPA. 2016. NFPA 13, Standard for the Installation of Sprinkler Systems.
Quincy, MA: National Fire Protection Association.
u
ro
G

3-1 What US organizations issue codes and standards for piping systems and com-
ponents?
3-2
t
What is the allowable working pressure (ASTM A53B to 4000°F) for each of
the following steel pipe diameters:
e
Nominal Size and Pipe OD Schedule Working Pressure, psig
10 in., D = 10.75 40XS
gp
20 in., D = 20 30XS
4 in., D = 4.5 40ST
8 in., D = 8.625 80XS
3-3 List three joining methods used with copper tubing.

3-4 List five methods of joining metal pipe.
pE
3-5 Name and briefly describe the two main categories of plastic piping materials,
and list at least three of each type of plastic pipe.
3-6 List the ASME B31 HDSs (psi at 73°F) for each of the following plastic pipe
materials:
Material ASME B31 Hydrostatic Design Stress (psi at 73°F)

PB 2110, Type II, Gl
u
ABS 1210 TI, G2
3-7 List the five methods of corrosion control.

3-8
ro
List the K values for each of the following screwed pipe fittings:
Nominal Pipe Diameter, in. Fitting Type K Value

1.25 90° ell long
0.5 Globe valve
3 Tee branch
G
3-9 What is the function of a backflow prevention device?

3-10 Determine the pipe sizing and total PD for the piping system shown below. You
are given that the system consists of:
• 1 gate valve: 2 in.
• 1 control valve: 2 in. rated at 35 gpm at 9 ft of pressure drop

• 1 balance valve: 2 in. rated at 40 gpm at 2.0 ft of pressure drop
• 2 tee branches
• 4 elbows
• Total pipe length = 200 ft (schedule 40 screwed pipe)
e t
gp
u pE
ro
G
G
ro
upE
gp
et
Centrifugal Pumps
e t
Study Objectives
gp
After studying the material in this chapter you should be able to
❑ understand how a centrifugal pump changes the flow and head of a
hydronic system,
❑ name six types of centrifugal pumps and where they are used,
❑ know how the head–capacity curve of a centrifugal pump is determined,
❑ understand the difference between family and individual pump curves,
know what characteristics can be learned from a manufacturer’s pump curve,
pE
❑
❑ determine horsepower for a centrifugal pump,
❑ know what pump efficiency is,
❑ understand the radial thrust of a centrifugal pump,
❑ understand pump affinity laws and how to apply them, and
❑ know what NPSH and NPSHR are.
Instructions
u
Types of Pumps
ro
This chapter deals with pumps: the types of pumps available for water systems,
the factors that must be considered in pump selection, and how to integrate
pumps into a system to achieve maximum energy efficiency.
Common Applications
G
Some have said that the pump is next to humanity’s most important of inven-
tions: the wedge, the lever, and the wheel. Many types of pumps have been devel-
oped, but the centrifugal pump is the most widely used in HVAC applications.
Centrifugal pumps circulate hot water in heating systems and chilled water
(CHW) for cooling systems to satisfy a predetermined rate of flow between the
boiler or chiller and their respective space-conditioning terminal units to satisfy
the load. This is illustrated for CHW by the system shown in Figure 4-1.
94 Chapter 4 Centrifugal Pumps
Other pump applications on hydronic systems include condenser water

circuits to cooling towers (Figure 4-2) and water-source heat pumps. When
a cooling tower rejects heat for a CHW plant, the condenser water pumps
are selected on the basis of the flow rate required by the load, the selected
refrigeration equipment, and the physical location of the tower relative to
the pump and condenser. Figure 4-2 shows a typical example of flow rates
and load temperature differences for a chiller’s evaporator and condensing
loops.
e t
gp
pE
Figure 4-1 Diagram of a CHW system.
u
ro
G
Figure 4-2 Condenser water circuits to cooling tower.

When applying a steam boiler, boiler feed pumps and condensate return
pumps are normally required in the system (Figure 4-3). Boiler manufacturers
recommend specific pumping requirements.
Operation
A pump is made up of three key elements: a volute, which has a rotating
t
impeller within it; a driving mechanism, which connects the shaft of a driver to
the pump impeller; and the driver itself (see Figure 4-4). Typically, the driver in
e
HVAC centrifugal pumps is an electric motor. Its output energy is transferred
via a coupler to the impeller’s shaft. As the impeller is rotated, centrifugal
gp
pE
Figure 4-3 Boiler feed and condensate return pumping.
u
ro
G
Figure 4-4 Driver for HVAC pumps.

force imparts energy into the fluid being pumped. The fluid is hurled away
from the eye of the impeller, then directed through the impeller by vanes
within. As can be seen in Figures 4-5 and 4-6, the vanes are backward inclined
and slap at the fluid; the vanes do not bite into the fluid. The water leaves the
impeller at a high velocity and, as can be seen in Figure 4-7, the energy is cap-
tured and directed through the impeller enclosure (the volute).
The rotation of the impeller adds energy to the fluid after it enters the eye
t
of the impeller, accelerating it in the volute (Figure 4-8). The fluid is then acted
upon by the outward centrifugal force and the rotational or tip speed force, as
e
shown in the vector diagram in Figure 4-9. These two forces result in a pres-
sure energy increase due to the velocity increase of the fluid. The water pas-
gp
u pE
Figure 4-5 Vanes directing fluid.
ro
G
Figure 4-6 Vanes are inclined backward.

e t
Figure 4-7
gp
Energy captured and directed through the impeller enclosure, called a volute.
u pE
ro
G
Figure 4-8 Centrifugal pump, impeller, and volute.

e t
gp
Figure 4-9 Impeller action on fluid.
pE
sage flow area is increased due to the design of the volute and the diameter of
the impeller (Figure 4-8). The pump casing or volute collects the liquid as it
leaves the impeller blades by the centrifugal force and guides it out the volute
throat into the discharge nozzle.
For the pump to operate properly, water must enter the volute suction in an
even and symmetrical manner. The pump shown in Figure 4-4 is a single-suction
frame (base) mounted centrifugal pump, which has a suction accessory to serve
this purpose. The suction accessory is not a requirement, as it serves the same pur-
u
pose as having several pipe diameters of straight pipe entering the pump suction.
Accessories such as this, though, are often used because there is generally not
enough space for the required length of pipe. How much straight pipe is required
is a function of entering pipe geometry (e.g., the number of changes in axis due to
ro
elbows and fittings, and velocity), with five pipe diameters being the typical
industry application recommendation. However, in conditions where there are
many changes in fluid direction, this number can be as high as 20 pipe diameters.
Using the Bernoulli equation, with modification for energy input and
energy loss, typically yields the following:
G
2 2
V 1 P1 V P
------ + ------ + Z 1 + E p = ------2 + -----2- + Z 2 + h L (4-1)
2g 2g 2
When considering a closed piping system, on a given elevation datum line

of pipe, if two points of operation were considered that were very close to each
other (where the pipe size does not change, there is no change in fluid density,
and so little change in pressure as to be unnoticeable), then it can be seen that
the amount of energy required of the pump to move the fluid through the sys-
tem is equal to the friction head losses associated with circulating the fluid in
the pipe. As a result, the pressure energy added by the pump in the system
behaves in the following manner:
• it overcomes the resistance due to friction loss caused by the flow through
the heating and air-conditioning equipment, such as the piping, valves,
coils, chillers, or boilers; or
t
• it raises the water in open systems to higher elevations (such as to the top of
a cooling tower above the tower basin water level) and overcomes the resis-
e
tance through the piping and the tower spray heads.
Equipment
gp
There are many types of centrifugal pumps available, and selection depends
on the application, head (feet), capacity (gpm), head–capacity characteristics,
piping arrangement, and other requirements, as illustrated in the next several
pages. Why so many options for pumps? In part, manufacturers are always
attempting to build a better mousetrap! However, there are some physical limi-
tations associated with the strengths of the materials used to build the pump as
well as the physics of imparting the energy to the pump. Figure 4-10 shows a
pE
variety of impeller shroud shape characteristics (as defined by the Hydraulic
Institute [1994]) and their specific speeds. Note that in the operating characteris-
tics, maximum pump efficiency is limited due in part to flow rate and head. As
pumps get larger, maximum efficiency tends to increase; however, it is not a
good idea to apply a large pump to a small flow requirement based on the mis-
taken thought that higher efficiencies will be attained. It is always best to match
the pump head and flow to attain the best efficiency in its class.
There are many designs of shrouds and vanes, and they are not limited to
u
what is shown. It is quite common to see double-suction impellers (Figure 4-11),
typically on larger high-flow and head pumps, although they are certainly not
limited to these. The double-suction characteristic of the impeller balances the
forces of the fluid entering the impeller, reducing the bearing loads associated
ro
with support of the impeller shaft, potentially allowing for a lighter trim (shaft
size, bearing size, etc.) as compared to a similar flow and head single-suction
impeller. Of course, this change does not come without some trade-off.
Implied in Figures 4-4 through 4-11 is the idea that a pump has a rotating ele-
ment (the impeller) operating within a stationary environment (the volute). Fluid
in the interior will be at a higher pressure than the atmosphere in which the
G
assembly is operated and will attempt to move through the casework at any place
that will allow a flow from high to low pressure. The rotating shaft is one such
place, and to greatly reduce the flow of fluid a mechanical seal is typically
applied to the shaft and the volute. These seals have a stationary element made
up of a material that resists wear and the effects of heat due to friction. The mate-
rials are often ceramics with a rotating element of similar characteristics, such as
graphite. The two surfaces mate flat to each other through special finishing pro-
e t
gp
pE
Figure 4-10 Variety of impeller shroud shape characteristics.
Source: HI 1994.
u
ro
G
Figure 4-11 Single- and double-suction impellers.

cedures and are designed to maintain a microscopic layer of fluid between the
two surfaces so that the heat generated from high-speed operation is dissipated as
an unseen vapor as the fluid transmits from the interior to the exterior of the
pump. Typically, the yearly fluid use for this lubrication is less than a few ounces
per year. Referring back to the application of a double- or single-suction impel-
ler, the trade-off from a manufacturing perspective is that the double-suction
impeller requires two seals and two sets of bearings to support the shaft, as com-
pared to the single set in a single-suction impeller. This tends to mean that double-
t
suction designs are seen more often on larger pumps based on flow and head.
e
Typical Pump Configurations
Circulator pump: Circulator pumps have the following characteristics: motor-
mounted, inline pumps with a close-coupled shaft connection requiring overall
gp
compact size; vertical suction and discharge piping; and found in small residential
systems, domestic water, or terminal-recirculation applications (see Figure 4-12).
The more traditional booster pump tends to be a built-up assemblage of volute,
coupler, and dry motor. Generally, pump flow rates are less than 60 gpm and heads
are less than 40 ft. Pump curves are considered steep; however, when compared to
other styles of small circulators (such as the more typical circulator of today), their
curves may actually appear to be flatter. The more typically applied circulator
pE
pump of today tends to be a more unitary design with the motor rotor exposed to
the fluid stream. Their pump characteristic curves tend to be steeper in design.
Close-coupled end-suction pump: Close-coupled end-suction pumps
have the following characteristics: motor-mounted pumps, overhung impeller,
close-coupled pump shaft connections, single stage, foot mounting (motor) or
pump casing mounting, compact size, and horizontal suction and vertical dis-
charge piping (see Figure 4-13).
Frame-mounted end-suction pump: Frame-mounted end-suction pumps
u
have the following characteristics: motor and pump mounted to common base,
overhung impeller, separately coupled, single stage, frame mounted on base-
plate, and horizontal suction and vertical discharge piping. In special cases
there is a centerline-mounted pump on a vertical baseplate with vertical suction
ro
and horizontal discharge piping (see Figure 4-14).

Base-mounted horizontal split-case pump: Base-mounted horizontal split-
case pumps have the following characteristics: motor and pump mounted to a
common base with an impeller between bearings; separately coupled; availability
with single and multistage impellers; and horizontal suction and discharge piping,
or, in some applications, vertical split case with vertical suction and horizontal dis-
G
charge piping. Split-case construction enables disassembly of casing to service

pump bearings and the impeller without disturbing the piping (see Figure 4-15).
Vertical inline pump: Vertical inline pumps have the following character-
istics: overhung impeller, close or separately coupled; and single stage, inline
(horizontal suction and discharge piping) with flexible or rigid pump shaft cou-
pling. Piping supports pump and requires proper pipe supports or hangers (see
Figure 4-16).
e t
gp
pE
Figure 4-12 Circulator pump.
u
ro
Figure 4-13 Close-coupled end-suction pump.

G
Vertical turbine single or multistage pump: Vertical turbine single or

multistage pumps have the following characteristics: large capacity for well or
wet-sump type applications; availability in variety of motor arrangements: sub-
mersible, barrel, or can or axial flow types; vertical submerged suction and
horizontal discharge; barrel type has horizontal suction and discharge pipes;
and barrel casing below the level of the floor (see Figure 4-17).
e t
Figure 4-14 Frame-mounted end-suction pump.
gp
u pE
Figure 4-15 Vertical inline pump.
ro
G
Figure 4-16 Base-mounted horizontal split-case pump.

Pump Nomenclature
The Hydraulic Institute has published a list of terms of parts and pump compo-
nents (HI 1994). Where possible, these terms and those in the 2012 ASHRAE
Handbook—HVAC Systems and Equipment (ASHRAE 2012) will be used in
this course.
Pump Selection
t
Centrifugal Pump Characteristics
e
In observing a manufacturer’s pump curve (as shown in Figure 4-18), there
are several points of data that are presented specific to that manufactured
gp
pump. The curve should specify the manufacturer, and the pump’s model num-
ber, size, and operating speed. It is not uncommon for a pump volute to have
u pE
ro
G
Figure 4-17 Vertical turbine pump, wet sump arrangement.

the capability to operate with a range of pump impeller sizes, and the pump
curve shows these different sizes with common operating points such as horse-
power and efficiency. The pump curve also specifies an operating pressure
requirement for net positive suction head (NPSH).
The characteristic of a centrifugal pump’s output head versus flow is shown
as a unique head–capacity curve by the manufacturer (Figures 4-18 and 4-19)
at a stated rotating speed (revolutions per minute, rpm). The head–capacity
curve of a pump model is determined by the pump manufacturer based on a set
t
of standard tests. ANSI/HI Standard 14.6 (HI 2011) designates the standard
testing and reporting methods. Typically, the discharge head (ft) decreases as
e
the pump capacity or flow (gpm) increases, as shown in Figures 4-18 and 4-19.
In some instances, there may be a slight hump and then decrease in the pump
curve as flow moves from right to left along the curve, as it approaches zero
gp
flow. The abscissa specifies pump flow capacity in gpm. The ordinate specifies
pump energy for that flow rate, which are in feet head. Head is total energy
produced by the pump (e.g., Bernoulli’s V2/2g + P/ ). Another way to con-
sider this is as the work statement of foot-pounds work per pound of fluid
pumped.
The pump characteristic may be further described as having a flat or a steep
characteristic (Bell & Gossett 1968) (Figure 4-20), and also as having a normal
rising curve, a drooping curve, or a steeply rising curve (Garay 1990). Flat char-
u pE
ro
G
Figure 4-18 Manufacturer’s pump curve.

et
gp
pE
Figure 4-19 Head–capacity curve.
u
ro
G
Figure 4-20 Flat versus steep pump curve.

acteristic pumps are applied on closed piping systems with modulating control
valves. Steep characteristic pumps are usually applied on open piping systems
(cooling towers), where high head and constant flow are usually desired.
The capacity of a centrifugal pump varies based on speed, impeller diame-
ter, and volute size. Manufacturers may combine all the curves for a particular
pump volute model and given speed (rpm) and various impeller diameter sizes
in a family of performance curves or envelope (Figure 4-21). Once the family
t
characteristic is determined, the individual pump curve (Figure 4-22) must be
carefully studied for its required performance characteristics. In Figure 4-22, the
e
pump manufacturer shows head–capacity curves for various impeller sizes (5.0
to 7.0 in. diameter shown). A sloping line shows the motor horsepower required
for various pump operating points. The pump efficiency (Equation 4-6) is
shown for various operating points along the head–capacity curve.
gp
Chapter 3, discusses how a system head loss is calculated by sizing the pres-
sure drop through the piping, components, and fittings to meet a design flow. The
pump is selected to provide this flow and pump head to meet the design condi-
tions. It was also discussed how to develop a system characteristic curve with that
calculated data. Combining the pump curve with the system curve allows the
designer to visualize the operation of the two system components. The pump
capacity curve and the system curve are plotted on the same graph, as shown in
pE
Figure 4-23. The intersection of these two curves is the system operating point,
u
ro
G
Figure 4-21 Characteristic curves for pump models at given speed.

e t
gp
pE
Figure 4-22 Selected pump head–capacity curve.
u
ro
G
Figure 4-23 System curve and pump head–capacity curve.

and the system will always operate on the pump curve when properly selected and
adjusted. If the system curve does not intersect the pump curve, the system will
operate in an area of unsatisfactory operation for the pump, potentially causing
damage. This may include pump cavitation, where fluid is transformed to vapor
and then back to fluid as pressures stabilize around the fluid vapor pressure point.
Pump Horsepower
t
Water Horsepower
e
The theoretical power to circulate water is the water horsepower (WHP) and is
shown as
lb liquid pumped/min total head in feet

WHP = ------------------------------------------------------------------------------------------------------ (4-2)
gp
33,000 ft/lb/min/hp
Because water has a specific weight of 62.3 lb/ft3 at 68°F (standard), the
formula is
Flow gpm Head feet

WHP = ---------------------------------------------------------------- (4-3)
3960
pE
Figure 4-24 compares the increase of WHP as the pump increases flow.
At other water temperatures or fluids, the formula is corrected for specific
gravity (sg) as
u
ro
G
Figure 4-24 Comparing the increase of WHP as the pump increases flow.
Flow gpm Head feet sg

WHP = -------------------------------------------------------------------------------- (4-4)
3960
Brake Horsepower
The brake horsepower (BHP) required to operate the pump is determined
by the manufacturer’s test of an actual pump running under standard condi-
tions to produce the required flow (gpm) and head (ft). The pump manufac-
t
turer furnishes the required BHP on the pump curve above the pump head
curve (Figure 4-25).
e
BHP = Flow
gpm Head feet
---------------------------------------------------------------- (4-5)
3960 P
gp
Pump Efficiency
The pump efficiency is determined by the manufacturer’s test by compar-
ing the theoretical output power (WHP) to the input power (BHP):
Output WHP
Efficiency = ---------------- = ------------- 100 % (4-6)
Input BHP
u pE
ro
G
Figure 4-25 BHP on the pump curve above the pump head curve.
Figure 4-26 shows the plotted pump efficiency, as determined by actual

test, for each impeller’s operating point. Note how the pump efficiency
increases with increasing flow until the best efficiency point (BEP) is reached
and then decreases as flow increases further.
The pump manufacturer will plot lines of constant efficiency for a given
pump volute and the applicable impeller sizes that provide rated perfor-
mance, as shown in Figure 4-27. Note the increase in pump efficiency until it
reaches the maximum efficiency and then decreases. It is at this BEP of
t
pump operation that the exit angle of the water as it leaves the impeller
closely coincides with the angle of the cut water or tongue in the body casing
e
(Figure 4-8). The so-called “eye” of these curves is the BEP, and it is consid-
ered to be the operating point with the best life-cycle cost when accounting
for operating and maintenance costs over the product life. Ideally, the pump’s
gp
BEPs should be in the middle of the pump head curve operating range. How-
ever, pump BEPs shift with respect to impeller location based on the operat-
ing speed of the pump. Pumps with a speed of 1100 rpm have a BEP that
tends to be closer to shutoff head. Conversely, 3500 rpm speed pumps have a
BEP that runs further out on the curve, approaching a point of unsatisfactory
operation. For this reason, BEP is an important evaluation but not necessarily
u pE
ro
G
Figure 4-26 Increase of pumping power required with pump flow.

e t
gp
pE
Figure 4-27 Pump efficiency curves.
u
the most important consideration. It is equally, if not more, important to
ensure that the design and pump selection allow for flow to be throttled in
operation and that head be adjusted to just that which is necessary for proper
ro
system operation.
The optimum selection area or region to use when selecting a centrifugal
pump is displayed in Figure 4-28. The areas bounded by lines AD to AE repre-
sent operating points that lie within the preferred selection range, but this area
does not always contain the needed points, as an infinite number of pump sizes are
G
not economical to produce and selection limits of 66% to 115% of flow at the
BEP are suggested. The satisfactory selection range is that portion of a pump’s
performance curve where the combined effect of circulatory flow, turbulence, and
friction losses due to the liquid being handled are at a minimum. Pump selections
to the left of the BEP are recommended where possible, as overstated data for pipe
friction and other system equipment tend toward lower actual system heads, caus-
ing overpumping and possible operation in the turbulent region of the pump.
e t
gp
pE
Figure 4-28 Recommended selection regions.
u
Pump manufacturers usually publish catalog pump performance data in the
form shown in Figure 4-29—in this particular case, a double-suction pump hav-
ing 10 in. discharge and 12 in. suction connections and operating at 1750 rpm.
ro
The performance curve illustrated is referred to as a flat curve type in that the
head at shutoff is approximately 25% higher than the head at the BEP. The
curves include other performance values based on trimmed impeller diameters
from 11 to 9 in., which is the minimum available diameter.
The horsepower lines are superimposed on the head–capacity curves to
indicate what motor ratings would be required. The NPSH section of this chap-
G
ter discusses the NPSH curve. Note that as the impeller diameter is reduced,
the overall efficiency suffers because of increased recirculation around the
body casing due to the additional radial clearance between the impeller outside
diameter (OD) and the cut water in the pump body (Figure 4-8).
The particular pump curve shown in Figure 4-29 has a wide impeller that is
well suited for use on hot-water or CHW systems in that the horsepower lines
are reasonably tangent to the head–capacity curve, which means that an over-
sized motor does not have to be used to protect from overload should the system
head be less than calculated. In this case, the 125 hp motor will be adequate.
Also in Figure 4-29, note that at the BEP the capacity is 4000 gpm. With the dis-
charge flange of 10 in., this results in a velocity of 16 fps (v = Q/A) and, with its
12 in. suction flange, an 11.5 fps velocity on the input side (which is equivalent
to a 4 ft/100 ft drop for 12 in. pipe). If the pump is located in a central power
plant, this potential velocity noise may be acceptable. However, if these mains
were in a building where the operating sound level might be critical, then the
t
next larger size (14 in. pipe) would reduce the velocity to 9.5 fps and would be a
more satisfactory selection. In some cases, manufacturers have designed both
e
narrow and wide (thickness) impellers to be interchangeable within the same
body volute to provide both steep or flat performance curves and extend the
application range of the basic pump model. In this case, the narrow impeller
gp
(not shown) would indicate a 3000 gpm capacity for the 11 in. impeller at the
BEP instead of the 4000 gpm shown.
u pE
ro
G
Figure 4-29 Pump performance data.

Radial Thrust
In a single-volute centrifugal pump, uniform or near-uniform pressures act
on the impeller at design capacity (Figure 4-30), which coincides with the BEP
(Garay 1990). But at other capacities, the pressures around the impeller are not
even and there is a resultant radial reaction. Figure 4-31 shows a typical change
in radial thrust with changes in the pumping rate. Specifically, radial thrust
decreases from shutoff to the design capacity (if chosen at the BEP) and then
t
increases with overcapacity. With overcapacity, the reaction is roughly in the
opposite direction than with partial capacity. Note that the force is greatest at
e
shutoff. The radial forces resulting from extreme low-flow operation can cause
severe impeller shaft deflection and ultimately shaft breakage, with the danger
becoming more severe with large head pumps. Later, we will discuss this con-
cern with variable-speed operation.
gp
NPSH
Particular attention must be given to the condition (pressure and tempera-
ture) of the water as it enters the suction of a pump, especially condenser-tower
applications, steam condensate return, and steam boiler feed systems. If the
absolute pressure on the liquid at the suction nozzle is approaching the vapor
pressure of the liquid, there is a high potential for vapor pockets to form in the
pE
impeller passages. As the fluid traverses the interior of the impeller, it will go
through changes in pressure due to shock losses associated with directional
change and the friction loss of the impeller itself. When the vapor forms, it will
interfere with pump performance, and when the vapor pocket collapses
u
ro
G
Figure 4-30 Pressures on impeller causing radial thrust.

e t
gp
pE
Figure 4-31 Change in radial thrust versus pumping rate.
because of the input pump energy moving the pressure higher than the vapor
pressure, there will be destructive forces that erode the metal of the pump. The
u
collapse of the vapor pockets (cavitation) will be noisy and typically destruc-
tive to the pump impeller. If a cavitating pump is not corrected, it will cause
permanent damage and eventually failure of the pump.
According to the Hydraulic Institute (HI 2012) definition, the NPSH
ro
required (NPSHR) of a pump is the NPSH that will cause the total head to be
reduced by 3% due to flow blockage from cavitation in the impeller vanes.
Cavitation does not begin at the NPSHR value; that starting point is called
incipient cavitation. The NPSH at incipient cavitation can be from 2 to
20 times greater than the 3% NPSHR value, depending on pump design, with
the higher values being associated with high-suction energy pumps.
G
The minimum recommended amount of pressure in excess of the vapor

pressure required to prevent the formation of vapor pockets is the NPSH margin
and is defined as the ratio of the NPSH available (NPSHA) at the pump inlet
divided by the NPSHR.
NPSHR is a characteristic of a given pump and varies with pump speed and
flow. It is determined by the manufacturer testing individual pumps and pro-
viding this data on the pump performance curve (Figure 4-32, upper curve).
e t
gp
pE
Figure 4-32 Selected pump curve showing NPSHR.
NPSHA is the total suction head available over the vapor pressure of the liq-
uid pumped to the centerline of the impeller. NPSHA is calculated as follows:
NPSHA = h ATM + h gs + h VS + z s – h VP – h f (4-7)
where (in feet of fluid head)

u
hATM = atmospheric pressure head
hf = pipe friction losses where applicable
hgs = suction gage head
ro
hVP = liquid vapor pressure head

hVS = suction velocity head
zs = suction elevation head (the valve is positive if the fluid is above the
suction centerline of the pump; the value is negative if the fluid is
below the centerline)
Particular attention must be given to NPSHR when a pump is operating
G
with hot liquids or applied to a circuit having a suction lift, such as a cooling
tower. The vapor pressure increases with water temperature and reduces the
NPSHA. Each pump has its own NPSHR (as shown on its curve), while the
installation has its own NPSHA, which is the total useful energy above the
vapor pressure at the pump suction connection. Note that the NPSHA is also
affected by the location of the expansion tank in closed systems. Figure 4-33
shows the determination of NPSHHA for a proposed installation.
e t
gp
pE
Figure 4-33 NPSHA in a proposed installation.
To determine NPSHA in an existing installation (Figure 4-34), use this for-

mula:
2
u
V
NPSHA = h ATM + h gs + ------ – h VPA (4-8)
2g
where (in feet of fluid head)

ro
hATM = atmospheric pressure head

hgs = head at suction flange corrected to centerline of pump; hgs is
negative if below atmospheric pressure
hVPA = head of fluid vapor pressure
V2/2g = velocity head at point of measurement
G
NPSHR is normally not a factor with hot and CHW pumps in closed-loop
pumping systems because pressure can be adjusted via the system fill pressure,
which is exerted on the pump suction. That said, sufficient margin for building
services, while variable for normal HVAC services as defined by the Hydraulic
Institute, tends to be in a range of 1.1 (generic pumping) to 1.5 (cooling tow-
ers). Figure 4-35 gives an example of a factory NPSHR test setup (Garay
1990).
e t
Figure 4-34
gp
NPSHA in an existing installation.
u pE
ro
G
Figure 4-35 Factory test setup to determine pump’s NPSHR.

Pump Selection Process

Every pump selection includes the following sequence of steps (Figure 4-36):
1. Determine the load to be pumped (heating or cooling) in Btu/h.

2. Determine the design t across the water side of the load heat transfer coil
or device. For example, 20°F drop for heating or 12°F rise for cooling. Cal-
culate the required flow in gpm for each load.
t
3. Total the zone load flows to determine the total flow in gpm.
e
a. Study the piping design to determine the highest circuit path friction
loss in feet of head (from pump discharge through distribution main,
zone riser or run outs, terminal piping, load coil, fittings, valves, and
return to pump suction).
gp
b. For primary/secondary systems, select the primary distribution pump for
the most resistant path up to and from the common or bridge and return.
4. For the secondary pump, select the most resistant path within the secondary
distribution and terminal piping.
5. Determine the method of mechanically mounting and supporting the pump
on a pad in the equipment room, inline in the piping, or within a well or wet
sump below floor level.
pE
6. With the total head (in feet drop) and capacity (in gpm) determined, select
a pump from the manufacturer’s family curves and the mounting required.
7. Refer to the manufacturer’s individual pump performance curve sheet show-
ing pump efficiency, impeller diameter size, brake horsepower, and NPSHR.
Select a flat curve pump for closed systems with control valves (to minimize
variation in head for a reduced load) and a steep curve pump for open sys-
tems or constant-flow systems without control valves. Select the most effi-
u
cient percent-pump operation to satisfy the head feet and capacity gpm.
° For constant-speed pumping with two-way valves, consider the varia-
tion in total flow, from design to minimum capacity, and select the
highest pump efficiency to lie between this range (for example, 50% to
ro
100% capacity). Select the motor horsepower to be greater than the

pump brake horsepower and motor efficiency over the range of opera-
tion. Check for overflow capacity when staging multiple pumps.
System Design Considerations

G
Similarity Relationships: The Affinity Laws

Any centrifugal machine that imparts velocity and converts velocity to pressure
can be categorized by a set of relationships that apply to any dynamic condition.
These relationships are referred to as the affinity laws (Garay 1990). They can be
described as similarity processes, which follow general rules within certain limits
and should always relate to a known system’s head curve.
e t
gp
u pE
ro
Figure 4-36 Pump selection process.
The affinity laws state the following:
• capacity varies as the rotating speed or the peripheral velocity of the impeller,
G
• head varies as the square of the rotating speed, and

• BHP varies as the cube of the rotating speed.
The affinity laws apply to centrifugal gas compressors as well as to centrif-

ugal pumps but are most distinctly useful for estimating pump performance at
different rotating speeds or impeller diameters, starting with pumps with
known characteristics.
Two basic variations can be analyzed by these relationships (see Table 4-1):
• By changing speed and maintaining constant impeller diameter, pump effi-

ciency will remain the same, but head, capacity, and brake horsepower will
vary according to the laws.
• By changing impeller diameter and maintaining constant speed, the effi-
ciency will not change, except for a diffuser-type pump where the effi-
t
ciency will change if the impeller size is reduced more than 5% (Garay
1990). (Note the change in efficiency will occur if the impeller size is
e
reduced sufficiently to affect the clearances between the casing and the
periphery of the impeller.)
gp
It is important to keep in mind that the use of the affinity laws assumes
that the system head curve is known and it behaves like a parabola where the
head varies as the square of the flow (h ~ Q2). The pump head curve is deter-
mined by the manufacturer’s test. The pump operating point is the intersec-
tion of the total system curve and the pump-head curve. As the affinity laws
are used to calculate a new condition due to flow or head change (such as a
reduced pump speed or impeller diameter), this new condition must follow
the same system curve. An exception to this would be a change in total flow
pE
due to the control valves reducing the system flow at constant pump speed.
In this case, the intersection of the system curve must follow the pump head
curve to the reduced flow condition, and then a new system head curve is
developed through this point. Another caution on using the affinity laws is
to take into account the system static head that is in addition to the system
head.
Table 4-1 Equations for Speed Change and Impeller Diameter Change
u
Speed Change (N) Impeller Diameter Change (D)
N D
Flow Q 2 = Q 1 ------2- Q 2 = Q 1 ------2
ro
N1 D1
N 2 D 2
Head h 2 = h 1 ------2- h 2 = h 1 ------2
N1 D1
N 3 D 3
BHP 2 = BHP 1 ------2- BHP 2 = BHP 1 ------2
G
Horsepower
N1 D1
BHP1 = original pump horsepower, hp N1 = original pump speed, rpm

BHP2 = new pump horsepower, hp N2 = new pump speed, rpm
D1 = original pump diameter, in. Q1 = original flow, gpm
D2 = new pump diameter, in. Q2 = new flow, gpm
h1 = original head, ft
h2 = new head, ft
Example 4-1
The pump is rated at 500 gpm at 60 ft head with a 1750 rpm electric motor.
What is the flow and head if used with an 1150 rpm motor? Assume no system
static head.
Solution Flow:
t
N 1150
e
Q 2 = Q 1 ------2- = 500 ------------ = 328 gpm
N1 1750
Head:
gp
N 2
1150 2
h 2 = h 1 ------2- = 60 ------------ = 26 ft
N1 1750
Parallel Pumping
pE
In considering the design of the system with respect to pump selection, providing
pumps in parallel is one option that designers should consider. Conservative
engineering practice often uses the installation of two pumps, one as primary and
the other as backup. As a result, two pumps of the same size are selected. They
are connected to the system in parallel, only from the perspective of pump failure
(e.g., when a pump is inoperable, it is hydraulically disconnected from the sys-
tem through valves and the other is connected so that the system may operate).
Parallel pumping takes this concept a step further, operating the two pumps
u
to provide design flow and head conditions and using a controller to turn one of
the pumps off at conditions of load less than design and complementary to the
flow required for heat transfer. The advantage to this design is that two smaller
pumps may be purchased, and when operating in single pump mode, pump
ro
energy may be reduced.

