$PNQVUFS
"QQMJDBUJPOT
JO

)ZESBVMJD
&OHJOFFSJOH
9th Edition

Brian E. Whitman
Thomas M. Walski
Thomas E. Barnard
S.
Rocky Durrans
Steve Lowry
Michael E. Meadows
$PNQVUFS
"QQMJDBUJPOT
JO
)ZESBVMJD
&OHJOFFSJOH

9th Edition

Copyright © 2021 Bentley Systems, Incorporated. All Rights Reserved.

Bentley, the “B” Bentley logo, and Bentley Institute Press are either registered or
unregistered trademarks or service marks of Bentley Systems, Incorporated or one of its
direct or indirect wholly owned subsidiaries. Other brands and product names are trademarks
of their respective owners.

The publisher does not warrant or guarantee any of the products described herein or perform
any independent analysis in connection with any of the product information contained herein.
The publisher also does not assume, and expressly disclaims, any obligation to obtain and
include information other than that provided to it by the manufacturer.

The reader is expressly warned to consider and adopt all safety precautions that might
be indicated by the activities herein and to avoid all potential hazards. By following the
instructions contained herein, the reader willingly assumes all risks in connection with such
instructions.

The publisher makes no representation or warranties of any kind, including but not limited
to, the warranties of fitness for particular purpose of merchantability, nor are any such
representations implied with respect to the material set forth herein, and the publisher takes
no responsibility with respect to such material. The publisher shall not be liable for any
special, consequential, or exemplary damages resulting, in whole or part, from the readers’
use of, or reliance upon, this material.

ISBN - 978-1-934493-70-0
ISBN (ePub) - 978-1-934493-71-7
LCCN – 2021930961

Published by:
Bentley Institute Press
Bentley Systems, Incorporated
685 Stockton Drive
Exton, PA 19341, United States
www.bentley.com
Printed in the United States
Authors and Editors

Brian E. Whitman, Ph.D. is a Professor of Environmental Engineering at

Wilkes University in Wilkes-Barre, Pennsylvania where he was the recipient of
the Wilkes University Outstanding Faculty Award in years 2002 and 2005. He
holds a M.S. in Civil Engineering and Ph.D. in Environmental Engineering
from Michigan Technological University. He has extensive experience teach-
ing hydraulic modeling software for use in water distribution system design,
wastewater collection system design, and water resources engineering. He is a
contributing author to four books, and has several peer reviewed publications.
Thomas M. Walski, Ph.D., P.E. is Senior Project Manager at Bentley, and
is registered as a professional engineer in Pennsylvania and Mississippi. He
received his Master’s Degree and Ph.D. in Environmental and Water Resources
Engineering from Vanderbilt University. He has written a number of authori-
tative books and several hundred journal and conference papers on the subject
of hydraulic modeling. Dr. Walski spent five years as an Associate Professor in
the Environmental Engineering Program at Wilkes University.

Computer ApplicationT in Hydraulic Engineering, 9th Edition iii

Contributing Authors

Bentley’s Civil Engineering Staff is an extremely diverse group of profes-

sionals from six continents with experience ranging from software develop-
ment and engineering consulting, to public works and academia. This broad
cross-section of expertise contributes to the development of the most com-
prehensive software and educational materials in the civil engineering indus-
try. This book was truly a team effort and many at Bentley (formally Haestad
Methods) contributed.
Thomas E. Barnard, Ph.D., P.E. had more than 25 years of experience in
environmental engineering working as a consultant, researcher, educator, and
author. He holds a B.S. in Civil Engineering from the University of Vermont,
an M.S. in environmental engineering from Utah State University, and a Ph.D.
in environmental engineering from Cornell University. His expertise includes
water and wastewater treatment, hazardous waste management, surface water
hydrology and water quality monitoring systems. He was registered profes-
sional engineer in Pennsylvania.
S. Rocky Durrans, Ph.D., P.E. was a Professor Emeritus of Civil and
Environmental Engineering at the University of Alabama, and was responsible
for the hydrologic and hydraulic engineering programs. He holds a Ph.D. from
University of Colorado-Boulder. Prior to his academic career, Dr. Durrans
spent many years in the consulting arena where he gained extensive experience
in design of stormwater and major urban drainage systems. He is regarded as
an authority on flood and rainfall analyses, and has many publications to his
credit.
Steve Lowry, P.E. is President of Steven G. Lowery & Associates, Inc. and is
a registered professional engineer in Colorado and Pennsylvania. He has over
30 years of civil engineering experience in all aspect of project management
and performance on project involving hydraulic, water quality, and transient
analysis of distribution systems; SCADA system design; hydrologic and safe
yield of water supply systems; computer modeling and computer program-
ming; and GIS applications. He has authored several publications regarding
water systems, modeling, and model calibration. He holds a B.S. and M.S. in
Civil Engineering from the Ohio State University.

Computer ApplicationT in Hydraulic Engineering, 9th Edition v

 Contributing Authors

Michael E. Meadows, Ph.D., P.E. is an Associate Professor in the Department

of Civil and Environmental Engineering at the University of South Carolina
where he instructs undergraduate and graduate courses on hydraulics and
hydrology. He holds a M.S. in Environmental Engineering and Ph.D. in Civil
Engineering from University of Tennessee. He has coauthored texts on storm-
water modeling and has published numerous papers on improved methods for
estimating stormwater discharge and evaluating drainage system performance.
Dr. Meadows has also served as special consultant to SC DOT, USACE,
USGS, EPA, and other state and municipal agencies in developing, evaluating,
and applying water resource modeling methodologies.

vi $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition

Revision History

During the development of each edition of CAiHE, we encourage feedback

from the hundreds of universities that have adopted this textbook and accom-
panying software. We are interested in learning more about what professors
and students want added to this publication and the way it is being utilized
within civil engineering curricula.
In the Second Edition, we added tutorial example problems with step-by-step
instructions for solving the problems using the included computer software.
More problems to be solved by students using the software were added. In
addition, updates were incorporated into the software on the CD-ROM.
In the Third Edition, we added two new chapters: one chapter on drainage
inlet design and the other on sanitary sewer design. These chapters included
tutorials and problems. Also, a brief discussion of weirs and orifices as an appli-
cation of the energy equation was added to Chapter 1. More problems were
added to the existing chapters. The CD-ROM enclosed with this text con-
tained new versions of FlowMaster and StormCAD, as well as a new product,
SewerCAD. In addition, updates were incorporated into other software on the
CD-ROM.
In the Fourth Edition, we added discussions on the use of hydrographs and
hydrologic routing to the sanitary sewer modeling chapter (Chapter 6). An
extended-period simulation problem was added to this chapter as well.
In the Fifth Edition, we added a chapter on basic hydrology and a chapter
on detention pond design, accompanied by the latest academic version of
PondPack. Tutorials throughout the text have been updated and expanded per
the latest software versions, and current updates have been incorporated into
the software on the CD-ROM. In response to requests from professors, the
capacity of the software on the CD-ROM has been increased so that students
may work with more complex problems.
In the Sixth Edition, we revised the tutorials to reflect revisions to our software.
In the Seventh Edition, we have revised the Pressure Piping and Water Quality
Analysis chapter (Chapter 6) and updated the material for our newest software.
We have also added a brand new Chapter 8 called Dynamic Modeling with
full tutorials and problems. Appendix C was deleted and replaced by the pre-
vious Appendix D. Appendix A was modified to provide users with updated

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition vii

 Revision History

instructions on how to download the software licenses for each program con-
tained on the academic CD.
In the Eighth Edition, we added chapter in transient analysis with tutori-
als using the HAMMER software (Chapter 7). Tutorials throughout the text
have been updated and expanded per the latest software versions, and current
updates have been incorporated into the software on the CD-ROM. CD-ROM
has been replaced by a procedure to download software from Bentley website.
In this Ninth Edition, we added material about Low Impact Development
(LID) for stormwater management to chapter 9, and the textbook CD-ROM
has been replaced by a procedure to download the latest software from Bentley
website. We took all the tutorials out of the print book and placed them on-line
so they can be easily updated as the software changes and is improved with new
features. We will also placed additional problems to go with all the chapters
on this same on-line location. We would like to thank Brandon Whitman who
completed, reviewed, and provided improvements for all the tutorials that were
updated to the latest Bentley CONNECT editions.
The following table lists the software that accompanies each chapter.
Chapter Accompanying Software
Chapter 1 FlowMaster
Chapter 2 not required
Chapter 3 StormCAD or SewerGEMS
Chapter 4 CulvertMaster
Chapter 5 PondPack
Chapter 6 WaterGEMS
Chapter 7 HAMMER
Chapter 8 SewerCAD or SewerGEMS
Chapter 9 SewerGEMS

viii $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition

Contacting Bentley

We welcome feedback regarding our products and services; we feel this is the
best way for us to continue providing software, training, publications, and
support that are tailored to the needs of professionals in the field of hydraulics
and hydrology today.

Included with this book is access to download and install academic versions
of our award-winning software, WaterGEMS, HAMMER, SewerGEMS,
SewerCAD, StormCAD, PondPack, CulvertMaster, and FlowMaster. They
have been included to assist you with working the problems contained in
this book and to introduce you to the latest developments in computer-aided
hydraulic and hydrologic modeling. Details on accessing the software is
included in Appendix A.

We hope that once you try the companion software packaged with CAiHE,
you’ll discover many ways in which it can enhance your work outside of this
book. You can get more products and upgrade information on any of our
software by calling our sales department or by visiting the Bentley’s OpenFlows
website:

Sales
https://www.bentley.com/en/about-us/contact-us/
sales-contact-requestorcalling

OpenFlows
https://www.bentley.com/en/products/brands/openflows

North America
1-800-BENTLEY (1 800 236 8539)
+1 610 458 5000

Europe, Middle East, and Africa

1800 948 368
+353 1 436 4600

Asia
400 842 7516
+86 10 5929 7000

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition ix

 Contacting Bentley

Bentley also works to bring the global civil engineering community together
with our online forums. Get in touch with fellow professionals worldwide and
discuss the latest tips and tricks in hydraulic and hydrologic modeling. Post
your engineering questions and share your unique modeling experiences with
an audience of thousands of professionals. Participants from around the world
include many industry-recognized experts and users like you. Get connected
today by visiting:

http://communities.bentley.com
https://communities.bentley.com/products/hydraulics___hydrology/

If you have any comments regarding this publication and/or any of Bentley
products and services, please contact us at:

Bentley Systems, Inc.

685 Stockton Drive
Exton, PA 19341 USA
Voice: 1-800-BENTLEY (1 800 236 8539)
Email: [email protected]
Internet: www.bentley.com

x $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition

Foreword

Bentley Systems, Inc. has developed hydrology and hydraulics software for civil
engineers and provides textbooks, training, and technical support to tens of
thousands of professional civil engineers, modelers, and universities.
Along the way, we have learned a lot about our clients and their professional
and educational backgrounds. Our continuing education program offers
courses for professionals who need to quickly get up-to-speed with various
numerical methods and practices.
Why is this important?
Our experience has shown us a great deal about the areas where engineers are
being trained, and it has also revealed a significant gap in this training—the
link between hydraulic theory and practical computer applications.
Hundreds of textbooks exist that offer enormous detail in the areas of engi-
neering history, equation derivations, and hand calculation methods. There
are also hundreds of published theses and articles that deal with computer
applications—unfortunately, most of these are highly research oriented and
are usually tied to a specific case study or an unusual set of circumstances. Both
of these publication types are very important to the civil engineering industry,
but in the majority of cases, they fail to address the issues most commonly
encountered by professionals in day-to-day practice.
This publication is intended as an introduction to the more common applica-
tions of water resources engineering software, and it demonstrates the types of
situations that an engineer will most likely come across on a daily basis in the
real world. It shows the true benefits of computer software: increased efficiency,
better flexibility, and—most important—an increased ability to try different
and better designs.
It is our hope that engineers, technicians, and students will find this book to
be challenging, but also easy to understand and very practical. Combined with
standard hydraulic references, we believe that this text provides many of the
tools needed to successfully proceed with a career in the fields of hydrology
and hydraulics.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition xi

Contents

Authors and Editors iii

Contributing Authors v
Revision History vii
Contacting Bentley ix
Foreword xi

Chapter 1: Basic Hydraulic Principles 1

Chapter 2: Basic Hydrology 29

Chapter 3: Inlets, Gravity Piping Systems, and Storm Sewer Design 81

Chapter 4: Culvert Hydraulics 113

Chapter 5: Detention Pond Design 127

Chapter 6: Pressure Piping Systems and Water Quality Analysis 161

Chapter 7: Hydraulic Transients in Distribution Systems 195

Chapter 8: Sanitary Sewer Design 211

Chapter 9: Low-Impact Development and Dynamic Modeling 235

Bibliography 261
Bentley Systems Software 267
Scenario Management 275
Index 289

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering, 9th Edition xiii

 1 Basic Hydraulic Principles

1.1 General Flow Characteristics

In hydraulics, as with any technical topic, a full understanding cannot come
without first becoming familiar with basic terminology and governing princi-
ples. The basic concepts discussed in the following pages lay the foundation for
the more complex analyses presented in later chapters.

System of Units
The basis for the problems in this book is real-world systems designed and con-
structed using U.S. customary units. Those units are retained in the problems.
In response to the goal to convert to a globally consistent system of units, many
agencies in the United States changed their regulations to require engineering
calculations and measurements in the International System of Units (SI). Even
though a new design may require SI units, likely the existing engineering sys-
tem, of which the new design is to become a part, was designed and constructed
using U.S. customary units. Consequently, proper system evaluation will require
the engineer to use both systems of units. This requirement will persist for many
years. Therefore it is essential that students, engineers, regulators, and planners
recognize and use contemporary units and, often, a dual system of units.

Flow Rate and Velocity

There are different measures for “how much” fluid is moving through a system.
In hydraulic systems, the most common measure is the volumetric flow rate.
This is a measure of the volume of fluid passing a location with time. Other
measures of flow rate can be determined by multiplying the volumetric flow
rate by the density of the fluid (mass flow rate), or the specific weight of the
fluid (weight flow rate).
The velocity of a fluid is a measure of “how fast” a fluid is moving through
a system. As shown in Figure 1.1, the velocity of a section is not constant
throughout the cross-sectional area. Instead, it varies with location. The veloc-
ity is zero where the fluid is in contact with the conduit wall.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 1

 1 Basic Hydraulic Principles
General Flow Characteristics

Conduit wall
Depth
Velocity distribution

Longitudinal section (profile)

Figure 1.1: Velocity Distribution.

Flow area

Wetted perimeter

Figure 1.2: Flow Area and Wetted Perimeter.

The variation of flow velocity within a cross section complicates the hydraulic
analysis, so the engineer usually simplifies the situation by looking at the aver-
age (mean) velocity of the section for analysis purposes. This average velocity is
defined as the total flow rate divided by the cross-sectional area and is in units
of length per time.
Q
V =
A
where
V = average velocity (m/s, ft/s)
Q = flow rate (m3/s, ft3/s)
A = area (m2, ft2).

Area, Wetted Perimeter, and Hydraulic Radius

The term area refers to the cross-sectional area of flow within a channel. When
a channel has a consistent cross-sectional shape, slope, and roughness, it is
called a prismatic channel.
A section’s wetted perimeter is defined as the portion of the channel in contact
with the flowing fluid. This definition is illustrated in Figure 1.2.
The hydraulic radius of a section is not a directly measurable characteristic,
but it is used frequently during calculations. It is defined as the area divided by
the wetted perimeter and therefore has units of length.

2 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
General Flow Characteristics

The hydraulic radius can often be related directly to the geometric properties
of the channel. For example, the hydraulic radius of a full circular pipe (such as
a pressure pipe) can be directly computed as:

A
R=
Pw

or
π ⋅ D2 /4 D
R circular = =
π⋅D 4

where
R = hydraulic radius (m, ft)
A = cross-sectional area (m2, ft2)
Pw = wetted perimeter (m, ft)
D = inside pipe diameter (m, ft).

Flow Conveyance
Water travels downhill from points of higher energy to points of lower energy
(unless forced to do otherwise) until it reaches a point of equilibrium, such as
an ocean. This tendency is facilitated by the presence of natural conveyance
channels such as brooks, streams, and rivers. The water’s journey may also be
aided by man-made structures such as drainage swales, pipes, culverts, and
canals. Hydraulic concepts can be applied equally to both man-made struc-
tures and natural features.
If the flow in a conveyance section is open to the atmosphere, such as in a
culvert flowing partially full or in a river, it is said to be open-channel flow or
free-surface flow. If a channel is flowing completely full, as with a water dis-
tribution pipe, it is said to be operating under full-flow conditions. Pressure
flow is a special type of full flow in which forces on the fluid cause it to push
against the top of the channel as well as the bottom and sides. These forces may
result from, for example, the weight of a column of water in a backed up sewer
manhole or elevated storage tank.

Steady Flow
Speaking in terms of “flow,” the word steady indicates that a constant flow
rate is assumed throughout an analysis. In other words, the flow velocity does
not change with respect to time at a given location. For most hydraulic calcu-
lations, this assumption is reasonable. A minimal increase in model accuracy
does not warrant the time and effort that would be required to perform an
analysis with changing (unsteady) flows over time.
When analyzing tributary and river networks, storm sewers, and other collec-
tion systems in which it is desirable to vary the flow rate at different locations

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 3

 1 Basic Hydraulic Principles
General Flow Characteristics

throughout the system, the network can often be broken into segments that
can be analyzed separately under steady flow conditions.

Uniform Flow
Uniform flow refers to the hydraulic condition in which the discharge and
cross-sectional area (and therefore velocity) are constant throughout the length
of the channel. For a pipe flowing full, the only required assumptions are that
the pipe be straight and have no contractions or expansions. For an open chan-
nel, additional assumptions include the following:
• The depth of flow must be constant [i.e., the hydraulic grade line (HGL)
must be parallel to the channel bed]. This depth of flow is called normal
depth.
• Because the velocity is constant, the velocity head does not change through
the length of the section; therefore the energy grade line (EGL) is parallel
to both the HGL and the channel bed.
In channels that are prismatic, the flow conditions will typically approach uni-
form flow if the channel is sufficiently long. When this occurs, the net force
on the fluid approaches zero because the gravitational force is equal to the
opposing friction forces from the channel bottom and walls. Nonuniform flow
is the hydraulic condition in which the discharge and cross-sectional area (and
therefore velocity) are not constant throughout the length of the channel. In
open channels, the velocity profile and cross-sectional area of flow will gradually
change. The analysis of gradually varied flow is described in Chapters 3.6 and 9.4

Laminar Flow, Turbulent Flow, and Reynolds Number

Laminar flow is characterized by smooth, predictable streamlines (the paths
of single fluid particles). An example of this type of flow is maple syrup being
poured. In turbulent flow, the streamlines are erratic and unpredictable.
Turbulent flow is characterized by the formation of eddies within the flow,
resulting in continuous mixing throughout the section (see Figure 1.3).

Laminar flow Turbulent flow

Figure 1.3: Instantaneous Velocity Distributions for Laminar and

Turbulent Flow.

4 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
General Flow Characteristics

Eddies result in varying velocity directions as well as magnitudes (varying

directions not depicted in Figure 1.3 for simplicity). At times, the eddies con-
tribute to the velocity of a given particle in the direction of flow and at other
times detract from it. The result is that velocity distributions captured at differ-
ent times will be quite different from one another and will be far more chaotic
than the velocity distribution of a laminar flow section.

By strict interpretation, the changing velocities in turbulent flow would cause

it to be classified as unsteady flow. Over time, however, the average velocity
at any given point within the section is essentially constant, so the flow is
assumed to be steady.

The velocity at any given point within the turbulent section will be closer to
the mean velocity of the entire section than with laminar flow conditions.
Turbulent flow velocities are closer to the mean velocity because of the con-
tinuous mixing of flow, particularly the mixing of low-velocity flow near the
channel walls with the higher velocity flow toward the center.

To classify flow as either turbulent or laminar, an index called the Reynolds

number is used. It is computed as follows:

4VR
Re =
ν

where
Re = Reynolds number (unitless)
V = average velocity (m/s, ft/s)
R = hydraulic radius (m, ft)
ν = kinematic viscosity (m2/s, ft2/s).
If the Reynolds number is below 2,000, the flow is generally laminar. For flow
in closed conduits, if the Reynolds number is above 4,000, the flow is gener-
ally turbulent. Between 2,000 and 4,000, the flow may be either laminar or
turbulent, depending on how insulated the flow is from outside disturbances.
In open channels, laminar flow occurs when the Reynolds number is less than
500 and turbulent flow occurs when it is above 2,000. Between 500 and 2,000,
the flow is transitional.

Example 1.1: Flow Characteristics

A rectangular concrete channel is 3-m wide and 2-m high. The water in the
channel is 1.5-m deep and is flowing at a rate of 30 m3/s. Determine the flow
area, wetted perimeter, and hydraulic radius. Is the flow laminar or turbulent?

Solution
From the section’s shape (rectangular), we can easily calculate the area as the
rectangle’s width multiplied by its depth. Note that the depth used should be
the actual depth of flow, not the total height of the cross section. The wetted
perimeter can also be found easily through simple geometry.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 5

 1 Basic Hydraulic Principles
Energy

A = 3.0 m × 1.5 m = 4.5 m2

Pw = 3.0 m + 2 × 1.5 m = 6.0 m
R = A/Pw = 4.5 m2/6.0 m = 0.75 m

In order to determine whether the flow is likely to be laminar or turbulent, we

must determine the Reynolds number. To do this, first find the velocity of the
section and a value for the kinematic viscosity.

V = Q/A = 30 m3/s/4.5 m2 = 6.67 m/s

From fluids reference tables, we find that the kinematic viscosity for water at
20°C is 1.00 × 106 m2/s. Substituting these values into the formula to compute
the Reynolds number results in

Re = (4 × 6.67 m/s × 0.75 m)/(1.00 × 10−6) = 2 × 107

This value is well above the Reynolds number minimum of 4,000 for turbulent
flow.

1.2 Energy
The Energy Principle
The first law of thermodynamics states that for any given system, the change
in energy (ΔE) is equal to the difference between the heat transferred to the
system (Q) and the work done by the system on its surroundings (W ) during
a given time interval.
The energy referred to in this principle represents the total energy of the sys-
tem, which is the sum of the potential energy, kinetic energy, and internal
(molecular) forms of energy such as electrical and chemical energy. Although
internal energy may be significant for thermodynamic analyses, it is commonly
neglected in hydraulic analyses because of its relatively small magnitude.
In hydraulic applications, energy values are often converted into units of
energy per unit weight, resulting in units of length. Using these length equiv-
alents gives engineers a better “feel” for the resulting behavior of the system.
When using these length equivalents, the engineer is expressing the energy
of the system in terms of “head.” The energy at any point within a hydraulic
system is often expressed in three parts, as shown in Figure 1.4:
• pressure head p/g
• elevation head z
• velocity head V 2 2 g
where
p = pressure (N/m2, lb/ft2)
γ = specific weight (N/m3, lb/ft3)

6 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Energy

V12 HL
2g Energy g
rade line V22
p1 Hydraulic grade line 2g
g p2
g

Z1 Z2

Datum

Longitudinal section (profile)

Figure 1.4: The Energy Principle.

z = elevation (m, ft)

V = velocity (m/s, ft/s).
Note that a point on the water surface of an open channel will have not only
a pressure head of zero but also a positive elevation head higher than that of a
point selected at the bottom of the channel for the same station.

The Energy Equation

In addition to pressure head, elevation head, and velocity head, energy may be
added to a system by a pump, for example, and removed from the system by
friction or other disturbances. These changes in energy are referred to as head
gains and head losses, respectively. Because energy is conserved, the energy
across any two points in the system must balance. This concept is demon-
strated by the energy equation:

p1 V2 p V2
+ z1 + 1 + HG = 2 + z2 + 2 + HL
g 2g γ 2g

where
p = pressure (N/m2, lb/ft2)
γ = specific weight of the fluid (N/m3, lb/ft3)
z = elevation above a datum (m, ft)
V = fluid velocity (m/s, ft/s)
g = gravitational acceleration (m/s2, ft/s2)
HG = head gain, such as from a pump (m, ft)
HL = combined head loss (m, ft).

Hydraulic Grade
The hydraulic grade is the sum of the pressure head (p/γ) and elevation head (z).
For open-channel flow (in which the pressure head is zero), the hydraulic grade

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 7

 1 Basic Hydraulic Principles
Energy

elevation is the same as the water surface elevation. For a pressure pipe, the
hydraulic grade represents the height to which a water column would rise in
a piezometer (a tube open to the atmosphere rising from the pipe). When the
hydraulic grade is plotted as a profile along the length of the conveyance sec-
tion, it is referred to as the HGL.

Energy Grade
The energy grade is the sum of the hydraulic grade and the velocity head
(V 2/2g). This grade is the height to which a column of water would rise in a
pitot tube (an apparatus similar to a piezometer, but also accounting for fluid
velocity). When plotted in profile, this parameter is often referred to as the
EGL. For a lake or reservoir in which the velocity is essentially zero, the EGL
is equal to the HGL.

Energy Losses and Gains

Energy (or head) losses (HL) in a system are due to a combination of several
factors. The primary cause of energy loss is usually the internal friction between
fluid particles traveling at different velocities. Secondary causes of energy loss
are localized areas of increased turbulence and disruption of the streamlines,
such as disruptions from valves and other fittings in a pressure pipe, or disrup-
tions from a changing section shape in a river.
The rate at which energy is lost along a given length of channel is called the
friction slope and is usually presented as a unitless value or in units of length
per length (ft/ft, m/m, etc.).
Energy is generally added to a system with a device such as a pump. Pumps are
discussed in more detail in Chapter 6.
Example 1.2: Energy Principles
A 1,200-mm diameter transmission pipe carries 126 l/s from an elevated stor-
age tank with a water surface elevation of 540 m. Two kilometers from the
tank, at an elevation of 434 m, a pressure meter reads 586 kPa. If there are no
pumps between the tank and the meter location, what is the rate of head loss
in the pipe? (Note: 1 kPa = 1,000 N/m2.)
Solution
Begin by simplifying the energy equation. Assume that the velocity within the
tank is negligible, and that the pressure head at the tank can be discounted
because it is open to the atmosphere. Rewriting the energy equation and enter-
ing the known values, we can solve for head loss. The velocity can be calculated
using the flow rate and pipe diameter.
Q = 126 l/s × (1 l/s/103 m3/s) = 0.126 m3/s
A = π × (0.6 m)2 = 1.13 m2
V = Q/A = 0.126 m3/s/1.13 m2 = 0.11 m/s
V 2/2g = (0.11 m/s)2/(2 × 9.81 m/s2) = 0.0006 m (negligible)

8 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Friction Losses

Neglecting the velocity simplifies the energy equation even further, and we can
now solve for head loss as
HL = 540 m − 434 m − (586,000 N/m2)/9,810 N/m3 = 46.27 m
The rate of head loss (or friction slope) can now be computed as
Friction slope = 46.27 m/(2 × 1,000 m) = 0.023 m/m, or 23 m/km

1.3 Friction Losses

There are many equations that approximate the friction losses associated with
the flow of a liquid through a given section. Commonly used methods include-
Manning’s equation
• Chézy’s (Kutter’s) equation
• Hazen–Williams equation
• Darcy–Weisbach (Colebrook–White) equation.
These equations can be described by a generalized friction equation:

V = kCR x S y
where
V = mean velocity
C = flow resistance factor
R = hydraulic radius
S = friction slope
x, y = exponents
k = factor to account for empirical constants, unit conversion, etc.
The lining material of the flow channel usually determines the flow resistance
or roughness factor, C. However, the ultimate value of C may be a function of
the channel shape, depth, and fluid velocity.

Manning’s Equation
Manning’s equation is the most commonly used open-channel flow equation.
The roughness component, C, is typically assumed to be constant over the full
range of flows and is represented by a Manning’s roughness value, n. These
n-values have been experimentally determined for various materials and should
not be used with fluids other than water. Manning’s equation is:

k 2 / 3 1/ 2
V = R S
n
where
V = mean velocity (m/s, ft/s)
k = 1.49 for U.S. customary units, 1.00 for SI units
n = Manning’s roughness value

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 9

 1 Basic Hydraulic Principles
Friction Losses

R = hydraulic radius (m, ft)

S = friction slope (m/m, ft/ft).

Chézy’s (Kutter’s) Equation

The Chézy equation, in conjunction with Kutter’s equation, is widely used in
sanitary sewer design and analysis. The roughness component, C, is a function
of the hydraulic radius, friction slope, and lining material of the channel. The
Chézy equation is:

V = C RS
where
V = mean velocity (m/s, ft/s)
C = roughness coefficient (see following calculation)
R = hydraulic radius (m, ft)
S = friction slope (m/m, ft/ft).
The roughness coefficient, C, is related to Kutter’s n through Kutter’s equa-
tion. Note that the n-values used in Kutter’s equation are actually the same as
Manning’s n coefficients.

U.S. Standard Units S.I. Units

0.00155 1
0.00281 1.811 23 + +
41.65 + + S n
S n C=
C = ⎛ ⎞
⎛ 0.00281 ⎞⎟ ⎜⎜23 + 0.00155 ⎟⎟n
⎜⎜41.65 + ⎟n ⎜⎝ S ⎟⎠
⎜⎝ S ⎟⎠ 1+
1+ R
R

where
C = roughness coefficient
n = Manning’s roughness value
R = hydraulic radius (m, ft)
S = friction slope (m/m, ft/ft).

Hazen–Williams Equation
The Hazen–Williams equation is most frequently used in the design and analysis
of pressure pipe systems. The equation was developed experimentally and there-
fore should not be used for fluids other than water (and only within temperatures
normally experienced in potable water systems). The Hazen–Williams equation is:

V = kCR 0.63S 0.54

where
V = mean velocity (m/s, ft/s)
k = 1.32 for U.S. customary units, or 0.85 for SI units

10 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Friction Losses

C = Hazen–Williams roughness coefficient (unitless)

R = hydraulic radius (m, ft)
S = friction slope (m/m, ft/ft).

Darcy–Weisbach (Colebrook–White) Equation

The Darcy–Weisbach equation is a theoretically based equation commonly
used in the analysis of pressure pipe systems. It applies equally well to any
flow rate and any incompressible fluid and is general enough to be applied to
open-channel flow systems. In fact, the ASCE Task Force on Friction Factors
in Open Channels (1963) supported the use of the Darcy–Weisbach equation
for free-surface flows. This recommendation has not yet been widely accepted
because the solution to the equation is difficult and not easily computed using
noncomputerized methods. With the increasing availability of computer solu-
tions, the Darcy–Weisbach equation will likely gain greater acceptance because
it successfully models the variability of effective channel roughness with chan-
nel material, geometry, and velocity.
The roughness component in the Darcy–Weisbach equation is a function of
both the channel material and the Reynolds number, which varies with veloc-
ity and hydraulic radius.

8g
V = RS
f

where
V = flow velocity (m/s, ft/s)
g = gravitational acceleration (m/s2, ft/s2)
f = Darcy–Weisbach friction factor (unitless)
R = hydraulic radius (m, ft)
S = friction slope (m/m, ft/ft).
The Darcy–Weisbach friction factor, f, can be found using the Colebrook–
White equation for fully developed turbulent flow, as follows:

Free surface Full flow (closed conduit)

⎛ k ⎞⎟ ⎛ k ⎞⎟
= −2 log ⎜⎜⎜
1 2.51
= −2 log ⎜⎜⎜
+ ⎟⎟ 1 2.51 ⎟⎟
⎟⎠ +
f ⎜⎝12 R Re f f ⎜⎝14.8R Re f ⎟⎠

where
k = roughness height (m, ft)
R = hydraulic radius (m, ft)
Re = Reynolds number (unitless).
This iterative search for the correct value of “f ” can become quite time-consuming
for hand computations and computerized solutions of many pipes. Another

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 11

 1 Basic Hydraulic Principles
Pressure Flow

method, developed by Swamee and Jain, solves directly for f in full-flowing cir-
cular pipes. This equation is:

1.325
f = 2
⎡ ⎛ k 5.74 ⎞⎟⎤
⎢ ln ⎜⎜ ⎥
⎢⎣ ⎜⎝ 3.7 D + Re0.9 ⎟⎟⎠⎥
⎦

where
f = friction factor (unitless)
k = roughness height (m, ft)
D = pipe diameter (m, ft)
Re = Reynolds number (unitless).

Typical Roughness Factors

Typical pipe roughness values for each of these methods are shown in Table 1.1.
These values will vary depending on the manufacturer, workmanship, age, and
other factors. For this reason, Table 1.1 should be used only as a guideline.

1.4 Pressure Flow

For pipes flowing full, many of the friction loss calculations are greatly sim-
plified because the flow area, wetted perimeter, and hydraulic radius are all
functions of pipe radius (or diameter). Table 1.2 presents the three pipe friction
loss equations that are commonly used to design pressure pipe systems.
There is much more information presented about pressure piping systems in
Chapter 6, including further discussion on pumping systems, minor losses,
and network analysis.
Example 1.3: Pressure Pipe Friction Losses
Use the FlowMaster program to compare the head loss computed by the
Hazen–Williams equation to the head loss computed by the Darcy–Weisbach
equation for a pressure pipe having the following characteristics: 12-in diameter
cast iron pipe (new) one mile in length with a flow rate of 1,200 gal/min (with
water at 65°F).
Solution
Although there are no elevations or pressures given, these values are not needed
to determine the head loss in the pipe. Setting up FlowMaster to solve for the
“Elevation at 1” allows us to use zero elevation and zero pressure assumptions
and fill in the rest of the pipe characteristics.
For the Hazen–Williams equation, a C coefficient of 130 is assumed. This
value results in 18.8 ft of head loss (which agrees with the computed 18.8-ft
elevation at point 1). Using Darcy–Weisbach, a roughness height of 0.00085 ft
is assumed. The solution indicates a head loss of 18.9 ft, which is only a 0.1-ft
difference from the value predicted by Hazen–Williams.

12 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Pressure Flow

Table 1.1: Typical Roughness Coefficients

Material Manning’s Hazen-Williams Darcy-Weisbach

Coefficient C Roughness Height
n
k (mm) k (ft)
Asbestos cement 0.011 140 0.0015 0.000005
Brass 0.011 135 0.0015 0.000005
Brick 0.015 100 0.6 0.002
Ductile (cast) iron, new 0.012 130 0.26 0.00085
Concrete:
Steel forms 0.011 140 0.18 0.006
Wooden forms 0.015 120 0.6 0.002
Centrifugally spun 0.013 135 0.36 0.0012
Copper 0.011 135 0.0015 0.000005
Corrugated metal 0.022 – 45 0.15
Galvanized iron 0.016 120 0.15 0.0005
Glass 0.011 140 0.0015 0.000005
Lead 0.011 135 0.0015 0.000005
Plastic 0.009 150 0.0015 0.000005
Steel:
Coal-tar enamel 0.010 148 0.0048 0.000016
New unlined 0.011 145 0.045 0.00015
Riveted 0.019 110 0.9 0.003
Wood stave 0.012 120 0.18 0.0006

Table 1.2: Three Pipe Friction Loss Equations

Equation Q (m3/s); D (m) Q (cfs); D (ft) Q (gpm); D (in.)

Darcy-Weisbach 0.083 fQ 2
0.025 fQ 2
0.031 fQ 2
Sf = Sf = Sf =
D5 D5 D5
Hazen-Williams 1.852 1.852 1.852
10.7 ⎛⎜Q ⎞⎟ 4.73 ⎛⎜Q ⎞⎟ 10.5 ⎛⎜Q ⎞⎟
Sf = ⎜ ⎟ Sf = ⎜ ⎟ Sf = ⎜ ⎟
D 4.87 ⎜⎝ C ⎟⎠ D 4.87 ⎜⎝ C ⎟⎠ D 4.87 ⎜⎝ C ⎟⎠

Manning 10.3(nQ )2 4.66(nQ )2 13.2(nQ )2

Sf = Sf = Sf =
D5.33 D5.33 D5.33

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 13

 1 Basic Hydraulic Principles
Open-Channel Flow

Discussion
If the same system is analyzed with 2,000–3,000 gal/min of flow, the difference
in head loss between the two equations becomes almost 10 ft.
Why such a big difference? For starters, the two methodologies are completely
unrelated, and the estimated roughness coefficients were taken from a list of
approximate values. If the Hazen–Williams equation is used with a roughness
value of 125, the results are much closer. This difference should emphasize the
fact that models are only as good as the data that is input into them, and the
engineer needs to fully understand all of the assumptions that are being made
before accepting the results.

1.5 Open-Channel Flow

Open-channel flow analysis is more complex than pressure flow analysis
because the flow area, wetted perimeter, and hydraulic radius are not neces-
sarily constant as they are in a uniform pipe section under full-flow condi-
tions. Because of this considerable difference, additional characteristics become
important when dealing with open-channel flow.

Specific Energy and Critical Flow

Of course, channels do not always flow at normal depth. If they did, it would
make the engineer’s task quite simple. A more in-depth at nonuniform flow is
presented in Chapters 3.6 and 9.4, but this chapter will continue by focusing
on another important concept—“specific energy.”
For any flow section, the specific energy is defined as the sum of the depth of
flow and the velocity head.

V2
E= y+
2g
where
E = specific energy (m, ft)
y = depth of flow (m, ft)
V = mean velocity (m/s, ft/s)
g = gravitational acceleration (m/s2, ft/s2).
If we assume the special case of an infinitely short section of open channel
(with essentially no friction losses and no change in elevation), we see that the
general energy equation can be reduced to an equality of specific energies. In
other words,

V12 V2
E1 = y1 + = y2 + 2 = E2
2g 2g

14 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Open-Channel Flow

Channel depth, y
45°
Subcritical flow, low velocity, F < 1

Critical depth, F = 1
Supercritical flow, high velocity, F > 1
Specific energy, E

Figure 1.5: Specific Energy.

Recall that the velocity of the section is directly related to the area of flow,
and that the area of flow is a function of channel depth. This means that, for a
given discharge, the specific energy at each point is solely a function of channel
depth and more than one depth may exist with the same specific energy. If the
channel depth is plotted against specific energy for a given flow rate, the result
is similar to the graph shown in Figure 1.5.
As this figure shows, a depth exists for which the specific energy is at a mini-
mum. This depth is called the critical depth. If the velocity is higher than crit-
ical velocity (that is, the depth is less than critical depth), the flow is considered
supercritical. If the velocity is lower than critical velocity (the depth is greater
than critical depth), the flow is subcritical.
The velocity at critical depth is equal to the wave celerity—the speed at which
waves will ripple outward from a pebble tossed into the water. A unitless value
called the Froude number, F, represents the ratio of actual fluid velocity to
wave celerity. The Froude number is computed as follows:

V
F =
gD
where
F = Froude number (unitless)
D = hydraulic depth of the channel, defined as A/T
A = flow area (m2, ft2)
T = top width of flow (m, ft)
V = fluid velocity (m/s, ft/s)
g = gravitational acceleration (m/s2, ft/s2).
By definition, when the flow is at critical depth (that is, the velocity is equal to
the wave celerity), the Froude number must be equal to 1. The equation can
therefore be rewritten and refactored to form the following equality:

A3 Q 2
=
T g

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 15

 1 Basic Hydraulic Principles
Open-Channel Flow

where
A = flow area (m2, ft2)
T = top width of flow (m, ft)
Q = channel flow rate (m3/s, ft3/s)
g = gravitational acceleration (m/s2, ft/s2).
This equation can now be used to determine the depth for which this equality
holds true, which is critical depth. For simple geometric shapes, the solution is
relatively easy to determine. However, quite a few iterations may be required to
find the solution for an irregularly shaped channel such as a natural streambed.
In fact, several valid critical depths may exist for irregular channels.
Example 1.4: Uniform Open-Channel Flow
A concrete trapezoidal channel has a bottom width of 4 m and 45-degree side
slopes. If the channel is on a 1% slope and is flowing at a depth of 1 m throughout
its length, how much flow is being carried (use Manning’s equation)? How much
flow would the same channel carry if it were a rectangular channel 4 m wide?
Solution
Because the channel is flowing at the same depth throughout, we can assume
that normal depth has been achieved (that is, the friction slope is equal to the
channel slope). We will assume a Manning’s n of 0.013 for concrete.
From the trapezoidal geometry, we can easily calculate the area and wetted
perimeter, and then the hydraulic radius, as follows:
A = (4 m × 1 m) + 2 × (0.5 × 1 m × 1 m) = 5.00 m2
Pw = 4 m + 2 × (1 m × 20.5) = 6.83 m
R = A/Pw = 5.00 m2/6.83 m = 0.73 m
Manning’s equation for velocity can then be solved. The discharge can be com-
puted as
V = (1.00/0.013) × 0.732/3 × 0.011/2 = 6.25 m/s
Q = V × A = 6.25 m/s × 5.00 m2 = 31.2 m3/s
To answer the second part of the question, we simply repeat the steps for a
rectangular section shape.
A = (4 m × 1 m) = 4 m2
Pw = 4 m + 2 × (1 m) = 6 m
R = 4 m2/6 m = 0.67 m
V = (1.00/0.013) × 0.672/3 × 0.011/2 = 5.87 m/s
Q = 5.87 m/s × 4 m2 = 23.5 m3/s
As we would expect, this discharge is less than the discharge of the trapezoidal
section.
Example 1.5: Critical Depth
What is the critical depth for a grassy triangular channel with 2H:1V side
slopes and a 0.5% slope when the flow is 3.00 m3/s? If the channel is actually
flowing at a depth of 1.2 m, is the flow critical, subcritical, or supercritical?

16 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Orifices and Weirs

Solution
For such a simple geometry, we can quickly create a relationship between the
flow area, top width, and depth of flow:
T = 4y, A = 0.5 × T × y = 0.5 × 4y × y = 2y2 m2
Inserting these values into the previous equation for critical depth, we can
algebraically solve for the channel depth:
(2y2)3/4y = Q2/g
8y6/4y = Q2/g
2y5 = Q2/g
y5 = Q2/2g
y = (Q2/2g)1/5 = [(3.00m3/s)2/(2 × 9.8 m/s2)]0.2 = (0.46 m5)0.2 = 0.86 m
The critical depth for this section is 0.86 m. The actual flow depth of 1.2 m is
greater than critical depth, so the flow is subcritical.

1.6 Orifices and Weirs

The energy equation serves as the foundation for calculating the flow through
and over hydraulic structures based on the size of the opening associated with
the structure and the difference in energy on either side of it. The flow exiting
the structure can be calculated by solving the energy equation for velocity, V 2,
and multiplying the resulting formula by the flow area and a coefficient to
account for different hydraulic and physical variables. These variables include
head loss, the shape and nature of the opening, the contraction of the flow after
it leaves the structure, and countless indefinable variables that are difficult to
measure but produce quantifiable effects.
Two common devices for which equations are derived in this manner are weirs
and orifices. They are important not only because of their widespread usage
in the industry, but also because the equations that describe them serve as
the foundation for mathematical descriptions of more complicated hydraulic
devices such as drainage inlets and culverts.

Orifices
Orifices are regularly shaped, submerged openings through which flow is pro-
pelled by the difference in energy between the upstream and downstream sides
of the opening. The stream of flow expelled from the orifice is called the jet.
When the jet exits the orifice, adverse velocity components cause it to contract
to a point after which the flow area remains relatively constant and the flow
lines become parallel (see Figure 1.6). This point is called the vena contracta.
Orifices and the orifice equations have the following applications:
• regulating the flow out of detention ponds,
• regulating the flow through channels in the form of radial and sluice gates,

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 17

 1 Basic Hydraulic Principles
Orifices and Weirs

Vena contracta

Jet

Figure 1.6: Cross-Sectional View of Typical Orifice Flow.

• approximating the interception capacity of submerged drainage inlets in

sag (see Chapter 3),
• approximating the flow allowed through a submerged culvert operating
under inlet control (see Chapter 4),
• measuring flow.
Example 1.6: The Orifice Equation
For the structure in Figure 1.7, derive the orifice equation for an orifice of area A.
Solution
First, start with the energy equation from Section 1.2:

p1 V2 p V2
+ z1 + 1 + HG = 2 + z2 + 2 + HL
g 2g γ 2g

List known variables and assumptions:

• The datum is at the centerline/centroid of the orifice.
• p1/γ = H.

• Point 2 occurs at the vena contracta.

• Elevation heads, z1 and z2, are equivalent.
• The velocity in the tank at point 1 is negligible.
• The jet is open to the air, so the pressure at point 2 is 0.
• There is no head gain.

18 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Orifices and Weirs

P
g =H

Datum 1 2

Figure 1.7: Orifice Example.

Taking these known variables and assumptions into account and solving for
V2, the energy equation becomes:

V2 = 2 g (H − HL )

To find the flow exiting the structure at point 2, multiply both sides of the
equation by the orifice area, A.

AV2 = Q = A 2 g (H − HL )

where Q = discharge (m3/s, ft3/s).

The point of discharge is the vena contracta, where the flow area is usually
contracted from the original orifice area. Also, computations can be simplified
by eliminating the head loss term, HL. Both of these variables are accounted
for by applying an orifice coefficient, C, to the right side of the equation. The
final form of the orifice equation becomes:

Q = CA 2 gH

where C = orifice coefficient.

When dealing with storm sewer design, the orifice coefficient is generally
about 0.6. For more in-depth information on orifice coefficients for different
situations, see Brater and King’s Handbook of Hydraulics (1996).

Weirs
Weirs are notches or gaps over which fluid flows. The lowest point of structure
surface or edge over which water flows is called the crest, whereas the stream of

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 19

 1 Basic Hydraulic Principles
Orifices and Weirs

water that exits over the weir is called the nappe (see Figure 1.8). Depending
on the weir design, flow may contract as it exits over the top of the weir, and,
as with orifices, the point of maximum contraction is called the vena contracta.
This contraction can be counteracted or suppressed by designing the weir such
that its shape conforms to the shape of the channel. This type of weir is called
a suppressed weir. With a contracted weir, the crest and nappe vary from the
channel to such a degree that a significant contraction of flow area does occur.
In addition to suppressed and contracted weir types, weirs are also distin-
guished as either sharp-crested or broad-crested. A sharp-crested weir has a
sharp upstream edge formed so that the nappe flows clear of the crest. Broad-
crested weirs have crests that extend horizontally in the direction of flow far
enough to support the nappe and fully develop hydrostatic pressures for at least
a short distance.
Weirs can also be distinguished by their shapes. The most common shapes are
shown in Figure 1.9. The effects of weir shape and other factors previously
mentioned are accounted for with modifications to the weir equation (derived
in Example 1.7), such as adjustments the weir coefficient. Table 1.3 contains
information on coefficients for V-notch weirs.
Weirs have the following applications:
• serving as emergency spillways for regulating high-return event flows over-
topping dams and detention ponds,
• regulating the flow in channels,
• measuring flow,
• approximating the flow over roadways acting as broad-crested weirs when
flow exceeds a culvert’s capacity (see Chapter 4),
• approximating the interception capacity of unsubmerged drainage inlets in
swales (see Chapter 3),
• approximating the flow allowed through an unsubmerged culvert operating
under inlet control (see Chapter 4).