When this is done, a composite pump curve is created by the designer and
compared with the design system curve. Pumps are selected for one-half of
the system design flow rate and the full design head of the system. The com-
posite curve is sketched on a single pump curve (as illustrated in Figure 4-37)
or by using a spreadsheet to plot the flow data by adding flow rates on a line
G
of constant head for the various points of the single pump curve (as illus-
trated in Figure 4-38). The points are then connected to represent a pump
curve when two pumps are operational. Generally, this must be done by the
system designer, as pump manufacturers publish their pump curves for single
pump operation
Chapter 8 covers how to draw the composite curve and discusses piping for
the application.
e t
gp
Figure 4-37 The old way: sketching the composite parallel pump curve by hand.
u pE
ro
Figure 4-38 The new way: sketching the composite parallel pump curve by spreadsheet.
G
Pumps in Series
Just as pumps may be piped in parallel, they may also be piped in series. In this
mode, when two pumps are operated together, the heads (not the flows) are
added. This makes application of this concept something to carefully consider
in HVAC systems. Generally, increasing the pressure by the addition of pump
energy tends to bring out more system-related problems than less in the tradi-
tional HVAC system. That is not to say that there are no application options.
Occasionally, systems have components that exercise distinct logical pressure
losses under specific piping configurations. An example of such might be a
constant-process cooling load where cooling water is circulated through the
dies of a machine to maintain product quality. One die might have a specific
flow rate with a low head loss, while the other die for manufacturing a different
product using the same machine might have the same cooling flow rate, but at a
t
far greater head loss.
e
In other cases, this concept might be applied temporarily to help operate a
poorly performing system. An example of such is shown in Figure 4-39, where
two full-sized pumps were selected for a design that improperly applied the use
of diversity factors and the pipe sized for it. The end result was that the system
gp
was undersized by 50%. When technicians attempted to rig the system to oper-
ate temporarily until an engineered solution could be determined, they reasoned
that with two half-size pumps all they had to do was operate the system with
both running to achieve design flow. The error in this logic is that while parallel
pumps are sized this way, the logic is based on the premise that the head loss of
the system piping accounts for this. In this particular case, pipes were sized for
one-half of the design flow rate, with associated head losses. Attempting to dou-
ble the flow of the fixed-size piping vis-à-vis the affinity laws requires that head
pE
be increased by a factor of four to create design flow and that brake horsepower
be increased by a factor of eight. Operating the two pumps in parallel increased
flow by only 4%, as illustrated in Figure 4-40. Had the technicians been able to
modify the piping so that the two pumps were operated in series (as shown in
u
ro
G
Figure 4-39 An example with two full-sized pumps.

Figures 4-40 and 4-41), a more significant increase in flow could have been had
due to the increase in developed pump head. It can be seen through interpreta-
tion of the composite pump and system curve that flow for this example would
have been increased 31% and, though this is not enough to provide complete
e t
gp
pE
Figure 4-40 Illustration of pump and system curves applied in parallel or in series and relative
system flow effects.
u
ro
G
Figure 4-41 Schematic modifications example to place pumps in series.

heat transfer, would have provided enough heat transfer to reach greater than
70% of design heat transfer and an operating point that would have been greater
than 80% of the operating hours. Of course, the best solution is to properly size
the pipes and pumps, but when in the field and trying to alleviate as-built condi-
tions until a satisfactory and proper solution can be dealt with, sometimes
remembering some of these core fundamentals may be helpful.
Variable-Speed Pumping
t
Advancements in modern electronics technology have significantly reduced
e
the costs of high-power motor speed control circuitry. In general, any centrifu-
gal pump may be controlled by a variable-speed drive with some of the horse-
power advantages associated with the affinity laws that are available with
respect to horsepower. This is not a given, though, because the application of
gp
variable-speed pumping often requires a separate pump control strategy that in
normal HVAC application relies on proportional control of a differential pres-
sure across some portion of the applied system.
Where a speed drive is simply applied to a centrifugal pump, the effect of
changing speed is illustrated by Figure 4-42. As can be seen, it is akin to hav-
ing an adjustable-sized impeller, so, for controlled systems with fixed system
curve characteristics (such as a coil controlling air temperature directly with a
pE
pump modulating flow to the coil instead of a control valve), the system rides
up and down the system curve for flow modulation rather than riding the pump
curve, as with a traditional throttling valve.
Figure 4-43 shows the typical application where a differential pressure con-
troller is used to maintain a constant differential pressure across the control
valve of the most hydraulically significant loop used to define the head selec-
u
ro
G
Figure 4-42 Example of pump curves at varying pump speeds.

e t
gp
pE
Figure 4-43 Schematic of simplified variable-speed-pump-controlled system.
tion of the pump and other coil flows are controlled by throttling valves. The
common expectation is that results will be similar to the application riding the
u
system curve; however, that is not the case.
In systems of this type, one must first understand that there is a separate
pump controller changing pump speed based on the control error of its manip-
ulated variable, typically differential pressure in hydronic systems. This con-
ro
trol loop sees only the variables associated with its discrete control loop,
namely differential pressure across the flow element used as a reference for the
entire system’s operation. This method of control raises several issues to ensure
complete and proper system operation, mainly all control issues related to the
building load, which are not addressed in this text. What is important to recog-
nize in this text is that because the separate control loop constantly modifies
G
the pump performance, the system curve construction changes. Were it a con-
stant-speed pump, the system curve would appear similar to the black line
labeled Balanced System Curve in Figure 4-44. We could also show the effects
of not balancing the system, indicated by the Unbalanced System Curve line,
which is based on the flow path circuits 1 and 2 not being adjusted to have the
same pressure loss as flow path circuit 3. The calculation of these different
curves is shown in Figure 4-45.
e t
gp
Figure 4-44 Typical constant-speed pump curve interaction with variable-speed pump.
u pE
ro
G
Figure 4-45 Flow coefficient method of calculating balanced and unbalanced conditions.
In the example system of Figure 4-43, the system requires 100 gpm at a
head loss of 100 ft. Because all of the coils, pipe valves and fittings, control
valves, and balance valves have the same head losses (load), and circuit 3 is the
furthest from the pump. Circuit 3 is the most significant hydraulic circuit, and
the pump is selected to provide design flow at the exact head necessary to get
water to the end and then back again to the suction of the pump. The distance
and friction losses of the riser pipes mean that circuit 2 and circuit 1 each have
less total head loss than circuit 3. However, without some sort of adjustment,
t
more water will flow through each circuit until the combined path system
curves intersect with the pump curve. That could be a lot or a little. It depends
e
on the losses, where they occur, and the shape of the pump curve. In this case,
total system flow is about 116 gpm, 16 gpm greater than the design flow
(Figure 4-44). The common approach is to proportionally balance a system,
which means that all water flow paths will have the same pressure loss. The dif-
gp
ference in circuit head loss could be made up by either a balancing valve of
some sort or extra pipe, which would imply reverse-return piping. In either case,
though, pipe or valve, all that is occurring is the addition of pressure loss to pre-
vent overflow.
The addition of the separate differential pressure controller and a variable-
speed drive to the pump does not modify how the pump works. These are sim-
ple devices that move water and make differential pressure. They do, however,
pE
modify the observed results of flow and head when under the influence of the
differential pressure controller.
In variable-speed pumping, the differential pressure control setpoint mimics
a fixed system head loss regardless of flow, similar to the elevation difference
experienced in the open-loop pumping application of the cooling tower. This
causes an immediate impact on the potential horsepower savings of applying a
variable-speed drive to the pump, as the controlled system will not follow the
theoretical relationship of the affinity laws. This is illustrated in Figure 4-46.
u
To achieve design flow, circuit 3 must have an available pressure differen-
tial of 37 ft of head. This comes from placing the differential pressure transmit-
ter across the branch piping entering connection from the supply pipe to the
branch leaving connection to the return pipe. The head losses of coil, valve,
ro
balancing device, and associated pipe, service valves, and fittings equal a total
head loss of 37 ft. As the branch has a temperature control valve reacting to a
controller that is maintaining the required flow for a heating or cooling load, as
the temperature control valve opens and closes (increasing or reducing flow),
the differential pressure across the branch will decrease on a required increase
in flow and increase on a required decrease in flow. The differential pressure
G
transmitter senses this change, and the controller then decreases or increases
pump speed to maintain a constant differential pressure across the branch. This
is a very important implied linkage of system operation. The separate zone
controller that opens and closes the temperature control valve has its own set of
controller adjustments (proportional, integral, and derivative settings), which
are based on the concept of the control valve representing a constant rate of
response to its controlled variable (temperature). When the load is 50% heat
e t
gp
Figure 4-46 System curve for variable-speed pumping.
transfer, the controller output is 50%, and enough water to be delivered to the
coil to provide 50% heat transfer is expected. The initial controller output is
based on the controller’s proportional setting. Therefore, if there is not enough
pE
differential pressure to provide design flow, (perhaps, only enough to provide a
maximum of 50% flow), then, if the controller were looking for 50% heat
transfer, it would go to a valve position that would only provide 25% (half of
what is necessary for proper operation and maintaining a controller that actu-
ally does control the temperature). This is why, in general, a constant differen-
tial pressure is maintained across the significant circuit.
That relationship also defines some other relationships by implications,
which are important to remember in variable-speed pump control. Note that all
u
branches have the same head losses, and the flows are equal. This is a very sim-
ple system example. Most real systems are not this way. The designer needs to
take great care in laying out the system and may need to use various combina-
ro
tions of piping and pumping to reduce how one circuit’s controlled flow may
interact with another circuit’s flow. The designer may also need to apply multi-
ple sensors to ensure that there is proper differential head where it is required
so that all system control loops operate properly.
As can be seen in Figure 4-46, as flow reduces, the pump controller and
drive reduce pump speed. At zero flow, the pump will still theoretically be put-
G
ting in enough energy to create 37 ft of head. That leaves 67 ft of head as vari-

able head loss, and (as can be seen from the graphic included in Figure 4-46) a
graphic system curve is constructed using the same relationship of flow and
2
head ( Q 2 Q 1 = h 2 h 1 ), as previously demonstrated.
A secondary issue that causes significant operating issues is that as control
valves open and close they modify flow rates and thus the available differential
pressure at each branch, thus affecting the flow at that branch and the others. This
is called control interaction and is due to the valve and the different device pres-
sure drops and their locations in the system. This topic could be addressed for sev-
eral volumes if we so chose to do so, but the intent here is to introduce the topic. If
the example system were proportionally balanced, then circuit 2 would have a bal-
ancing valve drop of 26 ft (21 + 5), and circuit 1 would have a drop of 47 ft, as
indicated in Figure 4-45. As long as all of the control valves open and close at rel-
atively the same time in response to load, all system flows remain proportional.
However, if there are extremes, such as valves being commanded to an open or
t
closed state without regard to the other system connected valves, flow can be
e
affected in far greater magnitude than a single valve might have as a design effect.
Figures 4-47 and 4-48 document the academic exercise of sketching what
is known as the control area of the variable-speed pump illustrated in the sche-
matic of Figure 4-43, when the system is balanced. Valves are closed in
gp
sequence starting with circuit 3, which is the controlling zone of the system
having the differential pressure transmitter across it. Valve 3 closes and, as a
result, the pump speed slows down to maintain just 37 ft of head. The friction
loss in the riser leading to circuit 3 goes to zero as flow has gone to zero, leav-
ing 37 ft of head to move water across circuit 2, which requires 58 ft at design
conditions. Lacking the extra 21 ft, flow is reduced to circuit 2 by about 20% to
26.6 gpm. That flow rate allows for a slight pressure loss in the riser
pE
sections BC and FG, making 37 ft plus 3.35 ft of differential pressure available
to circuit 1 for flow, which translates into about 24 gpm. The total system flow
with valve 3 closed is then about 51 gpm, substantially more than the 33 gpm
that valve 3 should control.
u
ro
G
Figure 4-47 Control area curve for example.

e t
gp
pE
Figure 4-48 Calculation of total dynamic head for area curve due to valve change of states.
Part of this effect is due to added pressure drops applied to the system to
proportionally balance the system through the use of hydraulic balancing valves
u
used to limit the flow in each branch to its design requirements. Hence, a long
discussion could be had about how all the attached valves react to system load
and what the system loads are for the attached loads with respect to time. Pro-
portional balancing may not be appropriate for this system. In that case, a
ro
dynamically responding balancing valve should be applied in the form of an

automatic flow-limiting valve, a pressure-independent control valve, or a system-
pressure-powered differential pressure control valve. Could the extra friction
losses required to balance be applied to resizing the temperature control valve?
The answer is not simple. Just as a manual proportional balancing valve would
be added to a circuit with a fixed resistance, moving that resistance to the con-
G
trol valve would have a similar effect on flow interaction. Ideally, when the
pump is providing less differential pressure, it is desirable to have the balancing
valve automatically open in response, keeping more system differential pressure
on the control valve rather than the balancing valve. This type of discussion
often leads to the control valve being looked upon as an automatic balancing
valve relying on the temperature controller to recognize that excessive flow to
the coil is causing excessive heat transfer and then relying on the controller inte-
gral action to make an adjustment to bring the hydronic distribution system back
into hydraulic balance. Suffice it to say that differences in time response and
controller adjustments, as well as the coil’s relationship of flow and heat transfer,
make that idea less practical. However, the pressure-independent control valve or
an additional differential-pressure control valve around the temperature control
valve can also alleviate that issue. Alternatively, the designer can just carefully
lay out the piping system so valve interaction is minimized.
As shown in Figure 4-49, variable-speed pumping offers a great opportu-
t
nity to save system pumping energy, but it requires careful examination of the
system as a whole as well as the performance of individual components to
e
achieve its energy savings potential. As can be seen from Figure 4-49, pump
speed will not be lower than about 58%—not bad when one considers that 58%
speed represents about 19% of the design horsepower of the pump. How low
gp
the pump speed can be taken is solely up to the designer based on the selection
of pipes, how well the coil and control valve work together, and how flow dis-
tributes through the system in operation so that all terminals are able to achieve
97% of the heat transfer that they require.
In the rare cases where the designer is actually able to significantly reduce
the pump head and thus the pump speed, the designer needs to be aware of the
pump maximum speed change. Some pump manufacturers have recommended
pump minimum speeds, normally in the 20% to 30% range, so that they can
pE
ensure proper pump seal lubrication.
u
ro
G
Figure 4-49 Control area curve with system curve overlaid.

Proper application of variable-speed pumping requires a strong understand-

ing of control valve application theory, proportional-integral-derivative (PID)
control theory, and sequence of operation impacts so that the complete system
will properly operate. Some of this is covered in other chapters, while other
portions will need to be independently studied through other texts.
The Next Step
t
In Chapter 5, you will examine terminal unit performance and control.
e
Summary
In this chapter, we covered the following:
gp
• How a centrifugal pump changes the flow and head of a hydronic system
• Six types of centrifugal pumps and where they are used
• How the head–capacity curve of a centrifugal pump is determined
• The difference between family and individual pump curves
• The characteristics that can be learned from a manufacturer’s pump curve
• How to determine horsepower for a centrifugal pump
pE
• Pump efficiency
• The radial thrust of a centrifugal pump
• Pump affinity laws and how to apply them
• NPSH, NPSHA, and NPSHR
References and Bibliography

u
Atlanta: ASHRAE.
Bell & Gossett. 1968. Pump and System Curve Selection and Application for
Centrifugal Pumps. Technical Report 375. Morton Grove, Illinois: Xylem
ro
Inc.
Garay, P. 1990. Pump Application Desk Book. Lilburn, GA: Fairmont Press.
HI. 1994. ANSI/HI 1.1-1.5:1994, Centrifugal Pumps—Nomenclature, Defini-
tions, Applications and Operation. Parsippany, NJ: Hydraulic Institute.
HI. 2011. ANSI/HI Standard 14.6-2011, Standard for Rotodynamic Pumps for
Hydraulic Performance Acceptance Tests. Parsippany, NJ: Hydraulic Insti-
G
tute.
HI. 2012. ANSI/HI Standard 9.6.1-2012, Rotodynamic Pumps Guideline for
NPSH Margin. Parsippany, NJ: Hydraulic Institute.
Stethem, W. 1988. Application of constant speed pumps to variable volume
systems. ASHRAE Transactions. 94(1):1074–84.

4-1 List three factors that influence the type of pump selected for a particular appli-
cation.
t
4-2 Label the components of the centrifugal pump shown below. Describe the
function of each component.
e
4-3 The suction flange gage of a pump reads 10 psi. If the temperature of water being
pumped is 220°F and the atmospheric pressure is 14.7 psia, what is the available
NPSH? At what temperature will the pump cavitate? (Water at 220°F vaporizes at
gp
17.2 psia, per ASHRAE Handbook—Fundamentals, Chapter 1, Table 3.)
4-4 What is the NPSH on the inlet to a 2 hp pump rated at 140 gpm at 32 ft for a
cooling tower application? The centerline of the pump inlet is to be 2.5 ft
below the tower sump water surface; assume tower water at 120°F and piping
equivalent to 60 ft of 2 in. pipe on pump suction. Assume atmospheric pressure
is 14.7 psia or 34.0 ft; assume friction head in 2 in. pipe is 20 ft (according to
Figure 2-15, 140 gpm flow in 2 in. pipe yields 33.2 ft/100 ft of pipe: 60/100 ×
pE
33.2 = 19.92, or 20 ft) and vapor pressure of water at 120°F is 1.69 psia (per
ASHRAE Handbook—Fundamentals, Chapter 1, Table 3) or 3.9 ft abs (1.69 psia
× 34 ft/14.7 psi = 3.9 ft abs). Pump curve shows 10 ft NPSHR.
4-5 What is radial thrust?
4-6 Explain what pump cavitation is and how it can be avoided.
4-7 Write the NPSHA formula for a proposed design and explain what each vari-
u
able represents.
ro
G
4-8 Write the pump affinity laws and explain how they are applied.
4-9 Speed Change Diameter Change:
Flow:
Head:
Horsepower:
4-10 Explain how to determine the horsepower for a centrifugal pump.
t
4-11 How does the capacity of a centrifugal pump change?
e
4-12 Name six types of centrifugal pumps and their mounting arrangements.
4-13 A pump is rated at 500 gpm at 60 ft of head. What are the flow and head if the
impeller size is changed to 85% of its original diameter? Assume there is no
gp
static head.
u pE
ro
G
G
ro
upE
gp
et
Terminal Unit
Performance
and Control
e t
Study Objectives for Chapter 5
gp
After studying the material in this chapter, you should be able to
❑ know what variables are involved with terminal control;

❑ understand what mechanical components are suggested at a terminal;
❑ know how a terminal control valve is selected;
❑ understand the four types of terminal control methods;
pE
❑ know the emission characteristics of heating terminals versus those of cool-
ing terminals;
❑ know what types of control valve characteristics are available and what
works best with a hydronic coil;
❑ understand valve authority, rangeability, and selection;
❑ know the advantages and disadvantages of two-way and three-way
valves;
u
❑ understand the principles of primary/secondary pumping systems and differ-
ent methods of control; and
❑ know what types of control methods vary airflow through a terminal.
ro
Instructions
Types of Terminals
G
In Chapter 1, boilers, chillers, heat exchangers, cooling towers, and heat

pumps were discussed as typical heating or cooling sources. These devices
are the source of the heating or cooling energy distributed by the air and
water systems to control the heating or cooling loads. The loads are condi-
tioned by a system of devices called terminals, or units that are designed to
transfer the load from the space to the heating or cooling water sources.
140 Chapter 5 Terminal Unit Performance and Control
Typical examples of load terminals include the following:
• Air-handling units with any of the following: preheat coils, heating coils,
cooling coils, or zone coils
• Variable-air-volume (VAV) air units (when they contain a heat transfer coil)
• Fan-coil units
• Induction-coil units
• Active or passive beams
t
• Radiant panels
e
• Baseboard radiation
• Convector radiation
• Heat exchangers
• Water-loop heat pumps
gp
• Unit ventilators
• Unit heaters
Water Supply Systems to Terminals

As discussed in Chapter 1, water supply systems to terminals are catego-
rized as two pipe, three pipe, or four pipe. They are similar in function and
pE
include capabilities to support year-round air conditioning but require careful
design analysis to apply properly.
Two-Pipe System
The system name is derived from the piping distribution design: two-pipe
systems have one supply and one return pipe for either chilled- or hot-water sup-
ply (Figure 5-1). This same two-pipe system may be used in a dual-temperature
configuration where the coil selection and design water temperatures are care-
u
fully selected to allow operation in either heating or cooling mode. The dual-
temperature two-pipe system may be limited during changeover season because
hot water may not be returned to the chiller or chilled water cannot be returned
ro
to the boiler. Typically, energy codes such as ANSI/ASHRAE/IES Standard 90.1

limit that type of cross-migration design (ASHRAE 2013); however, there is
also an equipment damage risk to the source equipment from thermal shock that
must not be overlooked.
Three-Pipe System
G
Note: This text is meant to provide historical perspective. It is not sug-

gested or implied that three-pipe systems should be designed.
The traditional three-pipe system had a chilled-water supply pipe, a hot-
water supply pipe, and a common return pipe. The coil supply was the common
port of a special three-way valve that allowed modulation of the chilled water,
had a dwell point of operation with no flow, and then allowed hot-water modu-
lation. Too often this valve allowed for some leakage of hot or cold water to
e t
gp
Figure 5-1 Two-pipe source/load concept.
mix prior to delivery to the coil. The return pipe had a mixture of chilled and
hot water during changeover operations when zones required both hot and cold
water, which causes excessive energy waste and can only be used with careful
pE
control design to prevent damage to the cooling or heating sources. The end
result of this is that this type of system is not allowed by energy codes such as
ANSI/ASHRAE/IES Standard 90.1 (ASHRAE 2013).
Four-Pipe System
The old four-pipe, single-coil system had two supply pipes (hot and chilled
water) and two return pipes to the respective source, with coordinated three-way
u
valves and reduced chance of damage to the source (Figure 5-2). Of course, the
problem of getting mixed cold and hot water exists due to the tolerances of the
valves, so this configuration is also no longer applied in HVAC.
What is applied today, and is often called a four-pipe system, is the applica-
ro
tion of two two-pipe systems providing distinct hot- and chilled-water systems
using separate zone coils for each, separate supply and return piping, separate
coils, and separate (but integrated) controls. Typically, the design and controls
coordinate the supply of heating and cooling so that heating is not applied to a
system being cooled or vice versa to provide comfort per energy design codes.
This is not always the case, though, as the classic reheat system is allowed by
G
exception in some jurisdictions where moisture and mold control is a climatic

concern. It may also be applied in certain process instances.
When considered carefully, zone comfort control systems tend to dictate
how energy is distributed from the source to the load and back again. Operating
characteristics of the zones versus the capabilities of the sources can lead to
unique methods of operating and distributing the flow of water to the system.
We will examine several of these.
e t
gp
pE
Figure 5-2 Four-pipe dual-temperature water system.
Performance and Control

A terminal unit can be considered a general description for any heat transfer
device, from a simple piece of hydronic radiation equipment to large heating or
cooling coils in an air-handling unit that delivers sensible heating or cooling to
u
the load.
Consider Figure 5-3; simplify the heat transfer with respect to controls.
As outdoor air changes temperature, more or less heat is required in the space
to maintain a set temperature. It is a simple linear function, a proportional
ro
function. To control the room temperature by sensing what is occurring and

reacting to it, a closed-loop control is established. As it cools outside, there
will be more energy transfer and the space will cool. A room controller,
known as a thermostat, senses this cooling and takes corrective action to add
more heat to the space. This action could be varying the airflow in the space,
or varying the temperature of the water in the space, or varying the amount of
G
water flowing to the coil. Throttling water flow has been the simplest and
most commonly applied method for control. Now, as it cools outside, a
higher flow rate of a fixed-temperature hot water flows to the terminal in the
room. As it warms up outside, flow is reduced (Figure 5-4). This is the
essence of proportional control. As long as operating characteristics stay
constant, the proportional controller on the wall adjusting the flow of water
can be adjusted so that it maintains the required space temperature. However,
e t
gp
pE
Figure 5-3 Space heat transfer in proportion to outdoor temperature.
u
ro
G
Figure 5-4 As outdoor temperature increases, hot-water flow decreases.

if variables change too much, the energy balance will be different, and the
controller will be hard pressed to maintain the set temperature. Therefore, it
is important to recognize that there are implied linkages among the various
design decisions made by the designer. In our case, we need to watch and
evaluate the different strategies that we may apply to the hydronic system.
For instance, as we will see, changing the differential pressure across the sys-
tem will change the control valves’ flow rate and the range of controlled
flow/heat transfer that the thermostat “thinks” it controls. This impacts the
t
proportional adjustment setting of the controller. Similarly, modification of
the supply water temperature to the terminal might be considered to save
e
some energy. This is typically done using a strategy known as outdoor reset.
Again, the range of controlled flow/heat transfer that the thermostat “thinks”
it controls through the valve is modified, impacting the proportional adjust-
ment setting of the controller. There is nothing wrong with these strategies,
gp
but they may be limited in their individual and collective application.
Improperly applied, they may also lead to the controller operating in a state
of noncontrol known as hunting, where, in our example, the valve would
cycle open and closed, leading to other problems.
Load control of terminal units and the relationship to proper distribution of
flows is an important consideration in hydronic system design. The designer
must determine what system method will be used and its control impact on
pE
both the thermal and hydraulic characteristics.
As illustrated in Figure 5-5, terminal control methods include controlling
the following:
• Water flow
• Water temperature
• Airflow
• Air temperature
u
As discussed in Chapter 13 of the 2012 ASHRAE Handbook—HVAC Systems
and Equipment, the basic principles of HVAC systems control are similar regard-
less of the control method (ASHRAE 2012). Every system uses a feedback con-
ro
trol loop where a sensor compares actual conditions to the set conditions. Based
on the results of this comparison, the system is instructed to deliver more heating
or cooling as required to make the actual conditions match the set conditions. This
is an extreme simplification, but it presents the founding principles on which all
control systems are based.
Controlling Water Flow

G
A common method of controlling load in a typical HVAC loop is varying

the water flow in a load coil by thermostatic positioning of a two-way control
valve (see Figure 5-5). Similarly, modern direct digital control (DDC) systems
use solid-state sensors to measure temperature and digitally control valve posi-
tioning. Either method controls the coil’s thermal output by controlling the
quantity of water flow in response to the sensed thermal load requirements.
e t
gp
pE
Figure 5-5 Flow control using two-way valve.
Types of HVAC Control Valves

u
Control of water flow through the coil should be accomplished using two-way
valves. In the past, it was very common to apply three-way valves (Figure 5-6).
ro
Conceptually, the two-way valve allows for modulation of the flow match-
ing required for part-load heat transfer, with the coil output and the flow rate to
achieve it. For systems where there is less concern over comfort, the same two-
way valve is used to essentially turn water on or off to the coil, relying on the
timing associated with varying the full flow to the coil and the turbulent mixing
of conditioned air from the coil with the still air of the conditioned space.
G
Traditional three-way valves operated in the same manner; however, they

would integrate a bypass flow port so that system water flow was kept constant.
We will see later that this was not the case; however, the idea was that the source,
be it chiller or boiler, was too sensitive to modifications of system flow rate,
therefore, the three-way valve application would allow for full system flow to
always be circulated. Energy design standards tend to limit the application of
three-way valves.
e t
gp
pE
Figure 5-6 Flow control using three-way valve.
Two-way valves are available in single-seat or double-seat bodies

(Figure 5-7). Single-seat bodies are most common but require adequate actua-
tor size to overcome water system differential pressures. The double-seat style
reduces the actuator force required due to the pressure counterbalance of the
u
forces in the two-seat area design.
Three-way valves, available as a mixing or a diverting pattern, are some-
times considered where continuous system flow is desired (Figure 5-8). How-
ro
ever, this has increased energy impacts on the system in pumping power and
the source energy load. Another application is where flow diversion is
required, such as condenser flow to or bypassing a cooling tower.
The three-way valves applied to control heating or cooling coils (Figure 5-6)
permit the supply water to bypass the coil when the thermostat is satisfied, provid-
ing continuous flow in the supply and return mains and the source. A balancing
G
valve is provided in the flow bypass line to set a pressure drop in the pipe equal to
the terminal drop when flow is on full bypass. The port controlling the flow
through the terminal should be chosen with an equal percentage characteristic and
the bypass port selected with a complimentary linear characteristic to maintain
nearly constant total flow. However, many if not all of the three-way valves applied
in U.S. HVAC service incorporate two linear characteristic control ports, which can
lead to increased system flow at valve positions nominally considered as part load.
e t
gp
Figure 5-7 Two-pipe control valve body–single and double seat.
u pE
ro
G
Figure 5-8 Three-way control valves—mixing and diverting types.
Aside from the valve providing throttled system flow or constant sys-
tem flow, there are several styles of valve body that are utilized in HVAC
systems. Some of the more common are discussed in the following subsec-
tions.
Globe Valves
Globe valves are the traditional HVAC control valve using linear actuation
that increases or decreases the flow control orifice of the valve by moving a plug
in and out of the hole. Figures 5-7 and 5-8 provide a generic view of a globe-style
valve. The plug is available in several varieties, but generally it is shaped to pro-
vide a specific flow characteristic relationship between valve position and valve
flow rate; this is discussed in the Valve Authority section. These types of valves
t
tend to be the traditional industry workhorse for flow control due to the nature of
the control system actuator and the forces required to move the valve.
e
Ball Valves
Actuated ball valves (Figure 5-9) have become a popular choice for flow
gp
control in HVAC applications due to their small size and the mechanical char-
acteristics of direct-drive electronic actuators. Generally, a ball valve is a sim-
ple device with a rotating sphere captured in the valve body that has a hole
drilled through it for water to pass through. In order to create the flow charac-
teristics required for HVAC, shaped plates are matched to the hole profile so
that the correct relationship of flow and position can be attained to control the
coil. A variation of the ball valve is the segmented ball valve, which is similar
to one half of a ball plug that can have a flow characteristic machined into its
pE
profile easier than a ball can.
Butterfly Valves
As flows become much larger, globe and ball valves become significantly
more expensive. Therefore, sometimes designers apply a butterfly valve, which is
essentially a rotating plate constricting the flow of water through the pipe. While
providing economic benefit from reduced material costs for the valve body, they
are not always a good selection for flow control, unless it is just coarse control.
u
ro
G
Figure 5-9 Actuated ball valve.

Other Valves
There are other types of valve bodies, plugs, and assemblies that exist to
serve specific system needs. They may also be hybrids of the basic valve types,
combining different throttling and control operating characteristics.
Hydronic Accessories
t
In addition to proper sizing of the control valve, hydronic accessories need to
be evaluated for good system operation and future servicing. These accesso-
e
ries include the following (see Figures 5-5 and 5-6):
• A service valve (gate or ball valve), which is used to isolate the terminal
unit, control valve, and strainer for servicing without shutting down the sys-
gp
tem or draining mains.
• A strainer, which is used to reduce foreign material from clogging flow in
the terminal unit or the control valve.
• A balancing valve, which is used to proportionally balance the terminal
flow in the test-and-balance portion of the project commissioning. It can be
closed for terminal servicing. The balancing valve permits measuring ter-
minal flow by reading P across its test ports and using the flow formula.
pE
The balancing valve is chosen for a minimum pressure drop (2.31 to 5 ft) in
wide-open position at the design flow using the flow formula.
• Unions, test ports, and flexible hoses.
Determining Valve Flow Rate

Control valve flow rate is determined by the load calculation, as indicated by
the coil heat transfer to be controlled. To attain the control, the valve is sized to
provide the design’s maximum flow through the control port based on design
u
(gpm) and the designer’s choice for pressure drop (psi). The valve is rated with a
flow coefficient, Cv, which is determined by the valve manufacturer after labora-
tory testing to ANSI/ISA Standard 75.02.01 specifications (ISA 2008). Valves
ro
are tested in a configuration using line size pipe into the valve and operation
under constant differential pressure. The test yields the flow coefficient for wide-
open and partial operating points. The test also determines other valve operating
characteristics. The equation then for flow coefficient calculation is
P
Q = C v ------- (5-1)
G
sg
where
Cv = valve flow coefficient, gpm at 1 psi drop
Q = design flow rate, gpm
sg = specific gravity of fluid (water = 1)
P = design pressure drop, psi
Because piping systems are usually sized on feet head drop, the water flow
formula can be shown as
Q = 0.67C v H
where
Cv = valve flow coefficient, gpm at 1 psi drop
Q = design flow rate, gpm
H = design pressure drop, feet of head
t
Control valve sizes are generally not the same as the pipe size; in fact, they
e
are often smaller. This needs to be monitored, though. Globe valves redirect flow
through the body in such a manner that there is a high degree of internal losses
as the flow of water is controlled. When comparing a ball valve or butterfly
valve in the same testing procedure as the globe, one will notice that the ball and
gp
butterfly valves have significantly higher flow coefficients than a similarly sized
globe valve. This is due to ball and butterfly valves having essentially straight
through-flow profiles, going through minimal flow directional changes and cor-
responding head losses. This can cause problems. A designer might select a ball
valve for the correct flow coefficient that is two to three valve pipe sizes smaller
than the corresponding pipe size. The corresponding change in velocity head
(V2/2g) when piped must be corrected for in the valve selection. The operating
valve flow coefficient will be lower than that stated by the manufacturer; hence,
pE
the flow required by the system under control will not be developed. Correction
factors are applied to the sizing equation, and the valve manufacturer must test
the valve to develop the correction factor. It cannot be assumed that correcting
for velocity head will simply transpose to the rated flow coefficient.
The Cv is based on the control manufacturer’s valve test data to a standardized
test. Typically ANSI/ISA Standard 75.02.01-2008 (ISA 2008), forms the basis
for this testing. However, different manufacturers of similar size and geometry
valves often have widely varying flow coefficient ratings.
u
System Control Characteristics

On the coil side, the energy emission from a hydronic coil has a nonlinear
ro
characteristic with respect to flow. Figure 5-10 illustrates this relationship for a
heating coil with a 20°F water temperature drop. The nonlinearity can be
improved by selecting a coil design with a higher water temperature design
drop (such as the 60°F shown in Figure 5-10). Recalling the previous discus-
sion about control linearity, the control valve and coil must give the appearance
to the controller that there is a linear relationship between valve position and
G
heat transfer. To integrate the nonlinear coil with the linear controller, the valve
is given a flow characteristic that complements the relationship of flow to heat
transfer, which is called equal percentage characteristic.
Flow characteristics refer to the normalized relationship of percentage design
flow to percentage valve position. There are three core flow characteristics—quick
opening, linear, and equal percentage—as illustrated in Figure 5-11. There are also
a host of modified characteristics to meet specific purposes or applications.
e t
gp
pE
Figure 5-10 Nonlinear heat transfer versus water flow in hydronic coil.
Source: ASHRAE Handbook—HVAC Systems and Equipment (2012), Chapter 13.
u
ro
G
Figure 5-11 Core flow characteristics for a valve.

Equal Percentage Characteristic

Generally, for temperature control by directly modulating the flow to the
coil, equal percentage valves are applied. The flow through the valve may be
simulated through the equation
l–1
C v = C vmax R (5-2)
t
where
e
Cv = flow coefficient at partial operating height, gpm/psi
Cvmax = flow coefficient at full operating height, gpm/psi
l = partial lift of the valve, a value from 0 to 1 equivalent to
gp
percentage lift
R = rangeability
Why the use of the term equal percentage? The relative (percent) change in
flow rate with respect to a given change in the lift is a constant, hence the term
equal. Because of this feature, with a larger valve opening (or flow rate), the same
percent change in the lift will lead to a larger increase in the actual flow rate.
pE
A simple characteristic, based off Equation 5-2, is shown in Figure 5-12.
u
ro
G
Figure 5-12 Valve flow characteristic.

It can be observed that at 0% lift of the valve, there is about 3% flow indi-
cated by the equation. This is what leads to the application of modified equal
percentage valves, either by biasing the valve down or by modifying incremen-
tal points.
Valve Rangeability
Valve rangeability is the ratio of the minimum controllable flow of the
t
valve to the maximum flow at a constant differential pressure and is tested as
part of ANSI/ISA Standard 75.02.01 (ISA 2008). Typical rangeability of com-
e
mercial control valve bodies is from 7:1 for large flanged globe valves to 100:1
for finely machined and small globe valve bodies. That said, these numbers
specifically refer to the valve body. ANSI/ISA Standard 75.02.01 allows for
gp
determination of the rangeability as an actuated body, which can lead to a sig-
nificant increase in stated rangeability. Rangeability higher than 300:1 can be
found depending on the manufacturer’s specification and actuator design.
The impact of rangeability is seen in the associated flow rate and corre-
sponding heat transfer. A valve with 7:1 rangeability is capable of predictably
controlling to about 6.5% minimum flow. If a hot-water coil were under con-
trol, the minimum controllable heat transfer would be about 30%.
pE
Valve Authority
From a design standpoint, it is very important to select a control valve pres-
sure drop that provides linearity to the combined coil and valve characteristic.
That means that for an equal percentage valve to maintain its relative shape, the
valve must have a significant portion of the head required to flow water at the
design flow rate through the valve. The general guidance on this is that at least
30% to 50% of the system loop pressure drop is necessary to provide as high a
u
valve authority as feasible to maintain the curve shape.
Valve authority index is defined as the ratio
ro
P min
= ---------------
- (5-3)
P max
where
= valve authority
Pmax = pressure drop across the system, subsystem, or the controlled
G
valve subsystem
Pmin = pressure drop across the control valve
Valve pressure drop in the authority calculation is straightforward. Across
the system, however, it is less so. In a constant-speed pumping system, the sys-
tem pressure drop is the pump head. However, not all systems are constant
speed, and they may also have specific control operations that control the sys-
tem differential pressure in the circuit of the valve. This means that the use of
the indexing variable can change based on the system under control. Another
issue with the use of the indexing number is that as a calculation it does not
yield usable data for explaining part-load system performance.
Figure 5-13 displays the distortion effect of low control valve authority due
to low pressure drop and how it can be improved by selecting a higher pressure
drop with an equal percentage control valve characteristic.
To develop these curves, a different form of the valve authority equation
t
is used that relies on the fundamentals of the system curve discussed in
Chapter 3 and developed further in Chapter 8. This equation is based on the
e
same concept as the equation used for valve sizing, Q = Cv × SQRT P.
When one considers the system curve, it becomes apparent that there is a
shared reference point of flow at 2.31 ft or 1 psi that is the same as the flow
coefficient for a valve. The difference, if there is one, is that a valve flow coef-
gp
ficient is developed from a standard test. Thus, a flow coefficient may be
implied for the system, and P may be considered as a function of flow and
flow coefficient such that
Q 2
P = ------- (5-4)
Cv
u pE
ro
G
Figure 5-13 Distortion effect of low control valve authority caused by low pressure drop.
Source: ASHRAE Handbook—HVAC Systems and Equipment (2012), Chapter 47.
So we know that when we are sizing the system, we calculate total head loss
P ( h) by summing all of the head losses in the pipe and components so that the
system head loss is the sum of the components. Using this concept, we could add
differential pressures for the system as flows divided by flow coefficients squared
for all the components and the control valve. Skipping a lot of math, we determine
that the system flow coefficient would be the sum of the components
1
1 - = -------- 1 1
------- + -------- + + -------- (5-5)
t
2 2 2 2
C vs C v1 C v2 C vn
e
for a number of component’s series, and when we then equate this to determin-
ing the system flow coefficient based on having a control valve with a unique
flow coefficient for each position and the other system components that have
fixed coefficients, the equation for the system coefficient becomes
gp
Cv – v Cv – C
C vs = ------------------------------------
- (5-6)
2 2
Cv – v + Cv – C
where
Cvs = composite or equivalent system coefficient
Cv – C = the component flow coefficient calculated from the pressure drop
pE
of the components (not including the control valve)
Cv – v = control valve flow coefficient at each incremental position
The distortion curves are calculated from this equation. Of course to see the
full effects of the distortion on the inherent equal percentage characteristic,
either through calculation or by laboratory test, the calculation must be per-
formed several times, making it ideal for a spreadsheet. Having the system
flow coefficient for each incremental valve position makes simulating system
flow control straightforward by creating system curves for the values and inter-
u
preting the corresponding intersection on the pump curve.
Modulating Control
ro
Regarding the 20°F design drop curve, note that as the flow is reduced from
full flow to 50% the coil heat emission reduces to 90%, and when reducing the
flow to 10%, the coil output is 45%.
A similar characteristic is observed for a chilled-water coil and is shown
in Figure 5-14 (ASHRAE 2015). Note that the flow must be above 40% to
begin moisture removal (latent heat), and this varies as a function of coil con-
G
struction. The typical control valve characteristics are available (for example,
quick-opening, linear, and equal percentage are shown in Figure 5-11) at a con-
stant pressure drop (ASHRAE 2012). The valve characteristic is determined by
the design or shape of the valve plug related to its travel in the valve port.
Figure 5-15 shows the ideal combination of the equal percentage valve curve with
the water coil emission curve to result in a nearly linear coil output versus valve
position (at constant pressure drop) (ASHRAE 2012, Chapter 47).
e t
gp
pE
Figure 5-14 Chilled-water coil heat transfer versus water flow.
Source: ASHRAE Handbook—HVAC Systems and Equipment (2015).
u
ro
G
Figure 5-15 Ideal combination of equal percentage valve curve with water coil emission
curve.
Using the technique described for calculation of flow coefficient and a

spreadsheet, the control effect of valve authority may be observed, as shown in
Figure 5-16.
The resultant control line of lift versus heat transfer at 100% valve author-
ity is essentially linear. The flow distortion due to valve authority at 30% still
has a bit of equal percentage curve to it; however, the effect on the controller is
seen in the resultant lift versus heat transfer 30% plot. A good bit of the line is
linear, so the controller proportional band can be adjusted to provide control
t
response; however, the slope of the response is steeper than the theoretical
equal percentage valve, and the result will be achieving design flow conditions
e
at loads around 70%. This is too early and causes excess energy use.
Two-Position Control
gp
In a terminal control application (Figure 5-4), if the control is two-position action
(Figure 5-17) (Landis and Gyr n.d.), then the valve is either open or closed and
the coil discharge temperature is abruptly changed (on/off). This cycling of dis-
charge temperature can be uncomfortable to occupants. Two-position control can
also cost extra energy use because valves tend to be open longer, allowing for
higher likelihood of all valves being open at one time, thus demanding design
flow from the pump and the source providing fluid reconditioning. Two-position
pE
control is still proportional control, as the gain in a proportional controller can be
adjusted so that the controller acts in a two-position mode.
u
ro
G
Figure 5-16 Control effect of valve authority.

Proportional and Proportional Integral Control

Where people occupy the load space, a modulating proportional control
action (Figure 5-18) is usually recommended to gradually adjust the flow in the
load coil in proportion to the space temperature difference above or below the
e t
gp
pE
Figure 5-17 Two-position control.
u
ro
G
Figure 5-18 Modulating proportional control.

desired room setpoint (ASHRAE 2015). A gradual change of water flow in the
coil reduces the extreme change in the coil discharge temperature and is more
comfortable to the occupants. This control mode also uses the least amount of
energy, as flow is provided only at the level required for load. As can be seen
from the nonlinear relationships of the equal percentage valve and the coil, 60%
of the load requires only 20% of the design flow, potentially allowing the pump
to operate at a much slower speed. More importantly, when compared to many
load calculations, it can be see that about 80% of the yearly hours of operation
t
will work with less than 60% heat transfer. That is an excellent opportunity for
energy savings, all just by designing the system to operate properly.
e
Achieving this performance requires use of proportional and integral con-
trol actions to reduce proportional control offset. The proportional integral (PI)
control is illustrated in Figure 5-19. Depending on the load conditions, the
space temperature will experience an offset above or below the desired temper-
gp
ature setpoint. The purpose of the integral action added to a proportional con-
troller is to reduce this offset to a minimum. This will keep the actual
temperature closer to the setpoint to improve occupant comfort and reduce
energy consumption by reducing the cycling of the space temperature. The
control mode will also compensate for minor deviations in other variables that
affect flow and temperature control, such as changes in air volume, water flow
rate, and water temperature.
pE
System Control Configurations
Controlling water temperature in a primary/secondary pumping arrange-
ment with a two-way valve is another approach to improving valve coil control,
as shown in Figure 5-20. This permits constant flow at all times at a variable
temperature in the coil circuit at the design velocity to maximize coil heat
transfer. Properly applied, the secondary pump operates independently from
u
the primary pump and reduces the pressure drop on the primary or distribution
pumping.
ro
G
Figure 5-19 Proportional integral (PI) control.

e t
gp
pE
Figure 5-20 Primary/secondary pumping with two-way valve.
Source: Bell & Gossett (2013).
The key is to provide a pipe common to both the primary and secondary
u
circuits with no pressure drop. As shown in Figure 5-20, this common pipe is
usually located in a bridge between the supply and return mains of the pri-
mary. The common pipe is selected with no pressure drop to either the sec-
ondary or primary circuits. As the temperature controller calls for an increase
ro
in cooling, the two-way valve opens, permitting primary water to mix into
the secondary. An equal amount of secondary return water is displaced into
the return main. When V-1 is fully open, chilled water flows from supply
main a into the bridge and into the secondary at b, through the load, and
returns to the bridge at c, where it flows through V-1 to the return main. As
valve V-1 is throttled by the zone thermostat, less chilled water is supplied to
G
the bridge at b and the secondary pump draws the balance of flow from the
common from c to b, thereby causing a mixing action. When the zone ther-
mostat is completely satisfied, the control valve V-1 is closed and all the sec-
ondary flow is recirculated from c to b.
A balancing valve is provided in the bridge return after the control valve.
This valve is set proportionally for the design flow in the testing and balancing
of the system. Note that the control valve is the major resistance in the bridge
e t
gp
pE
Figure 5-21 Primary/secondary with check valve in common.
circuit and should be selected for at least 50% of the secondary loop pressure
drop. The secondary pump is selected for the coil design flow and the pressure
drop of the coil and the secondary loop.
u
A balancing valve may also be installed in the secondary circuit for testing
and setting this flow. The secondary circuit can be limited to a unique coil or it can
supply a zone or riser in a large system. In a large zone, the zone water (second-
ary) temperature can be controlled independently from the primary by locating its
ro
sensor in the secondary return, or it can be a combination of zone supply and

return temperatures to control differential temperature. The secondary circuit’s
flow may be set so that a higher differential temperature can provide different
design conditions. It is important that the common piping be correctly installed so
that improper mixing does not occur. This is especially true on chilled-water sys-
tems where the available tonnage capacity may be limited.
G
Variations to the primary/secondary system include the following:
• Adding a check valve to the common to prevent the possibility of short-

circuit flow in the common, as shown in Figure 5-21. When the check
valve closes, this causes the primary pump to go into series with the sec-
ondary pump causing an overflow in the secondary and a possible waste of
primary pumping power.
• Placing a control valve in the common controlled by the zone thermostat or

secondary water supply or return as shown in Figure 5-22. A differential
pressure sensor controller senses a small flow across the valve in the com-
mon and positions the secondary zone valve in the bridge return. (This is a
proprietary design and may require patent licensing or purchasing from a
single source.) Placing a valve in the common adds resistance and forces
the primary pump in series with the secondary pump; this may waste pri-
mary pump head in the secondary circuit and possibly cause overflow of
t
the secondary.
e
• Including temperature sensors in the supply and secondary piping, as shown
in Figure 5-23. This permits the common to behave in its normal fashion. Ide-
ally, the chilled-water supply is drawn into the secondary by the action of the
gp
zone thermostat positioning valve V-1. As valve V-1 is throttled by the zone
thermostat, less chilled water is supplied to the bridge at b, and the secondary
pump draws the balance of flow from the common from c to b, thereby caus-
ing a mixing action.
u pE
ro
G
Figure 5-22 Primary/secondary with valve in common and differential pressure transmitter
monitoring flow.
Source: Mannion (1988).
When the zone thermostat is completely satisfied, valve V-1 is closed and
all the secondary flow is recirculated from c to b. The purpose of sensors T-1
and T-2 is to minimize bridge flow from b to c for large flow situations by
maintaining T-1 – T-2 < 1°F by overriding valve V-1. Optional sensors T-4
and T-3 are only required for large zones where the secondary pump is sup-
plying a number of terminals to maintain a low limit to open and override V-1
if T-4 > T-3 = 1°F. The monitoring of the zone room sensor and the zone water
temperatures can be accomplished by a DDC cabinet or a dedicated micropro-
t
cessor controller.
• Using a face bypass damper (Figure 5-24). The airflow can be throttled to
e
meet load conditions by mixing terminal coil discharge airflow at the termi-
nal coil and bypass airflow to maintain constant airflow. The control valve
is sequenced to open before the face damper is opened. On an outdoor air
gp
application, a low-temperature detection thermostat is located on the leaving
air side of the coil in the return pipe from the coil. Another concept is to open
the coil valve when the outdoor air drops below 40°F. The face bypass princi-
ple can be used on the preheat coils of air-handling units to reduce the possi-
bility of coil freeze up.
u pE
ro
G
Figure 5-23 Primary/secondary with temperature sensors in bridge.

e t
Figure 5-24
gp
Terminal with face bypass control.
pE
• Mixing the airflow through the coil face with bypass air from the return
air duct in an HVAC supply unit cooling coil (Figure 5-25). The bypass
air is preconditioned return air to reduce additional moisture load from
outdoor air/mixed air at partial loads.
• Varying the supply water temperature to a heating coil for part-load condi-
tions (Figure 5-26). Note that for a given entering air temperature, a flow of
u
100% at a baseline of 160°F delivers 100% heat transfer. At 180°F, 120%
of the heat transfer is provided, and at 200°F, 140% heat transfer is estab-
lished. In all cases, the coil maintains a relatively curved shape similar to
the base coil characteristic. This can be contrasted with a similar coil oper-
ro
ation that controls through changing entering water temperature. Flow is

kept at 100% to the coil through a circulating pump, and capacity is varied
as a function of entering water temperature. Note the linear control charac-
teristic that is established in Figure 5-27.
G
Pressure-Independent Control Valves

An alternative form of hybrid control valve that is discussed in Chapter 8 is
the pressure-independent control valve (PICV). Throughout this section, the
working premise of proper control valve operation and selection comes from the
implied relationship of a stable pump selection. Traditionally, this means that in
larger systems, a relatively flat performance curve for the pump is selected and as
e t
Figure 5-25
gp
Terminal with face bypass control and conditioned bypass air.
u pE
ro
G
Figure 5-26 Variable supply water temperature for part-load conditions.

e t
gp
pE
Figure 5-27 Linear control characteristic.
control valves throttle, reducing flow, the relatively minor increases in system
head attributed to the pump increase yield a very minor increase in flow across the
throttled control valves.
u
In variable-speed pumping systems, as flow decreases yielding an increase
in system differential pressure, pump speed is reduced and greatly reduces sys-
tem differential pressure. In theory, energy savings are harvested through the
ro
pump speed reduction. However, these same differential head savings may also
cause zone control loop instability due to the hydraulic interactions of the con-
trol valve and pump. One method of limiting or eliminating the pump speed
change as a source of system instability is to add a pressure control valve to the
circuit of the temperature control valve or of a number of temperature control
valves.
G
Figure 5-28 schematically shows a fluid-system-powered regulator (a

self-contained control device) that opens and closes in response to changes in
system pressure. It acts to maintain a constant differential pressure available
for the temperature control valve orifice. Doing this stabilizes available dif-
ferential pressure for the control valve (as long as the pump differential pres-
sure setpoint is properly controlled), which makes heat transfer control
potential capability constant. In this way, if the controller were to react to a
e t
gp
Figure 5-28 Fluid-system-powered regulator.
pE
zone disturbance, the capability to get the required flow for control is there.
As a by-product of the regulator, this also stabilizes the inherent characteristic
of the control valve, reducing or eliminating hydraulic interaction effects
associated with control valve authority.
What is making these valves more noteworthy is that manufacturing tech-
nologies and materials have been developed that allow the differential pressure
regulator to be implemented in the same body as the temperature control valve,
down to a small level (Figure 5-29). The valve concept has been available for a
u
long time; however, it was implementation size and cost that generally made
the concept less applicable.
All things considered, these actions are good things from a control stand-
point. On the other hand, the regulating device adds an extra pressure drop in
ro
the circuit, so it does use some energy (head). Most of these types of valves
offer the capability to limit the maximum flow that the valve controls. In this
regard, they eliminate one of the functions of a balancing valve. However, they
do not allow for proportional balancing unless special control considerations
are taken in the programming of the control system. This is discussed with
variable-speed pumping in Chapter 8.
G
Variable-Speed Circulator on Coil

Figure 5-30 shows another method of zone automatic temperature control
and is a variation of the primary/secondary/tertiary pumping concept, applying a
small circulating pump to directly control the coil (Green 1994). This can be
done with either a variable-speed circulating pump or an on/off circulating pump.
e t
Figure 5-29
gp
Differential pressure regulator in same body as temperature control valve.
u pE
ro
G
Figure 5-30 Variable-speed circulator on coil.
Research into the control method has shown that the temperature control can be
maintained quite close to the setpoint. On the other hand, and especially when
applied to hot-water heating systems, care must be exercised to prevent gravity
circulation from the supply or return main through the coil. This is typically
accommodated through application of a flow control check valve.
The Next Step

In the next chapter, we will discuss expansion tanks and air elimination.
Summary
t
• The variables involved with terminal control
• What mechanical components are suggested at a terminal
e
• How a terminal control valve is selected
• The four types of terminal control action (off/on position, proportional [P],
proportional integral [PI], and proportional-integral-derivative [PID])
gp
• The emission characteristics of heating versus cooling terminals
• The types of control valve characteristics available and what works best
with a hydronic coil
• Valve authority, rangeability, and selection
• The advantages and disadvantages of two-way and three-way valves
• The principles of primary/secondary pumping systems and different meth-
ods of control
pE
• What types of control method vary the flow of air through a terminal
References
Atlanta: ASHRAE.
ASHRAE. 2013. ANSI/ASHRAE/IES Standard 90.1-2013, Energy Standard
u
ASHRAE. 2015. ASHRAE Handbook—HVAC Applications. Atlanta:
ASHRAE.
Bell & Gossett. 2013. Primary secondary pumping application manual. Tech-
nical Report 775A. Morton Grove, IL: Xylem Inc.
ro
Green, R.H. 1994. An air-conditioning control system using variable-speed

water pumps. ASHRAE Transactions 94(1):463–70.
ISA. 2008. ANSI/ISA Standard 75.02.01-2008 (IEC 60534-2-3 Mod), Control
Valve Capacity Test Procedures. Research Triangle Park, NC: International
Society of Automation.
Landis and Gyr. n.d. Fundamentals of Control. L&G Appl. Bul.
G
Mannion, G. 1988. High temperature rise piping design for variable volume
systems: Key to chiller energy management. ASHRAE Transactions
94(2):1427–43.

5-1 A typical fan-coil terminal requires 4 gpm. What valve Cv should be selected if
a control valve is specified for a 9 ft drop?
t
5-2 What piping components should be specified at the terminal?
e
5-3 What type of control action should be considered to reduce discharge tempera-
ture cycle?
5-4
gp
A control valve is to be selected for a 44 gpm terminal coil; coil drop is 18 ft.
Select the correct size of control valve if the valve is specified for 50% of the
coil drop, and the typical valve sizes and Cv available (Cv is in parentheses) are
• 0.5 in. (2.5), 0.75 in. (6), 1 in. (10)

• 1.25 in. (16), 1.5 in. (20), 2 in. (36)
pE
5-5 What control valve flow characteristic should be specified for proportional
control of a hot water heating control?
5-6 An on/off thermostat controls a cabinet heater in a hallway. What valve flow
characteristic should be specified?
5-7 A three-way valve is to be applied to a refrigeration condenser and cooling

tower to maintain a 95°F condensing temperature. What type of three-way
u
valve arrangement should be applied?
5-8 It is desirable to control flow in a chilled-water coil down to a minimum of 5%

of design flow before close off. In addition to proper valve sizing for design
ro
flow capacity and proportional control, what else should be specified?
5-9 What should be specified in the bypass circuit of a three-way valve?
5-10 What type of control method varies airflow through a terminal coil?
G
5-11 Explain the difference between primary and secondary pumping systems.
5-12 Define valve authority, rangeability, and selection.
5-13 How is terminal control valve size selected?
5-14 What are the three types of terminal control action?

Expansion Tanks and
Air Elimination
e t
Study Objectives
gp
❑ understand the differences between open and closed systems,
❑ understand where air comes from in a hydronic system,
❑ know what maintains minimum and maximum pressures in a closed
hydronic system and how they can be set,
❑ know the types of compression tanks and where they are located in a
pE
hydronic system,
❑ understand what solubility of air is in a hydronic system and what factors
determine its increase,
❑ know what factors are needed to size and select a compression tank, and
❑ know where the point of no pressure change is.
Instructions
u
Open and Closed Water Systems
The fundamental difference between an open and a closed water system is the
ro
interface of the water with a compressible gas (such as air) or an elastic surface
(such as a diaphragm).
An open system (such as a cooling tower system) has at least two points of
interface with the air. In a cooling tower, these points of interface are the sur-
face of the tower basin water and the discharge pipe or nozzles entering the
tower. A closed water system has only one point of interface with a compress-
G
ible gas or surface, and flow may not be caused by elevation differences. Both
open and closed systems demand different design considerations.
Typical Open System
An open system is shown in Figure 6-1, illustrating a water-cooled refrigeration
condenser employing a cooling tower to provide atmospheric cooling of the
condenser water. A float-type valve is used in the tower sump to provide
172 Chapter 6 Expansion Tanks and Air Elimination
e t
gp
pE
Figure 6-1 Typical open water system.
makeup water to maintain a predetermined water level. In addition, water treat-

ing and screening equipment are provided to maintain proper water flow condi-
tions. Local building codes should be checked to ensure compliance of the
system design with local requirements.
u
Typical Closed System
A typical hydronic heating or cooling system, as shown in Figure 6-2, is
fundamentally a closed system. In addition to the source of heating or cooling,
ro
the distribution pumping, and the piping arrangement, it must include a means
of system pressure control.
In closed systems, the objectives of system pressure control are to
• limit the pressure of all system equipment to its allowable working pres-
sure;
G
• maintain minimum pressure for all normal operating temperatures, vent air,
and prevent cavitation of the pump suction and boiling of system water; and
• accomplish these objectives with a minimum addition of new water. (Air is
often admitted into closed hydronic heating or cooling systems as makeup
water is introduced. The makeup water may be at a different temperature
than the system water and subject to expansion or contraction as it is heated
or cooled.)
e t
Figure 6-2 Typical hydronic system.
gp
pE
Hydronic Accessories
System pressure control is affected using the following hydronic accessories:
• ASME-rated pressure relief valve

• Automatic pressure-reducing fill valve
u
• Expansion tank
• Manual or automatic air-venting equipment
ro
Pressure Relief Valve

A pressure relief valve is selected to limit the maximum operating pressure
of the piping system and the source. The main function is to prevent danger
to occupants, operating personnel, and equipment. In boilers, the relief
valve size and capacity are usually recommended by the boiler manufac-
G
turer. However, the valve is sized for the pressure and thermal capacity of
the boiler and in keeping with the ASME Boiler and Pressure Vessel Code
(2015). Boilers are not the only source that require a relief valve; chillers
also require a relief to prevent damage when the chiller is turned off and the
contained water temperature rises to ambient conditions. Reliefs are meant
to be used only as safety devices and should not normally operate. As such,
the pressure required will account for system static pressure, operating
increase of the source thermally, and the increase in head related to the
location with respect to the pump. If a relief valve is improperly sized, it
may begin to relieve as a normal part of system operation, opening to
relieve internal pressure. It is possible that the fluid flowing through the
relief could carry particulate and suspended mineral solids that could pre-
cipitate across the valve seat and prevent its proper operation, possibly
clogging the port and keeping it closed when it should open. This obviously
causes a potentially dangerous problem where the weakest part of the sys-
t
tem could rupture, possibly causing injury.
e
Pressure-Reducing Valve
gp
A pressure-reducing valve (PRV) is used for system fill and where required for
reducing the pressure of the city main to the level required in the system. An
example is shown in Figure 6-3. As can be seen, there is an adjusting screw
compressing a spring, which serves as a setpoint for a simple system-powered
regulator. The spring exerts force on the upper section of a diaphragm. Under-
neath the diaphragm is a complementary plate connected to the plug of a globe
valve. Porting connects the downstream portion of the valve under the dia-
phragm, transferring the pressure to counteract the force of the spring and thus
pE
controlling the pressure.
u
ro
G
Figure 6-3 Pressure-reducing valve.

Expansion Tank
An expansion tank is a partially filled tank operating at or above atmo-
spheric pressure and located in a water system to accommodate the volume
expansion and contraction of water (Figure 6-4).
Expansion tanks are of three basic configurations:
• Open tank. An open tank is a tank open to atmosphere. These were typi-
t
cally found in older residential heating systems (pre-1950s) and may have
been in larger systems of a similar time period. It is not good practice to
e
base designs on this type of system today.
• Compression tank. A compression tank (ASME rated) (see Figures 6-4
and 6-5) contains a captured volume of system air compressed by the
gp
expansion of water in the tank with an air/water interface (often called a
u pE
ro
G
Figure 6-4 Expansion tanks.

compression tank) and is just a plain steel tank. Use of the compression
tank makes the system an air management system, as air that is in the sys-
tem needs to be captured at one common point and directed to the tank for
the system pressure control. Automatic air vents may not be used in this
system. In addition to the tank, an interface fitting (Figure 6-5) is required
that directs the air into the tank while providing a restriction to prevent
gravity thermal circulation of the tank fluid with the system fluid.
t
• Diaphragm or bladder tanks. A diaphragm or bladder expansion tank
(ASME rated) is one in which a flexible membrane is located between the
e
air and the water. Diaphragm tanks represent air elimination systems as no
air needs to be captured and directed to the tank. As a result, air manage-
ment devices with automatic vents may be used in as many places as
gp
required for the system. That said, it should be noted that if automatic air
vents are used, they must always be placed in locations that have pressure
greater than atmospheric pressure so that air is not drawn into the piping
system.
In the open tank and the compression tank, air can enter the water through
the interface and can affect system performance over long periods of time.
u pE
ro
G
Figure 6-5 Closed tank.

As defined, a closed system should have only one expansion tank. The
presence of more than one tank can cause the closed system to behave in unex-
pected ways and can cause damage from water hammer or shock waves.
An expansion tank is required to serve thermal and hydraulic functions:
• In its thermal function, the expansion tank provides a space into which the
noncompressible liquid can expand or contract as the liquid undergoes vol-
umetric changes due to temperature. For example, at 40°F, 1 lb of water
t
occupies 0.01602 ft3, and at 220°F it occupies 0.01677 ft3. Figure 6-6
shows a graph of the expansion of water above 40°F.
e
gp
u pE
ro
G
Figure 6-6 Expansion of water above 40°F.