Nappe

Crest

Figure 1.8: Front View of Common Weir.

20 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Orifices and Weirs

Weir type Figure Equation Coefficients

L Contracted Metric
Q = C(L– 0.1iH) H 3/2 C = 1.84
Rectangular H
Suppressed English
Q = CLH 3/2 C = 3.367
i = number of iterations
Sharp crested

C varies between
q H 3/2 0.611 and 0.570
V-notch Q = C 8 2g tanq H
15 depending on
2
H and Q*

L Metric Metric
Q = CLH3/2 C = 1.86
Cipolletti H 4:1
English English
Q = CLH 3/2 C = 3.367

Cd is a function
Hr
crested

Broad hr of Ht, ht and Lt

sharp-
Non

(side view) Q = CdLHr 3/2 ranging between

Lr 1.25 and 3.1*

*Refer to FlowMaster help documentation for more information.

Figure 1.9: Front View of Common Weir.

Table 1.3: V-Notch Weir Coefficients of Discharge—English Units

Head (ft) Weir Angle, θ (°)
22.5 30 45 60 90 120
0.5 0.611 0.605 0.569 0.590 0.584 0.581
1.0 0.593 0.590 0.583 0.580 0.576 0.575
1.5 0.586 0.583 0.578 0.575 0.572 0.572
2.0 0.583 0.580 0.576 0.573 0.571 0.571
2.5 0.580 0.578 0.574 0.572 0.570 0.570
3.0 0.579 0.577 0.574 0.571 0.570 0.570
Derived from: Van Havern, Bruce P. Water Resources Measurements,
American Water Works Association, 1986

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 21

 1 Basic Hydraulic Principles
Orifices and Weirs

Example 1.7: The Weir Equation

Derive the weir equation for the rectangular weir with a crest of length L and
head H, which discharges from a free outfall as shown in Figure 1.10.
Solution
Begin with the energy equation:

p1 V2 p V2
+ z1 + 1 + HG = 2 + z2 + 2 + HL
γ 2g γ 2g

List the known variables and assumptions:

• The datum is at the crest of the weir.
• P1/γ = H.
• Point 2 occurs at the vena contracta.
• The elevation heads, z1 and z2, are equivalent.
• Assume the velocity in the tank is negligible.
• The exiting stream pressure at point 2 is 0.
• There is no head gain.
After applying the known variables and assumptions and solving for V2, the
energy equation becomes:

V2 = 2 g (H − HL )

To find the flow, multiply both sides by the flow area, which in this case is the
length of the weir, L, multiplied by the height of the head, H.

P
=H
g
Datum 1 2

Figure 1.10: Weir Example.

22 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Computer Applications

AV2 = Q = LH 2 g ( H − HL )

where L = length of weir (m, ft).

To account for head loss, contraction, and other variables, a weir coefficient,
C, is applied. Unlike the orifice coefficient, the weir coefficient generally takes
into account the constant 2g. Finally, the weir equation becomes:

32
Q = CLH

where C = weir coefficient (m1/2/s, ft1/2/s)

Unlike the orifice coefficient, the weir coefficient is not unitless. Care has to be
taken that the correct coefficient is applied when using a specific unit system.
Typical coefficient values for different weir shapes, as well as variations in the
equation, can be seen in the previously shown in Figure 1.9.

1.7 Computer Applications

It is very important for students (and practicing engineers) to fully understand
the methodologies behind hydraulic computations. Once these concepts are
understood, the solution process can become repetitive and tedious—the type
of procedure that is well suited to computer analysis.
There are several advantages to using computerized solutions for common
hydraulic problems:
• The amount of time to perform an analysis can be greatly reduced.
• Computer solutions can be more detailed than hand calculations. Performing
a solution manually often requires many simplifying assumptions.
• The solution process may be less error-prone. Unit conversion and the
rewriting of equations to solve for any variable are just two examples of
mistakes frequently introduced with hand calculations. A well-tested com-
puter program helps to avoid these algebraic and numeric errors.
• The solution is easily documented and reproducible.
• Because of the speed and accuracy of a computer model, more comparisons
and design trials can be performed. The result is the exploration of more
design options, which eventually leads to better, more efficient designs.
In order to prevent an “overload” of data, this chapter deals primarily with
steady-state computations. After all, an introduction to hydraulic calculations
is tricky enough without throwing in the added complexity of a constantly
changing system.
The assumption that a system is under steady-state conditions is often perfectly
acceptable. Minor changes that occur over time or irregularities in a channel

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 23

 1 Basic Hydraulic Principles
FlowMaster and Tutorials

cross section are frequently negligible, and a more detailed analysis may not be
the most efficient or effective use of time and resources.
There are circumstances when an engineer may be called upon to provide a
more detailed analysis, including unsteady flow computations. For a storm
sewer, the flows may rise and fall over time as a storm builds and subsides. For
water distribution piping, a pressure wave may travel through the system when
a valve is closed abruptly (the same “water-hammer” effect can probably be
heard in your house if you close a faucet quickly).
As an engineer, it is important to understand the purpose of an analysis; other-
wise, appropriate methods and tools to meet that purpose cannot be selected.

1.8 FlowMaster and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software.

What Does FlowMaster Do?

FlowMaster is an easy-to-use program that helps civil engineers with the hydrau-
lic design and analysis of pipes, gutters, inlets, ditches, open channels, weirs,
and orifices. FlowMaster computes flows and pressures in conduits and chan-
nels using common head loss equations such as Darcy–Weisbach, Manning’s,
Kutter’s, and Hazen–Williams. The program’s flexibility allows the user to choose
an unknown variable and automatically compute the solution after entering
known parameters. FlowMaster also calculates rating tables and plots curves and
cross sections. You can view the output on the screen, copy it to the Windows
clipboard, save it to a file, or print it on any standard printer. FlowMaster data
can also be viewed and edited using tabular reports called FlexTables.
FlowMaster enables you to create an unlimited number of worksheets to ana-
lyze uniform pressure pipe or open-channel sections, including irregular sec-
tions (such as natural streams or odd-shaped man-made sections). FlowMaster
does not work with networked systems such as a storm sewer network or a
pressure pipe network. For these types of analyses, StormCAD, WaterGEMS,
or SewerCAD should be used instead.
The theory and background used by FlowMaster have been reviewed in this
chapter and can be accessed via the FlowMaster on-line help system. General
information about installing and running Bentley Systems software can be
found in Appendix A.

24 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Problems

FlowMaster replaces solutions such as nomographs, spreadsheets, and BASIC

programs. Because FlowMaster gives you immediate results, you can quickly
generate output for a large number of situations. For example, you can use
FlowMaster to:
• analyze various hydraulic designs,
• evaluate different kinds of flow elements,
• generate professional-looking reports for clients and review agencies.

1.9 Problems
Solve the following problems using the FlowMaster computer program. Unless
stated otherwise, assume water is at 20°C.
1. The cross section of a rough, rectangular, concrete (k = 0.2 × 10−2 ft)
channel measures 6 × 6 ft. The channel slope is 0.02 ft/ft. Using the
Darcy–Weisbach friction method, determine the maximum allowable
flow rate through the channel to maintain one foot of freeboard (free-
board is the vertical distance from the water surface to the overtopping
level of the channel). For these conditions, find the following characteris-
tics (note that FlowMaster may not directly report all of these):
(a) Flow area
(b) Wetted perimeter
(c) Hydraulic radius
(d) Velocity
(e) Froude number
2. A 450-mm circular concrete (n = 0.013) pipe constructed on a 0.6%
slope carries 0.1 m3/s.
(a) Using Manning’s equation and normal depth assumptions, what are
the depth and velocity of flow?
(b) What would the velocity and depth be if the pipe were constructed
of corrugated metal (n = 0.024) instead of concrete?
3. A trapezoidal channel carries 2.55 m3/s at a depth of 0.52 m. The channel
has a bottom width of 5 m, a slope of 1.00%, and 2H:1V side slopes.
(a) What is the appropriate Manning’s roughness coefficient?
(b) How deep would the water be if the channel carried 5 m3/s?
4. Use Manning’s equation to analyze an existing brick-in-mortar (n = 0.015)
triangular channel with 3H:1V side slopes and a 0.05 longitudinal slope.
The channel is intended to carry 7 cfs during a storm event.
(a) If the maximum depth in the channel is 6 in., is the existing design
acceptable?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 25

 1 Basic Hydraulic Principles
Problems

(b) What would happen if the channel were replaced by a concrete

(n = 0.013) channel with the same geometry?
5. A pipe manufacturer reports that it can achieve Manning’s roughness val-
ues of 0.011 for its concrete pipes, which is lower than the 0.013 reported
by its competitors. Using Kutter’s equation, determine the difference in
flow for a 310-mm circular pipe with a slope of 2.5% flowing at one-half
of the full depth.
6. A grass drainage swale is trapezoidal, with a bottom width of 6 ft and
2H:1V side slopes. Using the friction method you feel is appropriate,
answer the following questions:
(a) What is the discharge in the swale if the depth of flow is 1 ft and the
channel slope is 0.005 ft/ft?
(b) What would the discharge be with a slope of 0.010 ft/ft?

7. A paved highway drainage channel has the geometry shown in the follow-
ing figure. The maximum allowable flow depth is 0.75 ft (to prevent the
flow from encroaching on traffic), and the Manning’s n-value is 0.018 for
the type of pavement used.

20 ft 5 ft 10 ft

1 ft

0.5 ft
Not to scale

Figure for Problem 7

(a) What is the capacity of the channel given a 2% longitudinal slope?
(b) Create a rating curve to demonstrate how the capacity varies as the
channel slope varies from 0.5% to 5%. Choose an increment that
will generate a reasonably smooth curve.

8. Using the Hazen–Williams equation, determine the minimum diameter of

a new cast iron pipe (C = 130) for the following conditions: the upstream
end is 51.8 m higher than the downstream end, which is 2.25 km away.
The upstream pressure is 500 kPa, and the desired downstream pressure
and flow rate are 420 kPa and 11,000 l/min, respectively. What is the min-
imum diameter needed? Assume pipes are available in 50-mm increments.

9. 2,000 gal of water per minute flow through a level, 320-yard-long,

8-in.-diameter cast iron pipe (C = 130, k = 2.5908e−4 m) to a large
industrial site. If the pressure at the upstream end of the pipe is 64 psi,

26 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydraulic Principles
Problems

what will the pressure be at the industry? Is there a significant difference

between the solutions produced by the Hazen–Williams method and the
Darcy–Weisbach method?

10. Develop a performance curve for the pipe in Problem 9 that shows the
available flow to the industry with residual pressures ranging from 20 to
80 psi (assume the source can maintain 64 psi regardless of flow rate).
Create similar curves for 10- and 12-in. diameter pipes and compare the
differences in flow.

11. Using the Darcy–Weisbach equation, find the minimum size of circular
corrugated metal storm drain (k = 1.0 × 10−3 ft) that will carry 1.5 cfs
with a maximum depth of 6 in. The drain carries water down a hill 3-ft
high to a pond with a free outfall 75 ft away. What pipe size should be
used? Assume pipes are available in 3-in. increments. What would the
maximum capacity of this pipe be? What would the capacity of the pipe
be when it is flowing full?
12. A channel with the cross section shown in the following figure has a
Manning’s coefficient of 0.040 from stations 0 to 3 and 0.054 from sta-
tions 3 to 8. The flow through the channel is 13 m3/s, and the water
surface is 1.7-m high. Find the following:
(a) Weighted Manning’s coefficient
(b) Slope of the channel
(c) Top width
(d) Wetted perimeter
(e) Flow regime (supercritical or subcritical)

2.00
Elevation, m

1.50
1.00
0.50
0.00
0 1 2 3 4 5 6 7 8
Station, m

Figure for Problems 12 and 13

13. A stream with the cross section shown in the previous figure has a flow
rate of 5 m3/s. The stream has a longitudinal slope of 0.002 m/m and a
natural stony bottom (n = 0.050, stations 0–8).
(a) Using Manning’s equation, what is the water surface elevation of the
stream?
(b) What is the maximum capacity of the channel?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 27

 1 Basic Hydraulic Principles
Problems

(c) How would the capacity of the channel be affected if you were to pave
the center of the channel (n = 0.013) between stations 3 and 5?
14. A rectangular concrete channel with a width of 1 m and a height of 0.5 m
is on a slope of 0.008 m/m. Design a concrete circular channel for which
the depth is half of the diameter and the flow area is the same as that of
the rectangular channel. Which channel is more efficient and by how
much?
15. A weir is placed in a rectangular channel to measure the flow. The dis-
charge from the rectangular channel enters a trapezoidal channel with a
stony bottom. The trapezoidal channel is 0.50-m wide at the base with
2:1 (H:V ) equal side slopes. The weir is a sharp-crested, V-notch weir
with a crest 0.43 m above the channel bottom, a weir coefficient of 0.58,
and a notch angle of 1.57 radians. The height of the water above the weir
is 0.70 m, and the depth of water in the trapezoidal channel is measured
to be 0.40 m.
What is the flow rate? What is the slope of the trapezoidal channel (using
Manning’s formula)? If the discharge is increased until the elevation of
the water surface in the trapezoidal channel reaches 0.61 m, what will the
headwater elevation be at the weir?
16. The outlet structure on a pond is used to regulate the flow out of the
pond for different storm events. An outlet structure must be designed
to discharge 2.20 m3/min when the water surface elevation in the pond
reaches 1.52 m, and 6.29 m3/min when the water surface elevation reaches
2.60 m. The outlet structure will be a circular orifice and a sharp-crested
rectangular weir combination, with the centroid of the orifice at an ele-
vation of 0.90 m and the weir crest at an elevation of 2.50 m. Both will
discharge to free outfall conditions.
Assume an orifice coefficient of 0.6. Find the orifice diameter needed to
supply the correct discharge when the water surface reaches the first spec-
ified elevation. What will the discharge from the orifice be when the water
surface reaches the second specified elevation? Find the width of the weir
needed to supply the extra discharge necessary to meet the requirement.
Use Manning’s formula where necessary.
17. An approximately trapezoidal, clean, natural stream carries the discharge
from a pond down a 0.001 slope. The maximum depth in the channel is
0.5 m. The channel has equal side slopes of 3.0 (H:V) and a bottom width
of 1.0 m. The pond discharges water through a circular orifice into the
channel. The centroid of the orifice is located 1.0 m above the bottom of
the channel.
Assume an orifice coefficient of 0.6. Design the orifice to discharge the
maximum flow rate possible without exceeding the maximum allowed
depth in the channel when the water surface in the pond reaches 4.6 m
above the channel bottom. Use Manning’s formula when necessary.

28 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 2 Basic Hydrology

Hydrology is the study of water on and below the Earth’s surface and in the
atmosphere. The discussion in this chapter is limited to rainfall and the result-
ing surface runoff. Data on surface runoff flow rates and volumes provides
the loads used to design and evaluate storm sewers, culverts, ditches, and
detention ponds. Several of these structures are covered in more detail in other
chapters.

2.1 Rainfall
Rainfall data is fundamental for determining the amount of stormwater gener-
ated by a particular storm event. This section describes the properties of rainfall
used in modeling.

Basic Rainfall Characteristics

For the design engineer, the most important characteristics of rainfall are
• the depth, or volume, of rainfall during a specified time interval (or its
average intensity over that time interval),
• the duration of the rainfall,
• the area over which the rainfall occurs,
• the average recurrence interval of a rainfall amount, and
• the temporal and spatial distributions of rainfall within the storm.

Return Period and Frequency

The probability that a storm event of a certain magnitude will occur in any
given year is expressed in terms of event frequency and return period.
The return period, or the recurrence interval, represents, on average, the
length of time expected to elapse between rainfall events of equal or greater

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 29

 2 Basic Hydrology
Rainfall

magnitude. Although a recurrence interval is expressed in years, it is actually

based on probability of occurrence.

For example, a 5-year return period represents a storm event that is expected
to occur once every 5 years on average. This does not mean that two storm
events of that size will not occur in the same year, nor does it mean that the
next storm event of that size will not occur for another 5 years. It just means
that the average will be once every 5 years (that is, there is a 20% chance of
occurrence in any given year).

The frequency, or exceedance probability, is a measure of how often a specific

rainfall event will be equaled or exceeded and is simply the inverse of the return
period.

Example 2.1: Computing a Recurrence Interval

What is the recurrence interval for a storm event that has a 20% probability
of being equaled or exceeded within any given year? For one that has a 2%
probability of exceedance?

Solution
For a storm event with a 20% probability of being equal to or exceeded in a
given year, the recurrence interval is computed as

1/0.20 = 5 years

For an event with a 2% probability, the recurrence interval is

1/0.02 = 50 years

Types of Rainfall Data

Historic rainfall data is compiled and analyzed to predict storm characteristics.
Rainfall data is available from a variety of sources, including governmental
organizations and agencies. The data can be presented in various formats,
including intensity–duration–frequency (IDF) curves, cumulative rainfall
depths, and rainfall hyetographs. In infrastructure design, a synthetic rainfall
distribution is often applied to the total rainfall depth for a storm of given
duration and recurrence frequency. In the United States, the most frequently
applied synthetic rainfall distributions are four 24-h distributions devel-
oped by the Natural Resources Conservation Service, U.S. Department of
Agriculture.

Intensity–Duration–Frequency Data
For a selected storm duration, a rainfall intensity exists that corresponds to
a given exceedance probability or recurrence interval. A rainfall IDF curve
illustrates the average rainfall intensities corresponding to a particular storm

30 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall

225

200

10
0
ye
ar
175

s
150

Intensity, mm/h
10
ye
125 ar
s
100 2y
ear
s
75

50

25

0
0 15 30 45 60
Duration, min

Figure 2.1: Example set of IDF curves.

Table 2.1: Example IDF Table

Rainfall Intensities (mm/h)
Storm Return Period
Durations 2 Years 10 Years 100 Years
5 min 88 135 204
10 min 75 114 168
15 min 65 97 142
30 min 44 66 100
60 min 24 37 56

recurrence interval for various storm durations (see Figure 2.1). These curves
are the result of the statistical analysis of rainfall data for a particular area.
Given the information on the graph shown in Figure 2.1, you can determine
that the average 1-h rainfall intensity expected to be equaled or exceeded, on
average, once every 100 years is 56.0 mm/h.
Although graphical rainfall curves are acceptable for hand calculations, they are
not well-suited to computer analyses. Data is therefore input into hydrologic
software either as equations or in a tabular format. Creating a rainfall table
from a set of IDF curves is a simple matter of manually picking values from
the curves. For example, Table 2.1 can be created from the data presented in
the IDF curves of Figure 2.1.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 31

 2 Basic Hydrology
Rainfall

Computer programs commonly access IDF data in the form of an equation.

Several forms have been developed to analytically describe the graphical IDF
relationships. The most common forms of these equations are:

Q A i = a + b(lnD ) + c (lnD )2 + d (lnD )3

V = R=
A Pw
where
i = intensity of rainfall (mm/h, in/h);
D = rainfall duration (min or h);
RP = return period (years); and
a, b, c, d, m, and n are coefficients.
When applying IDF data to system design, you must use data developed for
the specific geographic location where the system is to be constructed. A single
set of IDF curves can normally be used for areas as large as a city or small
county. Many drainage jurisdictions and agencies such as weather bureaus can
provide the engineer with IDF data recommended for their particular geo-
graphical location. Engineers should understand when and by whom the IDF
curves were created, as more recently updated resources may be available.
In the United States, such information can be found in several National
Weather Service publications. For example, the NWS publication TP 40
(Hersfield, 1961) presents maps showing precipitation depths over the United
States for storm durations from 30 min to 24 h and for recurrence intervals
from 1 to 100 years. TP 40 was partially superceded by a later publication
known as HYDRO-35 for the central and eastern United States (Frederick et
al., 1977), and by the NOAA Atlas 2 for the 11 coterminous western states
(Miller et al., 1973). Updated atlases for the Midwestern United States (Huff
and Angle, 1992) and the northeastern United States and southeastern Canada
(McKay and Wilkes, 1995) have also been published.

Temporal Distributions and Hyetographs for Design Storms

Some types of hydrologic analyses require the distribution of precipitation
over the duration of the storm. A temporal rainfall distribution, such as the
one shown in Figure 2.2, shows the cumulative progression of rainfall depth
throughout a storm. The y-axis is represented by a simple rain gauge that fills
over the period represented on the x-axis. Total depth is simply the final depth
in the gauge. The average intensity (i) during any time segment is represented
by the slope of the rainfall curve during that interval. The steeper the slope, the
greater the average intensity is for a given segment.
The temporal distribution shown in Figure 2.2 can also be represented using a bar
graph that shows how much of the total rainfall occurs within each time interval
during the course of an event. A graph of this nature is called a rainfall hyeto-
graph. Hyetographs can be displayed in terms of incremental rainfall depth meas-
ured within each time interval as shown in Figure 2.3, or as the average intensity
calculated for each interval by dividing incremental depth by the time interval.

32 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall

Rain depth, in

r
in/h
=5

1.0 in
.i
Fills Average intensity for

Avg
over a given time slice
time 0.2 hr can be computed using
the incemental depth
for that time slice.

0.0 0.5 1.0 1.5 2.0 2.5

Figure 2.2: Temporal distribution of rainfall.

0.7

0.6
Incrmental rainfall depth, in

0.5

0.4

0.3

0.2

0.1

0
0.0 0.10.20.3 0.4 0.50.60.7 0.80.9 1.01.11.21.3 1.4 1.5 1.61.71.81.92.02.1 2.22.32.42.5
Time, h

Figure 2.3: A hyetograph of incremental rainfall depth versus time.