• As a hydraulic device, the expansion tank serves as the reference pressure

point in the system, which is analogous to a ground in an electrical system.
Where the tank connects to the piping, the pressure equals the pressure of
the air in the tank plus or minus any fluid pressure due to the elevation dif-
ference between the tank liquid surface and the pipe (see Figure 6-7).
The point of tank interface often is referred to as a point of no pressure
change in a system (see Figure 6-8), making the pump suction the preferred
t
point of installation for the tank. In this way essentially all of the pump
energy is added to the system pressure, which is set via the tank. While this
e
tank location is preferred, it is not a requirement. Tank placement can be
gp
u pE
Figure 6-7 Tank pressure related to system pressure.
ro
G
Figure 6-8 Point of no pressure change.

managed as part of the design to manage the size of the required tank. For
example, in a chilled-water (CHW) system, it is not uncommon to install
the tank at the top of the piping system. This can reduce the size and
expense of the device.
Air Elimination
t
Excessive amounts of undissolved air in a piping system due to improper vent-
ing or removal can make the system operate poorly. It has been shown that air will
e
separate from water at low velocities, and it is recommended that a minimum
velocity of 1.5 to 2 ft/s in 2 in. pipe be used to reduce this risk (ASHRAE 2013a).
This leaves the designer with the choice of implementing either the air
management style of pressure control system, using a steel compression tank
gp
and appropriate fitting, or the air elimination style of system, using a dia-
phragm expansion tank.
When air management is selected, a common point of collection is selected for
the placement of an air separator. The air separator directs the air-entrained fluid to
the compression tank and tank fitting so that air collects in the tank. While termi-
nals are vented manually on start-up to remove large amounts of excess air, after
flow and fill are established, these are kept closed so that air in solution is directed
pE
towards the tank. Automatic air vents should not be used in the system.
When air elimination is selected, air separators may be located anywhere
and should be installed with automatic air vents
Figure 6-9 shows the internal operating mechanism for an example auto-
matic air vent. There are many varieties of construction. Figure 6-10 shows the
installation of automatic or manual air vents in the system zones or coils for
small pipe sizes. It is suggested to place the vent on the leaving side of a coil or
other heat transfer device and ahead of the control valve (on the return) to
enable constant pressure on the vent.
u
Air Separation
ro
On the larger supply or return mains, a dynamic air separator (see Figure 6-11)
is used to allow the air to separate centrifugally from the water. The lighter-
weight air mixture is collected by a perforated collector tube in the center and
flows up to the closed compression tank, displacing water downward (Figure
6-12). Some versions combine a strainer and blowdown connection in the
lower flange.
G
In chiller applications, the air separator, expansion tank, and distribution

pump may be located on the return main pumping into the chiller. The warmer
return water releases air more readily than the CHW.
An air separator is also used with a diaphragm tank and large-capacity
automatic air vent, as shown in Figure 6-13. The air vent should be chosen with
adequate capacity because the diaphragm tank stores the expanding water vol-
ume but not the released air.
e t
Figure 6-9
gp
Internal operating mechanism for automatic air vent.
u pE
ro
G
Figure 6-10 Automatic or manual air vents in system zones or coils for small pipe sizes.
e t
gp
pE
Figure 6-11 Dynamic air separator.
u
ro
G
Figure 6-12 Air separator.

Both Figure 6-13 for diaphragm expansion tanks and Figure 6-14 for com-
pression tanks show the accessories that are required and are important to sys-
tem operation:
• Relief valve protects for maximum system pressure.

• PRV is set to provide minimum system pressure.
• Service valve permits maintenance.
t
• When using a compression tank, a manual three-way hand valve permits
draining excess water from the expansion tank to provide proper air cush-
e
ion in the tank (if it is not part of the tank fitting). When the cushion is lost,
the system does not permit expansion and the relief valve will open, dump-
ing water down the drain. The PRV fill valve may add water and system
pressure may cycle, causing fill-dump-fill, etc.
gp
u pE
Figure 6-13 Diaphragm expansion tank.
ro
G
Figure 6-14 Compression tank.

° A glass sight gage might be added to the expansion tank to indicate

water level and air cushion and may be required by local code. In
some cases, this is required by code. This should be carefully consid-
ered, though, as the compression seals used to join the glass tube tank
fittings must be maintained so that the joint does not become a point
of air or water leakage, which would then defeat the purpose of hav-
ing the tank.
• Air separator should have a blowdown valve to drain collected sediment to
t
the sewer.
e
Other considerations related to the tank connection include the following
(refer to Figure 6-2):
gp
• A tank with an air/water interface is generally used with an air control sys-
tem that continually reverts air into the tank and should be located in a
place where air can best be released from the circulating pump.
• Within reason, the lower the overall pressure in a tank, the smaller the tank
(as will be seen in the tank size calculations); thus, in a vertical system, the
higher the tank is placed, the smaller it can be.
• In primary/secondary or compound pumping systems, the tank should
pE
only be at one location, preferably on the suction side of the source pump
(Figure 6-15).
u
ro
G
Figure 6-15 Tank location for primary/secondary or compound pumping systems.

Sizing Expansion Tanks

With a single tank on a system, and assuming isothermal conditions for the
air, the air pressure changes as a result of displacement by the water. Assum-
ing no water is being added or removed from the system, the only thing to
cause water to move in or out of the tank is the expansion or shrinkage of the
water on the system. Thus, in sizing the tank, thermal expansion is related to
the pressure extremes of the air in the tank (as will be seen in the tank calcu-
t
lations from the examples). The connection point of the tank to the system is
very important and should be based on the pressure requirements of the sys-
e
tem, remembering that the pressure at the tank connection will not change as
the pump is turned on or off. This is also commonly called the point of no
pressure change.
gp
For example, consider a system containing an expansion tank at 30 psig pres-
sure and a pump with a head of 23.1 ft or 10 psig, as shown in Figures 6-16a and
6-16b. Alternative locations for connecting the expansion tank are shown. In
either case, with the pump off the pressure will be 30 psig on both the pump suc-
tion and the discharge.
• With the tank connected to the pump suction side (Figure 6-16a), the pres-
sure increases on the pump discharge by an amount equal to the pump pres-
pE
sure (e.g., 30 + 10 = 40 psig).
• With the tank connected to the pump discharge side (Figure 6-16b), the
pressure decreases on the suction side by the same amount (e.g., 30 – 10 =
20 psig), and pressure at the pump discharge remains at 30 psi (point of no
pressure change).
Sizing the tank is the primary thermal consideration in incorporating a

tank into a system. However, prior to sizing the tank, the control or elimina-
u
tion of air must be considered. The amount of air that will be absorbed and
can be held in solution with the water is expressed by Henry’s equation
(ASHRAE 2012):
ro
P-
x = ---- (6-1)
H
where
H = Henry’s constant
G
P = absolute pressure, psia

x = solubility of air in water, percentage by volume
Henry’s constant varies with temperature and pressure. As shown in
Figure 6-17, it is only constant at a given temperature. While difficult to
determine molecularly, this concept of air absorption has been proven in
test. The solubility in percentage air by volume can be seen for system tem-
peratures and pressures in Figure 6-18. For example, if a system is at
e t
gp(a)
u pE
ro
G
(b)
Figure 6-16 Pressure effects of alternative tank locations: (a) pump suction side and (b)
pump discharge side.
e t
gp
pE
Figure 6-17 Henry’s constant versus temperature for air and water.
u
50 psig and 140°F, the water can contain about 6% air by volume; if the
pressure is dropped to 40 psig at the same temperature, the water contains
about 5% air by volume, or 1% by volume will be released and must be
ro
vented or work its way into the compression tank.

The equation for determining the size of a closed compression tank in a
closed hydronic system is
v2
----- – 1 – 3 t
v1
G
V t = V s -------------------------------------------
- (6-2a)
Pa Pa
------ – ------
P1 P2
where
Pa = atmospheric pressure, psia
P1 = lower-temperature pressure, psia
e t
gp
pE
Figure 6-18 Solubility versus temperature and pressure for air-water solutions.
u
P2 = higher-temperature pressure, psia
t1 = lower temperature, °F
ro
t2 = higher temperature, °F
v1 = specific volume of water (low temperature), ft3/1b
v2 = specific volume of water (high temperature), ft3/1b
Vs = volume of water in system, gal
G
Vt = volume of expansion tank, gal

= linear coefficient of thermal expansion, in/in·°F
( = 6.5 × 10–6 in/in·°F [steel])
( = 9.5 × 10–6 in/in·°F [copper])
t = (t2 – t1), °F
Vs is typically taken from tables such as Table 6-1 (for steel pipe and cop-
per tube); v1 and v2 are found in Table 3 of Chapter 1 of the 2013 ASHRAE
Handbook—Fundamentals (ASHRAE 2013a).
Similarly, the equation for determining the size of a diaphragm compres-
sion tank in a closed hydronic system is
v2
----- – 1 – 3 t
v1
t
V t = V s -------------------------------------------
- (6-2b)
P2
1 – ------
e
P1
Table 6-1 Volume of Water in Standard Pipe and Tube
gp
Standard Steel Pipe Type L Copper Tube
Nominal Pipe
Inside Diameter Volume Inside Diameter Volume
Size
Schedule
in. (mm) in. (mm) gal/ft (L/m) in. (mm) gal/ft (L/m)
No.
3/8 (10) — — — — — 0.430 (10.9) 0.0075 (0.09)
pE
1/2 (15) 40 0.622 (15.8) 0.0157 (0.19) 0.545 (13.8) 0.0121 (0.15)
5/8 (16) — — — — — 0.666 (16.9) 0.0181 (0.22)
3/4 (20) 40 0.824 (20.9) 0.0277 (0.34) 0.785 (19.9) 0.0251 (0.31)
1 (25) 40 1.049 (26.6) 0.0449 (0.56) 1.025 (26.0) 0.0429 (0.53)
1 1/4 (32) 40 1.380 (35.0) 0.0779 (0.97) 1.265 (32.1) 0.0653 (0.81)
1 1/2 (40) 40 1.610 (40.9) 0.106 (1.32) 1.505 (38.2) 0.0924 (1.15)
u
2 (50) 40 2.067 (52.5) 0.174 (2.16) 1.985 (50.4) 0.161 (2.00)
2 1/2 (65) 40 2.469 (62.7) 0.249 (3.09) 2.465 (62.6) 0.248 (3.08)
ro
3 (80) 40 3.068 (77.9) 0.384 (4.77) 2.945 (74.8) 0.354 (4.40)

3 1/2 (90) 40 3.548 (90.1) 0.514 (6.38) 3.425 (87.0) 0.479 (5.95)
4 (100) 40 4.026 (102.3) 0.661 (8.21) 3.905 (99.2) 0.622 (7.73)
5 (125) 40 5.047 (128.2) 1.04 (12.92) 4.875 (123.8) 0.970 (12.05)
G
6 (150) 40 6.065 (154.1) 1.50 (18.63) 5.845 (148.5) 1.39 (17.26)

8 (200) 30 8.071 (205.0) 2.66 (33.03) 7.725 (196.2) 2.43 (30.18)
10 (250) 30 10.136 (257.5) 4.19 (52.04) 9.625 (244.5) 3.78 (46.95)
12 (300) 30 12.090 (307.1) 5.96 (74.02) 11.565 (293.8) 5.46 (67.81)
Source: ASHRAE Pocket Guide for Air Conditioning, Heating, Ventilation, and Refrigeration (2013), Chapter 3, Table 3.7.
Example 6-1
Size an expansion tank for a heating system that will operate at 180°F to 220°F.
For the design, it is given that the minimum pressure is 10 psig (24.7 psia) and
the maximum pressure is 25 psig (39.7 psia). Atmospheric pressure is assumed
to be 14.7 psia. The steel piping in the system has an estimated volume of 3000
t
gal and will have minimum water fill temperature of 40°F.
Solution From the problem stated, we know
e
v1 = 0.01602 ft3/lb at 40°F (assumed lower temperature)
v2 = 0.01677 ft3/lb at 220°F (highest temperature per design)
Vs = 3000 gal
gp
Therefore, applying the equation,
v2
----- – 1 – 3 t
v1
V t = V s -------------------------------------------
-
Pa Pa
------ – ------
P1 P2
pE
0.01677 – 1 3 6.5 10 –6
------------------- – 220 – 40
0.01603
V t = 3000 --------------------------------------------------------------------------------------------------------
14.7 14.7
---------- – ----------
24.7 39.7
V t = 578 gal
u
Example 6-2
ro
Size a diaphragm tank for a heating system that will operate at 180°F to 220°F.
The system will operate with a minimum pressure of 10 psig (24.7 psia) and a
maximum pressure of 25 psig (39.7 psia). Atmospheric pressure is 14.7 psia.
The steel piping in the system has an estimated volume of 3000 gallons and
will have a minimum water-fill temperature of 40°F.
Solution We have been given Vs = 3000 gal.
G
Applying Equation 6.2 for a diaphragm tank:
v2
----- – 1 – 3 t
v1
V t = V s -------------------------------------------
-
P2
1 – ------
P1
0.01677 – 1 3 6.5 10 –6
------------------- – 220 – 40
0.01602
V t = 3000 --------------------------------------------------------------------------------------------------------
24.7
1 – ----------
39.7
V t = 344 gal
t
Pressure and Temperature Considerations
e
Selection of the lower and higher pressures P1 and P2 is critical in these
determinations. Pipe, tubing, boilers, chillers, and coils must be evaluated to
determine the total system volume (see Figure 6-19).
gp
The lower temperature for a heating system is normal ambient tempera-
ture at fill conditions (for example, 40°F to 50°F) and the higher temperature
is the operating supply water temperature for the system. For a CHW system,
the lower temperature is the design CHW supply and the higher temperature
is the ambient (86°F to 95°F). For a hot/CHW system, the lower is the CHW
design temperature and the higher is the heating water design supply temper-
ature. The specific volume data (v1 and v2) are found in the 2013 ASHRAE
Handbook—Fundamentals, Chapter 6, Table 1 (ASHRAE 2013a).
u pE
ro
G
Figure 6-19 Flowchart for sizing expansion tanks.

At the tank connection point, the pressure in closed tank systems increases
as the water temperature increases. Pressures at the expansion tank are gener-
ally set by the following parameters:
• The lower pressure is usually selected to hold a positive pressure at the

highest point in the system (set on the water PRV fill valve).
• The higher pressure is normally set by the safety relief valve selected for
the maximum pressure allowable at the location of the valve (without open-
t
ing the valve).
e
Other considerations are to ensure the following:
• The pressure at any point will not drop below the saturation pressure at
the operating temperature (again refer to the 2013 ASHRAE Handbook—
gp
Fundamentals, Chapter 6, Table 1 [ASHRAE 2013a]).
• All pumps have adequate net positive suction head (NPSH) available to
prevent cavitation.
The Next Step

In the next chapter, you will cover piping system development.
pE
Summary
• The differences between open and closed systems
• How air enters a hydronic system
• How minimum and maximum pressures are maintained and set in a closed
hydronic system
u
• The types of compression tanks and where they are located in a hydronic
system
• What solubility of air in a hydronic system is and what factors determine
its increase
ro
• The factors that are needed to size and select a compression tank
• Where the point of no pressure change is
References
ASHRAE. 1987. ASHRAE Handbook—HVAC Systems and Applications.
G
Atlanta: ASHRAE.
Atlanta: ASHRAE.
ASHRAE. 2013a. ASHRAE Handbook—Fundamentals. Atlanta: ASHRAE.
ASHRAE. 2013b. ASHRAE Pocket Guide for Air Conditioning, Heating, Venti-
lation, and Refrigeration. Atlanta: ASHRAE.

6-1 What maintains the maximum and minimum pressure limits of a hydronic system?
6-2 What must a closed water system have to permit the expansion and contraction
of the water volume?
t
6-3 What should be specified for terminal coil returns and high points in the piping
system to provide service for air in the hydronic system?
e
6-4 Where does air come from in a hydronic system?
6-5 What routine tasks should be performed by the building maintenance staff to
the hydronic systems prior to the heating and the cooling seasons?
gp
6-6 How much air can be present in water at 120°F and 30 psig?
6-7 A steel pipe system has 1000 gal total volume and will operate periodically in
the cooling season with 40°F water, but when it is off it will reach 100°F ambi-
ent; minimum pressure is 10 psig (24.7 psia) and maximum is 25 psig
(39.7 psia). What size diaphragm expansion tank is required?
6-8 A pump is selected for 100 gpm at 50 ft head, the system has 20 psig static pres-
sure when off, and the expansion tank is improperly located on the pump dis-
pE
charge. What will the gages read on the pump suction and discharge when the
pump is started? What will happen at the float-type air vents in the boiler room?
6-9 Explain what factors are needed to size and select an expansion tank.
6-10 What are the differences between open and closed systems?
6-11 What are the three types of expansion tanks?
6-12 Give three examples of where expansion tanks are used.
u
6-13 Why is it important to have the correct size expansion tank?
ro
G
Piping System
Development
e t
Study Objectives
gp
After studying the material in this chapter you should be able to
❑ know the factors to consider before starting a piping design,

❑ understand what piping equipment should be considered at a load coil,
❑ know which piping system design gives more uniform pressure drop and
why,
❑ understand how flow can be determined in a given loop,
pE
❑ know what type of pump curve should be selected if two-way valves are to
be used,
❑ know the two advantages of primary/secondary hydronic systems, and
❑ know how the possibility of freezing coils and piping in an HVAC system
can be reduced.
Instructions
u
Piping System Design

ro
To develop the piping system approach, the designer must consider several fac-
tors to arrive at a project solution. In short, there is no one way to satisfy all the
conditions; there are many ways. The purpose of this course is to help identify
possible design methods and to help select those options that best meet design
and energy efficiency requirements. Looking at a project, a designer must con-
G
sider some logical steps and questions to develop a design philosophy:
1. Know the building heat transfer load. In the initial approach, all aspects of
the load must be known. What is the total load? How is the load distributed
by time and location? How are controlled occupancy zones determined or
laid out? As the system designer, some of these calculations may be up to
you or may be determined by how the space to be conditioned is pro-
194 Chapter 7 Piping System Development
grammed. For example, in a simple office system, there may be a ring of

private offices surrounded by open floor space or vice versa. Each office
probably represents a separate point of temperature control for the occu-
pant, pointing the designer towards individual heating or cooling coils,
while the open areas present a gross overall area to condition. Often the
space occupancy program logically determines where the water system
loads (e.g., the terminal heat transfer coils) and the source (e.g., boilers and
chillers) are to be located.
t
There are several decisions to be made to determine the best design
e
approach depending on the building layout and type, whether it is a sin-
gle-story or multistory building and whether it is a single building or a
complex of buildings (such as a campus). Should there be a mechanical
room, or will the equipment be distributed throughout the structure?
gp
Where should the mechanical rooms be located? How will the piping and
ductwork be routed? All of these factors influence the design decisions
that must be made to specify equipment location. These design observa-
tions may also lead to modification of the space planning program, indi-
cating the importance of being involved in the initial project conceptual
phases, and also indicating the potential iterative nature of a design. Be
prepared for changes. Construct some of your initial design concepts with
pE
flexibility in mind, allowing for balance between first costs, design modi-
fication for the inevitable changes that will occur up to and during con-
struction, and changes to occupancy that will require building
modification.
u
ro
G
Figure 7-1 Typical building layout.

2. Determine the heating and cooling loads based upon occupancy, comfort
requirements, codes, and standards (see ANSI/ASHRAE/IES Standard 90.1
and 90.1 User’s Manual), and determine any special requirements for facili-
ties like computer rooms, laboratories, and clean rooms (see Figure 7-2).
3. Develop a concept for part-load control:
• Are large zones or individual rooms to be controlled?
• Are part-load heat transfer conditions to be obtained by varying water
t
flow, water temperature, or airflow and temperature?
• Will fan-coil terminals or large air-handling units (AHUs) be considered?
e
• Will the system satisfy the full-load conditions for all building zones or
must the designer consider some diversity factors?
• How is the source of cooling or heating to be operated at part-load
gp
conditions?
• Are control valves to be selected with adequate characteristics and
pressure drop to provide good coil control at various loads?
• Will the piping system design consider methods to measure, balance,
and adjust flows in each zone to ensure performance at various loads?
• Will backup be required? If so, which zones will be included?
pE
4. Develop the piping and pumping system concept (see Figure 7-3), such as
the following:
• Direct-return piping
• Reverse-return piping
• Primary/secondary piping
u
ro
G
Figure 7-2 Determine the loads and consult references.

e t
gp
pE
Figure 7-3 Develop piping/pumping system concept.
• Combinations of the above

• Constant-speed pumping
• Two-speed pumping
u
• Multiple pumps
• Primary/secondary pumps
• Variable-speed pumping
ro
• Distributed pumping
Consider modeling the system to determine the full- and part-load
flows, the pressure distribution required, and this effect on components.
5. Develop a first-cost analysis versus energy operating costs over the pro-
G
jected life of the system.

6. Determine the maintenance and operating requirements and if they match
with the personnel capabilities.
The piping system design may not address all of these issues, but the issues
need to be identified by the designer in the system layout and the project speci-
fications.
The flowchart in Figure 7-4 summarizes the key steps in the piping system
design process. In the section Introductory Concepts in Chapter 1, we intro-
duced the basic direct- and reverse-return piping concepts. In the section Basic
Considerations in Chapter 2, pressure drops and pipe sizing were discussed.
Direct-Return Analysis
t
Now, examine a model of a system requiring four AHUs with loads of 100 gpm
each at locations 100 ft apart in a square floor area configuration (see Figure 7-5).
e
At each load, the design is planned to control the load flow with two-way control
valves and to determine the effect that 10.7 and 5.3 ft drop coils have on the
piping design for direct-return and reverse-return layouts.
gp
u pE
ro
G
Figure 7-4 Piping system design flowchart.

e t
gp
Figure 7-5 System requiring four AHUs, each with 100 gpm load.
Each load coil should have a manual shutoff valve for servicing, a flow
pE
strainer on its supply, an automatic air vent at the coil return, a two-way control
valve (one size smaller than the branch pipe) on the leaving side of the coil, and
a manual balancing valve with test ports and shutoff (line size) after the control
valve (see Figure 7-6). In Figure 7-7, for a 100 gpm flow condition, 3 in. pipe
has a head loss over 2 ft/100 ft at 4+ fps velocity; a 2.5 in. pipe has 7 ft/100 ft
head loss at 7 fps, so a 3 in. pipe is a good starting point.
The general guideline for velocity is between 1 to 4 fps. Air separates from
the water below 1.5 fps, and above 4 to 5 fps noise might be noticed. Similarly,
u
200 gpm in 4 in. pipe is over 2 ft/100 ft (5 fps), 300 gpm in 4 in. pipe is about
5 ft/100 ft (7.5 fps), and 400 gpm in 5 in. pipe is nearly 3 ft/100 ft (6.5 fps).
These drops will be checked against the values in the Hydraulic Institute’s Engi-
neering Data Book (see Tables 7-1 to 7-3) (HI 1990).
ro
It is important to note that the selection of pipe size and the resulting veloc-
ity is based on the designer’s judgment and experience. For the example,
assume 4 in. pipe for main supply and return connections and 3 in. pipe for the
rest of the system. To assess each unit’s piping for pressure drop from the main
to the coil see Table 7-4.
G
Similarly, we check units 2, 3, and 4 in the model (Figure 7-5) and find
they look close, so 2 ft for supply and 9 ft for return will be assumed for this
example. Now look at the system pressure drops for a direct-return piping lay-
out (in Figure 7-8) using 10.7 ft drop coils and the control valve selection of
5.27 ft drop (assuming one size smaller than branch size). For direct return, see
Table 7-5.
The highest pressure drop path is unit 3, with a differential of 29.66 ft
required at A-F to provide 100 gpm flow.
e t
Figure 7-6 Coil connections.
gp
pE
Figure 7-7 Friction loss, schedule 40 steel pipe.
u
Because the other units in the model have lower pressure drops, this means
they will have a greater flow than design unless balanced. The unbalanced
ro
branches can be estimated by the use of the Darcy-Weisbach relationship,

where Q2 ~ h.
2 h
Q2 = Q 1 --------2-
h1
G
2 29.66 ft
Unit 1 = 100 ------------------- = 112.1
23.06 ft
2 29.66 ft
Unit 2 = 100 ------------------- = 103.2
27.84 ft
2 29.66 ft
Unit 4 = 100 ------------------- = 109.2
24.88 ft
Table 7-1 Schedule 40 Steel Pipe (3 in. Nominal Discharge)

ID = 3.066 in.
D = 0.000587
V, V2/2g, Pressure Drop,
ft3/s gpm
ft/s ft per 100 ft of pipe
ft
t
0.0111 5 0.217 0.000732 0.0112
0.0223 10 0.434 0.00293 0.0372
e
0.0334 15 0.651 0.0659 0.0762
0.0446 20 0.868 0.0117 0.126
gp
0.0557 25 1.085 0.0183 0.189
0.0668 30 1.3 0.0263 0.262
0.078 35 1.52 0.0359 0.347
0.0891 40 1.74 0.0468 0.443
0.1 45 1.95 0.0593 0.547
0.111 50 2.17 0.0732 0.662
pE
0.123 55 2.39 0.0885 0.789
0.134 60 2.6 0.105 0.924
0.145 65 2.82 0.124 1.07
0.156 70 3.04 0.143 1.22
0.167 75 3.25 0.165 1.39
0.178 80 3.47 0.187 1.57
u
0.189 85 3.69 0.211 1.76
0.201 90 3.91 0.237 1.96
ro
0.212 95 4.12 0.264 2.17

0.223 100 4.34 0.2927 2.39
0.245 110 4.77 0.354 2.86
0.267 120 5.21 0.421 3.37
0.29 130 5.64 0.495 3.92
G
0.312 140 6.08 0.574 4.51

0.334 150 6.51 0.659 5.14
0.356 160 6.94 0.749 5.81
0.379 170 7.38 0.846 6.53
0.401 180 7.81 0.948 7.28
Table 7-1 Schedule 40 Steel Pipe (3 in. Nominal Discharge) (Continued)

ID = 3.066 in.
/D = 0.000587
ft3/s gpm
ft
t
0.423 190 8.25 1.06 8.07
0.446 200 8.68 1.17 8.9
e
0.49 220 9.55 1.42 10.7
0.535 240 10.4 1.69 12.6
gp
0.579 260 11.3 1.98 14.7
0.624 280 12.2 2.29 16.9
0.668 300 13 2.63 19.2
0.713 320 13.9 3 22
0.758 340 14.8 3.38 24.8
0.802 360 15.6 3.79 27.7
pE
0.847 380 16.5 4.23 30.7
0.891 400 17.4 4.68 33.9
0.936 420 18.2 5.16 37.3
0.98 440 19.1 5.67 40.9
1.025 460 20 6.19 44.6
1.069 480 20.8 6.74 48.5
u
1.114 500 21.7 7.32 52.5
1.225 550 23.9 8.85 63.2
ro
1.337 600 26 10.5 74.8

1.448 650 26.2 12.4 87.5
1.56 700 30.4 14.3 101
1.671 750 32.5 16.5 116
1.782 800 34.7 18.7 131
G
1.894 850 36.9 21.1 148

2.005 900 39.1 23.7 165
2.117 950 41.2 26.4 184
2.228 1000 43.4 29.27 204

ID = 4.0286 in.
/D = 0.0004
ft3/s gpm
ft
t
0.0111 5 0.126 0.000247 0.0031
0.0223 10 0.252 0.000987 0.01017
e
0.0446 20 0.504 0.00395 0.0344
0.0668 30 0.756 0.00888 0.0702
gp
0.0891 40 1.01 0.0158 0.118
0.111 50 1.26 0.0247 0.176
0.134 60 1.51 0.0355 0.245
0.156 70 1.76 0.0484 0.325
0.178 80 2.02 0.0632 0.415
0.201 90 2.27 0.08 0.515
pE
0.223 100 2.52 0.0987 0.624
0.245 110 2.77 0.199 0.744
0.267 120 3.02 0.142 0.877
0.29 130 3.28 0.167 1.017
0.312 140 3.53 0.193 1.165
0.334 150 3.78 0.222 1.32
u
0.356 160 4.03 0.253 1.49
0.379 170 4.28 0.285 1.67
0.401 180 4.54 0.32 1.86
ro
0.423 190 4.79 0.356 2.06

0.446 200 5.04 0.395 2.27
0.49 220 5.54 0.478 2.72
0.535 240 6.05 0.569 3.21
G
0.579 260 6.55 0.667 3.74

0.624 280 7.06 0.774 4.3
0.668 300 7.56 0.888 4.89
0.713 320 8.06 1.01 5.51
0.758 340 8.57 1.14 6.19

ID = 4.0286 in.
/D = 0.0004
ft3/s gpm
ft
t
0.802 360 9.07 1.28 6.92
0.847 380 9.58 1.43 7.68
e
0.891 400 10.1 1.58 8.47
0.936 420 10.6 1.74 9.3
gp
0.98 440 11.1 1.91 10.2
1.025 460 11.6 2.09 11.1
1.069 480 12.1 2.27 12
1.114 500 12.6 2.47 13
1.225 550 13.9 2.99 15.7
1.337 600 15.1 3.55 18.6
pE
1.448 650 16.4 4.17 21.7
1.56 700 17.6 4.84 25
1.671 750 18.9 5.55 28.6
1.782 800 20.2 6.32 32.4
1.894 850 21.4 7.13 36.5
2.005 900 22.7 8 40.8
u
2.117 950 23.9 8.91 45.3
2.228 1000 25.2 9.87 50.2
2.451 1100 27.7 11.9 60.5
ro
2.674 1200 30.2 14.2 72

2.896 1300 32.8 16.7 84.3
3.119 1400 35.3 19.3 97.6
3.342 1500 37.8 22.2 112
G
3.565 1600 40.3 25.3 127

3.788 1700 42.8 28.5 143
4.01 1800 45.4 32 160
4.233 1900 47.9 35.6 178
4.456 2000 50.4 39.5 196

ID = 5.047 in.
/D = 0.000357
ft3/s gpm
ft
t
0.0111 5 0.0802 0.0000999 0.00107
0.0223 10 0.16 0.0004 0.00348
e
0.0446 20 0.321 0.0016 0.0116
0.0668 30 0.481 0.0036 0.0237
0.0891 40 0.641 0.00639 0.0395
gp
0.111 50 0.802 0.00999 0.0587
0.134 60 0.962 0.0144 0.0814
0.156 70 1.12 0.0196 0.1076
0.178 80 1.28 0.0256 0.137
0.201 90 1.44 0.0324 0.169
pE
0.223 100 1.6 0.04 0.204
0.267 120 1.92 0.0576 0.286
0.312 140 2.25 0.0783 0.38
0.356 160 2.57 0.102 0.487
0.401 180 2.89 0.129 0.606
0.446 200 3.21 0.16 0.736
u
0.49 220 3.53 0.193 0.879
0.535 240 3.85 0.23 1.035
0.579 260 4.17 0.27 1.2
ro
0.624 280 4.49 0.313 1.38

0.668 300 4.81 0.36 1.58
0.713 320 5.13 0.409 1.78
0.758 340 5.45 0.462 2
0.802 360 5.77 0.518 2.22
G
0.847 380 6.09 0.577 2.46

0.891 400 6.41 0.639 2.72
0.936 420 6.74 0.705 2.98
0.98 440 7.06 0.774 3.26

ID = 5.047 in.
/D = 0.000357
ft3/s gpm
ft
t
1.02 460 7.38 0.846 3.55
1.07 480 7.7 0.921 3.85
e
1.11 500 8.02 0.999 4.16
1.23 550 8.82 1.21 4.98
1.34 600 9.62 1.44 5.88
gp
1.45 650 10.4 1.69 6.87
1.56 700 11.2 1.96 7.93
1.67 750 12 2.25 9.05
1.78 800 12.8 2.56 10.22
1.89 850 13.6 2.89 11.5
pE
2.01 900 14.4 3.24 12.9
2.12 950 15.2 3.61 14.3
2.23 1000 16 4 15.8
2.45 1100 17.6 4.84 19
2.67 1200 19.2 5.76 22.5
2.9 1300 20.8 6.75 26.3
u
3.12 1400 22.5 7.83 30.4
3.34 1500 24.1 8.99 34.8
3.56 1600 25.7 10.2 39.5
ro
3.79 1700 27.3 11.6 44.5

4.01 1800 38.8 12.9 49.7
4.23 1900 30.5 14.4 55.2
4.46 2000 32.1 16 61
4.68 2100 33.7 17.6 67.1
G
4.9 2200 35.3 19.3 73.5

5.12 2300 36.9 21.1 80.1
5.35 2400 38.5 23 87
5.57 2500 40.1 25 94.2
Table 7-4 Unit 1 Supply and Return Sides

Length, ft Unit 1: Supply Side (100 gpm) Pressure Drop, ft
1 in. 4 in. tee branch 0.4
1 in. 4 × 3 in. bushing 0.1
1 in. 3 in. butterfly valve 0.44
t
3 in. 3 in. ells @ 0.1 0.3
1 in. 3 in. strainer 0.35
e
25 in. 3 in. pipe (2.39 ft/100 ft) 0.5975
Supply branch 2.1875
(report as 2.0 ft)
gp
Length, ft Unit 1: Return Side (100 gpm) Pressure Drop, ft
1 in. 4 × 3 in. tee branch 0.4
1 in. 4 × 3 in. bushing 0.02
3 in. 3 in. ells @ 0.1 0.3
1 in. 2½ in. control valve 5.27
pE
2 in. 3 × 2½ in. bushing at 0.02 0.04
1 in. 3 in. balancing and service valve (open) 1.9
25 in. 3 in. pipe (2.39 ft/100 ft) 0.5975
Return branch 8.5275
(report as 9 ft)
u
In summary, for direct-return piping and 10.7 ft coils:
Unit 1 = 112 gpm
Unit 2 = 103 gpm
ro
Unit 3 = 100 gpm

Unit 4 = 109 gpm
Total = 424 gpm
This example shows the effect of the uneven flow in a direct-return piping
example without balancing.
G
If a lower coil pressure drop is selected in the example (by changing the
coil design), say 5.3 ft, the unbalance gets worse. The calculations are per-
formed as seen in Table 7-6.
As you can see, unit 3 is still the highest pressure drop path, with 24.26 ft
required at A-F to provide it with a design flow of 100 gpm.
Again applying the Darcy-Weisbach relationship Q2, the overflows can be
approximated:
e t
Figure 7-8 Direct-return piping layout.
gp
pE
2 h
Q2 = Q 1 --------2-
h1
2 24.26 ft
Unit 1 = 100 ------------------- = 117.2
17.66 ft
u
2 24.26 ft
Unit 2 = 100 ------------------- = 104
22.44 ft
ro
2 24.26 ft
Unit 4 = 100 ------------------- = 111.6
19.48 ft
In summary, for direct-return piping and 5.3 ft coils:
Unit 1 = 117.2 gpm
G
Unit 2 = 104.0 gpm

Unit 3 = 100.0 gpm
Unit 4 = 111.6 gpm
Total = 432.8 gpm
The summary shows a greater flow unbalance in the direct-return piping
due to the lower coil drop.
Table 7-5 Direct Return for Units 1, 2, 3, and 4

Length, ft Unit 1 (Path A-B-B'-F): Supply Main Pressure Drop, ft
30 Supply main: 4 in. pipe, 200 gpm = 2.27 ft/100 ft 0.681
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
t
30 Return main: 4 in. pipe, 200 gpm = 2.27 ft/100 ft 0.681
e
Total 23.06
Length, ft Unit 2 (Path A-B-C-C'-B'-F) Pressure Drop, ft
gp
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
pE
Total 27.84
Length, ft Unit 3 (Path A-E-D-D'-E'-F) Pressure Drop, ft
Supply branch 2
Coil (100 gpm) 10.7
u
Return branch 9
ro

Total 29.66
Length, ft Unit 4 (Path A-E-E'-F) Pressure Drop, ft
Supply branch 2
G
Coil (100 gpm) 10.7

Return branch 9
Total 24.88
Table 7-6 Lower Coil Pressure Drop Selected for Units 1, 2, 3, and 4
Supply branch 2
Coil (100 gpm) 5.3
Return branch 9
t
e
Total 17.66
gp
Supply branch 2
Coil (100 gpm) 5.3
Return branch 9
pE
Total 22.44
Supply branch 2
Coil (100 gpm) 5.3
u
Return branch 9
ro

Total 24.26
Supply branch 2
G
Coil (100 gpm) 5.3

Return branch 9
Total 19.48
Reverse-Return Analysis
Let’s look at a reverse-return piping arrangement for the same example, per
Figure 7-9. We will assume the supply and return branches are also 2.0 ft and 9.0 ft.
As you can see in Table 7-7, units 2 and 3 are the highest pressure-drop paths,
with 36.47 ft required at A-F to provide them with a design flow of 100 gpm
each. Again applying the Darcy-Weisbach relationship Q2 ~ h , we can esti-
mate the overflow in the other coils:
e t
2 h
Q2 =Q 1 --------2-
h1
In summary, for reverse-return piping and 10.7 ft coils:
gp
2 36.47 ft
Unit 1 = 100 -------------------- = 103.6
34.0 ft
Unit 2 = 100
Unit 3 = 100
2 36.47 ft
Unit 4 = 100 -------------------- = 103.6
pE
34.0 ft
u
ro
G
Figure 7-9 Reverse-return piping layout.

Table 7-7 Reverse-Return Piping Arrangement for Units 1, 2, 3, and 4

30 Supply main: 5 in. pipe, 400 gpm at 2.72 ft/100 ft 0.82
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
100 Return main: 3 in. pipe, 100 gpm at 2.39 ft/100 ft 2.39
t
e
Total 34.0
Length, ft Unit 2 (Path A-B-C-C'-D'-E'-F) Pressure Drop, ft
gp
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
pE
Total 36.5
Length, ft Unit 3 (Path A-B-C-D-D'-E'-F) Pressure Drop, ft
u
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
ro

Total 36.5
Length, ft Unit 4 (Path A-B-C-D-E-E'-F) Pressure Drop, ft
G
Supply branch 2
Coil (100 gpm) 10.7
Return branch 9
70 Return main: 5 in. pipe, 400 gpm at 2.72 ft/100 ft Total 34.0
Total = 407.2 gpm
Note that the reverse-return lowered the total flow from 424 gpm to
407 gpm, and each branch’s pressure drop is more balanced.
If the lower pressure drop coil of 5.3 ft is substituted, the results are as
shown in Table 7-8.
Estimating the coil flows for the reverse-return piping and 5.3 ft coils, we
see a reduced flow as compared to direct-return piping:
t
2 31.1 ft
Unit 1 = 100 ---------------- = 104.3
e
28.6 ft
Unit 2 = 100
Unit 3 = 100
gp
2 31.1 ft
Unit 4 = 100 ---------------- = 104.3
28.6 ft
Total = 408.6 gpm
Summary
pE
By going through these examples for direct-return and reverse-return piping
system designs, we can see how the use of the reverse-return design lowers the
overall flow requirement and the magnitude of imbalance in the circuits. Fur-
thermore, we can see that coil flow selection (lower flow rating) increases this
flow imbalance for the direct-return system but has only a small effect for the
reverse-return system.
Summary of Pumping Horsepower—

u
Direct Return Versus Reverse Return
Using the pump horsepower equation discussed in Chapter 4 (Equation 4-3), we
can make a judgment on the relative power used in this example and Table 7-9:
ro
Flow (gpm) Head (ft)

WHP = --------------------------------------------------------, hp
3960
Table 7-9 shows the theoretical power at 100% flow and does not take into
account the pump or motor efficiencies, which would increase the values. Direct
G
return uses less horsepower than reverse return but is more unbalanced at 100%
flow. A further comparison might be made at part-load conditions to determine
operating hours versus percent of flow based on local weather patterns to get a
closer look at pumping horsepower. Further, decreasing pipe diameter selection
to increase flow velocities could be examined, as well as pressure drop selected
for the control valves. A basic comparison is to look at the piping lengths of the
direct-return versus reverse-return piping in this example:
Table 7-8 Lower Pressure Drop Coil for Units 1, 2, 3, and 4

30 ft Supply main: 5 in. pipe 400 gpm at 2.72 ft/100 0.82 ft
Supply branch 2 ft
Coil (100 gpm) 5.3 ft
Return branch 9 ft
100 ft Return main: 3 in. pipe 100 gpm at 2.39 ft/100 2.39 ft
t
e
Total 28.6 ft
Length, ft Unit 2 (Path A-B-C-C'-D'-E'-F) Pressure Drop, ft
gp
Supply branch 2 ft
Return branch 9 ft
pE
Total 31.1 ft
Length, ft Unit 3 (Path A-B-C-D-D'-E'-F): Pressure Drop, ft
Supply branch 2 ft
u
Return branch 9 ft
ro

Total 31.1 ft
Length, ft Unit 4 (Path A-B-C-D-E-E'-F): Pressure Drop, ft
G
Supply branch 2 ft
Return branch 9 ft
Total 28.6 ft
Direct Return:
Supply main (4 in.) = 30 + 70 = 100 ft
Return main (4 in.) = 30 + 70 = 100 ft
Supply main (3 in.) = 100 + 100 = 200 ft
Return main (3 in.) = 100 + 100 = 200 ft
Unit branches (3 in.) = 8 × 25 = 200 ft
t
Total = 800 ft
Reverse Return:
e
Return mains
Main (5 in.) = 30 + 70 = 100 ft
gp
Main (4 in.) = 100 + 100 + 100 + 100 = 400 ft
Main (3 in.) = 100 + 100 = 200 ft
Unit branches (3 in.) = 8 × 25 = 200 ft
Total = 900 ft
See Table 7-10 for a summary.
pE
Table 7-9 Direct and Reverse Return with Balanced and Unbalanced Flow
Direct Return (Unbalanced Flow)
424 gpm and 30 ft = 3.2 hp
10.7 ft coils
433 gpm and 24 ft = 2.6 hp
Direct Return (Balanced Flow)
400 gpm and 30 ft = 3.0 hp
u
5.3 ft coils
400 gpm and 24 ft = 2.4 hp
Reverse Return (Unbalanced Flow)
407 gpm and 36 ft = 3.7 hp
ro
10.7 ft coils
409 gpm and 31 ft = 3.2 hp
Reverse Return (Balanced Flow)
400 gpm and 36 ft = 3.6 hp
5.3 ft coils
400 gpm and 31 ft = 3.1 hp
G
Table 7-10 Summary of Direct and Reverse Return

5 in. 4 in. 3 in. Total
Direct — 200 ft 600 ft 800 ft
Reverse 100 ft 400 ft 400 ft 900 ft
As you can see, the analysis of the piping layout takes time, but it is valu-
able to determine the design options available, amount of piping needed, and
how to reduce pumping horsepower.
There are many design trade-offs to be considered. The direct-return system
requires less piping and theoretically less horsepower than the reverse-return sys-
tem. However, the reverse-return system balances the system flow better, which
translates into better efficiency and performance in load distribution. If the imbal-
ance in the system is left unchecked, uneven distribution of flow to the loads may
t
result, producing poor performance. In the next section, we explore alternative
designs that eliminate some of these performance shortcomings.
e
Primary/Secondary Analysis
gp
Another consideration in the piping design strategy is to select a secondary
pump for each unit coil and move the two-way valve to the connecting bridge
return (see Figure 7-10). One advantage of this concept is to reduce the burden
of the pressure drop of the unit coil from the distribution pump and allow con-
stant flow in the unit coil to improve coil heat transfer and response to load.
Adjusting the coil flow permits flexibility of higher coil ts as compared to the
primary system.
Reviewing the previous example, modified for primary/secondary design
pE
but limiting the study to the 10.7 ft drop coil and a direct-return piping system
(see Figure 7-11), we again assess each unit’s piping for pressure drop from the
main to the secondary bridge (see Table 7-11).
Unit 3 is the highest pressure drop path, with 19.86 ft required at A-F to
provide a design flow of 100 gpm.
Again applying the Darcy-Weisbach relationship, Q2 ~ h:
u
ro
G
Figure 7-10 Two-way valve in connecting bridge return with secondary pump.
e t
gp
pE
Figure 7-11 Primary/secondary pumping.
2 h
Q2 = Q 1 --------2-
h1
u
2 19.86 ft
Unit 1 = 100 ------------------- = 122.4
13.26 ft
ro
2 19.86 ft
Unit 2 = 100 ------------------- = 104.9
18.04 ft
2 24.26 ft
Unit 4 = 100 ------------------- = 114.8
15.08 ft
G
In summary:
Unit 1 = 122.4 gpm
Unit 2 = 104.9 gpm
Unit 3 = 100.0 gpm
Unit 4 = 114.8 gpm
Total = 442.1 gpm
Again using the water horsepower equation, the primary loop power used
at 442 gpm and 19.86 ft is 2.2 hp, which is less than the similar unbalanced
direct-return case above at 3.2 hp. When the system is balanced, the primary
loop uses 2 hp of energy, versus 3 hp for the direct return.
Table 7-11 Pressure Drop from Main to Secondary Bridge for Units 1, 2, 3, and 4
Length, ft Unit 1: Supply Side (100 gpm) Pressure Drop, ft
1 4 in. tee branch 0.4
t
1 4 × 3 in. bushing 0.1
e
1 3 in. butterfly valve 0.44
3 3 in. ells at 0.1 0.3
1 3 in. strainer 0.35
1 3 in. tee-thru 0.22
gp
25 3 in. pipe (2.39 ft/100 ft) 0.60
Supply branch Total 2.4
Length, ft Unit 1: Return Side (100 gpm) Pressure Drop, ft
1 3 in. tee-thru 0.22
3 3 in. ells at 0.1 0.3
1 2 1/2 in. control valve 5.27
pE
2 3 × 2 1/2 in. bushing at 0.41 0.82
1 3 in. balancing and service valve (open) 1.9
1 4 × 3 in. bushing 0.2
1 4 in. tee branch 0.4
25 3 in. pipe (2.39 ft/100 ft) 0.60
Return branch Total 9.49
u
Supply branch 2.4
Common pipe 0
Return branch 9.5
ro

Total 13.26
G
Supply branch 2.4

Common pipe 0
Return branch 9.5
Total 18.04
Table 7-11 Pressure Drop from Main to Secondary Bridge for Units 1, 2, 3, and 4 (Continued)
Supply branch 2.4
Common pipe 0
t
Return branch 9.5
e
Total 19.86
gp
Supply branch 2.4
Common pipe 0
Return branch 9.5
Total 15.08
pE
Table 7-12 Pressure Drop in Typical Secondary Loop
Length, ft Pressure Drop, ft
4 3 in. ells at 0.1 0.4
1 Coil 10.7
50 3 in. pipe 1.2
Total 12.3
u
The pressure drop in a typical secondary loop is shown in Table 7-12.
At 100 gpm coil flow, 0.31 hp pumping power is required to handle the sec-
ondary loop and coil. Comparing total horsepower, primary/secondary pumping
ro
would use 2 + 4(0.31) = 3.24 hp, as compared to 3 hp for the direct return. Why
use primary/secondary then in this type of application? Primarily for heat trans-
fer. Keeping full flow on the coil at all times maximizes heat transfer and main-
tains design differential temperature on the coil. For CHW applications, the coil
always has full latent energy capability for the maintained inlet temperature con-
dition. In other examples, primary/secondary pumping would be used to manage
G
system differential pressure, organize supply water distribution to systems, and

provide for flexibility with respect to future system additions.
Types of Pumps and Valves

Pump performance must be considered not only at the design point but across
its entire characteristic curve. Centrifugal pumps are available with steep
curves that drop from high head at low flow to low head at high flow versus
those with flat curves that show a small change in head between shutoff to
design flow (see Figure 7-12). Some designers like to limit this to a 15% to
25% rise-to-shutoff curve. These flat-curve pumps are always recommended
where two-way valves are applied to unit terminals. At part loads, the valves
will be operating at lower flows and this will move the system operating differ-
ential pressure up the pump curve.
Another factor is that when the pressure drop through the terminals is low
and the system balance is less than ideal, there may be a tendency to short-circuit
t
the flow in the units closest to the pump. With a steep-curve pump, this results in
a drop in head and less flow is available to the units farthest from the pump. A
e
flat-curve pump will show a minimum drop in head and give better flow to the
remote units.
Effects of Control Valves
gp
Factors affecting control valve performance in a typical load’s control loop are
as follow:
• Valve size. It is most important that a valve be sized for the required load
flow in gallons per minute and adequate pressure drop. Some size a valve
for one size less than coil inlet/outlet size, but this is not accurate enough.
u pE
ro
G
Figure 7-12 Flat versus steep pump characteristics.

Higher pressure drops yield higher valve authority. Ideally, the designer
calculates the coil characteristic and the valve characteristic to understand
the system control gain. The closer to linear the better for proportionally
throttled modulating control. The designer can also consider alternative
control strategies, such as using open/close control valves that are propor-
tionally controlled through the open and close time.
• Control valve actuator. An adequate control valve actuator with sufficient
power to hold the valve’s commanded position at maximum pressure drop
t
should be selected and specified.
e
• Valve characteristic. The valve characteristic selected for throttled
hydronic heating or cooling units should be an equal percentage character-
istic to give a linear output of Btu emission in relation to valve stroke.
• Valve authority. Valve authority is determined by the valve pressure drop
gp
at full load compared to the load coil and piping pressure drop. The pres-
sure drop selected should be at least 30% to 50% of the controlled loop
pressure drop (which may be the system head or some other controlled
pressure drop). In the direct-return example, a valve drop of 5.27 ft was
selected as compared to the loop drop of 29.66 ft, which results in 5.27/
29.7 = 17.7%. When the coil drop of 5.3 ft was selected, the authority
increased to 5.27/24.26 = 21.7%; an authority of 25% would require a drop
pE
of about 9 ft for the 29.7 ft loop pressure drop.
• Valve body rangeability. The valve body rangeability should be 30:1 or
more. That is the ratio of its maximum controlled flow; say, 100 gpm to a
minimum flow of at least 3 gpm (100/3 = 30:1). The designer may not be
given the valve body rangeability, with the manufacturer opting to formulate
a combined value of valve body and actuator. In these cases, a minimum
rangeability of 75:1 should be considered.
• Valve body style. In general, a two-way valve is recommended because it
u
modulates the volume of flow in relation to the load. Use of three-way
valves on terminals is not recommended because the flow is bypassed and
does not reduce the pumping power at low loads. Traditionally, globe
valves are used for proportionally throttled control systems, especially in
ro
hot-water systems where cavitation can become a concern. Segmented ball

valves are also an option for throttled control, although with all ball valve
installations, cavitation and pipe/valve size correction factors should be
carefully examined when specifying the valve.
• Balancing valve. The balancing valve is a multifunction valve that is used to
measure flow, stop flow for servicing, and proportionally balance the flow in
G
circuits that have excessive flow. This valve should be selected for the design
flow when wide open and a minimum but measurable pressure drop. When
applying variable-speed pumping, automatic flow-limiting valves are gener-
ally a better choice instead of manual balancing valves. This is due to the
changing system differential pressures as the pump differential pressure con-
troller changes speed in relation to the controlled pressure of the system. Care
should be taken in the selection of these devices for design flow at a minimum
pressure drop. Also, a shutoff valve is required in the return for servicing, and
a venturi or orifice may be required to determine the flow in the field.
Primary/Secondary Application Study

The use of a secondary pump to reduce the pump head of the main distribution
pump is always an option when designing a hydronic system. Each system
t
should be evaluated on its own requirements and the resulting pumping energy
reductions that may be obtained.
e
Figure 7-13 shows a four-zone heating system using a single 7.5 hp pump
delivering 230 gpm at 60 ft head with a 20°F t.
The zones are as follows:
gp
• One-pipe heating zone: 30 gpm at 12 ft
• Two-pipe reverse-return heating zone: 60 gpm at 20 ft
• Two-pipe direct-return heating zone: 100 gpm at 40 ft
• Heat exchanger zone: 40 gpm at 15 ft
The distribution main has a total friction loss of 30 ft from the distribution
pump through the mains and returning through the boiler to the pump suction.
pE
The pump has to deliver the total flow of 230 gpm and must overcome the
worst-case pressure drop of 60 ft. The zone 3 head loss calculations are shown
in Table 7-13.
u
ro
G
Figure 7-13 Four-zone heating system.

A redesign of the system to a primary/secondary concept is shown in Fig-

ure 7-14. Each zone is connected to a primary/secondary bridge and a second-
ary pump is selected for each zone’s load.
Because the primary/secondary bridge has a minimal pressure drop, the dis-
tribution pump is required to flow 230 gpm at 30 ft instead of 60 ft, so a 3 hp
motor can be selected, as shown in Table 7-14. Note that the flow and head are
large enough that a larger pump is selected and that there is a duty point opera-
t
tion of 2.3 hp. Each zone is pumped by a circulating pump that is significantly
smaller, in both size and cost but also flow and head. Small circulating pumps do
e
not come with a published duty point of operation, as the motors are generally
small, fractional horsepower and designed for non-overloading operation for any
potential operation of the pump. This means the original 7.5 hp pump is replaced
by five pumps totaling about 6 hp, which gives a reduction of 12% in connected
gp
pumping energy. If the system operates properly, the primary/secondary system
Table 7-13 Four-Zone Heating System

Zone, ft Mains, ft Total Head Required
Zone 1 12 + 0 = = 12 ft
Zone 2 20 + 5+5 = = 30 ft
pE
Zone 3 40 + 10 + 10 = = 60 ft
Zone 4 15 + 15 + 15 = = 45 ft
u
ro
G
Figure 7-14 Primary/secondary pumping, four-zone heating system.

will probably operate with slightly less energy than the 4.75 hp of a primary-only
pump, even though the duty point of the primary pump and the nominal horse-
power of the secondary pumps total 5.25 hp. This example assumes the load
flows are based on a 20°F drop; further study of the primary/secondary bridge
flows shows an ability to have different ts as follows for a typical secondary
bridge (see Figure 7-15):
t
Table 7-14 3 hp Motor Selected
e
Required Power, hp Operating Power, hp Motor Power, hp
Zone 1 30 gpm at 12 ft 0.09 0.3 0.3
Zone 2 60 gpm at 20 ft 0.30 0.75 0.75
gp
Zone 3 100 gpm at 40 ft 1.01 1.5 1.5
Zone 4 40 gpm at 15 ft 0.15 0.4 0.4
Secondary 230 gpm 1.56 2.95 2.95
Primary 230 gpm at 30 ft 1.74 2.3 3
Total 3.30 5.25 5.95
pE
Primary only 230 gpm at 60 ft 3.48 4.73 7.5
u
ro
G
Figure 7-15 Primary/secondary bridge energy.

Energy In = Energy Out

Primary Energy = Secondary Energy
500 GPM t = 500 GPM t
Here is the benefit of a primary/secondary system from an energy perspec-
tive: If we design for a greater drop of 40°F in the primary and maintain a 20°F
drop for the design flow in the secondary loops (as shown in Figure 7-15), we
t
see a 50% flow reduction in the primary supply required. The 50% reduction in
the supply flow means one-fourth of the original main pressure drop, so the
e
primary distribution pump size and horsepower are reduced to 115 gpm at
7.5 ft, which means the supply pump can be a 0.5 hp inline circulator. Now our
summary becomes the layout and data shown in Figure 7-15 and Table 7-15.
gp
The original design requiring a single 7.5 hp has been reduced to a total of
3.5 hp using the primary/secondary design principles per Figure 7-16. A fur-
ther review of pipe sizing may provide additional economies. Another review
may yield further flexibilities in the zoning capabilities to provide diversity
and energy economy by treating the secondary circuit as a controlled supply
temperature and t when used with a two-way valve in the bridge return.
Note: By increasing t in the primary piping circuits, a 50% reduction in sup-
ply flow was achieved. This allowed significant energy savings by reducing
pE
the total pump horsepower requirements.
Primary/secondary pumping techniques not only provide this powerful
design technique but also have other important design abilities. Implicit in the
discussion is that because of the common pipe, which is shared between the
two systems (primary and secondary), the two systems are hydraulically iso-
lated from each other. This allows for system pressures to be managed vis-à-vis
the design process. Managing pressures is ultimately what system balance is all
about—getting the design flow where it is needed when it is needed. It also
u
allows for changes to be made to parts of the system without necessarily affect-
Table 7-15 Primary/Secondary Bridge Energy

ro
Required Power, hp Operating Power, hp Motor Power, hp

Zone 1 30 gpm at 12 ft 0.09 0.3 0.3
Zone 2 60 gpm at 20 ft 0.30 0.75 0.75
Zone 3 100 gpm at 40 ft 1.01 1.5 1.5
G
Zone 4 40 gpm at 15 ft 0.15 0.4 0.4

Secondary 230 gpm 1.56 2.95 2.95
Primary 115 gpm at 8 ft 0.22 0.5 0.5
Total 1.77 3.45 3.45
Primary only 230 gpm at 60 ft 3.48 4.73 7.5
e t
Figure 7-16
gp
Primary/secondary pumping, four-zone heating system.
pE
ing the other connected systems. For example, a chiller plant is designed with
primary/secondary techniques and it is decided to add more loads to the sys-
tem, requiring another chiller on the primary side. Primary/secondary design
allows for the primary plant to be modified without having an impact on the
secondary side. In a primary-only system, chances are there would have to be a
complete modification of the pumping system.
The counter to this, though, is that there is disparate control of zone loads/
u
flow and the affect it has on the pump. If one zone is closed or closes while oth-
ers are open to flow, there can be a large flow interaction affecting all attached
zone circuits. When zones are laid out so that all zone valves are on similar ther-
mally reacting zones, acting in the same way opening and closing, flow control
ro
tends to stay proportional without large swings in flow. This specifically was one
of the early problems of variable-speed pumping being applied to primary/sec-
ondary CHW plants. Running a variable-speed, variable-flow secondary pump-
ing system with a constant-speed primary pump allowed for poor system t (and
associated plant efficiency) to be established at some operating conditions. At
others it created chiller sequencing problems. In large part, this was due to chill-
G
ers that required constant flow even at part load due to the inherent operating
characteristics of the chiller control. As chiller controls evolved, they too allowed
for flow to vary through them; however, designers failed to follow along with
variable-primary/variable-secondary systems to accommodate maximizing t
and system efficiency. As a design technique, primary/secondary is still quite
applicable, even in the world of variable-speed pumping, and brings the benefit
of being able to break up a system into smaller manageable hydraulic zones.
Antifreeze Solutions for Low-Temperature Applications

Another topic that relates to piping system development is how to treat
low-temperature applications. In most cases, protection is required in the form
of circulating an antifreeze solution that is capable of preventing bursting of
coils, fittings, and piping. Figure 7-17 shows a dedicated heat exchanger and
pump for a glycol subsystem complete with an expansion tank and detachable
t
water makeup (or other water makeup devices to meet local code require-
ments). Other approaches assume constant water flow in a coil by treating it as
e
a secondary pumping circuit with constant flow to provide 4 to 6 ft/s tube
velocities.
On the air side, face bypass dampers are required to provide temperature
gp
control to blend warm air from the coil side with bypass air to maintain a
controlled discharge temperature (see Figure 7-18). In addition, a separate
low-temperature thermostat with a long sensing element (within which the
coldest 1 ft length controls) set at 40°F is recommended on the coil air dis-
charge surface. An alternative method is a bulb inserted in the leaving water
to provide an alarm to the operating personnel before the discharge or water
temperature reaches freezing conditions. Sometimes this is wired to the fan
pE
circuit to shut down the fan and close outside dampers to reduce the possibil-
ity of damage.
u
ro
G
Figure 7-17 Coil with glycol heat exchanger and pump for low temperatures.
When applying water/glycol solutions, it is important to recognize the

reduced heat carrying capacity due to the lower specific heat and higher den-
sity of the mixture. Previously, in Chapters 1 and 2, we discussed that the heat
transferred to or from water qw (Btu/h) is expressed as
q w = mc p t (7-1)
where
t
cp = specific heat, 1.0 Btu/lb·°F
e
m = mass flow of water, lb/h
t = temperature change across unit, °F
If water flow Qw is expressed in gpm,
gp
q w = 8.02 w c p Qw t
where w is density of water, 62.4 lb/ft3.

For water/glycol mixtures, the specific heat cp and density w of the mix-
ture have to be used. For example, a 35% propylene water/glycol solution at
180°F with 20°F drop will transfer as follows:
qw = 8.02 wcpQw t
u pE
ro
G
Figure 7-18 Pumped coil with face bypass dampers for low-temperature primary/secondary
pumping.
= 8.02(62.2)(0.95)(1)(20)
= 9478 Btu/h
In other words, at 20°F t, the mixture will transfer 9478 Btu/h per 1 gpm,,
as compared to 10,000 Btu/h per gpm for water.
Chapter 31 in the 2013 ASHRAE Handbook—Fundamentals shows complete
reference tables of freezing points and graphs of specific heat versus percentage
t
water mixture and temperature (see Figures 7-19 and 7-20) (ASHRAE 2013b).
In addition to specific heat, the Handbook displays density, viscosity, and ther-
e
mal conductivity for ethylene/glycol and propylene/glycol water mixtures.
The chapter also states that for winterizing coils in HVAC systems, a 30%
ethylene/glycol or 35% propylene/glycol mixture with water can be used. It
states that as the fluid freezes, it forms a slush that expands and flows to any
gp
available space. Therefore, expansion volume must be included with this type of
protection. It also recommends that if the application requires the fluid to remain
entirely liquid, a concentration with a freezing point 5°F below the lowest
expected temperature should be chosen. Further information regarding corrosion
inhibition is also provided (ASHRAE 2013b). Without inhibitors, glycols oxi-
dize into acidic end products.
u pE
ro
G
Figure 7-19 Specific heats of aqueous solutions of industrially inhibited ethylene glycol (per-
cent by volume).
Provision should be made for testing and filling glycol and glycol inhibitor
into the piping system. Figure 7-21 is one design of a one-shot chemical feeder
that should be considered.
Pumping Design Factors

By completing this chapter, you should have learned the following general
design concepts:
t
• The higher the head losses through the terminal units of a hydronic system,
e
relative to the main piping losses, the closer the system comes to a natural
balance.
• Reverse-return systems are closer to a natural balance of flows than
gp
direct-return systems.
• If automatic control valves are employed, the design pressure drop selected
should be as high as practical. A pressure drop at least equal to the drop in
the terminal unit coil is a desirable goal. The valve should be sized for the
u pE
ro
G
Figure 7-20 Specific heats of aqueous solutions of industrially inhibited propylene glycol (per-
cent by volume).
e t
gp
pE
Figure 7-21 One-shot chemical feeder.
design flow with the Cv flow formula (Equation 5-1), which may not be the
same size as the coil inlet piping.
• Centrifugal pumps with flat characteristics should be selected for systems
with control valves.
• Two-way valves should be considered over three-way valves because
u
they vary the volume of water flow in direct relationship with the con-
trol signal. Three-way valves provide a continuous flow regardless of
the load and are not suitable with variable-volume pumping systems.
ro
• Manual balancing valves should be chosen for a minimal pressure drop

and provide the means to measure flows in various loops in the field as
well as provide a shutoff valve for coil servicing.
• Performance is best ensured by requiring proportional balancing after
the system is operating. Variable-volume pumping systems should be
checked and adjusted for balance at 50%, 75%, and 100% design flows.
G
The Next Step

In Chapter 8, you will learn how to match pumps to systems.
Summary
• The factors that must be considered before starting a piping design

• What piping equipment should be considered at a load coil
• Which piping system design gives more uniform pressure drop and why
• How flow can be determined in a given loop
• What type of pump curve should be selected if two-way valves are to be
used
• Two advantages of primary/secondary hydronic systems
t
• How the possibility of freezing coils and piping in an HVAC system can be
reduced
e
gp
Atlanta: ASHRAE.
ASHRAE. 2013a. ANSI/ASHRAE/IES Standard 90.1-2013, Energy Standard
ASHRAE. 2013b. ASHRAE Handbook—Fundamentals. Atlanta: ASHRAE.
ASHRAE. 2014. 90.1-2013 User’s Manual. Atlanta: ASHRAE.
Beaty, F. 1987. Sourcebook of HVAC Details. New York: McGraw-Hill Book Co.
pE
HI. 1990. Engineering Data Book, 2d ed. Parsippany, NJ: Hydraulic Institute.
u
ro
G

7-1 In the direct-return example (see Figure 7-8), which unit’s piping path dictates
the pump head from A to F? How much head is required?
7-2 What size pump capacity and head would be required to handle the four AHUs
t
in Figure 7-8 for the conditions shown from A to F?
e
7-3 What size pump capacity and head would be required in Figure 7-8 (direct
return for supplying four identical floors), assuming 10 ft between floors and
allowing a 25% head drop in the main for fittings and a 20 ft drop across the
chiller?
gp
7-4 In the reverse-return example in Figure 7-9, which unit’s piping path dictates
7-6 What size pump capacity and head would be required in Figure 7-9, assuming
pE
10 ft between floors and allowing a 25% head drop in the main for fittings and
a 20 ft drop across the chiller?
7-7 In the primary/secondary examples in Figures 7-10 and 7-11, which unit’s pip-
ing path dictates the pump head from A to F? How much head is required?
7-8 What size distribution pump capacity and head would be required to handle the
four AHUs in the Figure 7-11 primary/secondary example for the conditions
shown from A to F?
u
7-9 What size distribution pump capacity and head would be required for supply-
ing four identical floors (similar to the Figure 7-11 primary/secondary exam-
ple), assuming 10 ft between floors and allowing a 25% head drop in the main
ro
for fittings and a 20 ft drop across the chiller?