The design of stormwater management facilities typically requires a complete rain-

fall hyetograph. For such situations, engineers commonly use synthetic temporal
distributions of rainfall, which are essentially systematic and reproducible methods
for varying the rainfall intensity throughout a design event.
The selected length Δt of the time increment between the data points used to
construct a temporal rainfall distribution depends on the size (area) and other
characteristics of the drainage basin. As a rule of thumb, the time increment
should be no larger than about one-fourth to one-fifth of the basin lag time (tl),
or about one-sixth of the time of concentration (tc) of the basin. In any case,
the smallest time increment for which rainfall data is generally available is

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 33

 2 Basic Hydrology
Rainfall

about 5 min. In small urban drainage basins where it is often necessary to use
time increments as small as 1 or 2 min, the data must be extrapolated.
In addition to selecting an appropriate Δt, the engineer must select the total
duration to be used when developing a design storm hyetograph. In many
cases, the storm duration will be specified by the review agency having jurisdic-
tion over the area in which a stormwater conveyance facility will be built; this
approach promotes consistency from one design to another.
Many methods have been proposed for distributing a total rainfall depth
throughout a storm to develop a design storm hyetograph. The NRCS devel-
oped one of the more commonly used methods in the United States (SCS,
1986). With this method, Table 2.2 is used to find fractions of the total

Table 2.2: SCS Dimensionless Storm Distributions (SCS, 1986)

t (h) Type I Type IA Type II Type III
0 0.000 0.000 0.000 0.000
1 0.017 0.020 0.011 0.010
2 0.035 0.050 0.022 0.020
3 0.055 0.082 0.035 0.031
4 0.076 0.116 0.048 0.043
5 0.099 0.156 0.063 0.057
6 0.126 0.206 0.080 0.072
7 0.156 0.268 0.098 0.091
8 0.194 0.425 0.120 0.114
9 0.254 0.520 0.147 0.146
10 0.515 0.577 0.181 0.189
11 0.624 0.624 0.235 0.250
12 0.682 0.664 0.663 0.500
13 0.728 0.701 0.772 0.750
14 0.766 0.736 0.820 0.811
15 0.799 0.769 0.854 0.854
16 0.830 0.800 0.880 0.886
17 0.857 0.830 0.902 0.910
18 0.882 0.858 0.921 0.928
19 0.905 0.884 0.937 0.943
20 0.926 0.908 0.952 0.957
21 0.946 0.932 0.965 0.969
22 0.965 0.956 0.978 0.981
23 0.983 0.978 0.989 0.991
24 1.000 1.000 1.000 1.000

34 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall

1.0
0.9
II
0.8
0.7

Rain fraction
0.6
III
0.5
0.4
IA I
0.3
0.2
0.1
0.0
0 3 6 9 12 15 18 21 24
Time, h

Figure 2.4: Graphical representation of NRCS (SCS) rainfall distributions.

accumulated rainfall depth for storms with 24-h durations. (Figure 2.4 depicts
Table 2.2 graphically.) The storms are classified into various types, with each
type being recommended for use in a certain U.S. geographical region, as
shown in Figure 2.5. If necessary, interpolation may be used to obtain values
not shown in Table 2.2. Nonlinear interpolation methods are recommended
for this purpose.

Example 2.2: Developing a Design Storm Hyetograph from

SCS Distributions
Develop a design storm hyetograph for a 50-year, 24-h storm in Boston,
Massachusetts. Assume that Δt = 0.1 h is a reasonable choice for the drainage
basin to which the design storm will be applied.

Solution
Figure 2.5 illustrates that a Type III storm distribution is a reasonable choice
for Boston. From TP 40, the total depth of a 50-year, 24-h storm in Boston
is estimated to be 6.0 in. Table 2.3 illustrates the calculation of the storm
hyetograph.
The first column of the table is the time, in hours, since the beginning of the
storm, and is tabulated in 1-h increments for the total storm duration of 24 h.
(In actuality, the Δt used in the calculations would be 0.1 h; the 1-h increment
is used here for brevity.) The second column is the fraction of the total storm
depth that has accumulated at each time during the storm. These values are
obtained by interpolation from Table 2.2 for the Type III storm distribution.
The third column contains the cumulative rainfall depths for each time during
the storm and is obtained by multiplying each fraction in the second column
by the total storm depth of 6.0 in. The fourth column contains the incremental

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 35

 2 Basic Hydrology
Rainfall

IA

III

III

I
Rainfall
distribution
Type I
Type IA
Type II
Type III
III
III

Figure 2.5: Coverage of NRCS (SCS) rainfall distributions (Adapted from SCS, 1986).

Table 2.3: 50-Year, 24-H Storm Hyetograph for Boston, Massachusetts

t (h) Fraction Cum. P (in) Incr. P (in)
0 0.000 0.000
1 0.010 0.060 0.060
2 0.020 0.120 0.060
3 0.031 0.186 0.066
4 0.043 0.258 0.072
5 0.057 0.342 0.084
6 0.072 0.432 0.090
7 0.091 0.546 0.114
8 0.114 0.684 0.138
9 0.146 0.876 0.192
10 0.189 1.134 0.258
11 0.250 1.500 0.366
12 0.500 3.000 1.500

36 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall

Table 2.3: (Continued)

13 0.750 4.500 1.500
14 0.811 4.866 0.366
15 0.854 5.124 0.258
16 0.886 5.316 0.192
17 0.910 5.460 0.144
18 0.928 5.568 0.108
19 0.943 5.658 0.090
20 0.957 5.742 0.084
21 0.969 5.814 0.072
22 0.981 5.886 0.072
23 0.991 5.946 0.060
24 1.000 6.000 0.054

5
Precipition, in

0
0 4 8 12 16 20 24
Time, h

Figure 2.6: Graph of derived design storm cumulative precipitation.

depths of rainfall within each time interval during the storm; these values are
computed as the difference between the current and preceding values in the
third column.
The resulting graph of cumulative precipitation is shown in Figure 2.6, and the
hyetograph is shown in Figure 2.7. The height of each bar on the hyetograph
is the average rainfall intensity during that time interval, and the area of each
bar is the incremental rainfall depth during that time interval. Because the time
increment is 1 h (1-h increment is shown for simplicity; actual Δt is 0.1 h), the
value for the height of the bar (in units of in./h) is equal to the incremental
depth for that time increment (in in).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 37

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

1.6
1.4

1.2

Intensity, in/h
1.0
0.8

0.6
0.4
0.2
0.0
0 to 1

2 to 3

4 to 5

6 to 7

8 to 9

10 to 11

12 to 13

14 to 15

16 to 17

18 to 19

20 to 21

22 to 23
Time, h

Figure 2.7: Derived design storm hyetograph.

2.2 Rainfall Abstractions and Runoff Volume

Only a portion of the total rainfall occurring over a drainage basin contributes
to surface runoff and stream flow. In fact, a simple comparison of rainfall and
runoff records for most locations in the United States shows that the equivalent
depth of runoff is typically about 30%–50% of the precipitation depth.
To obtain the loading information necessary to design and model structures
such as storm sewers, culverts, and detention facilities, it is necessary to calculate
the runoff volumes and/or flow rates resulting from the storm events of inter-
est. This section presents three techniques for computing total runoff volume:
(1) the Horton equation, (2) the runoff coefficient, and (3) the NRCS curve
number equation.

Watershed Area
A watershed is an area that drains to a single point of discharge. The first step
in computing runoff at a point of interest (such as a culvert, inlet, or property
outfall) is to delineate its watershed on a contour map. Because water flows
downhill, delineating a watershed is simply a matter of identifying an outfall
and locating the watershed boundary such that any rain that falls within the
boundary will be directed toward that point of discharge. The delineated area
is then measured using a planimeter, or graphical or computer-aided meth-
ods. Because river and stream systems collect water, a watershed may have any
number of subwatersheds within it. The focus of the analysis and the determi-
nation of how large a watershed must be to be analyzed separately depend on
the scope and purpose of the project at hand. Figure 2.8 shows the collection
channels for two typical natural watersheds.

38 Computer Application in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Discharge
points
(outfalls)

Figure 2.8: Typical natural watersheds with collection channels.

After the watershed is delineated and measured, the modeler can proceed to
the next step of computing rainfall abstractions and runoff depth. To deter-
mine the total runoff volume, runoff depth is multiplied by the watershed area.

Rainfall Abstractions
Rainfall that does not contribute to direct surface runoff may be intercepted
by vegetation, infiltrated into the ground surface, retained as depression stor-
age in puddles and small irregularities in the land surface, or returned to the
atmosphere through transpiration and evaporation. Collectively, these losses of
rainfall are called abstractions. The rainfall that remains after abstractions have
occurred comprises the surface runoff and is called effective precipitation.
The four basic types of abstractions are interception, depression storage, infil-
tration, and evaporation. Evaporation, however, is not typically considered in
modeling stormwater conveyances.

Interception
Interception refers to the capture of rainfall on the leaves and stems of vege-
tation before it reaches the ground surface. Water intercepted by vegetation is
returned to the atmosphere by evaporation during dry-weather periods.

Depression Storage
Excess water begins to pond on the land surface when the rainfall intensity
exceeds the infiltration capacity of a soil during a storm event. The ponded
water fills small depressions and irregularities in the ground surface, and
additional water is held on the surface through the phenomenon of sur-
face tension. The water held in depressions is called depression storage, and
it either evaporates during dry-weather periods or infiltrates into the soil.
Noninfiltrated rainfall that remains after surface depressions and irregularities
have been filled contributes to surface runoff.

Infiltration
When rainfall occurs on a pervious surface, some of the rainwater infiltrates
into the ground. The infiltrated water may contribute to groundwater recharge,
or it may be taken up by the roots of vegetation and subsequently transpired
through the leaves of plants. Infiltrated water also may be evaporated from the

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 39

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

soil during dry-weather periods between storm events, or it may move laterally
through the near-surface soils and reappear as surface water in a stream.

Determining Runoff Volume

In its most general form, the relationship of runoff (effective precipitation) to
infiltration and abstractions is (see Figure 2.9):

Dr = Dp−Dli−Di−Ds−De for Dp>Dli and Dr>Ds

Dr = 0 for Dp≤Dli or Dr≤Ds

where
Dr = total depth of runoff (effective precipitation),
Dp = total depth of precipitation (rainfall),
Dli = total initial loss (interception),
Di = total depth infiltrated after initial losses,
Ds = total depression storage depth, and
De = transpiration and evaporation losses (often ignored for short-duration
stormwater events).
Volume of runoff is then expressed by

Vr = Dr × A
where
Vr = runoff volume,
A = contributing drainage area, and
Dr = total depth of runoff.

De
Dr
Drainage area
Ds
Dp
Dli

Di
Volume of runnoff, Vr

Figure 2.9: Precipitation, abstraction, and runoff volume.

40 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Note: Some hydrology literature refers to runoff volume in terms of depth

(in in. or mm), in which case the depth would be multiplied by the contribut-
ing area to get the true volume.
Example 2.3: Computing Runoff Volume
A 2.5-h storm deposits 2.0 in. of rainfall over a catchment (watershed) area
of 3 ac. If the interception capacity is 0.3 in., the depression storage is 0.2 in.,
the calculated total infiltration is 0.7 in., and evaporation is negligible, find
the total depth of effective precipitation and the total runoff volume for the
drainage area.

Solution
Total runoff depth is computed as

Dr = 2.0−0.3−0.7−0.2−0 = 0.8 in

And total runoff volume is

Vr = (0.8 in.÷12 in/ft) × 3 ac = 0.2 ac. ft = 8,712 ft3

Although the concept of computing runoff volume just described is a simple

one, the problem of determining the amount of rainfall lost to abstractions
remains. The following subsections describe a few of the methods used in
accounting for abstractions—the Horton equation, the runoff coefficient, and
the NRCS (SCS) curve number equation.

Horton Infiltration Equation

A widely used method of representing the infiltration capacity of a soil is the
Horton equation (Horton, 1939). The Horton equation models a decreasing
rate of infiltration over time, which implies that the rate of infiltration decreases
as the soil becomes more saturated. For conditions in which the rainfall inten-
sity is always greater than the infiltration capacity (that is, the rainwater supply
for infiltration is not limiting), this method expresses the infiltration rate as

f (t ) = fc + ( f0 − fc )e−k (t −t0 )

where
f (t) = the infiltration rate (in/h or mm/h) at time t (min),
f c = a steady-state infiltration rate that occurs for sufficiently large t,
f 0 = the initial infiltration rate at the time that infiltration begins (that is, at
time t = t0), and
k (min−1) = a decay coefficient.
It can be shown theoretically that the steady-state infiltration rate fc is equal to
the saturated vertical hydraulic conductivity of the soil.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 41

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Table 2.4: Typical Values of Horton Infiltration Parameters (Rawls et al.,

1976)
Soil Type f0 (in/h) fc (in/h) k (min−1)
Alphalpha loamy sand 19.0 1.4 0.64
Carnegie sandy loam 14.8 1.8 0.33
Dothan loamy sand 3.5 2.6 0.02
Fuquay pebbly loamy sand 6.2 2.4 0.08
Leefield loamy sand 11.3 1.7 0.13
Tooup sand 23.0 1.8 0.55

Estimation of the parameters fc , f0, and k in the preceding equation can be

difficult because of the natural variabilities in antecedent moisture condi-
tions and soil properties. Table 2.4 provides some values recommended by
Rawls et al. (1976), although such tabulations should be used with caution.
Singh (1992) recommends that f0 be taken as roughly five times the value of fc.
For large rain events with a high rainfall intensity that is greater than the infil-
tration rate, it is possible to estimate the runoff by integrating the Horton
equation. The area under the Horton curve would be the maximum potential
infiltration, and then by subtracting the maximum potential infiltration from
the rainfall depth, the runoff can be estimated. The integrated form of the
Horton equation to estimate the maximum potential infiltration is

fo − fc
Fmax = fc t +
k
(1 − e−kt )

where Fmax = maximum potential infiltration depth (in. or mm).

Often, the rainfall intensity during the early part of a storm is less than the
potential infiltration capacity of the soil; thus the supply of rainwater is a lim-
iting factor on the infiltration rate. During the time period when the water
supply is limiting, the actual infiltration rate is equal to the rate at which rain-
water is supplied to the ground surface. This effect is illustrated in Figure 2.10.
Later in the storm when the rainfall rate is greater than the infiltration rate, the
actual infiltration rate will be greater than that predicted by the preceding equa-
tion because infiltration was limited early in the storm. In this case, the runoff
is estimated by integration of the rainfall profile above the infiltration curve.

Land Use and Cover Considerations

When applying a physically based equation such as Horton’s, care should be
taken to consider the impacts of land use and cover conditions. If the param-
eters used to solve for infiltration consider soil conditions only, the computed
runoff may not be applicable.

42 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Rainfall intensity, infiltration rate

Total rainfall = P

Effective rainfall Loss rate

Abstractions and infiltration

Time

Figure 2.10: Rainfall intensity can limit infiltration rate during the early
part of the storm.

Example 2.4: Using Horton’s Equation to Determine Infiltration Rate

Using Horton’s equation, find the infiltration rate at time 2.0 h for Dothan
loamy sand. The rainfall duration is 4 h, and for this event, it takes 0.5 h to
reach the interception capacity, at which time infiltration begins.

Solution
From Table 2.4, f0 = 3.5 in/h; fc = 2.6 in/h; and k = 0.02 min−1. Thus f (2 h) =
2.6 in/h + (3.5−2.6 in/h)e−0.02(2−0.5) = 3.47 in/h.

Example 2.5: Using Horton’s Equation to Estimate Runoff

Using integrated form of Horton’s equation, estimate the runoff depth for
Dothan loamy sand for a 1-h rain event that produced 3.55 in. of rainfall.
Assume the rainfall intensity is always greater than the infiltration rate.

Solution
From Table 2.4, f0 = 3.5 in/h; fc = 2.6 in/h; and k = 0.02 min−1. Thus:

3.5 in/h − 2.6 in/h

Fmax = 2.6 in/h × 1hr +
0.02/min × 60 min/h
(
1− e
−0.02 /miin×60 min/hr×1 hr
)
= 3.12 inches

Runoff = 3.55 in − 3.12 in = 0.43 inches.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 43

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Runoff Coefficient
The concept of the runoff coefficient assumes that the effective rainfall intensity
ie is a fraction of the gross rainfall intensity i, or

ie = Ci

where C = runoff coefficient (0≤C≤1).

A good way to visualize the C coefficient is to treat it in terms of percent
of rainfall. For example, a C of 0.85 would yield a runoff amount (effective
precipitation) that is 85% of the gross rainfall. The C coefficient is used with
the rational method for peak runoff rate estimation, which is described in
Section 2.3.
Example 2.6: Using the Runoff Coefficient to Compute Effective Rainfall
Intensity
For a rainfall event with an average intensity of 2.0 in/h falling onto a drainage
area having a C coefficient = 0.70, determine the effective rainfall intensity.

Solution
ie = Ci = 0.70 × 2.0 in/h = 1.4 in/h

Table 2.5 is a listing of recommended runoff coefficients corresponding to var-

ious types of land uses. It should be noted, however, that some locales have
developed runoff coefficient tables that also consider soil type. Coefficients
should be selected carefully for proper application to the particular locale and
soil conditions.
When a drainage basin consists of a mixture of land uses, a composite runoff
coefficient may be computed for the basin by weighting individual runoff
coefficients for each land use by their respective areas, as demonstrated in
Example 2.7.

Table 2.5: Runoff Coefficients for Use in the Rational Method (Schaake
et al., 1967)
Type of Area or Development C
Types of Development
Urban business 0.70–0.95
Commercial office 0.50–0.70
Residential development
Single-family homes 0.30–0.50
Condominiums 0.40–0.60

44 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Table 2.5: (Continued)

Type of Area or Development C
Apartments 0.60–0.80
Suburban residential 0.25–0.40
Industrial development
Light industry 0.50–0.80
Heavy industry 0.60–0.90
Parks, greenbelts, cemeteries 0.10–0.30
Railroad yards, playgrounds 0.20–0.40
Unimproved grassland or pasture 0.10–0.30
Types of Surface Areas
Asphalt or concrete pavement 0.70–0.95
Brick paving 0.70–0.80
Roofs of buildings 0.80–0.95
Grass-covered sandy soils
Slopes 2% or less 0.05–0.10
Slopes 2%–8% 0.10–0.16
Slopes over 8% 0.16–0.20
Grass-covered clay soils
Slopes 2% or less 0.10–0.16
Slopes 2%–8% 0.17–0.25
Slopes over 8% 0.26–0.36

Example 2.7: Determining the Weighted Runoff Coefficient

Estimate the runoff coefficient for a drainage basin that is made up of 15 ac of
park and 30 ac of medium-density, single-family housing.

Solution
From Table 2.5, the runoff coefficients for the park and residential areas are
estimated to be 0.20 and 0.40, respectively. The composite runoff coefficient
for the entire drainage basin of 45 ac is therefore

C = [15(0.20) + 30(0.40)]/45 = 0.33

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 45

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

NRCS (SCS) Curve Number Method

In the 1950s, the U.S. Department of Agriculture Soil Conservation Service
(SCS; now the Natural Resource Conservation Service or NRCS) developed a
procedure to partition the total depth of rainfall represented by a design storm
hyetograph into initial abstractions, retention, and effective rainfall (SCS,
1969).
Initial rainfall abstractions consist of all rainfall losses occurring before the
beginning of surface runoff, including interception, infiltration, and depres-
sion storage. Retention refers to the continuing rainfall losses following the
initiation of surface runoff, which are predominantly due to continuing infil-
tration. Conservation of mass requires that

F = P−Ia−Pe

where
F = equivalent depth of retention (in., mm),
P = total rainfall depth in storm (in., mm),
Ia = equivalent depth of initial abstractions (in., mm), and
Pe = depth of effective precipitation (in., mm).
An assumption made in the development of the curve number method is

F Pe
=
S P − Ia

where S = maximum possible retention (in., mm).

The value of the maximum possible retention, S, depends on the soil type and
land use condition in the drainage basin. The value of S does not include Ia.
In essence, the assumption represented by this equation is that the ratio of
actual retention to maximum possible retention of water during a storm is
equal to the ratio of effective rainfall to maximum possible effective rainfall
(total rainfall less initial abstractions). Substitution of F = P−Ia−Pe into the
previous equation yields

( P − I a )2
Pe =
( P − Ia ) + S

for values of P > Ia. Data analyzed by the NRCS indicated that Ia is related
to S and on average supported the use of the relationship Ia = 0.2S. Thus the
equation becomes

( P − 0.2S )2
Pe =
P + 0.8S

46 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

where P > 0.2S (Pe = 0 when P ≤ 0.2S). Because the initial abstraction Ia
consists of interception, depression storage, and infiltration prior to the onset
of direct runoff, it may be appropriate in some applications to assume that
Ia = 0.1S or Ia = 0.3S instead of Ia = 0.2S. For example, the relationship
Ia = 0.1S might be appropriate in a heavily urbanized area where there is little
opportunity for initial abstractions to occur. The preceding equation must be
modified when the relationship between Ia and S is assumed to be different
from Ia = 0.2S.
The use of this equation for estimating the depth of effective rainfall during
a storm requires an estimate of the maximum possible retention S. NRCS
conducted research to approximate S for various soil and cover conditions.
In order to provide engineers with tables having a manageable range of coeffi-
cients that varying between 0 and 100, the original values for S were rearranged
using the following simple relationship:

1000
CN =
S + 10
where
CN = runoff curve number and
S = maximum possible retention (in).
Rearranging, S is related to the curve number, CN, as

1000
S= − 10
CN

Practical values of CN range from about 30 to 98, with larger values being
associated with more impervious land surfaces. The NRCS has tabulated curve
numbers as a function of soil type, land use, hydrologic condition of the drain-
age basin, and antecedent moisture condition.

Soil Group
Soils in the United States have been classified by the NRCS into four hydro-
logic groups: Groups A, B, C, and D. Group A soils have high infiltration rates
(low runoff potential), even when they are thoroughly wetted. Typical Group A
soils are well-drained sands and gravels. Group D soils are at the opposite end
of the spectrum, having low infiltration rates (high runoff potential). Typical
Group D soils are clays, shallow soils over nearly impervious material, and soils
with a high water table. Group B and Group C soils are in the midrange of
the spectrum.
In the United States, hydrologic soil group information can be determined
from the NRCS soil survey of the county in which the project is located. Note
that when a drainage basin undergoes urbanization, the hydrologic soil group
may change due to compaction of the soil by heavy construction equipment or
mixing of soils as a consequence of grading operations.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 47

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Cover Type
The surface conditions of a drainage area have a significant impact on runoff.
For example, in the exaggerated case of a sandy (Group A) soil completely
paved with asphalt, the soil itself will have no impact on the amount of runoff.
Even for pervious conditions, cover type plays a significant role in the amount
of runoff from a site. For example, a heavily forested area will yield runoff
volumes that differ from those of a lawn or plowed field.

Hydrologic Condition
The hydrologic condition of rangeland, meadow, or pasture is defined to be
“good” if it is lightly grazed and has vegetative cover on more than 75% of the
area. Conversely, a “poor” hydrologic condition corresponds to a heavily grazed
area with vegetation covering less than 50% of the surface.

Antecedent Moisture Condition

When rainfall events occur in quick succession, the time period between
storms may be too short for the soils to dry to their average or normal mois-
ture conditions. When rainfall occurs on soils that are already wet, the net
result is that runoff volumes and peaks will be higher than normal. The NRCS
(SCS) method accounts for this possibility by allowing the curve number to
depend on an antecedent moisture (AMC) condition. Three AMC classifica-
tions exist. Normal conditions correspond to AMC-II. AMC-I corresponds to
a drier condition, and AMC-III to a wetter condition.
Curve numbers corresponding to AMC-I and AMC-III conditions can be
computed from AMC-II condition curve numbers using the following equa-
tions (Chow, 1959). The computed CN should be rounded to the nearest
whole number.

4.2CN II
CN I =
10 − 0.058CN II

23CN II
CN III =
10 + 0.13CN II

where CNI , CNII , and CNIII = curve numbers for AMC-I, -II, and -III,
respectively.

Curve Number Tables

Table 2.6 provides a listing of curve numbers that account for both cover con-
ditions and soil type for normal antecedent moisture conditions (AMC-II).
The curve numbers shown for the urban and suburban land use conditions
are based on the percentages of directly connected impervious areas in the
drainage basin as shown in the table and should be used with caution when
the actual percentage of imperviousness in a drainage basin differs from this

48 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Rainfall Abstractions and Runoff Volume

Table 2.6: Runoff Curve Numbers for Urban Areas (SCS, 1986)1

Average Percent Curve Numbers for

Cover Type and Hydrologic Condition Impervious Area² Hydrologic Soil Group
A B C D
Fully developed urban areas (vegetation established)
Open space (lawns, parks, golf courses, cemeteries, etc.)³:
Poor condition (grass cover<50%) 68 79 86 89
Fair condition (grass cover 50%–75%) 49 69 79 84
Good condition (grass cover>75%) 39 61 74 80
Impervious areas:
Paved parking lots, roofs, driveways, etc. (excluding 98 98 98 98
right-of-way)
Streets and roads:
Paved; curbs and storm sewers (excluding right-of-way) 98 98 98 98
Paved; open ditches (including right-of-way) 83 89 92 93
Gravel (including right-of-way) 76 85 89 91
Dirt (including right-of-way) 72 82 87 89
Western desert urban areas:
Natural desert landscaping (pervious area only)4 63 77 85 88
Artificial desert landscaping (impervious weed barrier, desert 96 96 96 96
shrub with 1–2 in. sand or gravel mulch and basin borders)
Urban districts:
Commercial and business 85 89 92 94 95
Industrial 72 81 88 91 93
Residential districts by average lot size:
1/8 ac or less (town houses) 65 77 85 90 92
1/4 ac 38 61 75 83 87
1/3 ac 30 57 72 81 86
1/2 ac 25 54 70 80 85
1 ac 20 51 68 79 84
2 ac 12 46 65 77 82
Developing urban areas
Newly graded area (pervious areas only no vegetation)5 77 86 91 94
Idle lands (CNs are determined using cover types similar to those in 2-2(c) (SCS, 1986)
1
Average runoff condition, and Ia = 0.2 s.
2
The average percent impervious area shown was used to develop the composite CNs. Other assumptions are as follows:
impervious areas are directly connected to the drainage system, impervious areas have a CN of 98, and pervious areas are
considered equivalent to open space in good hydrologic condition.
3
CNs shown are equivalent to those of pasture. Composite CNs may be computed for other combinations of open space cover type.
4
Composite CNs for natural desert landscaping should be computed using Figures 2.3 or 2.4 (in SCS, 1986) based on the
impervious area percentage (CN = 98) and the pervious area CN. The pervious area CNs are assumed equivalent to desert
shrub in poor hydrologic condition.
5
Composite CNs to use for the design of temporary measures during grading and construction should be computed using
Figures 2.3 or 2.4 (in SCS, 1986) based on the degree of development (impervious area percentage) and the CNs for the
newly graded pervious areas.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 49

 2 Basic Hydrology
Rainfall Abstractions and Runoff Volume

assumed value. When necessary, a composite curve number can be developed

as an area-weighted average of individual curve numbers. The composite CN
should be rounded to the nearest whole number.
Additional information on hydrologic condition and curve numbers for land
uses other than those contained in Table 2.6 can be found in the National
Engineering Handbook, Section 4: Hydrology (NEH-4) (SCS, 1969) and
TR-55 (SCS, 1986).

Example 2.8: Estimating Runoff Depth and Volume Using the NRCS
(SCS) Curve Number Method (Modified from SCS, 1986)
Estimate the curve number, depth of runoff (effective precipitation), and
runoff volume for a 1,000-ac drainage basin if the total depth of precipitation
is 5.0 in. All soils in the basin are in hydrologic soil group C. The proposed land
use is 50% detached houses with 0.25-ac lots; 10% townhouses with 0.125-
ac lots; 25% schools, parking lots, plazas, and streets with curbs and gutters;
and 15% open space, parks, and schoolyards with good grass cover. Use an
antecedent soil moisture condition of AMC-III. The detached housing and
townhouse areas have directly connected impervious area percentages similar
to those assumed in Table 2.6.

Solution
The composite curve number corresponding to AMC-II conditions is com-
puted as a weighted average, as shown in Table 2.7.
Thus the composite CN is computed as

CN = 86,100 ÷ 1,000 = 86

The curve number corresponding to AMC-III moisture conditions is found

to be

23(86 )
CN III = = 93
10 + 0.13(86 )

Table 2.7: Calculations for Example 2.8

Land Use Area (ac) CN Product
Detached houses (0.25-ac lots) 500 83 41,500
Townhouses (0.125-ac lots) 100 90 9,000
Streets, plazas, etc. 250 98 24,500
Open space, parks, etc. 150 74 11,100
Sum 1,000 86,100

50 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

The maximum possible retention for this basin at AMC-III is

S = (1,000/93) − 10 = 0.75 in

Initial abstractions are estimated to be

Ia = 0.2S = 0.15 in

Because P > Ia, the depth of runoff (effective precipitation) is estimated as

Pe = [5.0 − 0.2(0.75)]2/(5.0 + 0.8(0.75)] = 4.20 in

The total volume of runoff is therefore

Vr = (4.2 in/12 in/ft) (1,000 ac) = 350 ac. ft

2.3 Computing Peak Runoff Flow Rate

To design stormwater conveyance or detention systems, the engineer must
obtain information on runoff flow rates. Two basic levels of analysis exist. The
first level is a peak flow calculation to determine the maximum runoff flow rate
at a given point resulting from a storm event. This level of analysis is often suffi-
cient for designing storm sewers and culverts whose only function is to convey
runoff away from areas where it is unwanted. The second level, which is more
complex, consists of the generation of a runoff hydrograph, which provides
information on flow rate versus time and runoff volume. This type of informa-
tion is necessary in the analysis of detention and retention facilities where time
and volume considerations are critical. In some situations, hydrograph analysis
is also required for storm sewer analyses.
After you determine the required analysis approach, you must choose tech-
niques for computing how much of the rainfall will contribute to surface run-
off and, in the case of a hydrograph analysis, how this flow will be distributed
through time. These techniques are sometimes dictated by local regulations.

Time of Concentration
The maximum amount of flow coming from any watershed is related to the
amount of time it takes for the entire watershed to be contributing to the flow.
In other words, it may start raining right now, but it could be several minutes
(or even hours) before the water that falls on some parts of the watershed
actually makes its way to the discharge point.
Some places in a watershed are hydraulically closer to the discharge point
than others, but for peak flow generation only the most hydraulically remote

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 51

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

location is considered crucial. The amount of time that it takes for the first
drop of water falling on this location to work its way to the discharge point is
called the time of concentration.
Many methods for calculating time of concentration are given in various
private, federal, and local publications. Although each of these methods
is different (in some cases only slightly), they are all based on the type of
ground cover, the slope of the land, and the distance along the flow path.
In most locales, there is also a minimum time of concentration (typically
5–10 min) recommended for small watersheds, such as a section of a park-
ing lot draining to a storm sewer. Some methods predict the response time
directly, whereas others predict the velocity of flow. The predicted velocity
coupled with estimates of the flow path length can then be used to estimate
the response time.
With few exceptions, methods for prediction of basin response time are
empirical in nature. Consequently, large errors in response time estimates
can be expected to occur if these methods are not carefully selected and
applied, and these errors can significantly affect peak runoff estimates. The
method selected for estimation of tc should be one that was developed for
basin conditions similar to those in the drainage basin for which an estimate
is desired.
Table 2.8 lists several commonly used methods for estimation of the time of
concentration of a drainage basin. Figure 2.11 illustrates average overland flow
velocities as a function of land use characteristics and surface slope. When flows

Table 2.8: Commonly Used Methods for Estimation of the Time of Concentration, in minutes
Equation Source Remarks
0.4
tc = 60LA /DS
0.2 Williams (1922) L = basin length (mi), A = basin area (mi2),
D = diameter (mi) of a circular basin of
area A, and S = basin slope (%). The basin
area should be smaller than 50 mi2

tc = KL
0.77
/S n Kirpich (1940) Developed for small drainage basins in
Tennessee and Pennsylvania, with basin
areas from 1 to 112 ac. L = basin length
(ft), S = basin slope (ft/ft), K = 0.0078
and n = 0.385 for Tennessee; K = 0.0013
and n = 0.5 for Pennsylvania. The esti-
mated tc should be multiplied by 0.4 if the
overland flow path is concrete or asphalt,
or by 0.2 if the channel is concrete-lined

52 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

Table 2.8: (Continued)

Equation Source Remarks

tc = (2 LN /3S
0.5 0.47
) Hathaway (1945), Kerby (1959) Drainage basins with areas of less than 10 ac
and slopes of less than 0.01. This is an over-
land flow method. L = overland flow length
from basin divide to a defined channel (ft),
S = overland flow path slope (ft/ft), and
N is a flow retardance factor (N = 0.02
for smooth impervious surfaces; 0.10 for
smooth, bare packed soil; 0.20 for poor
grass, row crops, or moderately rough bare
surfaces; 0.40 for pasture or average grass;
0.60 for deciduous timberland; and 0.80 for
conifer timberland, deciduous timberland
with deep ground litter, or dense grass)

tc = 300( L /S )0.5 Johnstone and Cross (1949) Developed for basins in the Scotie and
Sandusky River watersheds with areas
between 25 and 1,624 mi2. L = basin
length (mi), and S = basin slope (ft/mi)

tc = {(0.007 I + c )/S
0.33
}× Izzard (1946) Hydraulically derived formula. I = effec-
tive rainfall intensity (in/h), S = slope of
( IL /43, 200)−0.67 L /60
overland flow path (ft/ft), L = length of
overland flow path (ft), and c is a roughness
coefficient (c = 0.007 for smooth asphalt,
0.012 for concrete pavement, 0.017 for tar
and gravel pavement, and 0.060 for dense
bluegrass turf)

tc = 0.94 I −0.4 ( Ln /S 0.5 )0.6 Henderson and Wooding (1964) Based on kinematic wave theory for flow
on an overland flow plane. I = rainfall
intensity (in/h), L = length of overland
flow (ft), n = Manning’s roughness coeffi-
cient, S = overland flow plane slope (ft/ft)

AV2 = Q = A 2 g (H − HL ) Federal Aviation Agency (1970) Developed based on airfield drainage data. C =
Rational Method runoff coefficient, L = over-
land flow length (ft), and S = slope (%)

1 Soil Conservation Service (1986) Time of concentration is developed as a sum

tc =
60
∑ (L /V )i of individual travel times. L = length of an
individual flow path (ft) and V = velocity
of flow over an individual flow path (ft/s).
V may be estimated using Figure 2.11 or
using Manning’s equation

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 53

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

50

ow

n
tio

ow
ad

n
law
iva
me

rfl
t

tte
and
cul
and

d
20

hal way
un

gu
ge
re
ver

gro

low
tilla

r
stu

ate
xo

are
pa

ow sed w
m
nd
10

yb

ds
imu

ss,
rou

arl

) an
gr a
min

s
yg

hee Gra
Ne
Slope, %

ort
eav

or
5

tfl
Sh
low
,h
t

Fal
res

a (s
Fo

are
2

ed
Pav
1

0.5
0.1 0.2 0.5 1 2 5 10 20
Velocity, ft/s

Figure 2.11: Average overland flow velocities as a function of land use

characteristics and surface slope.

are channelized in gutters, open channels, or storm sewers, Manning’s equation

may be used to estimate the velocity of flow (see Chapter 1).
Example 2.9: Time of Concentration
An urbanized drainage basin is shown in Figure 2.12. Three types of flow con-
ditions exist from the furthest point of the drainage basin to its outlet. Estimate
the time of concentration based on the following data:

Reach Flow Description Slope (%) Length (ft)

A–B Overland (forest) 7 500
B–C Overland (shallow gutter) 2 900
Storm sewer with manholes, inlets, etc.
C–D 1.5 2,000
(n = 0.015, dia. = 3 ft)
Open channel, gunite-lined, trapezoidal
D–E 0.5 3,000
(B = 5 ft, y = 3 ft, z = 1:1, n = 0.019)

Solution
For the reach from A to B, the average flow velocity is V = 0.7 ft/s. The travel
time for that reach is therefore

tAB = L/V = 500/0.7 = 714 s

54 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

50
0
B 900

ft
ft C
2,0
00
ft
D

3,
00
0
ft
E

Figure 2.12: Flow paths in a drainage basin for calculation of tc.

Similarly, for the reach from B to C, the average flow velocity is V = 2.8 ft/s.
The travel time for that reach is therefore

tBC = L/V = 900/2.8 = 321 s

To compute the travel time in the storm sewer from C to D, Manning’s

equation is employed to compute the pipe-full velocity:

2/3 2/3
1.49 ⎛⎜ D ⎞⎟ 1.49 ⎛⎜ 3 ⎞⎟
V = ⎜ ⎟ S1/ 2 = ⎜⎜⎝ ⎟⎟⎠ (0.015) = 10 ft/s
1/ 2
n ⎜⎝ 4 ⎟⎠ 0.015 4

The travel time for that reach is therefore

tCD = L/V = 2,000/10 = 200 s

Travel time in the open channel from D to E is computed using the bank-full
velocity, again found via Manning’s equation:

1.49 2 / 3 1/ 2 1.49
V = R S = (1.78)2 / 3 (0.005)1/ 2 = 8.2 ft/s
n 0.019

The travel time for that reach is therefore

tDE = L/V = 3,000/8.2 = 366 s

The time of concentration is the sum of the four individual travel times and is

tc = 1,601 s = 0.44 h

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th 55

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

The Rational Method

The Rational Method is an equilibrium-based approach to peak flow estima-
tion that uses rainfall intensity data and watershed characteristics to predict
peak flows for a rainfall event. The Rational Method is a popular choice for
storm sewer design because this type of design usually considers only peak
flows, and because of the simplicity of the calculations involved.
At the most fundamental level, the Rational Method assumes that a steady
state is attained such that the rainfall inflow rate of water onto a drainage basin
is equal to the outflow rate of water from the basin. If one expresses the volu-
metric inflow rate as the product of the basin area A and the effective rainfall
intensity ie, the outflow rate Q is obtained as Q = ieA. Further, if one accounts
for abstractions, the effective intensity is a product of the actual rainfall inten-
sity and a runoff coefficient, resulting in

Q = CiA

where
Q = runoff rate (ac.in/h, ha mm/h),
C = runoff (abstractions) coefficient,
i = rainfall intensity (in/h, mm/h), and
A = drainage area (ac, ha).
Because 1 ac.in/hr = 1.008 cfs ≈ 1 cfs, engineers performing calculations by hand
typically ignore the conversion factor and simply assume that the discharge Q
is in units of cfs. In SI units, a conversion factor of 0.278 will yield Q in units
of m3/s.
The time of concentration is the smallest time for which the entire basin
is contributing runoff to the basin outlet; therefore the storm duration
must be at least as long as the time of concentration if a steady-state con-
dition is to be achieved. Steady-state conditions also dictate that the storm
intensity be spatially and temporally uniform. It is not reasonable to expect
that rainfall will be spatially uniform over a large drainage basin, or that it
will be temporally uniform over a duration at least as long as the time of
concentration when tc (and hence A) is large. Therefore these conditions
limit the applicability of the Rational Method to small drainage basins. An
upper limit of 200 ac has been suggested by some, but the limit should
really depend on the storm characteristics of the particular locale. These
local characteristics may limit the applicability of the Rational Method to
basins smaller than 10 ac.in some cases.
When several drainage basins (or subbasins) discharge to a common facility
such as a storm sewer or culvert, the time of concentration should be taken
as the longest of all the individual times of concentration and should include

56 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

pipe travel times when appropriate. Further, the total drainage area served (the
sum of the individual basin areas) should be no larger than the 200-ac limit (or
smaller where applicable) of the Rational Method.
The basic steps for applying the Rational Method are as follows:

Step 1: Apply IDF Data

Develop or obtain a set of IDF curves for the locale in which the drainage
basin resides. Assume that the storm duration is equal to the time of concen-
tration and determine the corresponding intensity for the recurrence interval
of interest. Note that the assumption that the storm duration and time of con-
centration are equal is conservative in that it represents the highest intensity for
which the entire drainage area can contribute.

Step 2: Compute Watershed Area

The basin area A can be estimated using topographic maps, computer tools
such as CAD or GIS software, or by field reconnaissance. The time of con-
centration may be estimated using the procedures discussed in the preceding
subsection.

Step 3: Choose C Coefficients

The runoff coefficient C may be estimated using Table 2.5 if the land use is
homogeneous in the basin, or a composite C value may be estimated if the land
use is heterogeneous (see Example 2.7).

Step 4: Solve Peak Flow

Finally, the peak runoff rate from the basin can be computed using the equa-
tion Q = CiA.
The following example illustrates the use of the Rational Method for several
subbasins draining into a common storm sewer system.

Example 2.10: Computing Flows for Multiple Subbasins with the Rational
Method
Figure 2.13 is a plan view of a storm sewer system draining three subbasins.
Use the Rational Method to determine the peak discharge in each pipe and
size each pipe assuming the pipes flow full. Assume also that the pipes will be
concrete with n = 0.013. Perform the calculations for a storm recurrence inter-
val of 25 years. Subbasin and pipe characteristics and IDF data for the 25-year
event are tabulated as follows:

Subbasin A (ac) C tc (min)

A 6.0 0.6 20
B 4.0 0.8 10
C 4.5 0.8 15

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 57

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

Subbasin A

Pipe 1
Subbasin B Subbasin C

Pipe 2
Pipe 3

To outfall

Figure 2.13: System for Example 2.10.

Pipe Length (ft) Slope (%)

1 500 1.0
2 400 1.2
3 500 0.9

Duration (min) Intensity (in/h)

5 8.40
10 7.02
15 5.96
20 5.26
30 4.42
60 2.97

58 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

Solution
Flow into Pipe 1 occurs from Subbasin A only. Using the time of concentra-
tion as the storm duration, the 25-year rainfall intensity is 5.26 in/h. The peak
discharge used in sizing Pipe 1 is therefore

Q = 0.6(5.26)(6.0) = 19 cfs

Assuming that Pipe 1 is flowing full, its required diameter D may be found
using Manning’s equation as follows:

3/8
⎛ Qn ⎞⎟3 / 8 ⎛⎜ 19(0.013) ⎞⎟
D = ⎜⎜ ⎟ = ⎜ ⎟ = 1.87 ft
⎝ 0.464S ⎟⎠
1 / 2
⎝⎜ 0.464(0.01) ⎠⎟
1 / 2

Rounding up to the next commercially available size, Pipe 1 should have a

diameter of 24 in. Use FlowMaster to determine that the depth of flow is
1.40 ft and the area of flow is 2.36 ft2.
Because the cross-sectional area of Pipe 1 is 2.36 ft2, the average velocity in
Pipe 1 is

V = Q/A = 19/2.36 = 8.05 ft/s

The travel time in Pipe 1 is

t = L/V = 500/8.05 = 62 s = 1.04 min

Pipe 2 is treated the same way as Pipe 1, recognizing that runoff from Subbasin
C only enters Pipe 2. The peak discharge from Subbasin C is Q = 22 cfs, and
the required diameter of Pipe 2 is D = 24 in. The travel time in Pipe 2 is
t = 45 s = 0.75 min.
Pipe 3 must be sized to handle the runoff from all three of the subbasins, which
have a total area of A = 14.5 ac. The runoff coefficient for the combined areas
is computed as a composite value and is

6(0.6 ) + 4(0.8) + 4.5(0.8)

C= = 0.72
14.5

The time of concentration is computed as the longest of the travel times to the
upstream end of Pipe 3. These travel times are (1) the time of concentration
of Subbasin B (10 min), (2) the time of concentration of Subbasin A plus the
travel time in Pipe 1 (20 + 1.0 = 21 min), and (3) the time of concentration
of Subbasin C plus the travel time in Pipe 2 (15 + 0.75 = 15.75 min). Thus
the time of concentration for Pipe 3 is 21 min, and the corresponding rainfall
intensity (by interpolation) is 5.17 in/h.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 59

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

The peak discharge for Pipe 3 is

Q = 0.72(5.17)(14.5) = 54 cfs

The required diameter of Pipe 3 (rounded to the nearest standard size) is 36 in.