7-10 What is the cooling capacity of 100 gpm of water flow with a 50% propylene-
glycol mixture at an average temperature of 50°F and a 10°F rise?
7-11 What is the cooling capacity of 100 gpm of water flow at a 10°F rise? How
many tons of cooling?
G
7-12 What is the pumping horsepower for the propylene-glycol mixture (specific
gravity = 1.05) in Exercise 7-10, compared to the plain water in Exercise 7-11,
if the coil pressure drop is 20 ft, assuming a pump efficiency of 75% and motor
efficiency of 85%?
7-13 How can the possibility of frozen coils and piping in an HVAC system be
reduced?
7-14 What are two advantages of primary and secondary hydronic systems?
7-15 What type of pump curves should be selected if two-way valves are to be used?
7-16 How can flow be determined in a given loop?
7-17 Which piping system design gives more uniform pressure drop? Why?
e t
gp
u pE
ro
G
G
ro
upE
gp
et
Matching Pumps
to Systems
e t
Study Objectives
gp
After studying the material in this chapter, you should be able to do the following:
• determine the operating point of a pump and a system by plotting the sys-
tem head curve and the pump head–capacity curve;
• determine static pressure in a system by turning the system pumps off and
reading the gage pressure at the pump;
• plot system head and pump capacity curves, including the incremental
pE
effect of static pressure;
• know what methods are available to match pumping operation to a system
for full- and part-load flow;
• develop operating curves for pumps connected in series and in parallel;
• plan for emergency flow in case of a pump failure;
• determine how many combinations of capacity and head two two-speed
pumps can provide when operating in parallel;
• identify what the variable-volume controller should measure to control
u
variable-speed pumps; and
• describe what methods are available to provide flow to buildings that are
remote from a central chiller plant.
ro
Instructions
Matching the Pump to the System

G
The section Pipe Selection in Chapter 3 discusses pressure drop in piping

systems and the system curve that results from plotting pressure drop versus
flow (see Figure 8-1).
Similarly, in the section Pump Selection in Chapter 4, the concept of pump
head curves was developed, noting that each type of pump has a unique curve
governed by its size and design. The concept of the pump operating at the
intersection point of the system curve was also introduced.
236 Chapter 8 Matching Pumps to Systems
e t
gp
pE
Figure 8-1 Typical open-system system curve.
u
Superimposing the system curve and pump head curve, the operating point
is defined by the intersection of the curves (see Figure 8-2). Under actual oper-
ating conditions, control valves are varied to meet changing load conditions
ro
and the system curve changes, moving the operating point along the pump
curve.
In this manual, the premise was that the intersection point was based on a
system curve for a closed-loop piping system. That makes sense considering the
energy implementation of the Bernoulli equation (Equation 2-1), which shows
that pump energy is equal to the friction loss of the piping in a closed-loop sys-
G
tem. However, Bernoulli refers to three layers of pressure, one of which is due to
changes in elevation. When the system is piped in such a manner that a change in
elevation occurs within the piping circuit, this too must be accounted for on the
system curve. One such example is with a cooling tower. Using Figure 8-3 as an
example, we will define that the pump is supposed to move 100 gpm and that
there is 30 ft of pipe friction loss. (We are not dealing with extra additions for
pipe aging in this case, and those allowances could be significant). From
e t
gp
pE
Figure 8-2 Pump curve and system curve intersection.
u
ro
G
Figure 8-3 System with cooling tower.

e t
gp
pE
Figure 8-4 Cooling tower curves.
Figure 8-4, we will define d1 and d2 as 10 and 20 ft, respectively. These distances
represent both the difference in elevation and the distance between the surface of
the water in the tower sump serving the pump and either the discharge of the pipe
(d1) or the maximum pipe elevation (d2). The pump selected is slightly oversized,
being provided with a 7 in. diameter pump impeller. Under ideal circumstances
u
and if the system ran continuously, the pump would always have to elevate the
water 10 ft, the difference between pipe discharge and sump height. This is a
constant head loss due to elevation, regardless of what the flow rate is and the
varying head loss due to friction found in the pipe. Plotting the 30 ft variable sys-
ro
tem curve from a starting point of 10 ft, the pump will provide 110 gpm. To get
to the design flow, either a balancing valve is throttled, adding an extra 7.5 ft of
head loss, making the pump flow at 100 gpm and 47.5 ft head, or the pump
impeller is trimmed to slightly larger than 6 1/2 in. and operating at 40 ft of head.
Why does curve 2 appear in Figure 8-4? In cooling towers this helps illustrate
a simple point: the pump has to first elevate the water to the top of the pipe, and it
G
must have the energy to do so. What happens on start-up in this case? We will
stipulate that there is 10 ft of elevation from the sump water level to the centerline
of the pump suction. With the pump off, water is in the sump, floods the pump
case, and goes up the supply pipe 10 ft. As the pump begins to add energy, water
is removed from the sump and the elevation goes down, increasing the elevation
difference as water starts to go up the pipe. An important question here is whether
there is enough sump capacity to fill the pipe and allow the period in time when
the system is off as the water backs up into the sump. As the pump works, and it
will happen rather slowly, water will eventually make it to the top of the pipe and
establish a flow with the discharge pipe (sometimes called a downcomer), eventu-
ally filling. Cooling tower start-up has always been a troublesome subject when it
comes to start-up and having enough pump energy to initiate the process of design
flow. Adding to the issues are sump volume and coordination with sump makeup
water control so that water is not wasted at start-up or shutdown. One suggested
option for reducing issues is leaving the piping system in a filled state (perhaps
t
incorporating a control valve in the discharge that closes prior to the pump being
turned off) and possibly a check valve so that flow moves from the sump to the
e
pump but not vice versa. Regardless, remember that there are effects such as this
that must be allowed for when dealing with open systems.
Independent head is the system static pressure with the pump off. A good
piping system design will match the system characteristics to the pump head
gp
curve to provide the best system performance with the best economics over the
life of the system.
System Curves
System curves can be a powerful tool for the designer in determining the effects
that various components can have on system performance. Chapter 13 of
pE
ASHRAE Handbook—HVAC Systems and Equipment (2012) defines the use of a
system coefficient as
Q
C s = ----------- (8-1)
P
and is the system version of the control valve flow coefficient Cv. As such, it rep-
resents a defined flow in gallons per minute at a pressure drop of 1 psi or a head
u
of 2.31. Combined then with the use of the Darcy flow and head relationship (see
Equation 2-2), one can imply flow coefficients for any device or component that
has a defined head loss at a specific flow and plot the component curve just as
one might for a system curve. Note: because the flow coefficient is based on units
ro
of pressure in psi, we will stipulate that as we discuss a change in pressure we

will treat it as an equal change in head by the factor of 2.31 ft/psi.
Series Flow Coefficients

If one were to take this a few more steps, then we could form a logical path
G
such as the following: To calculate system head loss, we add all of the pressure
(head) losses for devices and components such that
Ps = P1 + P2 + + Pn (8-2)
Pn = pressure drop of components

PS = pressure drop of system or subsystem
Algebraically, if the system pressure loss is equal to the sum of the compo-
nent losses, we could rearrange the equation such that
Q Q- 2
C s = ----------- P = ------ (8-3)
P Cs
and
t
2 2 2 2
e
Q Q Q Q
P = ------- = ---------- + ---------- + + ---------- (8-4)
Cs C s1 C s2 C sn
Knowing that the flow in a circuit Q is the same for all components under
gp
evaluation, then the equation for a system flow coefficient made up of a group
of components in series is
1 1 1 1
------ = -------- + -------- + + -------- (8-5)
2 2 2 2
Cs C s1 C s2 C sn
As shown in Chapter 5, this equation is used for calculation of the effects

pE
of control valve authority. Commonly behaving devices can be grouped into
one composite flow coefficient. Considering how a system operates, pipe, fit-
tings, coils, and other nonmoving pieces represent these types of compo-
nents. They all have a fixed flow coefficient. The premise of a control valve,
though, is that every incremental position has a unique flow coefficient so
that flow may be decreased or increased according to the valve position.
When tested in a flow laboratory, the control valve is tested at every position,
maintaining a constant differential pressure across the valve. In so doing, the
u
system effects of components are factored out of the test, so the true flow
capability of the valve is accounted for. In application, it is rare that differen-
tial pressure is controlled across the control valve, so the variable nature of
the valve flow coefficients must be weighed against the constant component
ro
flow coefficients. Equation 8-5 may be rearranged algebraically to yield the

following:
Cs – v Cs – C
C s = ------------------------------------
- (8-6)
2 2
Cs – v + Cs – C
G
where
Cs = composite system coefficient
Cs – C = component flow coefficient calculated from the pressure drop of
the components (not including the control valve)
Cs – v = control valve flow coefficient at each incremental position
Example 8-1
For an equal percentage control valve, calculate and plot the composite flow
characteristic for the valve operating in a system with 50%, 30%, and 10% con-
trol valve authority. For the example, use design parameters as follows:
Control valve rangeability = 35:1
t
System head = 100 ft
Control valve flow = 100 gpm
e
Solution There are a couple of equations presented in previous chapters that will be
repeated here for this example. To begin with, recall how a theoretical equal
percentage control valve operates. In general it follows the equation
gp
l–1
C v = C vmax R
where
Cv = partial operating height flow coefficient,
gpm/psi
Cvmax = flow coefficient at full operating height,
pE
gpm/psi
l = partial lift of the valve
R = rangeability of valve, dimensionless and
tested
In organizing a solution, it becomes obvious that we will need to do a lot of
repetitive calculations that reference each other. In this regard, using a spread-
sheet to calculate and organize the data makes a lot of sense. Therefore, step 1
has us calculating the theoretical flow coefficient required for a flow rate of
u
100 gpm using all available head, which would be 100 ft. To do that, we use the
form of the flow equation incorporating head as follows:
Q 100
ro
C v = -------------- -------------- = 15.2

h 100
---------- ----------
2.31 2.31
This sequence is repeated for each valve position, the results of which are
shown for increments of 10% position (raw valve Cv). Note that at 0% valve
position there is an incremental flow due to the math of the calculation. As
G
mentioned previously, most manufactured valves are a modified equal percent-

age characteristic of some sort. For the purposes of this problem, we will sim-
ply modify the percentage by subtracting the closed value from the 0% through
90% valve positions (biased valve Cv). The biased valve flow coefficient is
then normalized (% Cv) and is plotted (curve 1) as the control valve character-
istic. This is the reference equal percentage characteristic, as if it were tested in
a fluids laboratory. See Table 8-1 for the data.
GTheoretical Equal
Percentage
Table 8-1 Theoretical Equal Percentage Control Valve Data for Example 8-1
Equal Percentage 50% Authority Equal Percentage 30% Authority Equal Percentage 10% Authority
Design Flow = 100 Design Flow = 100 Design Flow = 100 Design Flow = 100
Design P = 100 ft Design P = 50 ft Design P = 30 ft Design P = 10 ft
Raw Biased Raw Biased Sys- Sys- Percent Raw Biased Sys- Sys- Percent Raw Biased Sys- Sys- Percent
Posi- Valve Valve
ro
%
Valve Valve tem tem System Valve Valve tem tem System Valve Valve tem tem System
tion C Cv
v Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv Cv
100% 15.20 15.20 100% 21.5 21.5 21.5 15.20 100% 27.7 27.7 18.2 15.20 100% 48.1 48.1 16.0 15.20 100%
90% 10.65 10.22 67%
80% 7.46 7.03 46%
u 15
11
14.4
9.9
21.5 11.99
21.5 9.02
79%
59%
19.45 18.7
13.63 12.8
18.2 13.01
18.2 10.48
86%
69%
33.68
23.60
32.3
22.2
16.0
16.0
14.35
13.00
94%
86%
70% 5.23 4.80 32% 7 6.8 21.5 6.47 43% 9.55 8.8 18.2 7.89 52% 16.54 15.2 16.0 11.01 72%
60% 3.67 3.23 21% 5 4.6 21.5 4.47 29% 6.69 5.9 18.2 5.61 37% 11.59 10.2 16.0 8.62 57%
50% 2.57 2.13 14% 4 3.0 21.5 2.99 20% 4.69 3.9 18.2 3.81 25% 8.12 6.8 16.0 6.22 41%
pE
40% 1.80 1.37 9% 3 1.9 21.5 1.92 13% 3.29 2.5 18.2 2.47 16% 5.69 4.3 16.0 4.17 27%
30% 1.26 0.83 5% 2 1.2 21.5 1.17 8% 2.30 1.5 18.2 1.51 10% 3.99 2.6 16.0 2.58 17%
20% 0.88 0.45 3% 1.25 0.6 21.5 0.64 4% 1.61 0.8 18.2 0.82 5% 2.80 1.4 16.0 1.42 9%
10% 0.62 0.19 1% 0.88 0.3 21.5 0.26 2% 1.13 0.3 18.2 0.34 2% 1.96 0.6 16.0 0.59 4%
0% 0.43 0.00 0% 0.61 0.0 21.5 0.00 0% 0.79 0.0 18.2 0.00 0% 1.37 0.0 16.0 0.00 0%
gp
242
e t
The next step in the process is to calculate each control valve based on the
specified authority. Recall that control valve authority is the ratio of the control
valve pressure drop to the controlled system drop. In this case, the controlled
system drop is the 100 ft of the system as specified. Therefore, a 50% authority
would yield 50 ft head across the control valve, 30% 30 ft on the valve and
70 ft on the system components, and 10% 10 ft on the valve and 90 ft on the
system components.
50% valve authority
e t
Q 100
C v = -------------- -------------- = 21.5
h 50
---------- ----------
2.31 2.31
gp
30% valve authority
Q 100 = 27.7
C v = -------------- --------------
h 30
---------- ----------
2.31 2.31
10% valve authority
pE
Q 100
C v = -------------- -------------- = 48.1
h 10
---------- ----------
2.31 2.31
Similar to the development of the theoretical equal percentage coefficient,
each valve is biased and incremental positions are calculated. However, to see
the effect on the system flow control, the pressure drop effect of the system
components must also be calculated. To do this, Equation 8-5 is used and is
u
shown here for the 50% authority control valve operating at 50% position
Cs – v Cs – C
C s = ------------------------------------
-
2 2
ro
Cs – v + Cs – C
3 21.5 64.5 64.5

C s = ----------------------------- = ------------------------------ = ---------- = 2.99
2 2 9 + 462.25 21.7
3 + 21.5
This value is about 20% of the total system composite flow coefficient and
G
provides about 50% more water flow than implied by the theoretical control
valve coefficient. The rest of the calculations are repeats of this same process.
When the normalized values are plotted, the graph shown in Figure 8-5 is
developed.
In Figure 8-5, characteristic 1 is the theoretical characteristic,
characteristic 2 is the 50% authority characteristic, characteristic 3 is the 30%
authority characteristic, and characteristic 4 is the 10% authority characteristic.
e t
gp
pE
Figure 8-5 Valve flow versus lift for Example 8-1.
u
Parallel Flow Coefficients

ro
Knowing how to calculate the flow coefficient when we have components

in series leads to dealing with parallel flow paths. For discussion purposes, Fig-
ure 8-6 shows head versus flow for a pump serving three circuits, which we
will treat this as three balanced paths (e.g., that all paths have the same pres-
sure drop). The system coefficient can be calculated just as it would be if we
were developing a single value number akin to the system curve.
G
If each circuit has flows as designated in the figure, with each pressure drop
being equal, then it stands to reason that we could also calculate a single flow
coefficient for the path.
In doing this, it becomes apparent that the system flow coefficient is the
sum of each single path coefficient as designated in Equation 8-7. If we wanted
to see what happens in the system, we could use the approach of the compo-
nents in series to relate what happens when adjustments are made to specific
e t
gp
Figure 8-6 Head versus flow for pump serving three circuits.
pE
components. When this is done, it is important to remember that shared piping
pressure drops are used with respect to each path’s individual flow rate to
establish the flow coefficient for that path under evaluation.
C s = C sP1 + C sP2 + + C sPn (8-7)
For example, in circuit 2 shown in Figure 8-6, the flow rate is 80 gpm. The
path shares pipe segments A-B, with a head loss of 4 (2 × 2), and B-C, with a
u
head loss of 12 (2 × 6). If we were to consider just the flow coefficient for seg-
ment A-B, we could calculate it as about CsA – B = 140. However, while the
pipe carries the full flow, from a circuit perspective the path head loss is calcu-
lated as the total head loss for the incremental flow path. So if we were calcu-
ro
lating the path flow coefficient from the sum of its components in series, the
coefficient would be close to 86 for CsA – B, as part of circuit 2.
Figure 8-6 shows how each path curve interacts with a pump curve chosen
for this system example and also shows how the system curve interacts. It is
based on the calculations shown in Table 8-2, which were done in a spread-
sheet. It is a balanced path example, meaning that all paths have the same head
G
loss, 42.5 ft.

What would have happened if the system had been left unbalanced? The
effect may be easily calculated by leaving the balancing valve pressure drop
effect out of the path calculation, as shown in the calculations in Table 8-2. Note
that for each path, the balanced flow coefficient is shown above the unbalanced,
and it is always less, indicating that unbalanced will have more flow as deter-
mined by the unbalanced system curve intersection point with the pump curve
246
G Table 8-2 Balanced Path Example

Balanced Path Flow
Path 1
A-B B-C C-D Comps Coil Valve Balance E-F F-G G-H Total
Flow 20 20 20 20 20 20 20 Cv
Head 2 2.5 12 12 12 2 42.5 4.6627
Flow
Coefficient
21.49
ro 19.22 8.77 8.77 8.77 21.49 Cv
2
1/ C S 0.0022 u 0.0027 0.013 0.013 0.013 0.0022 0.046 4.6627
Path 2
Chapter 8 Matching Pumps to Systems
Flow 80 80 80 80 80 80 80 80 80 Cv
Head 2 6 1 3 4 18.5 6 2 42.5 18.651
Flow
85.98 49.64 121.59 70.20 60.79 28.27 49.64 85.98 Cv
Coefficient
pE
2
1/ C S 0.0001 0.0004 7E-05 0.0002 0.0003 0.0013 0.0004 0.0001 0.0029 18.651
Path 3
Flow 30 30 30 30 30 30 30 30 30 30 30 Cv
Head 2 6 5 0.99 3.2 10 2.31 5 6 2 42.5 6.9941
gp
Flow
32.24 18.61 20.39 45.83 25.49 14.42 30.00 20.39 18.61 32.24 Cv
Coefficient
2
1/ C S 0.001 0.0029 0.0024 0.0005 0.0015 0.0048 0.0011 0.0024 0.0029 0.001 0.0204 6.9941
e t
(Table 8-3). This is then plotted so that it can be evaluated, as seen in Figure 8-7:
the unbalanced system flows about 155 gpm, which is about 20% more than
design flow.
Many scenarios can be altered to show different effects. In Figure 8-8, distri-
bution pipes were made larger, substantially reducing friction losses in part of
the system, and the coils and control valves were resized to reflect less pressure
loss. As can be seen, the system curve nearly goes off the pump curve to a point
of unsatisfactory operation of the pump. Similarly, we can make the curve come
t
closer to the actual system curve by redistributing where the head losses are
located in the piping network.
e
The idea being presented is not to prove a particular point, it is actually to
show a technique that is not difficult to implement in a spreadsheet that allows
the designer to evaluate and make system decisions to optimize the control of
gp
the system and minimize the energy use. With respect to balancing or not bal-
ancing a system, using normal design guidance on pipe selection with reason-
able pressure drops for control valves and coils (preferably more than the pipe
losses), systems that are unbalanced will typically overflow from 8% to 25%.
Of course, these examples are based on the idea that the pump head has
actually been calculated and selected for what is actually required. It is very
common in the field to see a variety of what can only be described as gross
errors, such as installing the wrong speed pump, unapproved modification of
pE
the pump selection through the addition of extra head or flow as a safety factor,
guesstimating the system head, using diversity factors improperly, and install-
ing components that are not what was called for in the design. This technique,
because it can establish a flow coefficient based on components, allows for sys-
tem evaluation in a simple and reasonably accurate way.
Parallel Pumping
u
As mentioned in Chapter 4, designers should consider pumps in parallel,
especially when the designer is installing two pumps, one as primary and the
other as backup. In backup mode, two pumps of the same size are selected.
They are connected to the system in parallel only from the perspective of
ro
pump failure (e.g., when one pump is inoperable, it is hydraulically discon-

nected from the system through valves and the other is connected so that the
system may operate).
Parallel pumping operates two pumps to provide design flow and head con-
ditions. A controller turns one of the pumps off at conditions of load less than
design and complementary to the flow required for heat transfer. The advantage
G
to this design is that two smaller pumps may be purchased, saving money. Also,
when it is operating in single pump mode, pump energy may be reduced.
When two smaller pumps are purchased and operated in parallel, a com-
posite pump curve is created by the designer and compared with the design
system curve. Pumps are selected for one-half of the system design flow rate
and the full design head of the system. The composite curve is sketched on a
pump curve or by using a spreadsheet by adding flow rates on a line of constant
Table 8-3 Unbalanced Path Example
248
Unbalanced Path Flow

G A-B B-C C-D Comps Coil
Path 1
Valve Balance E-F F-G G-H Total
Flow 20 20 20 20 20 20 20 Cv
Head 2 2.5 12 12 12 2 42.5 4.662743
Flow
21.49 19.22 8.77 8.77 21.49 Cv
Coefficient
ro
2
1/ C s 0.0022 0.0027 0.0013 0.0013 0.0022 0.033 5.5041
Path 2
A-B
u
B-C C-D Comps Coil Valve Balance E-F F-G G-H Total
Chapter 8 Matching Pumps to Systems
Flow 80 80 80 80 80 80 80 80 80 Cv
Head 2 6 1 3 4 18.5 6 2 42.5 18.65097
Flow
85.98 49.64 121.59 70.20 60.79 49.64 85.98 Cv
Coefficient
pE
2
1/ C s 0.0001 0.0004 7E-05 0.0002 0.0003 0.0004 0.0001 0.0016 24.819
Path 3
Flow 30 30 30 30 30 30 30 30 30 30 30 Cv
Head 2 6 5 0.99 3.2 10 2.31 5 6 2 42.5 6.994115
Flow
gp
32.24 18.61 20.39 45.83 25.49 14.42 20.39 18.61 32.24 Cv
Coefficient
2
1/ C s 0.001 0.0029 0.0024 0.0005 0.0015 0.0048 0.0024 0.0029 0.001 0.0193 7.1923
e t
e t
gp
pE
Figure 8-7 Unbalanced system curve intersection point with pump curve.
u
ro
G
Figure 8-8 Effects of larger distribution pipes and resized coils and control valves.
head for the various points of the single pump curve. The points are then con-
nected to represent a pump curve when two pumps are operational. At each
head, the horizontal vector of flow is added to the first as follows:
Y 1 + Y 2 = Total flow at B
and
t
X 1 + X 2 = Total flow at A
e
A new pump curve is developed (Figure 8-9) on the paralleled pump curve
line A-B-C.
Plotting a system curve across the parallel pump curve shows the operating
points for both single and parallel pump operation (see Figure 8-9). It is inter-
gp
esting to note that parallel pumping tends to flatten the combined pump curve
to make the system pressure more suitable for control valve operation.
Pumps should be identical in rating of flow and head, design, impeller
diameter, and speed. It may cause problems if one of the pump’s characteris-
tics is greater than those of the other (such as closing the discharge check
valve), thereby making start-up unpredictable. Pumps of unequal pressures
may result in one pump creating a pressure across the other pump in excess
pE
of its cutoff pressure, causing the flow through the second pump to reduce or
u
ro
G
Figure 8-9 Pump curve developed on paralleled pump curve line A-B-C.
cease. This can cause flow problems or pump damage. By examination of the
system and composite pump curves, when the system is in a design condition
(meaning that flow may go to all connected paths), running a single pump
yields more than 50% flow.
While dissimilar pumps may be pumped in parallel, care should be
taken so that when both pumps are operated together, the system flow does
not fall low enough that the operation of a pump in the shadow of a larger
pump occurs. When this occurs, it is possible for uneven forces and excess
t
radial thrusts to be forced on the smaller pump, with the potential for cavi-
tation. Typically two (or more) same-sized pumps are used and are pre-
e
ferred.
The system curve should intersect both pump curves as shown in
Figure 8-10. Where the system curve intersects the single pump curve, it
gp
will be further out on the curve than the nominal one-half design point con-
ditions, yielding a flow greater than one-half of the design flow. This inter-
section point is sometimes referred to as the standby flow, and generally
when a flat-curve pump is selected, the percentage of standby flow yields a
high degree of design heat transfer with respect to the system.
The system curve crosses the single pump curve to the left of the com-
bined pumps’ operating point. This factor leads to two important points:
pE
• The pump motor must be adequately sized to prevent overloading dur-
ing single pump operation (see Figure 8-10).
• The flow rate associated with where the pump system curve and the sin-
gle pump curve intersect divided by the design flow rate yields a per-
centage of design flow for the system in single pump operation. A
single pump may provide standby service as high as 80% of design
flow; the actual amount depends on the specific pump curve and the
system curve. Recall that in a heating system designed for 20°F T,
u
75% of design flow yields 97.5% heat transfer and 60% yields about
90%.
ro
The piping of parallel pumps (Figure 8-11) should be laid out with pro-
vision to run either pump with a bypass around the other. Hand valves (gate
or butterfly) must be manually positioned, or two-position automatic con-
trol valves can be tied in with the pump selection controls. The check valve
in the pump discharge closes when the pump is shut down. The alternate
pump draws from the return bypass and discharges into the supply bypass.
G
Series Pumping
When pumps are operated in series, each operates at the same flow rate and pro-
vides its share of the total pressure at that flow (see Figure 8-12). At each flow, the
vertical vector of head is added to the first (shown as Xl + X2 = total head at A, and
Yl + Y2 = total head at B, etc.) until a new pump curve is developed (curve C-A-B).
e t
gp
pE
Figure 8-10 Pump motor sized to prevent overloading during single pump operation.
u
ro
G
Figure 8-11 Piping schematic of parallel pumps.

e t
gp
pE
Figure 8-12 Pump curve for series operation.
The series pump curve is drawn with full flow at low head and then with dou-
ble the head of the single pump curve at each flow value to construct a similar
u
pump curve. A system curve plotted across the series pump curve shows the
operating points for both single and series pump operation (see Figure 8-13).
Note that the single pump provides up to about 80% flow as a standby and at a
lower power requirement.
ro
As with parallel pumps, it is important that pumps in series be identical in

rating of flow and head, design, impeller diameter, and speed. Pumps of differ-
ent flow capacity connected in series can result in problems; the pump of
greater capacity can overflow the pump of lesser capacity, causing damage in
the smaller pump due to cavitation. This can also cause a pressure drop rather
than a pressure rise across that pump.
G
It is important to discuss the proposed parallel, series, or multiple pump

selection with the pump manufacturer to prevent a potential problem and to get
the manufacturer’s suggestions and horsepower requirements.
The piping of series pumps (see Figure 8-14) should be laid out with provi-
sion to run either pump with a bypass (B-C in Figure 8-14) around the other.
Hand valves (gate or butterfly) must be manually positioned, or two-position
automatic control valves can be selected.
e t
gp
pE
Figure 8-13 Piping schematic of series pumps.
u
ro
G
Figure 8-14 Operating conditions for series pump installation.

Standby Pumps
It is always good practice to consider a backup pump of equal capacity and
proper valves to permit operation when the normal pump is inoperable. Usually
this is an application for a parallel pump (see Figure 8-15). Failure often occurs
at the worst possible time, meaning that heating pumps fail in extremely cold
weather or chilled-water or condenser water pumps fail in the middle of a hot
t
spell during the cooling season. The original investment costs of bypass piping
and pumps will be trivial compared to the inconvenience or damage for the
e
building occupants or the operator. Standby pumps are similar to parallel
pumps except that they are sized for full design flow operation. This is why siz-
ing for parallel operation is considered a convenience: it cuts down the pump
cost with almost no impact on system heat transfer or temperature control
gp
when properly considered.
Trimming Pump Impellers and Adjusting Pump Speed

The pump affinity laws were introduced in Chapter 4. You will remember that
• flow is directly related to pump impeller diameter,

pE
• head is related to the square of the pump impeller diameter, and
• power is related to the cube of the pump impeller diameter.
u
ro
G
Figure 8-15 Piping schematic of standby pump.

As shown in Table 4-1, Chapter 4,

BHP1 = original pump horsepower, hp
BHP2 = new pump horsepower, hp
D1 = original pump diameter, in.
D2 = new pump diameter, in.
h1 = original head, ft
t
h2 = new head, ft
N1 = original pump speed, rpm
e
N2 = new pump speed, rpm
Q1 = original flow, gpm
Q2 = new flow, gpm
gp
These conditions in Table 4-1 hold true as long as the pump head curve is
for a centrifugal pump and the system curve is followed.
Based on these relationships, optimizing the installed pump construction so
that the impeller matches the system (rather than matching the system to the
pump by providing excessive throttling losses) is something that should be
carefully considered. This can be done in two different ways, by varying either
the speed of the pump or the diameter of the impeller.
pE
Depending on the connected horsepower, variable-speed drives can be a
highly economical method of getting system adjustment with little commis-
sioning headache. In 2011, variable-speed drives could be purchased for as
little as $150 for a 1/4 hp drive or $700 for a 10 hp drive. This is a pittance,
especially considering that a 10 hp drive was thousands of dollars just a few
decades ago. Considering that energy codes tend to require the use of a drive
on the pump at about 5 hp as a rule of thumb (about $400), the speed drive,
even if left uncontrolled by a separate controller, allows for the maximum
u
pump speed to be limited to an amount equivalent to a reduced-size impeller,
thus saving horsepower after efficiencies are accounted for.
The alternative adjustment is to trim the pump impeller. In this operation,
the pump is operated, the system is commissioned and adjusted, and, when
ro
convenient, the system is shut down with the pump isolated. The pump is disas-
sembled, removing the impeller from the pump volute, and machined, reducing
the impeller diameter using the affinity laws. The pump is then reassembled,
and any throttling adjustments that were made during system commissioning
are removed from the system by opening the pump throttling device.
If we were to examine a modified pump curve to show the difference in
G
effects from speed adjustment versus impeller trimming, the curves would look
exactly the same. Either method is an acceptable method for reducing pump
horsepower to just that necessary to make the system operate properly.
Prior to any of these adjustments, it is imperative that the operating point of
a pump be considered when the system terminals operate at full load and part
load with thermostat control of two-way valves. In Figure 8-16, point 1 shows
the pump operating at the design flow and the calculated design pressure drop of
the system. Typically, the actual system curve is slightly different than the design
curve. As a result, the pump operates at point 2 and produces a flow rate higher
than design. In the process of balancing the system, a terminal balancing valve in
the most hydraulically significant circuit should be open, or throttled only
enough to provide a measurable differential pressure to calculate terminal flow.
All other circuits should be proportionally balanced with respect to this circuit.
To reduce the actual flow to the design flow at point 1, a balancing valve
downstream from the pump can be adjusted when all terminal valves are in a
t
wide-open position. This pump discharge balancing valve imposes a pressure
drop equal to the pressure difference between point 1 and point 3. This alternative
e
simply adds extra head loss to the system by throttling a pump discharge valve to
limit maximum pump flow after the system has been proportionally balanced.
Lest we think that this is not energy efficient, a simple comparison of oper-
gp
ating horsepower in the adjusted and nonadjusted states shows that whenever
flow is reduced in a centrifugal pumping system, operating energy use will be
reduced. However, that reduction is not optimal in the sense that getting the
u pE
ro
G
Figure 8-16 Pump operating points.

pump to provide only the energy required to the system without extra throttling
will save more pump energy than throttling alone. A note on this, though:
depending on just how flawed a system and its connected pump are, it may be
impossible to get the pump adjusted so that it provides only the minimum
energy required to create the design flow in the system. In these cases, minor
throttling losses and pump speed or impeller modification might be used to
match the connected system and the pump.
The manufacturer’s pump curve will dictate exactly how much the capacity
t
may be reduced by substituting a new impeller with a smaller diameter, trim-
e
ming the existing pump impeller, or adjusting the speed. The manufacturer’s
pump curve shows the boundary limits to impeller sizes in a particular volute.
Pump speed, on the other hand, can usually be taken down to a minimum of
30% speed, a larger range than normally is achieved by trimming the impeller.
gp
After adjustment, reopening the balancing valve in the pump discharge elimi-
nates artificial drop and the pump operates at point 3 (see Figure 8-16).
Points 3 and 4A demonstrate the effect that a trimmed impeller has on reducing
flow (Figure 8-16).
Two-Speed Pumping
pE
Multiple-speed motors are also an option to reduce system overpressure at
reduced flow. Standard two-speed motors are available in models with speeds
of 1750/1150 rpm, 1750/850 rpm, 1150/850 rpm, and 3500/1750 rpm. These
are common options on small fractional horsepower circulating pumps. In
larger pumps, as previously mentioned, application of a variable-speed drive
may be much more economical on a single-speed motor rather than purchasing
the two-speed motor.
However, as mentioned regarding the pump affinity laws (Chapter 4),
u
• flow is directly related to pump speed,

• head is related to the square of the pump speed, and
ro
• power is related to the cube of the pump speed.
Figure 8-17 shows the performance of a system with a 1750/1150 rpm

multiple-speed pump. In Figure 8-17, curve A shows the system’s response
when the pump runs at 1750 rpm. When the pump runs at 1150 rpm, opera-
G
tion is at point 1 and not at point 2 (as pump affinity laws calculate), because
the pump head curve (at 1150 rpm) and the system curve must be satisfied.
Another concern is if the system were designed to operate as shown in
curve B, the pump would operate at shutoff and be damaged if it was run at
1150 rpm. This example demonstrates that the designer must analyze the system
carefully to determine the pump’s limitations and the effect of lower speed on per-
formance.
Pumps with Two-Speed Motors (Stethem 1988)

If a pump has a two-speed motor, a simple approach to reducing pump
capacity is selecting a relay to direct the power to the high or low windings of
the pump motor. Figure 8-18 shows an example of a pump performance curve
for a 6 × 8 × 9.5 in. pump. If this application calls for a capacity of 1670 gpm
and 50 ft head (at point C in Figure 8-19), a two-speed motor of 1750/
t
1150 rpm will permit this flow to be reduced to 985 gpm and 26.3 ft (point A).
If the example further calls for a design increase of 500 gpm in the future,
e
the two-speed pump can be combined with a duplicate two-speed pump in par-
allel to give the results shown in Table 8-4.
If 2170 gpm represents the ultimate load, then you can see this arrangement
permits four flow selection steps to reduce pumping power at a minimum of
gp
equipment investment:
u pE
ro
G
Figure 8-17 Two-speed pumping.

e t
gp
pE
Figure 8-18 Typical performance curve—6 in. suction × 8 in. discharge × 9.5 in. impeller.
u
ro
G
Figure 8-19 Two-speed pumping example.