NRCS (SCS) Peak Flow Estimation

The NRCS (SCS) has developed a simple procedure for estimation of peak
runoff rates for small drainage basins. Graphical and tabular solutions to this
procedure are presented in Technical Release 55 (TR-55) (SCS, 1986), and
tabular and regression-based solutions are presented in the U.S. Department
of Transportation Federal Highway Administration’s Hydrologic Engineering
Circular No. 22 (HEC-22) (Brown et al., 2009). The HEC-22 regression pro-
cedure is described in this subsection.
For use with this method, the drainage basin should have a fairly homogeneous
distribution of CN, and the composite CN should be at least 40. The basin
should also have one main channel, although branches are acceptable if they
have approximately equal times of concentration. Further, this method was
developed for use with 24-h storm rainfall depths. Its use with other storm
durations is not advised.
This method is implemented by first applying the following equation to esti-
mate the depth of effective precipitation during a storm. The peak discharge
is estimated as

Qp = qu APe

where
Qp = the peak flow rate (cfs),
qu = the unit peak flow rate (cfs/mi2/in),
A = the drainage basin area (mi2), and
Pe = the depth of effective precipitation (in).
The unit peak flow rate is estimated as

qu = 10K

where

K = C0 + C1 log10 tc + C2 (log10 tc )2

C0, C1, and C2 are coefficients listed in Table 2.9 as a function of the SCS 24-h
design storm distribution type and the ratio Ia /P, and tc is the time of concentra-
tion (h). P is the 24-h rainfall depth, and Ia is the depth of initial abstractions.
For convenience, the ratio Ia /P is tabulated in Table 2.10 as a function of the

60 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

Table 2.9: Coefficients for SCS (NRCS) Peak Discharge Method (SCS, 1996)

Storm Type Ia/P C0 C1 C2

0.10 2.30550 −0.51429 −0.11750
0.15 2.27044 −0.50908 −0.10339
0.20 2.23537 −0.50387 −0.08929
0.25 2.18219 −0.48488 −0.06589
I 0.30 2.10624 −0.45696 −0.02835
0.35 2.00303 −0.40769 −0.01983
0.40 1.87733 −0.32274 0.05754
0.45 1.76312 −0.15644 0.00453
0.50 1.67889 −0.06930 0
0.10 2.03250 −0.31583 −0.13748
0.15 1.97614 −0.29899 −0.10384
0.20 1.91978 −0.28215 −0.07020
0.25 1.83842 −0.25543 −0.02597
IA 0.30 1.72657 −0.19826 0.02633
0.35 1.70347 −0.17145 0.01975
0.40 1.68037 −0.14463 0.01317
0.45 1.65727 −0.11782 0.00658
0.50 1.63417 −0.09100 0
0.10 2.55323 −0.61512 −0.16403
0.15 2.53125 −0.61698 −0.15217
0.20 2.50928 −0.61885 −0.14030
0.25 2.48730 −0.62071 −0.12844
II 0.30 2.46532 −0.62257 −0.11657
0.35 2.41896 −0.61594 −0.08820
0.40 2.36409 −0.59857 −0.05621
0.45 2.29238 −0.57005 −0.02281
0.50 2.20282 −0.51599 −0.01259
0.10 2.47317 −0.51848 −0.17083
0.15 2.45395 −0.51687 −0.16124
0.20 2.43473 −0.51525 −0.15164
0.25 2.41550 −0.51364 −0.14205
III 0.30 2.39628 −0.51202 −0.13245
0.35 2.35477 −0.49735 −0.11985
0.40 2.30726 −0.46541 −0.11094
0.45 2.24876 −0.41314 −0.11508
0.50 2.17772 −0.36803 −0.09525

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 61

 2 Basic Hydrology
Computing Peak Runoff Flow Rate

Table 2.10: Ia/P for Selected Rainfall Depths and Curve Numbers (Adapted from SCS, 1996)

Curve Number, CN
P (in)
40 45 50 55 60 65 70 75 80 85 90 95
0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.44 0.21
1.00 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.35 0.22 0.11
1.50 0.50 0.50 0.50 0.50 0.50 0.50 0.50 0.44 0.33 0.24 0.15 0.10
2.00 0.50 0.50 0.50 0.50 0.50 0.50 0.43 0.33 0.25 0.18 0.11 0.10
2.50 0.50 0.50 0.50 0.50 0.50 0.43 0.34 0.27 0.20 0.14 0.10 0.10
3.00 0.50 0.50 0.50 0.50 0.44 0.36 0.29 0.22 0.17 0.12 0.10 0.10
3.50 0.50 0.50 0.50 0.47 0.38 0.31 0.24 0.19 0.14 0.10 0.10 0.10
4.00 0.50 0.50 0.50 0.41 0.33 0.27 0.21 0.17 0.13 0.10 0.10 0.10
4.50 0.50 0.50 0.44 0.36 0.30 0.24 0.19 0.15 0.11 0.10 0.10 0.10
5.00 0.50 0.49 0.40 0.33 0.27 0.22 0.17 0.13 0.10 0.10 0.10 0.10
5.50 0.50 0.44 0.36 0.30 0.24 0.20 0.16 0.12 0.10 0.10 0.10 0.10
6.00 0.50 0.41 0.33 0.27 0.22 0.18 0.14 0.11 0.10 0.10 0.10 0.10
6.50 0.46 0.38 0.31 0.25 0.21 0.17 0.13 0.10 0.10 0.10 0.10 0.10
7.00 0.43 0.35 0.29 0.23 0.19 0.15 0.12 0.10 0.10 0.10 0.10 0.10
7.50 0.40 0.33 0.27 0.22 0.18 0.14 0.11 0.10 0.10 0.10 0.10 0.10
8.00 0.38 0.31 0.25 0.20 0.17 0.13 0.11 0.10 0.10 0.10 0.10 0.10
8.50 0.35 0.29 0.24 0.19 0.16 0.13 0.10 0.10 0.10 0.10 0.10 0.10
9.00 0.33 0.27 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10
9.50 0.32 0.26 0.21 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10
10.00 0.30 0.24 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10
10.50 0.29 0.23 0.19 0.16 0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.10
11.00 0.27 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10
11.50 0.26 0.21 0.17 0.14 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10
12.00 0.25 0.20 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10
12.50 0.24 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10
13.00 0.23 0.19 0.15 0.13 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
13.50 0.22 0.18 0.15 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
14.00 0.21 0.17 0.14 0.12 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
14.50 0.21 0.17 0.14 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10
15.00 0.20 0.16 0.13 0.11 0.10 0.10 0.10 0.10 0.10 0.10 0.10 0.10

62 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Peak Runoff Flow Rate

Table 2.11: Adjustment Factor, Fp, for

Ponds and Swampy Areas (SCS, 1996)
Pond and Swamp Area
Fp
(%)
0 1.00
0.2 0.97
1.0 0.87
3.0 0.75
5.0 0.72

curve number. The ratio Ia/P should not be less than 0.1, nor larger than 0.5.
The time of concentration should be in the range of 0.1 to 10 hours.
When ponding or swampy areas exist in a drainage basin, the peak discharge
should be reduced to account for the temporary storage of runoff. In this case,
an adjusted peak flow Qpa should be computed as

Qpa = FpQp

where Fp is an adjustment factor obtained from Table 2.11. Ponds and swampy
areas lying along the flow path used for computation of the time of concen-
tration should not be included in the area percentage in Table 2.11. Neither
channel nor reservoir routing can be accommodated with this method.
Example 2.11: Computing Peak Discharge Using the NRCS Method
Compute the peak discharge from the 1,000-ac (1.56 mi2) drainage basin
described in Example 2.8, but using AMC-II soil moisture conditions. The
rainfall distribution is an SCS Type II distribution, and the time of concen-
tration of the drainage basin is 0.9 h. There are no ponds or swampy areas in
the basin.
Solution
From Example 2.8, P = 5.0 in and CN = 86 for AMC-II. For this condition,
the maximum possible retention S = 1.63 in, and the effective precipitation is
Pe = 3.46 in.
From Table 2.10, the ratio Ia/P is 0.10; and from Table 2.9, the coefficients C0,
C1, and C2 are C0 = 2.55323, C1 = −0.61512, and C2 = −0.16403, respec-
tively. Therefore
K = 2.55323−0.61512 log10(0.9)−0.16403[log10(0.9)]2 = 2.58104

The unit peak flow rate is

qu = 10 2.58104 = 381 cfs/mi 2/in
The peak discharge from the drainage basin is computed as
Qp = 381(1.56)(3.46) = 2,060 cfs

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 63

 2 Basic Hydrology
Computing Hydrographs

2.4 Computing Hydrographs

A hydrograph represents the flow rate as it varies over time at a particular loca-
tion within a watershed. The integrated area under the hydrograph represents
the total hydrograph volume of water. Base flow represents subsurface flow
from groundwater that discharges into the conveyance channel. Many storm
conveyance channels are dry at the beginning of a rainfall event, which equates
to a base flow of zero as long as the water table does not rise into the channel
during the storm. Figure 2.14 displays the various components of a surface
runoff hydrograph, with a base flow of zero.
Estimation of a complete runoff hydrograph, as opposed to merely the peak
rate of runoff, is necessary to account for the effects of storage in a drainage
basin. Because a hydrograph accounts for volume and flow variations over an
entire rainfall event, it is useful for analyzing complex watersheds and design-
ing detention ponds. Hydrographs enable you to assess of the effects of stor-
age associated with natural ponds and lakes, and with artificial stormwater
detention and/or retention facilities. Hydrograph estimation is also necessary
in order to assess the impacts of storm duration and/or hyetograph shape on
runoff production, and in cases when two or more adjacent drainage basins (or
subbasins) are discharging to a common stream.
Procedures for estimating unit hydrographs and performing discrete convolu-
tion are presented in this and the following section. Unit hydrograph estima-
tion methods involve several basic steps:
1. Establish a design storm rainfall hyetograph.
2. Convert the design storm rainfall hyetograph into a runoff (effective rain-
fall) hyetograph by subtraction of rainfall abstractions.

De
Drainage area Dr
Ds
Dp
Dli
Volume runoff, Vr Di

Peak
flow
Flow rate

Runoff volume
(integrated area)

Time

Figure 2.14: Hydrograph definition sketch.

64 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

3. Estimate a unit hydrograph for the drainage basin of interest.

4. Merge the runoff (effective rainfall hyetograph) and the unit hydrograph
using discrete convolution to produce what is called a direct runoff
hydrograph, which represents the surface runoff hydrograph for the com-
plex storm pattern represented by the rainfall hyetograph.
5. If appropriate, add base flow to the direct runoff hydrograph to
obtain the total hydrograph representing both surface runoff and
base flow.

Creating Runoff (Effective Rainfall) Hyetographs

The first step in calculating a runoff hydrograph from a given rainfall hyetograph
is to calculate a runoff (effective rainfall) hyetograph. Provided that the data is
available, the most accurate way to do this is to develop a runoff hyetograph
by modeling abstractions and infiltration within each time step of a rainfall
event. To do this, subtract the interception from the beginning of a gross rain-
fall hyetograph, subtract infiltration from what remains after interception has
been accounted for, and subtract depression storage from what remains after
both interception and infiltration have been accounted for. This approach is
physically based and is not limited to use with any particular storm duration or
rainfall hyetograph shape.

Example 2.12: Computing an Effective Rainfall Hyetograph

Develop a runoff (effective rainfall) hyetograph for a watershed with an initial
loss (interception capacity) of 0.3 in, a depression storage capacity of 0.2 in,
and Horton infiltration parameters of f0 = 1.5 in/h, fc = 0.3 in/h, and k =
0.04 min−1. The rainfall hyetograph is tabulated as follows:

t (min) P (in)

0–10 0.24

10–20 0.46

20–30 1.17

30–40 0.58

40–50 0.35

50–60 0.17

Solution
The interception capacity of 0.3 in is subtracted first. Because 0.24 in of
rainfall occurs during the first 10 min of the storm, all of that rainfall plus
an additional 0.06 in of rainfall occurring in the second 10 min of the storm

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 65

 2 Basic Hydrology
Computing Hydrographs

is lost to interception. Thus, after accounting for interception, the rainfall

hyetograph is

t (min) P (in)

0–10 0

10–20 0.4

20–30 1.17

30–40 0.58

40–50 0.35

50–60 0.17

The infiltration rate, f(t), can be calculated and tabulated as a function of t

using the Horton equation (see Section 2.2), where t0 is the time at which
rainwater first begins to infiltrate (t0 = 10 min in this example, because rainfall
prior to that time is lost to interception and hence is not available for infiltra-
tion). The first column in the following table is the time since the beginning of
rainfall, and the second is the time since the beginning of infiltration. The third
column is the computed infiltration rate.
The fourth column contains incremental infiltration depths for each 10-min
period during the storm. For example, the first value of F = 0.22 in is com-
puted as the average of the current and preceding infiltration rates, multiplied
by the time interval of Δt = 10 min = 1/6 h (for example, 0.22 = [(1.50 +
1.10)/2]/6).

t (min) t−t0 (min) f(t) (in/h) Incr. F (in)

10 0 1.5

20 10 1.1 0.22

30 20 0.84 0.16

40 30 0.66 0.13

50 40 0.54 0.1

60 50 0.46 0.08

Subtraction of the infiltration depth in each time interval from the correspond-
ing rainfall depth remaining after interception leads to the following hyetograph
(any negative values produced should be set equal to zero):

66 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

t (min) P (in)

0–10 0

10–20 0.18

20–30 1.01

30–40 0.45

40–50 0.25

50–60 0.09

Finally, subtraction of the depression storage capacity of 0.2 in leads to the

following effective rainfall hyetograph:

t (min) Pe (in)
10–20 0
10–20 0
20–30 0.99
30–40 0.45
40–50 0.25
50–60 0.09

The rainfall hyetograph and the effective rainfall hyetograph are illustrated in
Figures 2.15 and 2.16. Note that the effective rainfall intensities are less than the
actual rainfall intensities. Also note that effective rainfall, and hence surface run-
off, does not begin (in this example) until 20 min after the beginning of the storm.

8
7

6
Intensity, in/h

4
3
2

1
0
0 to 10 10 to 20 20 to 30 30 to 40 40 to 50 50 to 60
Time, min

Figure 2.15: Rainfall hyetograph for Example 2.12.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 67

 2 Basic Hydrology
Computing Hydrographs

8
7

Effective intensity in/h

6
5
4
3
2
1
0
0 to 10 10 to 20 20 to 30 30 to 40 40 to 50 50 to 60
Time, min

Figure 2.16: Effective rainfall (runoff) hyetograph for Example 2.12.

Unit Hydrographs
A unit hydrograph is an important intermediate step in computing a full
runoff hydrograph for design or analysis purposes. After a unit hydrograph
is established for a particular basin, the data can be used to compute the
complete runoff hydrographs resulting from various rainfall events. The basic
theory rests on the assumption that the runoff response of a drainage basin
to an effective rainfall input is linear. Practically speaking, this assumption
means that the concepts of proportionality and superposition can be applied.
For example, it means that the direct runoff caused by 2 in. of effective rain-
fall in a given time interval is four times as great as that caused by 0.5 in. of
effective rainfall in the same amount of time. It also means that the total
amount of time for the basin to respond to each of these rainfall depths is the
same. Consequently, the base length, in hours, of each of the direct runoff
hydrographs is the same.

Visualizing and Developing Unit Hydrographs

By definition, the Δt-h (or Δt-min) unit hydrograph of a drainage basin is
the direct runoff hydrograph produced by 1 in. of effective rainfall falling uni-
formly in time and space over the drainage basin during a duration Δt time
units. By conservation of mass, the volume of runoff produced (represented by
the area under a graph of the unit hydrograph) must be equal to 1 in. of runoff
(effective rainfall) times the drainage basin area (see Figure 2.17). It may be
observed from this definition that the unit hydrographs for two drainage basins
must be different from one another, because the basin areas (as well as other
factors) are different.
Figure 2.18 provides a metaphor to help demonstrate the abstract principals of
a unit hydrograph. A sprinkler system is uniformly distributed over an entire
drainage area. Constant rainfall is modeled by using the sprinklers to apply a
constant rate of water until exactly 1 in of runoff is generated, at which time

68 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

tL

Effective intensity, discharge

Qp
Volume of runoff =
1 inch × basin area

Unit hydrograph of duration Δt

1 in of effective
rainfall in Δt time units

0 Δt tp tB
Time

Figure 2.17: Unit hydrograph resulting from 1 in. of effective rainfall

(runoff) over time (Δt).

Apply water for

duration Δt
until 1 in runoff
is generated.

1in
runoff Contributing area Flo
Ru w
no
Total ff Flow
volu
water me
applied
Abstractions
Runoff volume

Figure 2.18: Visualizing the unit hydrograph concept.

the sprinklers are turned off. Δt represents the duration of time that the sprin-
klers are on. A stream gauge is set on the downstream end of the contributing
drainage area to record the runoff hydrograph resulting from the 1 in of runoff.
The resulting hydrograph represents the unit hydrograph generated for the
drainage area.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 69

 2 Basic Hydrology
Computing Hydrographs

The time tL in Figure 2.17, which is the length of time from the center of Δt to the
peak discharge on the unit hydrograph, is the basin lag time. This response time
can be thought of as an approximate average of all possible travel times for runoff
in a drainage basin. In practice (and as shown in the figure), it is usually assumed
to be the amount of time between the center of mass of a pulse of effective rainfall
and the peak of the resultant hydrograph. The basin lag time is often used instead
of time of concentration when estimating the complete runoff hydrograph.
Two basic categories of unit hydrographs exist: (1) unit hydrographs for gauged
watersheds and (2) synthetic unit hydrographs. Most commonly, rainfall and
runoff records for a drainage basin do not exist, and one must resort to syn-
thesis of a unit hydrograph based on information that can be gathered about
the basin. Extensive literature exists on various ways to calculate synthetic unit
hydrographs, including procedures proposed by Clark, Snyder, and Singh. This
text introduces the development of synthetic unit hydrographs by applying the
NRCS (SCS) method. Further details on these and other synthetic unit hydro-
graphs methods can be found in Stormwater Conveyance Modeling and Design
(Haestad Methods and R. Durrans, 2003)

NRCS Synthetic Unit Hydrographs

The NRCS analyzed a large number of unit hydrographs derived from rainfall
and runoff records for a wide range of basins and basin locations and developed
an average dimensionless unit hydrograph (SCS, 1969). Times on the horizon-
tal axis of the dimensionless unit hydrograph are expressed in terms of the ratio
of time to time of peak discharge (t/tp), and discharges on the vertical axis are
expressed in terms of the ratio of discharge to peak discharge (Q/Qp). Figure 2.19
shows the NRCS dimensionless unit hydrograph, and Table 2.12 lists its ordinates.

1.2

1.0

0.8
Q / Qp

0.6

0.4

0.2

0.0
0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0
t / tp

Figure 2.19: The NRCS (SCS) dimensionless unit hydrograph.

70 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

Table 2.12: Ordinates of the SCS (NRCS)

Dimensionless Unit Hydrograph (SCS, 1969)

t/tp Q/Qp t/tp Q/Qp

0 0.000 1.7 0.460

0.1 0.030 1.8 0.390

0.2 0.100 1.9 0.330

0.3 0.190 2.0 0.280

0.4 0.310 2.2 0.207

0.5 0.470 2.4 0.147

0.6 0.660 2.6 0.107

0.7 0.820 2.8 0.077

0.8 0.930 3.0 0.055

0.9 0.990 3.2 0.040

1.0 1.000 3.4 0.029

1.1 0.990 3.6 0.021

1.2 0.930 3.8 0.015

1.3 0.860 4.0 0.011

1.4 0.780 4.5 0.005

1.5 0.680 5.0 0.000

1.6 0.560

Application of the dimensionless unit hydrograph method involves estimation

of the lag time tL of the drainage basin. Lag is typically approximated as 0.6tc in
NRCS (SCS) unit hydrograph computations. The time to peak of the synthetic
unit hydrograph of duration Δt is then computed as

Δt
tp = + tL
2

The NRCS recommends that Δt should be equal to 0.133tc, or equal to

0.222tL (SCS, 1969). A small variation from this is acceptable. The Δt cho-
sen for development of the synthetic unit hydrograph must be consistent
with the Δt chosen for development of the design storm and effective rain-
fall hyetographs.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 71

 2 Basic Hydrology
Computing Hydrographs

The peak discharge Qp (in cfs) of the synthetic unit hydrograph is calculated as

645.33 KAQ
Qp =
tp

where
645.33 = conversion factor = (1 h/3,600 s) (1 ft/12 in) [(5,280 ft)2/1 mi2];
K = 0.75 is constant based on geometric shape of dimensionless unit hydro-
graph (NEH-4);
Q = 1 in. is the runoff depth for unit hydrograph calculation;
A = the drainage basin area (mi2); and
tp = the time to peak (h).
Simplifying yields

484 A
Qp =
tp

The coefficient 484 appearing in the numerator of this equation includes

a unit conversion factor and is an average value for many drainage basins.
It may be reduced to about 300 for flat or swampy basins, or increased to
about 600 for steep or mountainous basins. Care needs to be taken when
the factor 484 is changed, as the base length and/or shape of the synthetic
unit hydrograph must also be changed to ensure that it represents a vol-
ume of water equivalent to one inch of effective rainfall over the drainage
basin area.
Once tp and Qp have been estimated using the preceding equations, the desired
synthetic unit hydrograph can be graphed and/or tabulated using the dimen-
sionless unit hydrograph shown in Figure 2.19 and Table 2.12.
Example 2.13: Developing an NRCS (SCS) Synthetic Unit Hydrograph
Develop a synthetic unit hydrograph for a 1.5-mi2 drainage basin in
Memphis, Tennessee having a time of concentration of 90 min. Assume that
the basin slopes are moderate so that the factor 484 can be applied in com-
puting Qp.

Solution
Employing the guidelines for estimation of Δt, its value should be no larger
than
0.133tc = 0.133(90) = 12 min

A duration of Δt = 10 min is selected, and the synthetic unit hydrograph will

be a 10-min unit hydrograph. The basin lag is estimated as

tL = 0.6tc = 54 min

72 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

Table 2.13: SCS (NRCS) Synthetic Unit Hydrograph Computations for Example 2.12

(1) (2) (3) (4) (5) (6) (7) (8)

t (min) t/tp Q/Qp Q (cfs) t (min) t/tp Q/Qp Q (cfs)

0 0.00 0.00 0 140 2.37 0.16 118

10 0.17 0.08 59 150 2.54 0.12 89

20 0.34 0.24 178 160 2.71 0.09 67

30 0.51 0.49 363 170 2.88 0.07 52

40 0.68 0.79 585 180 3.05 0.05 37

50 0.85 0.96 710 190 3.22 0.04 30

60 1.02 1.00 740 200 3.39 0.03 22

70 1.19 0.93 688 210 3.56 0.02 15

80 1.36 0.81 599 220 3.73 0.02 15

90 1.53 0.64 474 230 3.90 0.01 7

100 1.69 0.47 348 240 4.07 0.01 7

110 1.86 0.35 259 250 4.24 0.01 7

120 2.03 0.26 192 260 4.41 0.01 7

130 2.20 0.21 155 270 4.58 0.00 0

The time to peak for the synthetic unit hydrograph is

tp = (10/2) + 54 = 59 min = 0.98 h

The peak discharge is estimated as

Qp = 484(1.5)/0.98 = 740 cfs.

Ordinates of the synthetic unit hydrograph are given in Table 2.13. The first
column of the table is the time t, in minutes, and is tabulated in Δt = 10 min
intervals. The second column is the dimensionless time ratio t/tp, where tp =
59 min. The third column is the dimensionless discharge ratio and is deter-
mined using the dimensionless time ratio and interpolation from Table 2.12.
The fourth and last column contains the ordinates of the 10-min unit hydro-
graph, which are computed as the products of the dimensionless discharge
ratios and Qp = 740 cfs (Figure 2.20).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 73

 2 Basic Hydrology
Computing Hydrographs

800
700

Discharge, cfs
600
500
400
300
200
100
0
0 50 100 150 200 250 300
Time, min

Figure 2.20: Synthetic unit hydrograph for Example 2.13.

Discrete Convolution
A unit hydrograph represents the runoff hydrograph from a drainage basin sub-
jected to 1 in. of runoff (effective precipitation) applied over a duration of Δt.
However, a runoff (effective rainfall) hyetograph typically contains many
Δt time periods, each of which has its own associated runoff depth. In discrete
convolution, the direct runoff hydrograph resulting from a complete rainfall
hyetograph is computed by applying a unit hydrograph to each discrete time
step within the hyetograph.
To conceptualize discrete convolution, refer to Figures 2.21 and 2.22. The
first figure illustrates a runoff (effective rainfall) hyetograph consisting of
np = 3 rainfall pulses, each of duration Δt = 10 min, with the depth of each
pulse denoted by Pi (i = 1, 2, …, np). The second figure shows a unit hydro-
graph with nu = 11 nonzero ordinates shown at Δt time intervals, with the
nonzero ordinates denoted by Uj (j = 1, 2, …, nu). (Note that the ordinates
of the unit hydrograph are in units of cfs per inch of runoff, and thus the
notation of this subsection is different from the previous subsection in which
unit hydrograph ordinates were denoted by discharge Q.) The time incre-
ment Δt used for development of the runoff (effective rainfall) hyetograph
must be the same as the Δt duration of excess rainfall used to create the unit
hydrograph.
Because unit hydrograph theory assumes that drainage basins behave linearly,
the ordinates Qj of the runoff hydrograph produced by P1 inches of effective
rainfall must be equal to Qj = P1Uj. These ordinates are shown in the third
column of Table 2.14 (the first and second columns of the table reproduce the
information in Figure 2.22). These ordinates were calculated by multiplying
each unit hydrograph ordinate by the runoff depth for pulse P1 = 0.2 in. The
values in column three are shifted down such that the time at which the runoff
hydrograph in column three begins is the same as the time at which the first
pulse of effective rainfall begins.
The fourth and fifth columns in Table 2.14 are computed by multiplying the
unit hydrograph ordinates in the second column by the effective rainfall depths

74 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

Runoff hyetograph
1.6
1.4
1.2

Runoff depth, in
1.0
0.8 P2
0.6
0.4
P3
0.2
P1
0.0
0 10 20 30 40 50 60
Time, min

Figure 2.21: Effective rainfall (runoff) hyetograph.

Unit hydrograph
300

250 U4
Discharge, cfs/in

200 U5
U3
150
U6
100 U2 U7
U8
50 U1 U9
U10 U11
0
0 20 40 60 80 100 120
Time, min

Figure 2.22: Unit hydrograph ordinates.

Table 2.14: Discrete Convolution

(1) (2) (3) (4) (5) (6)
t (min) U (cfs/in) P1U (cfs) P2U (cfs) P3U (cfs) Q (cfs)
0 0 0
10 20 0
20 70 0 0
30 160 4 0 4
40 240 14 30 0 44
50 180 32 105 14 151
60 110 48 240 49 337

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 75

 2 Basic Hydrology
Computing Hydrographs

Table 2.14: (Continued)

(1) (2) (3) (4) (5) (6)
t (min) U (cfs/in) P1U (cfs) P2U (cfs) P3U (cfs) Q (cfs)
70 75 36 360 112 508
80 50 22 270 168 460
90 33 15 165 126 306
100 21 10 113 77 200
110 10 7 75 53 134
120 0 4 50 35 89
130 2 32 23 57
140 0 15 15 30
150 0 7 7
160 0 0

600

500
Total runoff

400
Flow, cfs

Pulse P2
300
Pulse P3
200
Pulse P1
100

0
0 20 40 60 80 100 120 140 160
Time, min

Figure 2.23: Summing of hydrographs from individual rainfall pulses.

P2 = 1.5 in and P3 = 0.7 in, respectively. Again, the time at which each runoff
pulse’s hydrograph begins corresponds to the time at which each runoff (effec-
tive rainfall) pulse begins. The sum of the runoff hydrograph ordinates in each
row of the table (shown in the sixth column) is an ordinate of the direct runoff
hydrograph caused by the complete rainfall event.
Figure 2.23 shows the runoff hydrograph resulting from each rainfall pulse and
the total runoff hydrograph for the entire rainfall event.
Example 2.14: Computing a Runoff Hydrograph
Compute the direct runoff hydrograph for the Memphis, Tennessee drain-
age basin described in Example 2.13. Use the unit hydrograph developed for

76 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Computing Hydrographs

the basin in that example, and use the runoff (effective rainfall) hyetograph
developed in Example 2.12.

Solution
There are a total of np = 4 runoff (effective rainfall) pulses, and a total of
nu = 26 nonzero unit hydrograph ordinates. Thus there will be a total of nq = 29
nonzero direct runoff hydrograph ordinates.
Table 2.15 illustrates the tabular calculation (discrete convolution). The first
column is the time since the beginning of rainfall in Δt = 10-min increments

Table 2.15: Discrete Convolution to Obtain Direct Runoff Hydrograph

(1) (2) (3) (4) (5) (6) (7)
t (min) U (cfs/in) 0.99U 0.45U 0.25U 0.09U Q (cfs)
0 0 0
10 59 0
20 178 0 0
30 363 58 0 58
40 585 176 27 0 203
50 710 359 80 15 0 454
60 740 579 163 45 5 792
70 688 703 263 91 16 1,073
80 599 733 320 146 33 1,231
90 474 681 333 178 53 1,244
100 348 593 310 185 64 1,152
110 259 469 270 172 67 977
120 192 345 213 150 62 769
130 155 256 157 119 54 585
140 118 190 117 87 43 436
150 89 153 86 65 31 336
160 67 117 70 48 23 258
170 52 88 53 39 17 197
180 37 66 40 30 14 150
190 30 51 30 22 11 115
200 22 37 23 17 8 85
210 15 30 17 13 6 65
220 15 22 14 9 5 49
230 7 15 10 8 3 36

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 77

 2 Basic Hydrology
Problems

Table 2.15: (Continued)

(1) (2) (3) (4) (5) (6) (7)
t (min) U (cfs/in) 0.99U 0.45U 0.25U 0.09U Q (cfs)
240 7 15 7 6 3 30
250 7 7 7 4 2 19
260 7 7 3 4 1 15
270 0 7 3 2 1 13
280 7 3 2 1 12
290 0 3 2 1 6
300 0 2 1 2
310 0 1 1
320 0 0

(the duration of the unit hydrograph and the duration of the effective rain-
fall pulses). The second column contains the ordinates of the unit hydrograph
from Example 2.13. From the third to the sixth columns are the unit hydro-
graph ordinates multiplied by the effective rainfall depths. Note that the first
entry (the first zero) in each column corresponds to the time at which the
corresponding effective rainfall pulse begins (t = 20 min for the first pulse,
t = 30 min for the second pulse, and t = 40 min for the third pulse; see Figure
2.16). The seventh column is the sum of the previous four and is the direct
runoff hydrograph.

2.5 Problems
1. The following data is from Huff and Angle (1992). The numbers indicate
the total rainfall (in) expected in Chicago, Illinois for storm recurrence
intervals of 2, 5, 10, 25, 50, and 100 years and durations from 5 min to
24 h. Plot a series of IDF curves showing the rainfall intensities for each
recurrence interval for storm durations between 5 min and 2 h.

Recurrence Interval (Years)

Duration
2 5 10 25 50 100
24 h 3.11 3.95 4.63 5.60 5.63 7.36
2h 1.83 2.33 2.74 3.31 3.86 4.47
1h 1.46 1.86 2.18 2.63 3.07 3.51
30 min 1.15 1.46 1.71 2.07 2.42 2.77
15 min 0.84 1.07 1.25 1.51 1.76 1.99
10 min 0.68 0.87 1.02 1.23 1.44 1.62
5 min 0.37 0.47 0.56 0.67 0.78 0.89

78 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Basic Hydrology
Problems

2. Use the data in Problem 1 and the SCS Type II distribution to develop a
25-year, 24-h storm for Chicago.
3. During a 40-mm rainfall event, the interception capacity of the water-
shed is estimated to be 11 mm, the depression storage is 9 mm, and the
infiltration is 7 mm. What is the total volume of runoff from the 2.56 ha
watershed?
4. A proposed development consists of the following land use areas:

Land Use Area (ac)

Condominiums 2.75
Greenbelt 0.89
Commercial 1.37
The time of concentration is estimated to be 25 min. Using the Rational
Method IDF data from Problem 1, calculate the peak runoff resulting from
a 10-year storm. Use the midpoint in the range of C coefficients in Table 2.5.
5. A suburban watershed has a time of concentration of 2.5 h and the fol-
lowing land use characteristics:

Land Cover Area (mi2) CNII

Commercial and business 4 92
1/4 ac housing 14 75
Parking lots, roofs, and driveways 0.78 98

(a) Use the NRCS (SCS) CN method to determine volume of runoff

resulting from a 6-in. storm. Assume that AMC-I applies.
(b) Use the SCS method to determine the peak discharge from the
watershed.
6. Develop a unit hydrograph for a drainage basin with an area of 3.5 mi2.
Use a time of concentration of 120 min. Assume that a shape factor of
484 applies and that the time of concentration is 2 h.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 79

 3 Inlets, Gravity Piping Systems,
and Storm Sewer Design

3.1 Inlet Design Overview

Road and highway designs must include sufficient drainage provisions in order
to minimize the danger resulting from stormwater runoff and to optimize traf-
fic efficiency under most weather conditions. The key objective when designing
inlets is to minimize the spread of water across a roadway and in the gutter. In
stormwater drainage, the gutter is the channel on the side of the road through
which stormwater runoff is conveyed to storm sewer inlets. Spread is the top
width of the flowing water on the road, measured from the curb.
The allowable spread length, which is generally determined by local or state
regulations, is based on the classification of the road. For example, a road with a
higher speed limit should have a smaller allowable spread than a road designed
for slower speeds because of the increased risk of hydroplaning. In addition
to spread width, roadway classification also dictates the return period of the
design storm to use in calculating the spread at a point. Table 3.1, from the

Table 3.1: Suggested Minimum Design Frequencies and Spreads for Gutter Sections on Grade
Road Classification Design Return Design Spreads
Period (Year)
High volume or <70 km/h (45 mph) 10 Shoulder + 1 m (3 ft)
bidirectional >70 km/h (45 mph) 10 Shoulder
Sag point 50 Shoulder + 1 m (3 ft)
Collector Low volume 10 1/2 Driving lane
High volume 10 Shoulder
Sag point 10 1/2 Driving lane
Local streets Low volume 5 1/2 Driving lane
High volume 10 1/2 Driving lane
Sag point 10 1/2 Driving lane

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 81

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Gutter Sections on Grade

Federal Highway Administration’s (FHWA) HEC-22 Urban Drainage Design

Manual (Brown et al., 2009), provides an overview of different road conditions
and the design criteria these conditions necessitate. We recommend you refer
to the FlowMaster or StormCAD online help for further explanations, as well
as the HEC-22 manual.
The incoming surface flow (and spread) observed can be controlled by the
efficiency and spacing of the inlets located upstream along the road. One addi-
tional factor to consider is whether the inlet is located on grade or in a sag, as
the design criteria and the equations involved differ. Inlets on grade are located
on a slope and intercept a portion of the water as it flows past. Inlets in sag
are located at a point where runoff from a given area will ultimately collect,
and these inlets are normally designed to capture 100% of the surface flow;
otherwise, flooding will occur in the surrounding area.

3.2 Gutter Sections on Grade

The main curb and gutter section types are the uniform section and the
composite section, as illustrated in Figure 3.1 with their defining variables.
Uniform gutter sections have a constant slope across the section. Composite
gutter sections are defined by a continuous gutter depression, a, measured
from the bottom of the curb to the projected road cross-slope at the curb. The
frontal flow, Qw, is normally defined as the flow in the depressed section of the
gutter or the flow over the grate width in the case of a grate inlet (grate inlets
are discussed later in this chapter).

T
T
Qs Qw

Sx
Sx
a Sw

Uniform road section Composite road section

Qw = Flow in depressed section (m3/s, ft3/s)

Qs = Side flow (m3/s, ft3/s)
Sw = Gutter cross-slope (m/m, ft/ft)
Sx = Road cross-slope (m/m, ft/ft)
T = Total width of flow or spread (m, ft)
a = Continuous gutter depression (mm, in)

Figure 3.1: Uniform and composite gutter sections.

82 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Gutter Sections on Grade

Runoff in gutters on grade is treated as open channel flow, and Manning’s

equation is applicable. Because the friction along the curb height is negli-
gible compared to the friction against the pavement width (spread) of flow,
Manning’s formula was modified, becoming the following equation for calcu-
lating flow in uniform gutter sections:

K c 1.67 0.5 2.67

Q= Sx S L T
n

where
Kc = 0.376 (0.56 in U.S. units),
n = Manning’s coefficient, and
SL = road longitudinal slope (m/m, ft/ft).
Using this equation, the spread, T, in a uniform gutter section can be explicitly
calculated for a given flow rate, Q. However, in the case of composite gutters,
T can no longer be expressed as an explicit function of Q. Rather, an iterative
process is required to calculate the spread.
On a road with a grade, the spread will be at a maximum just upstream of the
inlet. Note that the spread at this location is independent of the inlet’s ability to
capture flow and its efficiency. The spread for a specific roadway is a function of
the discharge in the gutter. To decrease the flow to an inlet, reduce inlet spacing
so that they serve as collection points for smaller watersheds.
For composite sections, the variable E0 is introduced to account for the added
conveyance of a depressed gutter. E0 is the ratio of flow in the depressed sec-
tion, Qw, to the total gutter flow, and is expressed as:

−1
⎧⎪ ⎡⎛ ⎞⎟2.67 ⎤⎫⎪⎪
⎪ S ⎢ ⎜ S S ⎥ ⎪
E0 = ⎪⎨1 + W ⎢⎜⎜1 +
W X ⎟
⎟ − 1⎥⎬
⎪⎪ SX ⎢⎜⎝ (T W ) − 1⎟⎟⎠ ⎥ ⎪⎪
⎪⎩ ⎢⎣ ⎥⎦ ⎪⎭

In the case of a grate inlet in a uniform section, the variable E0 is also useful,
with the frontal flow, Qw, now defined as the gutter flow contained in the
width of the grate. In this case, E0 becomes:

2.67
⎛ Wg ⎞⎟
E0 = 1 − ⎜⎜1 − ⎟⎟
⎜⎝ T ⎠

where Wg = width of grate inlet (m, ft).

Drainage inlets typically capture frontal flow more efficiently than side flow.
In drainage inlet design, it is good practice to maximize E0 by increasing the
gutter cross-slope, or by increasing the width of the gutter depression or the
grate extending into the road.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 83

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets on Grade

3.3 Inlets on Grade

Designs of inlets on grade are based on how much flow will be intercepted for
a given total flow (gutter discharge) to the inlet. Inlet design equations solve
for this efficiency.
Q
E= i
Q

where
E = inlet efficiency,
Qi = intercepted flow (m3/s, cfs), and
Q = total gutter flow (m3/s, cfs).
Flow that is not intercepted by a drainage inlet on grade is bypassed and carried
over to another inlet downstream or is “lost” to a stream or pond, for example.

Grate Inlets on Grade

Grate inlets, as shown in Figure 3.2, tend to be more efficient than curb inlets
when on a grade. Curb inlets are discussed in the next section of this chapter.
Two concerns must be addressed in the grate inlet design process. First, the engi-
neer needs to choose a grate type appropriate for the roadway being designed.
Bicycle traffic, for example, would limit the engineer to grate types with both
longitudinal and transverse bars in order to prevent bicycle accidents. Second,

W
L

Reticuline

P-50 mm P-50 mm × 100 mm

Figure 3.2: Grate inlet in a gutter and some typical grate types.

84 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets on Grade

grate inlets have a higher propensity to clog than other types of inlets. If debris
is prevalent in runoff at the point of design, adequate provisions must be made
to account for inlet clogging, such as utilizing a combination (grate and curb)
inlet at these points.
As shown in Figure 3.1, the total gutter flow is composed of frontal flow, Qw
(the flow in the depressed gutter or over the grate width), and side flow, Qs. The
total efficiency of the grate inlet is determined by calculating the grate’s ability
to capture frontal flow and side flow.
Rf is the ratio of intercepted frontal flow to the total frontal flow and is
expressed as:
R f = 1 − K cf (V − V0 )

where
Kcf = 0.295 (0.090 in U.S. units),
V = average velocity in the gutter at the location of the inlet
(m/s, ft/s), and
V0 = splash-over velocity of the inlet (m/s, ft/s).
The splash-over velocity is the minimum velocity of the gutter flow capable of
inducing enough momentum for some of the flow to skip over the grate open-
ing and be carried over downstream. The splash-over velocity is a function of the
grate type and the grate length and can be found in Appendix A of the HEC-22
manual (Brown et al., 2009) or obtained from the grate manufacturer. If the
gutter velocity is less than the splash-over velocity, all frontal flow is intercepted
and Rf equals 1.0.
The ratio, Rs, of side flow intercepted to total side flow is expressed as:
−1
⎛ K V 1.8 ⎞
Rs = ⎜⎜⎜1 + cs 2.3 ⎟⎟⎟
⎜⎝ Sx L ⎟⎠
where
Kcs = 0.0828 (0.15 in U.S. units) and
L = grate length (m, ft).
The total intercepted flow is expressed as:
Qi = Qw R f + Qs Rs

The total efficiency of the grate inlet on grade is expressed as:

E = R f E0 + Rs (1 − E0 )

Curb Inlets on Grade

Curb inlets are openings within the curb itself (see Figure 3.3), and they are
used in areas where grate inlets are prone to clogging.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 85

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets on Grade

Figure 3.3: Curb inlet.

Sx
a Sw
a’ S w’
Sw’ = Slope of local depression (m/m, ft/ft)
a = Gutter depression (mm, in)
a’ = Total depression (mm, in)

Figure 3.4: Local depression at a curb inlet.

The efficiency of a curb inlet is based on the ratio of the actual inlet length
to the inlet length necessary to capture 100% of the total runoff. Curb inlets
are often inset into the pavement to create a local depression. A local depres-
sion, as shown in Figure 3.4, is a depression of the gutter at the location of
the inlet only, as opposed to a gutter depression, which is continuous along
the curb.
The curb opening length, LT, that would be required to intercept 100% of a
flow, Q, on a roadway section with a uniform cross-slope is defined as:

⎛ 1 ⎞⎟0.6
LT = K C Q 0.42SL0.3 ⎜⎜ ⎟
⎜⎝ nSx ⎟⎟⎠

where Kc = 0.817 (0.60 in U.S. units).

To account for local depressions or gutter depressions, an additional composite
or equivalent slope, Se, is necessary:

Se = Sx + Sw ′ E0

86 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets on Grade

where
Sw′ = gutter cross-slope at the inlet location measured from the pavement
cross-slope, Sx (m/m, ft/ft) and
E0 = ratio of flow in the depressed section to the total gutter flow upstream of
the inlet (does not account for local depression).
To calculate LT with a composite gutter, replace the road cross-slope, Sx, with
the equivalent slope, Se, in the equation solving for LT.
The efficiency of a curb opening on grade shorter than the required length for
total interception is:

1.8
⎛ L⎞
E = 1 − ⎜⎜1 − ⎟⎟⎟
⎜⎝ LT ⎟⎠

Combination Inlets on Grade

Combination inlets, shown in Figure 3.5, consist of both grate and curb open-
ings. The curb inlet functions as a sweep, removing debris from the runoff
before it can clog the grate inlet.
If the curb inlet length is equal to the grate inlet length, the flow intercepted
by the combination inlet is assumed to be equivalent to that intercepted by the
grate inlet alone.

(A) Lc

W
e Lg
charg g
Dis

Lc
(B)

W Lg
e g
arg
D isch

Figure 3.5: Typical combination inlets.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 87

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets in Sag

The curb inlet is often extended upstream of the grate for more efficient
removal of debris than when the curb opening length equals the grate length.
The total flow intercepted by this configuration is calculated as the sum of
• the flow intercepted by the portion of the curb opening located upstream
of the grate and
• the flow that bypassed the upstream curb opening and is intercepted by
the grate alone. (the flow intercepted by the portion of the curb opening
adjacent to the grate is neglected.)

3.4 Inlets in Sag

Inlets located in sag are assumed to capture 100% of flow because, once col-
lected, the runoff in the sag has no other place to go. As opposed to inlets
located on a grade, the size and type of inlet directly affects the spread. As
shown in Table 3.1, the HEC-22 manual typically recommends employing
a larger rainfall return period for designing an inlet located in sag than for
designing an inlet located on grade.
The computations for calculating the amount of flow intercepted by inlets
in sag are based on the principles of weir flow and orifice flow discussed in
Chapter 1.
For an unsubmerged inlet operating as a weir, the flow capacity is calculated as:

Qiw = Cw Pd 1.5

where
Qiw = flow intercepted by the inlet operating as a weir (m3/s, ft3/s);
Cw = weir coefficient, which varies depending on the flow condition and inlet
structure;
P = perimeter of the inlet (m, ft); and
d = flow depth at the curb (m, ft).
Note that for a grate inlet, the perimeter does not include the length along the
curb. Also, if the gutter is depressed (locally or continuously), the perimeter, P,
of the grate is calculated as:

P = L + 1.8W

where
L = grate length (m, ft) and
W = grate width (m, ft).
The depth, d, for both types of inlets is measured from the projected pavement
cross-slope. For a curb inlet, the perimeter is equivalent to the length of the
inlet.

88 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlets in Sag

If the inlet is submerged and is operating as an orifice, its capacity becomes:

Qio = Co A (2 gd0 )
0.5

where
Qio = flow intercepted by the inlet operating as an orifice (m3/s, ft3/s),
C0 = orifice coefficient (varies based on the class of inlet and its configuration),
A = area of the opening (m2, ft2),
g = 9.81 m/s2 (32.16 ft/s2 in U.S. customary units), and
d0 = effective head at the orifice (m, ft).
Note that for a grate inlet, the effective head, d0, is simply the water depth
along the curb. For a curb inlet, the effective head is expressed as:

h
d0 = di − (sinθ)
2
where
di = depth at lip of curb opening (m, ft),
h = curb throat opening height (m, ft) (see Figure 3.6), and
θ = inclination of the curb throat measured from the vertical direction
as shown in Figure 3.6.

do
di
h

Horizontal throat
do = di – (h/2)
θ = p/2

do
θ
h
Inclined throat
do = di – h/2(sinθ)

do
Vertical throat
do = di
θ=0
h

Figure 3.6: Different curb inlet throat types.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 89

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Inlet Design Overview

Grates alone are not typically recommended for installation in sags because of
their propensity to clog and exacerbate ponding during severe weather. A com-
bination inlet may be a better alternative. At low flow depths, the capacity of a
combination inlet for which the grate inlet length equals the curb inlet length
is equivalent to the capacity of the grate inlet alone. At higher flow depths for
the same type of inlet, both the curb inlet and grate inlet act as orifices. The
total intercepted flow is then calculated as the sum of the flows intercepted by
the grate and curb openings.

3.5 Inlet Design Overview

This section provides a quick overview of features necessary to solve inlet prob-
lems using software from Bentley Systems. FlowMaster can be used to analyze
the hydraulic performance of individual inlets. StormCAD is used to model
inlets in storm sewer networks.

Inlet and Gutter Problems Using FlowMaster

FlowMaster can compute the capacities of gutters and drainage inlets using
the methodology put forth in the FHWA’s HEC-22 Urban Drainage Design
Manual (Brown et al., 2009). Given a type of inlet and design constraints, it
can also compute the inlet size required. With FlowMaster you can solve for
different variables associated with the following structures by selecting them
from the list of available worksheets:
• gutter sections on grade
• grate inlets
• curb inlets
• combination inlets
• slot inlets
• ditch inlets
For a more in-depth tutorial on the use of FlowMaster, see Chapter 1.
When creating a new inlet worksheet, you will be required to name the work-
sheet by providing a label. Additionally, you will need to specify whether the
inlet is on grade or in sag by selecting the appropriate radio button. When
editing an inlet’s worksheet, remember that there are multiple tabs on which
you have to enter data. For example, the combination inlet worksheet has a
Gutter tab, an Inlet tab, a Grate tab, and a Curb tab.
In the worksheet for a gutter section, you can provide a discharge and solve for
the resulting spread, or you can enter a spread and have FlowMaster solve for
the allowable discharge. In the worksheet for an inlet, you can either solve
for an inlet’s efficiency based upon the entered parameters, or you can size the
inlet to meet specified design criteria.

90 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Inlets, Gravity Piping Systems, and Storm Sewer Design
Gradually Varied Flow

Set the variable you want to solve for by selecting it from the pull-down menu
in the Solve For: field. Then, click Solve. The calculated values will appear in
the fields highlighted in yellow.

Inlet and Gutter Network Problems Using

StormCAD
StormCAD, like FlowMaster, calculates and designs drainage inlets by apply-
ing the HEC-22 methodology. Unlike FlowMaster, StormCAD will design an
entire gutter network as opposed to just one inlet at a time. The gutter network
dictates what happens to flow on the surface when it bypasses an inlet and is
routed downstream to another inlet.
To edit an inlet’s properties, first double-click on that inlet in the drawing
pane and select the Inlet tab at the top of the dialog. In the Inlet field, select
an inlet from the pull-down menu, which lists all the inlets currently defined
in the Inlet Library. To examine the characteristics of the inlet selected, click
the ellipsis (…) button next to this field. The Inlet Library dialog will appear.
The Inlet Library is a separately saved file and editable library allowing cus-
tomization for local regulations. Select the inlet of interest and click Edit…
to examine or change its properties, or click the Insert button to create a new
inlet. To exit an inlet’s properties dialog in the library, click OK or Cancel. To
exit the Inlet Library, click Close.
In the Inlet tab of an Inlet Editor dialog, the Inlet Opening section displays
variables for the type of inlet chosen. Gutter characteristics can be entered
under Inlet Section.
Finally, you have to specify an inlet location by clicking either the On Grade
or In Sag radio button. If the inlet is on grade, fill in the Longitudinal Slope
of the road and Manning’s n-value for the pavement. You must also select a
Bypass Target to establish the connectivity of the gutter network. The bypass
target is another inlet or outlet in the same project (the target element does not
have to be an element in the same pipe network) where the flows bypassing the
current inlet will be carried. If no bypass target is chosen, all bypass flow will
automatically carry over to the outlet of the sewer network that the inlet is part
of (in this case, the bypassed flow is not accounted for in the sewer network).
Click OK to exit the dialog. To view a graphical representation of the gutter
network in plan view, select Gutter Network from the Tools’ pull-down menu.