• 100% gpm at 100% power point D

• 77% gpm at 57% power point C
• 58% gpm at 27% power point B
• 45% gpm at 16% power point A
Variable-Speed Pumping
To understand what happens in a constant-speed pumping system, look at a
t
direct-return system with two-way control valves (see Figure 8-20). At full load,
all two-way control valves are wide open and the pump operates at point A (see
e
Figure 8-21) to deliver 1100 gpm at 100 ft total system head loss. The design
head loss is 80 ft for piping, coil, and fittings, and 20 ft for the control valve. As
gp
Table 8-4 Four Flow Selection Steps to Reduce Pumping Power
Pumps, rpm Capacity, gpm Head, ft bhp
Both at 1750 2170 75.0 48.0
One at 1750 1670 50.5 27.5
Both at 1150 1250 33.0 13.0
pE
One at 1150 985 26.3 7.5
u
ro
G
Figure 8-20 Typical direct-return system.

e t
gp
pE
Figure 8-21 Head loss with two-way valves at full load and 50% part load.
u
the load reduces, the two-way valves assume new positions (under command of
their zone or room thermostats), and the system flow is reduced to 50% (to
550 gpm). The system curve moves along the pump head curve from point A to
ro
point B. This means that the total system head loss is 110 ft, of which 90 ft must
be absorbed by the control valve and 20 ft is the head loss for piping, coil, and fit-
tings at 50% flow.
The control valve must be chosen to hold its position against this high dif-
ferential. If one wants the total system to operate down to extremely low flows
G
(10% to 25%), then the valve actuators have to be chosen for an even higher
h (100 to 110 ft). This is a serious concern because unit valves may be
mounted in limited space conditions of unit cabinets where larger-diameter
actuators may not fit. An increase in valve cost is another factor, and so the
problem requires other system design features. Staging fewer parallel pumps
off or selecting a lower pump speed should be considered. Staging fewer
pumps may result in an operating cost penalty.
Because variable-speed drives have become more economical, their use for
controlling pump speed is now a typical application. Generally, the economics are
so favorable that building energy codes now make them a requirement starting at a
very low horsepower, often around 5 hp. A pump controller is installed, typically
using a proportional-integral-derivative (PID) controller where differential pres-
sure becomes the controlled variable and pump speed is the manipulated variable
through the variable-speed drive. Differential pressure is the most common con-
trol application; however, other variables could be controlled to implement the
t
designer’s control concept. Differential pressure is most common because it
closely aligns with the primary functional response of the building system (space
e
temperature control for comfort), as discussed in Chapter 5. Differential pressure
is sensed at the most hydraulically significant point in the system, which is the
flow path with the greatest head loss used to select the pump. The controller acts
to maintain a constant differential pressure across a portion of the piping system
gp
that must include a modulating control valve. As the valve opens and closes, the
P controller senses the difference between the measured value and the design
head loss for the controlled segment of pipe and transmits a proportional control
signal (usually 0 to 10 vdc or 4 to 20 ma) to the variable-speed drive to vary the
pump speed proportionally to maintain the desired P setpoint. The device may
be calibrated in differential pressure for feet head or psi. The sensor/transmitter
continually sends a signal to the controller, making this a feedback control appli-
pE
cation. In some applications, multiple sensors/transmitters are located in different
zones and are used to maintain each zone’s setpoint. That said, with only one sig-
nal output to increase or decrease a pump speed, multiple zone setpoints would
suggest that only one could be satisfied at any given moment. Typically, control-
lers that incorporate multiple sensors select the zone with the greatest control
error, using that as the PID controller input for variable-speed drive adjustment.
Note: To ensure continuous signal monitoring within direct digital control
(DDC) systems, it is recommended that the differential pressure sensor be
u
hardwired to the PID controller rather than transmitted across a digital network
(see Figure 8-22).
In a basic direct-return system, a sensor/transmitter is usually located at the
end of the last riser (see Figure 8-23) or in a zone with the highest estimated
ro
drop to maintain a set differential, nominally 20 or 30 ft.

As the controller changes the pump speed to maintain the P ( H) setpoint,
it creates a control curve (see Figure 8-24) maintained by the control system as
the load on the system is varied by the zone or local loop temperature controls.
Reviewing the pump affinity laws, we see in Figure 8-25 how the pump
speed directly varies the pump head curve; if flow is reduced to 1/2, then the
G
head is reduced to (1/2)2 = 1/4. Figure 8-26 shows the horsepower reduction
with pump speed; as flow is reduced to 1/2, the power is reduced to (1/2)3 = 1/8.
We can see how the variable-speed pumping system then reduces high h at part
loads but also takes advantage of the pump laws to reduce the operating cost.
The electric motor should be selected to be equivalent to a National Electrical
Manufacturers Association (NEMA) class B design, and the motor manufacturer
should be consulted regarding minimum practical operating speeds. In general,
e t
gp
pE
Figure 8-22 Proportional controller and adjustable-frequency drive controlling pump.
u
ro
G
Figure 8-23 Direct-return system with sensor/transmitter located at end of last riser.
e t
gp
pE
Figure 8-24 Differential pressure control curve above piping friction loss.
minimum speeds are recommended to be 30% of the maximum speed. The con-
troller/drive is set for minimum pump speed to protect the pump motor and pro-
vide enough circulation to flush and cool the seals. In addition, as pumps become
larger where couplers are used between the motor and the pump shaft, the pump
manufacturer must be notified and the pump specified for variable-speed opera-
u
tion so that the appropriate coupler is applied. Constant-speed couplers tend to be
hard rubbers or plastic materials. Variable-speed pump couplers incorporate a
more flexible rubber material that is designed to absorb the speed changes. Typi-
cally, constant-speed couplers are not tolerant of constant changes in speed, and
ro
the coupler can be ground up as each individual shaft’s mating piece impacts the
coupler material. The more flexible coupler is noticeably larger and more expen-
sive and can absorb the twists that occur when the pump changes speed.
Many pump controllers are of microprocessor design and are available with
the following features:
• Staging multiple pumps to maintain the head-flow conditions to prevent a
G
pump from becoming overloaded before extending beyond its power rating
• Automatic bypass controls for mechanical or electrical failure or power
consumption/efficiency
• Operator alarm indications and warnings
Some of these functions are integrated with the building DDC system, and
some are stand-alone dedicated pump control systems.
e t
gp
pE
Figure 8-25 Pump curve showing head reduction with change in pump speed.
Source Distribution Pumping

u
Reviewing the basic concept of primary/secondary pumping (see Figure 8-27),
we can see how this principle is applied to the source (such as a group of chillers
or boilers) to meet the partial loads and bring on additional units to meet high
loads. The primary/secondary concept allows the distribution pumping of the
ro
source supply from a central pumping facility (see Figure 8-28) or distributing
the pumping to remote buildings or zones of a large facility (see Figure 8-29).
Each case must be studied by the designer to determine the most economical
operating cost versus the installed cost and the maintenance staff capabilities.
In some cases, the staff may be limited to a central boiler or chiller plant
operation. In other cases, remote locations of pumping stations may fit very
G
well if the individual buildings have adequate mechanical equipment rooms

and staff to maintain them.
The primary/secondary concept allows continual flow through the source
and still permits two-way valve control in the loads. There is flexibility in
dedicating a pump to a chiller or boiler or manifolding the pumps. This might
simplify the need for having backup pumps for every system. Pumps, mani-
folds, accessories, and associated pumping control may be assembled to
e t
gp
pE
Figure 8-26 Pump curve showing pumping power reduction with change in pump speed.
u
ro
G
Figure 8-27 Primary/secondary pumping concept.

e t
Figure 8-28
gp
Main-source primary/secondary variable-speed pumping.
u pE
ro
G
Figure 8-29 Distributed variable-speed pumping.

match installation constraints or can be factory prepackaged as an assembly

for a designated mounting location. The designer must weigh the pros and
cons of cost, flexibility, and installation requirements of the various concepts
to determine the best arrangement and cost for the project.
The Next Step

The next chapter covers water chillers and load control.
t
Summary
e
The pumps covered in this chapter include the following:
• Single pump: selected for a simple application
gp
• Single pump with trimmed impeller: optimizing pump capacity for a spe-
cific application
• Single pump with backup pump: in addition to a selected application, pro-
vides 100% backup
• Two-speed pump: provides limited variable-flow steps with an added
investment
• Parallel pumps: flexible capacity control without increasing system head,
pE
good for two-way valve control
• Series pumps: steep head change with limited flow change, two-way valves
would require high differential pressure operation and capability
• Primary/secondary pumping: flexible zoning approach with minimum
pumping energy
• Distributed pumps: special application of primary/secondary pumping
• Variable-speed pumps: applied to pumping systems to reduce power by
u
lowering pump speed to meet control differential pressure in selected loca-
tions, usually applied to parallel pumping distribution systems using pri-
mary/secondary or distributed pumping with two-way control valves
ro
• The methods to match pump operation to a system

• That the parallel pump curve is constructed by adding flow capacity at each
value of the pump head
• That the series pump curve is constructed by adding pump head at each
G
value of pump flow

• How to provide emergency flow in case of a pump failure
• How to take advantage of capacity and head combinations provided by
two-speed pumps
• What the pump controller measures to control pump speed
• The methods to provide flow to buildings that are remote from a chiller
plant

Atlanta: ASHRAE.
Bell & Gossett. 1985. VIS Pumping Fundamentals. Technical Report 685.
Morton Grove, IL: Xylem Inc.
Hegberg, R. 1991. Converting constant-speed hydronic pumping systems to
t
variable speed pumping. ASHRAE Transactions (97)1:739–45.
Rishel, J.B. 1991. Control of variable-speed pumps on hot- and chilled-water
e
systems. ASHRAE Transactions (97)1:746–50.
Stethem, W. 1988. Application of constant speed pumps to variable volume
systems. ASHRAE Transactions 94(2):1458–68.
gp
u pE
ro
G

Complete these questions by writing your answers on the worksheets at the
back of this book.
8-1 Assume two pumps (6 × 8 × 9.5 in., see Figure 8-14) are to be applied in paral-
lel for a 70 ft head system. What flow will result at 70 ft?
t
8-2 Assume the same two pumps (6 × 8 × 9.5 in., with a 9.6 in. impeller) are to be
piped in series. What will the resulting head be at 1600 gpm?
e
8-3 Assume the 6 × 8 × 9.5 in. pump is being considered for variable-speed
operation of a system designed at 1600 gpm at 70 ft head and the pump
manufacturer advised that the NEMA type B electric motor to be furnished
gp
can be run down to 40% of its furnished speed (1750 rpm). What is the min-
imum flow and head the pump can be run at following the same system
curve?
8-4 In Exercise 8-3, what is the theoretical brake horsepower reduction if the pump
is run at 40% speed?
8-5 What methods are used to provide flow to buildings that are remote from a
pE
chiller plant?
8-6 What does the pump controller measure to control pump speed?
8-7 Explain how combinations of two-speed pumps can be used to provide required
head and flow capacity with improved efficiency at part-load conditions.
8-8 What is a good way to provide emergency flow in case of a pump failure?
8-9 What methods are used to match pump operation to a system?
u
ro
G
G
ro
upE
gp
et
Water Chillers and
Load Control
e t
Study Objectives
gp
• identify a water chiller’s capability to reduce water temperature, dependent
on the tonnage rating and water flow rate;
• understand water temperature rise across a chilled-water (CHW) coil;
• understand temperature drop in condenser water temperature;
• understand the many types of refrigeration compressors used in chillers,
pE
depending on the manufacturer and the tonnage size;
• know that the theoretical horsepower used in a chiller is directly related
to the heat (enthalpy) absorbed in the evaporator, the weight rate of
refrigerant flow, the enthalpy change in the compressor, and the chiller
tonnage;
• understand that chillers can be piped in series or parallel but are commonly
in parallel to provide for expansion;
u
• understand piping of multiple chillers, using primary/secondary princi-
ples, and loading chillers evenly when locating the common bridge
between the load and the chillers; and
• know how to optimize chiller operation versus CHW supply temperature,
ro
taking into account the compressor power and the pumping power.
Instructions
G
Basic Water Chiller Components

Water chillers provide cooling for water, brines, or other secondary cool-
ants for air conditioning or refrigeration (ASHRAE 2014). The systems can be
either factory assembled and wired packages or shipped as component sections
and built up in the field. The basic components of a liquid chilling system
include the compressor, evaporator (liquid cooler), condenser, refrigerant flow
274 Chapter 9 Water Chillers and Load Control
control device, and a control center, as well as other auxiliary devices (receiver,
intercooler, oil separator, etc.) within the chiller package (see Figure 9-1).
Externally, the chiller must have a condensing water system (cooling tower,
pumps, diverting valve, etc.) to transfer the heat properly.
Figure 9-2 shows a schematic of a simple liquid chiller cooling a water sys-
tem from 54°F to 44°F and transferring that energy to a condenser water system
operating from 85°F to 95°F.
t
Refrigeration Cycle
e
A pressure-enthalpy (p-h) chart (also called a Mollier diagram) is a portrayal
of the pressure enthalpy values of the refrigerant. Figure 9-3 shows the low-
pressure (evaporation) side versus the high-pressure (condensing) side of a
gp
basic system and is useful for visualizing the refrigeration cycle of the water
chiller (Trane 2002). Exact values of the refrigerant pressures, temperatures,
enthalpy, density, and specific volumes are found in manufacturers’ tables or
in ASHRAE Handbook—Fundamentals (ASHRAE 2013).
To understand the power used in a water chiller system, it is important to
understand the basic refrigeration cycle and how compressor power and pump-
ing power interrelate. In the p-h chart (Figures 9-3 to 9-5), temperature lines
are constant with the pressure line between saturated liquid and the saturated
pE
vapor line, or the wet region. If the refrigerant is at point A in Figure 9-5, it
u
ro
G
Figure 9-1 Basic components of liquid chilling system.

e t
gp
Figure 9-2 Schematic of simple liquid chilling system.
u pE
ro
G
Figure 9-3 Pressure-enthalpy diagram for a refrigerant.

e t
gp
pE
Figure 9-4 Simplified pressure-enthalpy diagram for a refrigerant.
u
ro
G
Figure 9-5 Refrigeration cycle shown on simplified pressure-enthalpy diagram.

absorbs heat with no change in pressure and will begin to boil, and evaporation
will occur with no change in temperature. As heat is added at point D, the
refrigerant’s enthalpy increases until it becomes saturated at point B. Further
addition of heat at constant pressure moves the refrigerant condition into the
superheat region to point C.
Tracing the simple chiller shown in Figure 9-2 onto the p-h chart (Figure 9-3),
it is shown that liquid is throttled by the expansion valve from the condenser pres-
sure at point 4 in Figure 9-5 to the evaporator low pressure at point 1 at a constant
t
enthalpy. As the refrigerant flows through the evaporator, it absorbs the heat nec-
essary to completely vaporize it by point 2 (water chilling) and in practice is
e
superheated. The compressor raises the vapor from low pressure at point 2 to high
pressure at point 3. The high-pressure vapor is then condensed from point 3 to
point 4, where the heat picked up in the evaporator plus the superheat and the heat
of compression are transferred to the cooling-tower water.
gp
The Air-Conditioning, Heating, and Refrigeration Institute (AHRI) is the
industry trade association that rates chillers at their operating conditions. The
applicable standard is AHRI Standard 550 (2011). In this standard, operating
points for evaluation are designated so that different chillers and vendors may
be examined. AHRI rating points are as follows:
• Leaving CHW temperature: 44°F

pE
• CHW flow rate: 2.4 gpm/ton
• Entering condenser water temperature: 85°F
• Condenser water flow rate: 3 gpm/ton
• Evaporator fouling factor and 0.00025 condensing fouling factor
Heat Transfer Chiller

The size of the chiller is rated in tonnage or tons of refrigeration. The historic
u
definition of a ton comes from making 1 ton of ice in 24 hours.
2000 lb 144 Btu

1 ton = ------------------ ------------------- = 12,000 Btu/h (9-1)
ro
24 h lb
Reviewing Chapter 1, we remember that the heat transferred to or from
water is
Q = q 500 t ent – t lvg (9-2)

G
then
tons = 12,000 Btu/h = 500 q t (9-3)
q t q t
tons = ---------------------- = --------------- (9-4)
12,000 24
----------------
500
This may also be referred to in the following format, showing flow in gal-
lons per minute per ton of cooling:
q gpm
---------- = 24
------ (9-5)
ton t
This is a good formula for estimating flows versus t.

If the liquid is a brine or glycol mixture and not water only, then the for-
t
mula must include specific heat and specific gravity of the solution:
e
q t c p sg
tons = -------------------------------------
- (9-6)
24
gp
In the condenser, the heat transferred to the condenser water includes the
heat from the evaporator plus the heat of compression. For most practical com-
fort air-conditioning applications, a value of 14,400 Btu/h is used as the total
heat transferred to the condenser water (Trane 2002). This may be referred to
as a cooling tower ton.
Following the same logic as for the evaporator, the short form for the con-
denser is
pE
gpm
q ---------- = 28.8
---------- (9-7)
ton t
Example 9-1
What CHW flow will a 100 ton chiller handle for a 12°F rise in water tempera-
u
ture and a 8°F drop in tower water?
Solution First, determine the flow in the evaporator (chiller flow):

ro
Flow q gpm ---------- = 24 ------

ton t
100 tons 24
gpm = --------------------------------- = 200 gpm
12 F
G
Next, find the flow in the condenser:
gpm 28.8
Flow q ---------- = ----------
ton t
gpm = 100
tons 28.8
------------------------------------- = 360 gpm
8 F
Using these formulas, if you revisit AHRI rating points, you will find that
the evaporator is rated for a 10°F t with entering water temperature of 54°F.
Similarly for the condenser, a temperature difference of 9.6°F is calculated,
leaving a leaving condenser water temperature of 94.6°F. The temperature dif-
ference is often rounded to 10°F t.
Refrigeration Power
t
The theoretical power required by a water chiller compressor is a function of
e
the refrigerant flow rate and the change in enthalpy during compression.
The heat absorbed in the evaporator is
Q = W RE Btu/min
gp
where
RE = refrigerating effect
= heat absorbed in evaporator
= (h2 – h1) Btu/lb of refrigerant (see Figure 9-3)
Q = heat absorbed in the evaporator, Btu/min
W = weight rate of refrigerant flow, lb/min
pE
or
tons 200
W = ------------------------- lb/min
h2 – h1
The theoretical compressor power (THP) required is

u
h2 – h1 W
THP = --------------------------------
-HP
42.4
ro
The compressor brake horsepower (BHP) required is
THP
BHP = -------------------------------------------------- HP
overall efficiency*
* Efficiency from manufacturer’s data.

G
Chiller Types and Control

There are several types of chillers available for air conditioning: centrifu-
gal, screw, scroll, and reciprocating. A general guideline for choosing the type
depending on the chiller’s tonnage size is shown in Figure 9-6. Figure 9-7
shows a cross section of a centrifugal compressor. A section of a single-screw
refrigeration compressor is shown in Figure 9-8, and its compression process

sequence is shown in Figure 9-9. Figure 9-10 shows the components of a spe-
cific type of twin-screw compressor. The scroll compression process is shown
in Figure 9-11, and its components are shown in Figure 9-12. Figure 9-13
shows a reciprocating compressor refrigeration system.
e t
gp
pE
Figure 9-6 General guideline of the types of chillers available for air conditioning.
Source: ASHRAE Handbook—Refrigeration (2014).
u
ro
G
Figure 9-7 Centrifugal compressor cross section.

e t
Figure 9-8
gp
Section of single-screw refrigeration compressor.
u pE
Figure 9-9 Sequence of compression process in single-screw compressor.
ro
Absorption refrigeration machines are available for water chilling in capaci-

ties of 3 to 5 tons (ammonia), 3 to 30 tons (lithium bromide), and 50 to 1500 tons
(lithium bromide) (see Figures 9-14 and 9-15).
G
Reciprocating compressors are used in smaller systems of up to

100 tons. An interesting discussion is presented in the 2012 ASHRAE
Handbook—HVAC Systems and Equipment about the performance charac-
teristics of reciprocating compressors compared to those of centrifugal and
screw compressors (see Figure 9-16) (ASHRAE 2012). A distinguishing
feature of the reciprocating compressor is its pressure rise versus capacity
characteristic.
e t
gp
u pE
Figure 9-10 Vertical, discharge-cooled, hermetic twin-screw compressor.
Pressure rise has only a slight influence on the volume flow rate of the
ro
compressor; therefore, a reciprocating liquid chiller retains nearly full cooling

capacity even on days above design wet bulb. It is well suited for air-cooled
condenser applications and low-temperature refrigeration. Methods of capacity
control are furnished by
• unloading compressor cylinders,
G
• on/off cycling of compressors,

• hot-gas bypass,
• compressor speed control, and
• a combination of the above.
Figure 9-17 shows the relationship between system demand and com-
pressor performance with three equal steps of cylinder unloading. This illus-
trates the relationship between the system demand and performance of a
e t
gp
u pE
Figure 9-11 Scroll compression process.
Source: Purvis (1987).
ro
compressor with three steps of unloading. As cooling load drops to the left of
fully loaded compressor line A (Figure 9-17), compressor capacity is
reduced to that shown by line B, which produces the required refrigerant
flow. Because cooling load varies continuously, whereas machine capacity is
available in fixed increments, some compressor on/off cycling or successive
G
loading and unloading of cylinders is required to maintain fairly constant liq-

uid temperature. In practice, a good control system minimizes load/unload or
on/off cycling frequency while maintaining satisfactory temperature control.
Two basic piping arrangements for multiple chiller systems are parallel
and series CHW flow (as described in the 2014 ASHRAE Handbook—Refrig-
eration). In the parallel arrangement, the water to be chilled is divided among
the liquid chillers and combined again in a common header after chilling (see
e t
gp
u pE
Figure 9-12 Bearings and Other Components of Scroll Compressor
Source: Elson et al. (1990).
ro
Figure 9-18). As the cooling load decreases, one unit may be shut down, but
the remaining units must then provide colder-than-design CHW so that when
the streams combine, the design water supply is provided. Usually the idling
chiller’s pump is shut down when the chiller is stopped and a check valve
closes in this pump’s discharge to prevent a bypass flow.
G
In the case of water chilling designs above 45°F, all units should be controlled
by the combined leaving water temperature or by the return water temperature,
because overchilling will not cause a dangerously low water temperature in the
operating machine. In the case of water-chilling designs below 45°F (see
Figure 9-19), each machine should be controlled by its own CHW temperature,
both to prevent dangerously low evaporator temperatures and to avoid frequent
shutdowns by the low-temperature cutout.
e t
Figure 9-13
gp
Reciprocating compressor refrigeration system.
u pE
ro
G
Figure 9-14 Two-shell lithium bromide cycle water chiller.

e t
gp
pE
Figure 9-15 Schematic of double-effect, direct-fired absorption chiller with reverse parallel
flow cycle.
u
ro
G
Figure 9-16 Comparison of single-stage centrifugal, reciprocating, and screw compressor

performance.
e t
gp
pE
Figure 9-17 Reciprocating liquid chiller performance with three equal steps of unloading.
u
ro
G
Figure 9-18 Parallel arrangement: water to be chilled is divided among liquid chillers and
combined again in common header after chilling.
e t
gp
Figure 9-19 Series arrangement.
pE
In this case, the temperature differential setting of the return water tempera-
ture must be carefully adjusted to prevent short cycling caused by the step increase
in CHW temperature when one chiller is cycled off. No overchilling is required
and compressor power consumption is lower than it is for the parallel method at
part loads. In the series method, a valved piping bypass is suggested around the
lead and lag chillers to facilitate future servicing. This piping design philosophy
also applies to parallel chillers; it again gives the maintenance staff an opportunity
for service without a complete shutdown.
u
Chiller Piping Arrangements
In general, we see multiple chiller designs applied in parallel instead of series
because this permits adding chillers in the future due to single or multiple
ro
building additions, such as for campus-type site plans. The designer must allow
space in the chiller plant for the additions and the distribution mains that must
be sized for the future flows (see Figure 9-20).
Until the mid 1990s, chiller controllers were mainly pneumatic or electric
proportional operation, with switched interlocks. To protect the operating com-
ponents from freezing due to load changes, chillers were operated with con-
G
stant water flow, typically represented by the three-way valve. These

applications usually used three-way valves on unit terminals to permit contin-
ual flow in the chiller (see Figure 9-21) and to permit part-load terminal control
by shifting coil flow to bypass flow around the coil. The result is that the CHW
pumping power was theoretically constant and did not reduce with load. As the
size of the system increased, the distribution system pumping increased in size,
pumping a constant volume, and was not energy efficient.
e t
gp
pE
Figure 9-20 Chillers piped in parallel in primary production loop.
u
ro
G
Figure 9-21 Schematic showing constant- and variable-flow arrangements.

During the first energy crisis era of the early 1970s, application of the two-
way valve was given new opportunity to reduce the distribution pumping flow
rate with the load in large campus-oriented systems. Popular primary/secondary
pumping systems were installed with greater vigor, keeping flow to the chillers
constant through the primary pump while distribution flow to the system was
made variable through the secondary pumps by the closing of the valves riding
the pump curve for reduced horsepower. System flow is controlled through
engineering decisions that allow the valves to perform properly, reducing excess
t
flow due to high valve differential pressures by staging parallel pumps and judi-
cious pump selection and layout.
e
In the late 1980s, speed drives became cost-effective enough for large sys-
tems to apply variable-speed pumping control. The chillers continued to be piped
in parallel in a primary production loop (see Figure 9-20) with a common bridge
gp
to hydraulically decouple the chiller pumps from the distribution pumping. Dis-
tribution pumping was kept at constant speed and constant flow to maintain the
integrity of the chiller controller. Chiller staging was determined from recircula-
tion and at part-load if chillers should be shut down. The location of the common
bridge determines how the chillers will be loaded or unloaded (Coad 2011).
Since the 1980s, several simultaneous advances in the industry have
occurred, shifting system design thoughts. Direct digital control (DDC)
replaced pneumatic and electric proportional controls, ushering in new capabil-
pE
ities in advanced control strategies and applied monitoring of system operating
data. Adjustable-speed drive prices became much more economical due to ris-
ing energy costs and manufacturing economies of scale. In this era a renewed
focus on energy efficiency took hold, and with it new capabilities were added
to chillers and sources in general, allowing for better control at partial loads,
which allowed for variable flow to the source. With measurable data, known
temperature operating problems in the primary/secondary system occurring
from primary and secondary mixing are now catching the attention of the
u
industry.
Some of the difficulties can be seen in Figure 9-22, which represents a sys-
tem consisting of three chillers. Each chiller is fed by a constant-speed chiller
pump (with a check valve) that operates only when the chiller is on. Each
ro
chiller pump is sized to achieve a t of 16°F, and the chiller is controlled by a

discharge thermostat at 42°F.
With the common bridge located between the load and the production sections
(see Figure 9-23) and the system operating at part load (1200 gpm, 800 ton load)
due to the hydraulics inherent in the design, chillers 2 and 3 will load proportion-
ally and chiller 1 will shut down. Chillers 2 and 3 receive the same-temperature
G
water from the return main. The chillers load in proportion to the ratio of their
flow rates to the total load flow. Chiller 2 loads to 320 tons and chiller 3 loads to
480 tons (each is 64% of their full load output, because the load is 800/1250 =
64% of the combined chiller flow rate). Because the chiller pumps are constant
speed, chiller 2 delivers 750 gpm and chiller 3 delivers 1125 gpm, for a total of
1875 gpm. Because the load is calling for 1200 gpm, 675 gpm must flow in the
common bridge.
e t
gp
Figure 9-22 Three chillers in parallel, each with a dedicated constant-speed pump.
u pE
ro
Figure 9-23 Parallel arrangement with common bridge between load and production sections.
Mixing occurs at the return tee, creating a temperature of

G
675 42 F + 1200 58 F
-------------------------------------------------------------------------- = 52.24 F
1875
The load on chiller 2 is
750 52.24 F – 42 F
tons = ---------------------------------------------------------- = 320 tons
24
and on chiller 3 is
1125 52.24 F – 42 F
tons = ------------------------------------------------------------- = 480 tons
24
If the common bridge is located on the opposite end of the production sec-
tion (see Figure 9-24), the chillers will load unevenly due to the hydraulics.
Looking at this part-load example, chiller 3 loads fully:
t
58 F – 42 F
tons = 1125
e
------------------------------------------------------ = 750 tons
24
Chiller 2 receives 75 gpm (1200 to 1125) from the return main, which is
gp
mixed with 675 gpm (750 to 75) from the common bridge at 42°F. This water
enters chiller 2 at 43.6°F:
75 58 F + 675 42 F - = 43.6 F
-------------------------------------------------------------------
750
The load on chiller 2 is only

pE
750 43.6 F – 42 F
tons = ------------------------------------------------------- = 50 tons
24
u
ro
G
Figure 9-24 Parallel arrangement with common bridge at opposite end of production section.
Chiller 1 shuts down due to the 42°F return water. So you can see the
uneven loading that can occur and why the designer should consider the com-
mon bridge between the load section and the production section.
What is the best way to handle this? This is an introductory course, so
know that there are exceptionally sophisticated models and great debate about
system design and optimization, and there will be many fast-moving changes
in the industry that this text cannot keep up with. But the fundamentals of sys-
tem design are established here, and even in new and sophisticated plant
t
design, all of the techniques discussed are quite applicable.
Sources, and particularly chillers, still have some limits to their flexibility
e
with respect to operation. Where previous designs required constant flow,
today’s systems allow variable fluid flow, although it tends to be in a limited
range. In many cases, chiller controls are able to control with chiller tube
gp
velocities down to a range of 1 to 2 fps; however, concerns over tube fouling
often increase this low limit to around 3 fps to reduce the possibility of fouling.
Maximum velocity can often be stated as high as 10 fps; however, this is for
limited hours of operation. Chiller heat exchanger tubes are sophisticated
pieces of manufacturing engineering, often being rifled on the interior and
exterior to increase the heat transfer surface area. The result is great perfor-
mance; the proviso is that while 10 fps tube velocity for limited time periods is
acceptable, a more common operating limit of 8 fps is recommended to reduce
pE
tube erosion, performance degradation, and system maintenance. Still, if
chiller tubes are selected and operated in a range of 3 to 8 fps, there is a serious
capability to reduce energy consumption in the pumping system and the chiller
operating energy. A point will be reached where minimum flow must be main-
tained for the chiller to operate properly.
The simplest implementation to create a more efficient system, then, is
to consider the benefits of the primary/secondary system with the new tech-
nological twist, making it variable-speed primary/variable-speed secondary.
u
Primary/secondary pumping techniques do increase the cost of new systems
slightly because there are more pumps. On the other hand, primary/second-
ary techniques allow for the designer to manage system constraints due to
loads and hydraulics normally attributed to adapting the HVAC system to
ro
the facility to be controlled better than any other options. As previously

mentioned, primary/secondary systems allow various system components to
be hydraulically isolated from each other, and controlling system pressure
differences is the first step in attaining flow control performance at part load
that allows the chillers or other sources to operate at reduced loads and
energy consumption. There are a lot of plain systems that may not require
G
very sophisticated techniques; on the other hand it is often the case that the
larger systems become, the more techniques such as primary/secondary
become important tools in a designer’s tool box.
These are particularly noticeable whenever there is a possibility that system
size will be expanded. As seen in Figure 9-25, the primary/secondary allows
for simple addition of extra chillers and loads through proper up-front plan-
ning. Another option to this implementation is to manifold the pumping system
e t
gp
pE
Figure 9-25 Primary/secondary allows for simple addition of extra chillers and loads.
to the sources, giving additional flexibility for maintenance and failure opera-
tion modes. Often, pumps have the capability to provide more system flow, and
u
in the manifolded system the benefits of parallel pumping can be taken advan-
tage of, as diagrammed in Figure 9-26.
In these applications, the designer should determine the sequence of opera-
ro
tion as a combined temperature control and hydraulic problem as the staging of

chillers and pumps is considered. Properly maintained (and selected) hydraulic
relationships allow for zone controls to act independently and for the source
system to react to the loads as indicated by the demanded flow from the system.
Not all control points are diagrammed. Temperatures are shown in Figure 9-25
G
to specifically monitor and control against system degradation due to secondary

remix and resupply back to the system. Varying of chiller flow and minimum
flow control can be handled by the addition of flow, temperature, differential
pressure data, and closed-loop feedback control strategies, or open-loop control
tied in with the chiller control panel.
In smaller systems of one to two chillers, or systems where expansion will
not be required, the variable primary system is also an option.
e t
gp
pE
Figure 9-26 Pumping system manifolded to sources gives flexibility for maintenance and fail-
ure operation modes as well as capability to provide more system flow.
Dedicated operating pumps for the chillers are eliminated, and the distribu-
tion pumps are selected for a larger head to account for the chillers. The com-
mon bridge is maintained in the system, with a bypass control valve to ensure
u
that there is the capability to maintain the minimum flow required by the chiller.
There are several ways that minimum flow and chiller staging can be accom-
plished, either through addition to the diagrammed schematic of flow or differ-
ential pressure sensors. It is also recommended that minimal chiller loading be
ro
monitored in the control package to properly sequence pumps and chillers.
Chiller Energy Performance

The commonly applied evaluation of a chiller’s performance is kilowatt of
energy per ton of refrigeration (kW/ton). Plotted against the percentage of
G
design capacity of the chiller, as shown in Figure 9-27, which shows a real-
world example, this curve shows the improvement in efficiency that a particu-
lar manufacturer accomplished in redesign to make the chiller more energy
efficient over time. Older chiller systems have a challenge operating at less
than 1 kW/ton. Manufactured equipment today has embraced the application of
variable-speed drives on compressors and sophisticated control strategies, in
part because of competition and in part due to legislated requirements for
e t
gp
pE
Figure 9-27 Chiller performance improvement versus percent load.
u
energy efficiency improvement. Today’s chillers now normally achieve effi-
ciencies on the order of 0.56 kW/ton and, depending on their sophistication of
application, can operate as low as 0.2 kW/ton at reduced loads.
In a simple form, Figure 9-28 (Mannion 1988) shows a control and design
ro
strategy for higher ts, various control valves, pumping methods, and chiller
staging options for a hypothetical design.
These options always require a great deal of study and due diligence. It is not
at all uncommon to see detailed and complex modeling and mapping of the total
system cost and operating expense to determine the lowest life-cycle cost for sys-
tem selection. As one might imagine, the competition among those desiring this
G
type of work can be quite ferocious and lead to many claims and counterclaims. It
is also important while performing due diligence to be aware of proprietary fea-
tures and design strategies that may require the designer to consider patent licens-
ing. It is not at all uncommon to find specific operating strategies and control
sequences granted patent protection. What might seem like a well-meaning edu-
cational article in a trade journal about what might be considered common prac-
tice or a new idea could often be a veiled advertisement for a protected concept.
e t
gp
pE
Figure 9-28 Chiller, control, and pumping alternatives versus design temperature rises.
Source: Figure 10, Mannion (1988).
Energy performance of a CHW system should not be limited to the most

u
efficient chillers, piping, valves, and variable-speed pumping arrangements.
There is a need for skilled, trained operators working with some form of DDC
system to optimize the overall chiller system by optimizing the following (Cas-
cia 1988):
ro
• The staging on or off of multiple chillers

• Condenser water temperature and cooling-tower operation
• CHW temperature when variable-speed CHW pumps are used
Resetting CHW supply temperature upward under lighter load conditions

G
saves energy at the chiller due to lower refrigerant head requirements of the
compressor. This reset can be done through cascade control of the supply
influenced by the return water temperature or incorporation of outdoor temper-
ature reset (or both). However, increasing the CHW temperature can cause the
variable-speed pumps to increase in speed due to the two-way valves opening
to satisfy the load. How much that happens depends on the speed of response
of the system and how far the reset influence extends into the system. For
instance, under lighter loads, the designer might also have the zone controller
setpoint reset slightly higher (1°F to 2°F), which is barely noticeable to the
occupant, or might apply that criteria as a high load day as occupancy begins
to come to a close. Figure 9-29 shows the relationship between chiller power
consumption and the variable-speed pump power consumption and how the
optimal CHW temperature can be found.
Thermal Storage
t
Another important design consideration is to provide a thermal storage facility
e
as part of the CHW system to reduce construction costs and operating costs.
The principles of thermal storage can be applied with the primary/secondary
chiller piping and variable-speed pumping discussed.
It is also suggested that ASHRAE publications such as Cool Storage Mod-
gp
eling and Design (1989) and “Achieving Energy Conservation with Ice-Based
Thermal Storage” (Brady 1994) be studied for further design considerations.
Overall improvement in water chiller performance and selection for a particu-
lar design requires careful study by the designer, not only in various chiller
manufacturers’ performance data, pumping and piping design arrangements,
control valve selections, and pump speed control, but also in how they perform
together from part load to full load.
u pE
ro
G
Figure 9-29 Relationship between chiller power consumption and variable-speed pump
power consumption.
An often overlooked thermal storage concept that requires little if any addi-
tional cost is application of the building mass, or fabric, as the storage vessel.
Thermal storage often does not save energy in the form of kilowatt-hours. It does,
however, reduce energy cost by shifting use from peak electrical demand points to
off peak and, in the case of a chiller, a usually lighter operating load condition
(e.g., hottest time of the day with solar loading to night and about 20°F cooler). In
the building mass concept, the idea is to run the chiller and HVAC plant off peak
(usually at night) to precool the building and lower the mass temperature. The dis-
t
charge of the stored energy is controlled by manipulation of the zone controller’s
setpoint, operating cooler in the morning and slowly raising the zone setpoint
e
going into the loaded portion of the day (afternoon). In the facility that the author
is familiar with and where this was implemented, occupants were made aware of
the strategy and told that if they were too cool in the morning they could wear a
sweater. The facility was built with two side-by-side, equal but separate buildings
gp
that shared a connecting interstitial space. To reduce costs, no excess or safety
capacity was installed in the CHW plant. Changing facility uses and requirements
made the owner consider installation of a full-sized backup CHW plant to be
shared by the two buildings. The ultimate result of the storage implementation
was that during an extended one-week period when design conditions were
exceeded by greater than 20°F, the tested building was able to operate at peak
operating load with 75% of the plant’s installed capacity while maintaining occu-
pE
pancy setpoint. The owner was able to revise the backup plan to a simpler piping
cross-connection so that each individual operating plant could back up the other in
case of required backup.
Figure 9-30 shows a more typical example of a thermal storage concept
using an ice builder with a CHW system. This is often the way most designers
think of thermal storage, and it too helps reduce installed system costs and
operating expense. Thermal storage, regardless of the style employed, should
always be considered as part of the base system design.
u
Summary
ro
• Identification of a water chiller’s capability to reduce water temperature,

dependent on the tonnage rating and water flow rate
• Temperature rise in water temperature across a CHW coil
• Temperature drop in condenser water temperature
G
• The types of refrigeration compressors used in chillers, depending on the

manufacturer and the tonnage size
• That the theoretical horsepower used in a chiller is directly related to the
heat (enthalpy) absorbed in the evaporator, the weight rate of refrigerant
flow, the enthalpy change in the compressor, and the chiller tonnage
• That chillers can be piped in series or parallel but are commonly in parallel
to provide for expansion
e t
gp
pE
Figure 9-30 Schematic of thermal storage concept using ice builder with a chilled-water system.
• Piping of multiple chillers, using primary/secondary principles and load-

u
ing chillers evenly when locating the common bridge between the load
and the chillers
• How to optimize chiller operation versus CHW supply temperature, taking
ro
into account the compressor power and the pumping power
References
AHRI. 2011. AHRI Standard 550/590, Standard for Performance Rating of
Water-Chilling and Heat Pump Water-Heating Packages Using the Vapor
G
Compression Cycle. Arlington, VA: Air-Conditioning, Heating, and Refrig-

eration Institute.
ASHRAE. 1989. Cool Storage Modeling and Design. Technical Data Bulletin.
Atlanta: ASHRAE.
Atlanta: ASHRAE.
ASHRAE. 2014. ASHRAE Handbook—Refrigeration. Atlanta: ASHRAE.

Brady, T. 1994. Achieving energy conservation with ice-based thermal storage.
ASHRAE Transactions 100(1):1735–45.
Cascia, M. 1988. Optimizing chiller plant energy savings using adaptive DDC
algorithms. ASHRAE Transactions 94(2):1937–46.
Coad, W. 2011. Hydronic Systems. Technical data bulletin. Atlanta: ASHRAE.
Elson, J., G. Hundy, and K. Monnier. 1990. Scroll compressor design and
application characteristics for air conditioning, heat pump, and refrigera-
t
tion applications. Proceedings of the Institute of Refrigeration, 2.1–2.10.
Mannion, G. 1988. High temperature rise piping design for variable volume
e
systems: Key to chiller energy management. ASHRAE Transactions
94(2):1424–43.
Purvis, E. 1987. Scroll compressor technology. Heat Pump Conference, New
Orleans.
gp
Trane. 2002. Air Conditioning Manual. Atlanta: Trane Company.
u pE
ro
G

9-1 What pumping flow is required for a 400 ton CHW system using 20% ethylene
glycol and water at 44°F supply and 12°F rise? (Per Table 6 of Chapter 31 of
the 2013 Handbook—Fundamentals, the density of a 20% ethylene glycol/
t
water solution at 44°F is 64.6 lb/ft3.)
e
9-2 What is the pumping flow in Exercise 9-1, except using propylene glycol and
water at the same freezing conditions? (Per Table 9 of Chapter 29 of the 2013
ASHRAE Handbook—Fundamentals, the density of a 20% propylene glycol/
gp
9-3 Estimate the volumetric flow rate of condensing water to be pumped for the
condenser of an R-22 water-cooled unit operating at a condensing temperature
of 110°F and an evaporating temperature of 40°F with a 10°F liquid subcooling
and 10°F suction superheat. Water enters the condenser at 85°F and exits at
95°F; the load is 400 tons (use the figure below).
u pE
ro
G
9-4 In Figure 9-17, at what load does chiller 1 shut down?
9-5 The capacity of a chiller is dependent on what two basic load factors?
9-6 What are three advantages of a DDC system for a multiple-chiller plant?
9-7 What type of control valves should be used with variable-speed CHW pumps?
9-8 Name several methods that manufacturers furnish for refrigeration capacity
control.
9-9 Name at least four types of chillers.
e t
gp
u pE
ro
G
G
ro
upE
gp
et
Skill Development
Exercises
To receive full continuing education credit, all questions must be answered and
submitted at www.ashrae.org/sdlonline. Please log in using your student ID
t
number and the SDL number. Your student ID number is composed of the last
five digits of your Social Security Number or another unique five-digit number
e
you create when first registering online. The SDL number for this course can
be located near the top of the copyright page of this book.
gp
u pE
ro
G
G
ro
upE
gp
et
Fundamentals of Water System Design I-P
Chapter 1 Skill Development Exercises

Total number of questions: 11
1-1 Water systems that convey heat to or from a conditioned space or process with
t
hot or chilled water are frequently called _____________________________.
e
1-2 What is the fundamental difference between closed and open types of water
systems?
1-3
gp
A cooling tower has at least two points of interface with air. Where are they?
pE
1-4 What is the maximum working pressure for LTW boiler systems?
u
ro
1-5 What is a CHW system? How is it different from a CW system?

G
1-6 What are the fundamental components of a closed hydronic system?

Skill Development Exercises Chapter 1
1-7 Explain the most common source devices for heating and cooling systems.
t
1-8 Explain what load means.
e
gp
1-9 What factors influence the heating and cooling load requirements?
pE
1-10 Define sensible heat transfer.
u
1-11 Name five heating load devices and describe how each is used in system appli-
ro
cations.
G

2-1 What causes unbalanced flow rates in direct-return piping arrangements?
e t
gp
2-2 What is the most often used equation that relates to pressure drop?
pE
2-3 Fluids can flow through a pipe under two different conditions. Name them.
Explain the differences between these flow conditions.
u
ro
2-4 Explain friction factor.

G
2-5 In commercial installations, it is suggested that _____% should be added to the

friction loss to allow for aging.
e t
2-6 What is the Bernoulli principle?
2-7
gp
What factors determine pressure drop in piping?
pE
2-8 What methods allow thermal expansion?
u
ro
2-9 What is the minimum distance upstream and downstream for a water flow-
measuring device (in pipe diameters)?
G

3-1 What US organizations issue codes and standards for piping systems and com-
t
ponents?
e
gp
3-2 What is the allowable working pressure (ASTM A53B to 4000°F) for each of
the following steel pipe diameters:
Nominal Size and Pipe OD Schedule Working Pressure, psig

pE
10 in., D = 10.75 40XS
20 in., D = 20 30XS
4 in., D = 4.5 40ST
8 in., D = 8.625 80XS
3-3
u
List three joining methods used with copper tubing.
ro
3-4 List five methods of joining metal pipe.

G
3-5 Name and briefly describe the two main categories of plastic piping materials,
and list at least three of each type of plastic pipe.
e t
3-6 List the ASME B31 HDS (psi at 73°F) for each of the following plastic pipe
materials:
Material ASME B31 Hydrostatic Design Stress (psi at 73°F)
gp
PB 2110, Type II, Gl
ABS 1210 TI, G2
3-07 List the five methods of corrosion control.

pE
3-8 List the k values for each of the following screwed pipe fittings:
u
Nominal Pipe Diameter, in. Fitting Type k Value
1.25 90° ell long
0.5 Globe valve
ro
3 Tee branch
3-9 What is the function of a backflow prevention device?

G
3-10

Determine the pipe sizing and total PD for the piping system shown below. You
are given that the system consists of:
• 1 gate valve: 2 in.
• 1 control valve: 2 in. rated at 35 gpm at 9 ft PD
• 1 balance valve: 2 in. rated at 40 gpm at 2.0 ft PD
t
• 2 tee branches
• 4 elbows
e
Total pipe length = 200 ft (schedule 40 screwed pipe)
gp
u pE
ro
Diagram for exercise 3-10.

G
G
ro
upE
gp
et

4-1 List three factors that influence the type of pump selected for a particular appli-
t
cation.
e
gp
4-2 Label the components of the centrifugal pump shown below. Describe the
function of each component.
pE
4-3 The suction flange gage of a pump reads 10 psi. If the temperature of water being
pumped is 220°F and the atmospheric pressure is 14.7 psia, what is the available
NPSH? At what temperature will the pump cavitate? (Water at 220°F vaporizes at
17.2 psia, per ASHRAE Handbook—Fundamentals [2013], Table 3, p. 6.10.)
u
ro
G
Centrifugal Pump.
4-4 What is the NPSH on the inlet to a 2 hp pump rated at 140 gpm at 32 ft for a
cooling tower application? The centerline of a pump inlet is to be 2.5 ft below
the tower sump water surface; assume tower water at 120°F and piping equiv-
alent to 60 ft of 2 in. pipe on pump suction. Assume atmospheric pressure is
14.7 psia or 34.0 ft; assume friction head in 2 in. pipe is 20 ft (according to
Figure 2-15, 140 gpm flow in 2 in. pipe yields 33.2 ft/100 ft of pipe: 60/100 ×
33.2 = 19.92, or 20 ft) and vapor pressure of water at 120°F is 1.69 psia (per
ASHRAE Handbook—Fundamentals, Table 3, Chapter 1) or 3.9 ft abs (1.69 psia
t
× 34 ft/14.7 psi = 3.9 ft abs). Pump curve shows 10 ft NPSHR.
e
gp
4-5 What is radial thrust?
pE
4-6 Explain what pump cavitation is and how it can be avoided.
u
ro
4-7 Write the NPSHA formula for a proposed design and explain what each vari-
able represents.
G
4-8 Write the pump affinity laws and explain how they are applied.

t
4-9 Speed Change Diameter Change:
e
Flow:
gp
Head:
Horsepower:
pE
4-10 Explain how to determine the horsepower for a centrifugal pump.
u
ro
4-11 How does the capacity of a centrifugal pump change?

G
4-12 Name six types of centrifugal pumps and their mounting arrangements.
t
4-13 A pump is rated at 500 gpm at 60 ft of head. What are the flow and head if the
e
impeller size is changed to 85% of its original diameter? Assume there is no
static head.
gp
u pE
ro
G

5-1 A typical fan-coil terminal requires 4 gpm. What valve Cv should be selected if
t
a control valve is specified for a 9 ft drop?
e
gp
5-2 What piping components should be specified at the terminal?
pE
5-3 What type of control action should be considered to reduce discharge tempera-
ture cycle?
u
ro
5-4 A control valve is to be selected for a 44 gpm terminal coil; coil drop is 18 ft.
Select the correct size of control valve if the valve is specified for 50% of the
coil drop, and the typical valve sizes and Cv available (Cv is in parentheses) are
G
• 0.5 in. (2.5), 0.75 in. (6), 1 in. (10)

• 1.25 in. (16), 1.5 in. (20), 2 in. (36)
5-5 What control valve flow characteristic should be specified for proportional
control of a hot water heating control?
e t
5-6 An on/off thermostat controls a cabinet heater in a hallway. What valve flow
characteristic should be specified?
5-7
gp
A three-way valve is to be applied to a refrigeration condenser and cooling
pE
tower to maintain a 95°F condensing temperature. What type of three-way
valve arrangement should be applied?
u
5-8 It is desirable to control flow in a chilled-water coil down to a minimum of 5%

of design flow before close off. In addition to proper valve sizing for design
ro
flow capacity and proportional control, what else should be specified?

G
5-9 What should be specified in the bypass circuit of a three-way valve?

5-10 What type of control method varies airflow through a terminal coil?

t
5-11 Explain the difference between primary and secondary pumping systems.
e
gp
5-12 Define valve authority, rangeability, and selection.
pE
5-13 How is terminal control valve size selected?
u
ro
5-14 What are the three types of terminal control action?

G
G
ro
upE
gp
et

6-1 What maintains the maximum and minimum pressure limits of a hydronic system?
e t
gp
6-2 What must a closed water system have to permit the expansion and contraction
of the water volume?
pE
6-3 What should be specified for terminal coil returns and high points in the piping
system to provide service for air in the hydronic system?
u
6-4 Where does air come from in a hydronic system?

ro
6-5
G
What routine tasks should be performed by the building maintenance staff to

the hydronic systems prior to the heating and the cooling seasons?
6-6 How much air can be present in water at 120°F and 30 psig?
6-7
t
A steel pipe system has 1000 gal total volume and will operate periodically in
the cooling season with 40°F water, but when it is off it will reach 100°F ambi-
ent; minimum pressure is 10 psig (24.7 psia) and maximum is 25 psig (39.7
e
psia). What size diaphragm expansion tank is required?
gp
6-8 A pump is selected for 100 gpm at 50 ft head, the system has 20 psig static pres-
sure when off, and the expansion tank is improperly located on the pump dis-
charge. What will the gages read on the pump suction and discharge when the
pump is started? What will happen at the float-type air vents in the boiler room?
pE
6-9
u
Explain what factors are needed to size and select an expansion tank.
ro
6-10 What are the differences between open and closed systems?
G
6-11 What are the three types of expansion tanks?

t
6-12 Give three examples of where expansion tanks are used.
e
gp
6-13 Why is it important to have the correct size expansion tank?
u pE
ro
G
G
ro
upE
gp
et

7-1 In the direct-return example (see Figure 7-8), which unit’s piping path dictates
t
e
gp
pE
7-3 What size pump capacity and head would be required in Figure 7-8 (direct
return for supplying four identical floors), assuming 10 ft between floors and
allowing a 25% head drop in the main for fittings and a 20 ft drop across the
u
chiller?
ro
7-4 In the reverse-return example in Figure 7-9, which unit’s piping path dictates
G

e t
7-6 What size pump capacity and head would be required in Figure 7-9, assuming
10 ft between floors and allowing a 25% head drop in the main for fittings and
a 20 ft drop across the chiller?
gp
pE
7-7 In the primary/secondary examples in Figures 7-10 and 7-11, which unit’s pip-
ing path dictates the pump head from A to F? How much head is required?
u
7-8 What size distribution pump capacity and head would be required to handle the
four AHUs in the Figure 7-11 primary/secondary example for the conditions
ro
shown from A to F?
G
7-9 What size distribution pump capacity and head would be required for supply-
ing four identical floors (similar to the Figure 7-11 primary/secondary exam-

ple), assuming 10 ft between floors and allowing a 25% head drop in the main
for fittings and a 20 ft drop across the chiller?
e t
7-10 What is the cooling capacity of 100 gpm of water flow with 50% propylene-
glycol mixture at an average temperature of 50°F and a 10°F rise?
gp
pE
7-11 What is the cooling capacity of 100 gpm of water flow at a 10°F rise? How
many tons of cooling?
u
ro
7-12 What is the pumping horsepower for the propylene-glycol mixture (specific
gravity = 1.05) in Exercise 7-10, compared to the plain water in Exercise 7-11,
if the coil pressure drop is 20 ft, assuming a pump efficiency of 75% and motor
efficiency of 85%?
G
7-13 How can the possibility of frozen coils and piping in an HVAC system be
reduced?
e t
7-14 What are two advantages of primary and secondary hydronic systems?
7-15
gp
What type of pump curves should be selected if two-way valves are to be used?
pE
7-16 How can flow be determined in a given loop?
u
ro
7-17 Which piping system design gives more uniform pressure drop? Why?
G

8-1 Assume two pumps (6 × 8 × 9.5 in., see Figure 8-14) are to be applied in paral-
t
lel for a 70 ft head system. What flow will result at 70 ft?
e
gp
8-2 Assume the same two pumps (6 × 8 × 9.5 in., with a 9.6 in. impeller) are to be
piped in series. What will the resulting head be at 1600 gpm?
pE
8-3 Assume the 6 × 8 × 9.5 in. pump is being considered for variable-speed
operation of a system designed at 1600 gpm at 70 ft head and the pump
manufacturer advised that the NEMA type B electric motor to be furnished
u
can be run down to 40% of its furnished speed (1750 rpm). What is the min-
imum flow and head the pump can be run at following the same system
curve?
ro
G
8-4 In Exercise 8-3, what is the theoretical brake horsepower reduction if the pump
is run at 40% speed?
8-5 What methods are used to provide flow to buildings that are remote from a
chiller plant?
e t
8-6 What does the pump controller measure to control pump speed?
8-7
gp
Explain how combinations of two-speed pumps can be used to provide required
head and flow capacity with improved efficiency at part-load conditions.
pE
8-8 What is a good way to provide emergency flow in case of a pump failure?
u
ro
8-9 What methods are used to match pump operation to a system?

G

9-1 What pumping flow is required for a 400 ton CHW system using 20% ethylene
t
glycol and water at 44°F supply and 12°F rise? (Per Table 6 of Chapter 31 of
the 2013 Handbook—Fundamentals, the density of a 20% ethylene glycol/
e
gp
9-2 What is the pumping flow in Exercise 9-1, except using propylene glycol and
water at the same freezing conditions? (Per Table 9 of Chapter 29 of the 2013
ASHRAE Handbook—Fundamentals, the density of a 20% propylene glycol/
pE
9-3 Estimate the volumetric flow rate of condensing water to be pumped for the
condenser of an R-22 water-cooled unit operating at a condensing temperature
u
of 110°F and an evaporating temperature of 40°F with a 10°F liquid subcooling
and 10°F suction superheat. Water enters the condenser at 85°F and exits at
95°F; the load is 400 tons (use the figure below).
ro
9-4
G
In Figure 9-17, at what load does chiller 1 shut down?

e t
gp
Figure 9-1 Diagram for Exercise 9-3.
9-5 The capacity of a chiller is dependent on what two basic load factors?
pE
9-6 What are three advantages of a DDC system for a multiple-chiller plant?
u
ro
9-7 What type of control valves should be used with variable-speed CHW pumps?
G
9-8 Name several methods that manufacturers furnish for refrigeration capacity
control.

e t
9-9 Name at least four types of chillers.
gp
u pE
ro
G
G
ro
upE
gp
et
ASHRAE LEARNING INSTITUTE
Self-Directed Learning Course Evaluation Form
Course Title: Fundamentals of Water System Design (I-P Edition), Second Edition
On a scale of 1 to 5, circle the number that corresponds to your feeling about the statements below.
(1 = strongly agree, 5 = strongly disagree, 3 = undecided)
t
Strongly Strongly
Course Content Agree Undecided Disagree
e
1. The objectives of the course were clearly stated. 1 2 3 4 5
2. The course content supported the stated objectives. 1 2 3 4 5
3. The content quality and format of the course material make it valuable as a 1 2 3 4 5
future reference.
gp
4. The quality and clarity of the charts and diagrams enhanced your ability 1 2 3 4 5
to understand the course concepts.
5. The organization of course material supported effective mastery of the topic. 1 2 3 4 5
6. The material presented will be of practical use to you in your work. 1 2 3 4 5
7. The degree of difficulty (level) of this course was correct to meet your 1 2 3 4 5
needs and expectations.
General
pE
1. Which description best characterizes your primary job function?
_____Architect* _____Developer _____Manufacturer _____Sales

_____Code Agency _____Educator/Research _____Marketing _____Specifier
_____Consultant _____Energy Conservation _____Owner _____Student
_____Contractor/Installer _____Facilities Engineer _____Plant Engineer _____Utilities
_____Consumer/User _____Government _____Policy Maker/Regulator
_____Other (please specify) _______________________________________________________________________
u
*Are you a registered architect? ___No ___Yes, AIA Membership Number (required): _____________________
2. Which describes your educational background?

ro
_____High School _____Master's Degree—Engineering

_____Associates Degree/Certificate Program _____Master's Degree—Other Than Engineering
_____Bachelor's Degree—Engineering Technology _____Doctoral Degree—Engineering
_____Bachelor's Degree—Engineering _____Doctoral Degree—Other Than Engineering
_____Bachelor's Degree—Other Than Engineering
_____Other (please specify) _______________________________________________________________________
G
3. Approximately how many hours did it take you to complete this course?
_____10 hours _____20 hours _____30 hours _____40 hours _____Other (please specify)___________
4. What topics would you suggest for future courses? ______________________________________________________________

General comments regarding any aspect of the course, including suggestions for improvement:
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
t
_________________________________________________________________________________________________________
e
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
gp
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
pE
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
u
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
ro
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
_________________________________________________________________________________________________________
G
Name (optional) __________________________________________________________________________________________
Phone (optional) __________________________________________________________________________________________
E-mail (optional) __________________________________________________________________________________________
Return to: ASHRAE, Education Department, 1791 Tullie Circle NE, Atlanta, GA 30329
Fax: 404-321-5478
Flexible and Effective Continuing Education
for HVAC&R Professionals
Key Knowledge for

Successful Hydronic System Design
Water, or hydronic, systems are a key foundation technology of HVAC system design.
t
They represent an important tool in the designer’s toolbox for providing effective and ef-
ficient HVAC systems to buildings around the world.
pe
Fundamentals of Water System Design, Second Edition, is an entry-level text on the principles
of hydronic HVAC system design and calculations. The fundamentals presented in this text
can be used to successfully design the modern hydronic system, and they provide an equally
sound footing for work in hydronic system commissioning, operation, and troubleshooting.
This course book’s nine chapters provide detailed information on the following topics:
•
•
•
•
•
•
•
Eg
Components of closed and open hydronic systems
Pipe materials and fittings
Piping system design
Pumps and their operation
Matching pumps to systems
Control of terminal units
Optimizing water-chiller operation
Each chapter includes detailed diagrams, examples, and calculations, plus exercises at the
up
end of each chapter to develop skills in using the knowledge provided.
This book is part of the ASHRAE Learning Institute (ALI) Fundamentals of HVAC&R Series
of self-directed or group learning courses. For information on topics covered in the series,
visit www.ashrae.org.
1791 Tullie Circle

ro
Atlanta, GA 30329-2305
Telephone: 404/636-8400
Fax: 404/321-5478
E-mail: edu@ashrae.org
www.ashrae.org/ali
G
978-1-936504-66-4
9 781933 74271 7
Product Code: 98020 11/15
SDL_cover_I-P.indd 2 10/22/2015 1:52:53 PM

ASHRAE Fundamentals of Water System PDF

Uploaded by

Document Information

Original Title

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

ASHRAE Fundamentals of Water System PDF

Uploaded by

Copyright:

Available Formats

Fundamentals of

Water System Design

A Course Book for

Includes Skill Development Exercises

© 1996, 2000, 2015 ASHRAE

For course information or to order additional materials, please contact:

ASHRAE Learning Institute Telephone: 404/636-8400

Atlanta, GA 30329 Web: www.ashrae.org/ali

Any updates/errata to this publication will be posted on the

We hope your educational experience is satisfying and successful.

American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

Each course book includes the following:

DDC Controls Standard 90.1 for Architects

Types of Terminals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 139

Open and Closed Water Systems . . . . . . . . . . . . . . . . . . . . . . . . 171

Fundamentals of Water System Design, Second Edition is an entry level

what later became the professional development series. The industry is

❑ know what determines the load;

Given the design requirements, it is the designer’s task to evaluate the

Figure 1-1 Source/load.

Figure 1-3 Source–distribution–part-load.

• Low-temperature water (LTW) system. This hydronic heating system

Figure 1-4 Hydronic system fundamentals (closed system).

• Medium-temperature water (MTW) system. This hydronic heating sys-

Figure 1-5 Cooling tower (open system).

Figure 1-7 Medium- or high-temperature water system.

• Chilled-water (CHW) system. A hydronic cooling system normally oper-

Figure 1-8 Chilled-water system—direct return piping.

Antifreeze or brine solutions may be used for applications (process appli-

Figure 1-11 Distribution orientation.

The design should consider source protection to prevent temperature

Figure 1-12 Closed hydronic system fundamental components.

• Heating-source devices: Hot-water generator or boiler, steam-to-water

• Cooling-source devices: Electric compression chiller (Figure 1-14), thermal

Figure 1-14 Multiple chiller variable-flow chilled-water system.

The reciprocal of the turndown ratio is sometimes used. For example, a

The diversity factor, as a ratio of actual load/design load, can be applied

Specific methods of load control for CHW plants are discussed in

Heat Transfer in Hydronic Systems

Figure 1-15 Psychrometric chart chilled-water example.

Qa = airflow rate, cfm

Figure 1-16 Example system with heating coil.

Next, using LMTD, find q.

Figure 1-17 Example system with heating coil.

Latent Cooling and Dehumidification of Air

h = enthalpy difference of entering and leaving conditions of cooled

Heat Transferred to or from Water

Assume qt (cooling coil) = 384,750 Btu/h and t = 12°F.

It is at this point that further consideration of the coil selection becomes

Figure 1-20 Fan-coil unit.

• Space design conditions are 75°F/50% rh

In this example, we have to incorporate our basic knowledge of psychromet-

The process could be sketched psychrometrically per Figure 1-22.

52°F/90% rh (about 50°F wb).

Q = 1.08 cfm t cfm

The ADP is an average temperature representing the coil surface and is

Coil EWT = 2 ADP – T WBLVG

ufacturer’s coil selection program is shown in Figure 1-23.

Q COIL = 4.5 cfm h = 4.5 1950 31.2 – 21 = 89,500 Btu/h

A variety of options are up to the designer in coil selection. Typical tube

Figure 1-23 Example of manufacturer’s coil selection program.

Table 1-1 Coil Depth and Velocity to Achieve Dehumidification

Figure 1-24 Heat transfer versus water flow for a coil.