3.6 Gradually Varied Flow

When performing an open channel hydraulic analysis, the change in water
depth along the length of a section may be so gradual that the entire section
can be assumed to have a constant water depth (such as normal depth) without
any loss of accuracy. In other systems, however, such as storm sewers, there may
be some sort of restriction that prevents the water depth from equaling normal

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 91

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Gradually Varied Flow

Water surface elevation

Normal depth

Critical depth

Figure 3.7: Nonuniform flow in an open channel.

depth throughout the length. For example, a high tailwater elevation may force
the depth to be above normal depth at the downstream end of a pipe, as shown
in Figure 3.7.
When the difference in water depth from one end of a conduit to the other is sig-
nificant, a gradually varied flow profile analysis is required. Gradually varied flow
profile analysis is the process of splitting a channel length into segments and ana-
lyzing each segment separately. It is based on the following several assumptions:
• The head loss within any given calculation segment is the same as for uni-
form flow conditions.
• The velocity is the same across the entire cross section.
• The slope of the conduit is less than 10%.
• The roughness coefficient is constant throughout the reach under consider-
ation and is independent of the depth of flow.
• The depth of flow changes gradually along the length of the conduit, start-
ing from some controlling boundary condition (usually the tailwater eleva-
tion); there are no sudden increases or decreases in depth.
• If the pipe is sufficiently long, the depth of flow will approach normal depth.

Flow Classification
The first step in performing a gradually varied flow analysis is to identify the
flow classification that is expected to occur in the conduit based on the slope of
the channel, the normal depth, the critical depth, and the controlling bound-
ary condition.

Slope Classification
Once the normal depth and critical depth have been computed for the section,
the conduit slope can be determined. If the normal depth is above critical

92 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Gradually Varied Flow

depth, the slope is said to be mild. If the normal depth is equal to the critical
depth, the slope is said to be critical. If normal depth is below critical depth,
the slope is said to be steep.
There are two other slope types to consider: a horizontal channel and a chan-
nel on an adverse slope (an “uphill”-sloping channel). For these slope types,
normal depth is undefined.
For a gradually varied flow profile, the first letter of the slope type is used in
the identification of the profile. For example, a channel with a hydraulically
steep slope is a type S, a channel with a hydraulically mild slope is a type M,
and so forth.

Flow Zone Classification

In a gradually varied flow analysis, the flow is assumed to stay within the same
zone throughout the length of the conduit. For example, the flow will not
jump from subcritical to supercritical within the same profile type. There are
three zones where flow profiles can occur:
• Zone 1: Flow depth is above both normal and critical depths.
• Zone 2: Flow depth is between normal depth and critical depth.
• Zone 3: Flow depth is below both normal and critical depths.
A given profile will exist in only one of these zones. Because normal depth is
undefined for horizontal slopes and adverse slopes, Zone 1 flow does not exist
for these profile types.

Profile Classifications
Once the slope classification and flow zone have been determined, the profile
type can be defined, and the engineer can determine how to proceed with the
hydraulic grade computations. Figure 3.8 shows the basic profile types.
To perform the computations, the engineer must determine from the profile
type whether the flow is subcritical or supercritical (based on the location of
actual depths compared to critical depth). In order to prevent excessive veloc-
ities that could cause pipe scour or channel erosion, most storm sewers are
designed with mild slopes to carry subcritical flows. For this type of flow, the
hydraulic control is at the downstream end of the section, and the profile cal-
culation proceeds toward the upstream end. When the flow depth is above
normal depth (as in an M1 profile), this type of analysis is called a backwater
analysis. When the flow depth is between critical depth and normal depth, it is
called a drawdown analysis.
When supercritical flows are encountered, the controlling section is at the
upstream end of the conduit, and the computations proceed from upstream to
downstream. This calculation type is a frontwater analysis.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 93

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Gradually Varied Flow

Zone 1 profiles Zone 2 profiles Zone 3 profiles

y > yn; y > yc yn ≥ y ≥ yc or yn ≤ y ≤ yc y < yn; y < yc

M1 M2 M3
yn yn yn
Mild slope
yn > yc

yc yc yc

C1 C2 unstable C3
Critical slope

yc,n yc,n yc,n

yn = yc

S1 S2 S3
yc yc yc
Steep slope
y n < yc

yn yn yn

H1 none H2 H3
Horizontal slope

yc yc yc

A1 none A2 A3
Adverse slope

yc yc yc

Figure 3.8: Gradually varied flow classifications.

Energy Balance
Even for gradually varied flow, the solution is still a matter of balancing the
total energy between the two ends of each segment. The energy equation as it
relates to each end of a segment can be written as (note that the pressures for
both ends are zero, because it is free surface flow):

V12 V2
Z1 + = Z 2 + 2 + HL
2g 2g

94 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Inlets, Gravity Piping Systems, and Storm Sewer Design
Mixed Flow Profiles

where
Z1 = hydraulic grade at the upstream end of the segment (m, ft),
V1 = velocity at the upstream end (m/s, ft/s),
Z2 = hydraulic grade at the downstream end of the segment (m, ft),
V2 = velocity at the downstream end (m/s, ft/s),
HL = head loss due to friction (other losses are assumed to be zero) (m, ft), and
g = gravitational acceleration constant (m/s2, ft/s2).
The friction loss is computed based on the average rate of friction loss along the
segment and the segment length. This relationship is:

S1 + S2
HL = SAvg 3Δx = Δx
2

where
HL = loss across the segment (m, ft),
SAvg = average friction loss (m/m, ft/ft),
S1 = friction slope at the upstream end of the segment (m/m, ft/ft),
S2 = friction slope at the downstream end of the segment (m/m, ft/ft), and
Δx = length of the segment being analyzed (m, ft).
The conditions at one end of the segment are known (through assumption or
from a previous calculation). Because the friction slope is a function of velocity,
which is a function of depth, the depth at the other end of the segment can be
found through iteration. There are two primary methods to obtain the solu-
tion: the standard-step method and the direct-step method.

Standard-Step Method
This method involves dividing the channel into segments of equal known
lengths and solving for the unknown depth at one end of the segment (starting
with a known or assumed depth at the other end). The standard-step method
is the most popular method of determining the flow profile because it can be
applied to any channel, not just prismatic channels.

Direct-Step Method
The direct-step method is based on the same basic energy principles as the
standard-step method, but it takes a slightly different approach to the solution.
Instead of assuming a segment length and solving for the depth at the end of
the segment, the direct-step method assumes a depth and then solves for the
segment length.

3.7 Mixed Flow Profiles

So far, only conditions for which the entire channel has the same profile type,
resulting in the smooth curves of a gradually varied flow analysis, have been

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 95

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Storm Sewer Applications

Hydraulic jump S1
S2

Critical depth

Normal depth

Figure 3.9: Plot of hydraulic jump.

discussed. This section explores those cases in which profile types are mixed within
the same section and the steps that can be taken to analyze these occurrences.

Sealing Conditions
There may be conditions whereby part of a pipe or culvert section is flowing
full, while part of the flow remains open. These conditions are called sealing
conditions and are analyzed in separate parts. The portion of the section flow-
ing full is analyzed as pressure flow, and the remaining portion is analyzed as
gradually varied flow.

Rapidly Varied Flow

Rapidly varied flow is turbulent flow resulting from the abrupt and pronounced
curvature of flow streamlines into or out of a hydraulic control. Examples of
rapidly varied flow include hydraulic jumps, bends, and bridge contractions.

Hydraulic Jumps
When flow passes rapidly from supercritical to subcritical flow, a phenomenon
called a hydraulic jump occurs. In storm sewer networks, this often occurs when
a steep pipe discharges into a particularly high tailwater, as shown in Figure 3.9.
There are significant losses associated with hydraulic jumps due to the amount
of mixing and turbulence that occur. These forces are also highly erosive, so
engineers typically try to prevent jumps from occurring in storm sewer sys-
tems, or at least predict the location of jumps in order to provide adequate
channel, pipe, or structure protection.

3.8 Storm Sewer Applications

Storm sewer analysis occurs in two basic calculation sequences:
• Hydrology: Watersheds are analyzed and flows are accumulated from
upstream inlets toward the system outlet.
• Hydraulics: A tailwater condition is assumed at the outlet, and the flow
values (from the hydrology calculations) are used to compute hydraulic
grades from the outlet toward the upstream inlets.

96 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Storm Sewer Applications

Hydrology Model
As the runoff from a storm event travels through a storm sewer, it combines
with other flows, and the resulting flows are based on the overall watershed
characteristics. As with a single watershed, the peak flow at any location within
the storm sewer is assumed to occur when all parts of the watershed are con-
tributing to the flow. Therefore the rainfall intensity that produces the largest
peak flow at a given location is based on the controlling system time at that
same location.
The controlling system time is the larger of the local time of concentration
(to a single inlet) and the total upstream system time (including pipe travel
times). The controlling time is used for computing the intensity (and therefore
the flow) in the combined system.

Example 3.1: Flow Accumulation

A storm sewer inlet has a local time of concentration of 8 min for a water-
shed with a weighted CA (i.e., the weighted runoff coefficient times the total
drainage area) of 1.23 ac. This inlet discharges through a pipe to another storm
sewer inlet with a weighted CA of 0.84 ac and a local time of concentration of
9 min. If the travel time in the pipe is 2 min, what is the overall system CA and
corresponding storm duration?

Solution
The total CA can be found by simply summing the CA values from the two
inlets. The storm duration, however, must be found by comparing the local
time of concentration at the second inlet to the total time for flow from the
upstream inlet to reach the downstream inlet.
Total CA = 1.23 ac + 0.84 ac = 2.07 ac
Upstream time = 8 min + 2 min = 10 min
The total upstream flow time of 10 min is greater than the local time of con-
centration at the downstream inlet (9 min). The 10-min value is therefore the
controlling time and should be used as the duration of the storm event.

Other Sources of Water

There may be a number of flow sources for a storm sewer system other than
direct watershed inflow. Flow may be piped into an inlet from an external con-
nection, or there could be flow entering an inlet that is the carryover (bypass)
flow from another storm sewer inlet.
Design practices for handling carryover flows vary from jurisdiction to jurisdic-
tion. Some local regulations may require that pipes be sized to include all flow
that arrives at an inlet, whereas other locales may specify that pipes be sized to
accommodate only the flow that is actually intercepted by the upstream inlets.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 97

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
StormCAD, SewerGEMS and Tutorials

It is the responsibility of the design engineer to ensure that the design is in

agreement with the local criteria and policies.
The hydraulics of a storm sewer are often computed as described previously in
this chapter. Normal depth may occur in portions of the system, whereas other
areas may experience pressure (submerged) conditions. Gradually varied flow
and rapidly varied flow may also occur.
Computations start at the system outlet, where a tailwater condition is
assumed. There are four basic assumptions for tailwater conditions:
• Normal depth: The depth at the outfall of the farthest downstream pipe
is assumed to be equal to normal depth, as in a sufficiently long S2 profile.
• Critical depth: The depth at the pipe outfall is assumed to be critical
depth, as in subcritical flow to a free discharge.
• Crown elevation: The depth is set to the crown (top) of the pipe for free
outfall.
• User-specified tailwater: A fixed tailwater depth can also be used, as when
there is a known pond or river water surface elevation at the outfall of the
storm sewer.
Care should be taken to choose an accurate tailwater condition, because this
value can affect the hydraulics of much of the system. The designer of a storm
sewer system should consider the tailwater depth during storm conditions.
An outlet may be above the receiving stream during dry weather but may be
submerged during the design storm event.

3.9 StormCAD, SewerGEMS and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software. Information
about SewerGEMS can be found in Chapter 9.

What Does StormCAD Do?

StormCAD is an extremely powerful, easy-to-use program that helps civil engi-
neers design and analyze storm sewer systems. Just draw your network on the
screen by using the tool palette, double-click any element to enter data, and
click the GO button to calculate the network.
Rainfall information is calculated using rainfall tables, equations, or the
National Weather Service’s HYDRO-35 data. StormCAD also plots intensity–
duration–frequency (IDF) curves. You have a choice of conveyance elements

98 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
StormCAD, SewerGEMS and Tutorials

that include circular pipes, pipe arches, boxes, and more. Flow calculations
handle pressure and varied flow situations, including hydraulic jumps, back-
water, and drawdown curves. StormCAD’s flexible reporting features allow you
to customize and print the design and analysis results in report format or as a
graphical plot.

How Can You Use StormCAD?

StormCAD is so flexible that you can use it for all phases of your project, from
the feasibility report to the final design drawings and analysis of existing net-
works. During the feasibility phase, you can use StormCAD to create several
different system layouts with an AutoCAD® or MicroStation™ drawing as the
background. For the final design, StormCAD lets you complete detailed draw-
ings with notes that can be used to develop construction plans. In summary,
you can use StormCAD to:
• design multiple storm sewer systems with constraint-based design;
• design/analyze inlets based on HEC-22 methodology;
• use AASHTO, HEC-22 Energy, standard, absolute, or user-specified
(generic) methods to compute structure losses;
• analyze various design scenarios for storm sewer systems;
• import and export AutoCAD® and MicroStation™ DXF files;
• predict rainfall runoff rates;
• generate professional-looking reports for clients and review agencies; and
• generate plan and profile plots of the network.
The theory and background used in StormCAD are presented in more detail in
the StormCAD online help system.

Analysis and Design

StormCAD’s automatic design feature allows you to design whole or part of
a storm sewer system based on a set of user-defined design constraints. These
constraints include minimum/maximum velocity, slope, and cover; choice of
pipe invert or crown matching at structures; inlet efficiency; and gutter spread
and depth. StormCAD will automatically design the invert elevations and
diameters of pipes, as well as the size of a drainage inlet necessary to maintain a
given spread (for inlets in sag) or capture efficiency (for inlets on grade).

Profiles
StormCAD also includes an option to automatically generate storm sewer
profiles—longitudinal plots of the storm sewer. Profiles allow the design engineer,
the reviewing agency, the contractor, and others to visualize the storm system. They

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 99

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

are useful for viewing the hydraulic grade line and determining if the proposed
storm sewer is in conflict with other existing or proposed underground utilities.

3.10 Problems
Solve Problems 1–3 using FlowMaster and Problems 4–11 using StormCAD.
1. The spread on a proposed road paved with smooth asphalt and having a
1.0% longitudinal slope is limited by local regulations to 3.0 m for the
sake of automobile safety. The road cross-slope is 2.5%.
(a) What is the maximum allowable discharge on the road if the gutter
section is uniform? What is the depth of the flow measured at the
curb for this rate of discharge?
(b) An alternative design proposes the addition of a gutter depression
with a width of 1.5 m and a gutter cross-slope of 4.0%. What is the
maximum allowable discharge in the gutter under these conditions?
What is the depth of the flow at this rate of discharge?
(c) Does the addition of the gutter depression increase or decrease the
capacity of the gutter? Why?
2. A proposed gutter section has a road cross-slope of 0.015 m/m and a
gutter cross-slope of 0.045 m/m. The gutter is 1.20-m wide and is on a
longitudinal slope of 0.02 m/m. The Manning’s coefficient for the road is
0.013. The predetermined runoff in the gutter is 0.05 m3/s.
(a) What is the efficiency of a P-50 mm×100 mm grate inlet that is
0.70-m long and 0.5-m wide? Assume no clogging. What is the
intercepted flow? What is the bypassed flow?
(b) What is the efficiency of a 0.70-m-long curb inlet on the same gutter
section with no local depression? How does this efficiency compare
to the efficiency of the grate inlet in part (a)?
(c) For the curb inlet in part (b), add a local depression that is 20-mm
deep and 0.7-m wide. What is the efficiency of the inlet? How does
this efficiency compare to the grate inlet in part (a) and the curb
inlet in part (b)?
3. An inlet in sag must collect the 50-year peak runoff of 8.0 cfs. The gut-
ter section has a road cross-slope of 2.0% and a gutter cross-slope of
4.0%. The gutter width is 3.0 ft. The local depression is 2.0-in. deep and
3.0-ft wide.
(a) What length combination inlet is necessary to maintain a spread of
8.0 ft at the section? The grate is a 3.0-ft wide, P-50-mm type, and
the curb inlet has an opening height of 0.7 ft and a vertical throat
configuration. Assume the length of the grate will equal the length
of the curb opening.

100 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

(b) How big would the curb inlet from part (a), acting alone, have to be
to maintain the spread width of 8.0 ft under the same conditions?
How long would the grate inlet in part (a) have to be to maintain the
same spread? Which inlet type is the best solution?
Solve the following problems using StormCAD.
4. Lay out the storm sewer system shown below in StormCAD and enter
the data for the network from the tables below. Calculate the results using
Manning’s equation and the 10-year storm event data tables that follow.
Inlet I-1 is on grade with a longitudinal slope of 3%, whereas inlet I-2 is
in sag. The tailwater condition at the outlet is free outfall. Assume that
there is no clogging of the inlets.

I-1 P-1 I-2 P-2 O-1

Schematic for Problem 4

Pipe Data for Problem 4
Pipe Circular Section Upstream Downstream Length Pipe
Size (mm) Invert Invert (m) Material
Elevation (m) Elevation (m) (n-value)
P-1 300 115.40 115.10 46 Concrete
(0.013)
P-2 300 115.05 114.60 61 Concrete
(0.013)

Node Data for Problem 4

Nodes Ground Sump Head Loss Head Loss
Elevation (m) Elevation (m) Method Coefficient
I-1 117.40 115.40 Standard 0.5
I-2 117.15 115.05 Standard 0.5
O-1 118.00 114.60 – –

Inlet Catchment Data for Problem 4

Inlet Time of Additional Area (ha) Inlet C
Concentration Carryover
(min) (m3/s)
I-1 6.3 0 0.45 0.75
I-2 5.2 0 0.22 0.80

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 101

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

Inlet Data for Problem 4 Grate DI-1

(Provided in case missing from inlet library)
Inlet Inlet Type Grate Road Cross- Bypass Manning’s
Length Slope (m/m) Target n Structure width = 0.67 m
(m) Structure length = 0.67 m
I-1 Grate DI-1 1.1 0.02 I-2 0.012 Grate Type = P-50 mm × 100 mm
I-2 Grate DI-1 1.1 0.02 – – Width = 0.76 m Standard Length = 0.76 m

Rainfall Data for Problem 4

Duration (min) Rainfall Intensities (mm/h)
5 Years 10 Years 50 Years
5 69.8 78.7 99.6
10 54.6 61.0 77.5
20 40.6 45.2 56.6
30 31.7 36.2 45.7

Fill out an answer table like the one below for each of the following
situations:
(a) Assume uniform gutters with a slope of 0.02 m/m.
(b) Assume continuously depressed gutters with a road cross-slope of
0.02 m/m, a gutter cross-slope of 0.04 m/m, and a gutter width
of 0.8 m.
(c) Assume continuously depressed gutters as described in (b), as well as
inlet lengths increased from 1.1 to 1.6 m.
(d) Explain the reasons for the differences between the three resulting
tables.

Answer Table for Problem 4

Inlet Gutter Total Flow Intercepted Bypassed Efficiency
Spread to Inlet Inlet Flow Inlet Flow (%)
(m) (m3/s) (m3/s) (m3/s)
I-1
I-2

5. Enter and calculate the storm sewer network shown below using
Manning’s equation and the rainfall data used in Problem 4. The tailwa-
ter condition is free outfall. Assume no clogging of the inlets. Answer the
questions that follow.

102 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

I-3

P-2

I-2

P-3

P-1

O-1

I-1

Schematic for Problem 5

Pipe Data for Problem 5
Pipe Circular Section Upstream Invert Downstream Invert Length Pipe Material
Size (mm) Elevation (m) Elevation (m) (m) (n-value)
P-1 300 425.40 424.70 45 Concrete (0.013)
P-2 450 424.70 424.00 40 Concrete (0.013)
P-3 300 425.00 424.70 54 Concrete (0.013)

Node Data for Problem 5

Node Ground Sump Elevation
Elevation (m) (m)
I-1 428.00 425.40
I-2 427.50 424.70
I-3 428.60 425.00
O-1 427.10 424.00

Inlet Data for Problem 5

Inlet Inlet Type Inlet Road Cross- Inlet Bypass Long. Manning’s
Length Slope (m/m) Location Target Slope n
(m) (m/m)
I-1 Curb DI-3A 1 0.02 On Grade I-2 0.01 0.012
I-2 Curb DI-3A 1 0.02 In Sag – – –
I-3 CurbDI-3A 1 0.02 In Sag – – –

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 103

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

Inlet Catchment Data for Problem 5

Inlet Time of Additional Area (ha) Inlet
Concentration (min) Carryover (m3/s) C
I-1 5.5 0 0.67 0.90
I-2 5.0 0 0.70 0.80
I-3 6.0 0 0.80 0.90
(a) Given the design criteria shown in Table 3.1, what minimum gutter
cross-slope should the 3-m-wide shoulders have (within 1%) on this
high-volume, bidirectional road if the speed limit is 90 km/h?
(b) As an alternative, set all gutter cross-slopes to 4% and replace the
curb inlet at I-2 with a grate inlet called “Test.” You will need to
add this new inlet to the program’s Inlet Library. The inlet has the
characteristics given below. What is the minimum grate length for
I-2 that would satisfy the 50-year design criteria shown in Table 3.1,
assuming no clogging?
Test Inlet Data
Structure width and length 1.2 m
Grate type P-50 mm
Grate width 1m
Grate lengths start at 1 m and are available in increments of 0.5 m.
(c) Using the same data and inlet designed in part (b), calculate the
spread at I-2 assuming 50% clogging.
Solve the following problems using StormCAD. Assume all inlets are
Generic Default 100%, which means that they are assumed to capture
100% of the surface flow.
6. The data that follows describes the existing storm sewer system shown
below. For runoff calculations, assume C = 0.3 for pervious land cover
and C = 0.9 for impervious cover. The ground elevation at the system
discharge point is 17.0 m. All pipes are concrete (n = 0.013) and circular.
Apply a standard head loss coefficient of 0.5 to inlet I-3.
(a) Analyze the system for a design return period of 10 years. Assume a
free outfall condition. Provide output tables summarizing pipe flow
conditions and hydraulic grade lines at the inlets. How is this system
performing?
(b) Increase the size of pipe P-3 to 450 mm. Rerun the analysis and pres-
ent the results. How does the system perform with this improvement?
(c) Local design regulations require that storm sewer systems handle
25-year return periods without flooding. Rerun the analysis for the
improved system in part (b). Does the system meet this performance
requirement?

104 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

(d) The above analyses are run using a default Manning’s n of 0.013.
Many drainage design manuals propose a less conservative design
roughness of 0.012. Reanalyze the improved system under 25-year
flows using n = 0.012. How does this change influence the pre-
dicted performance of the system?
I2

P2

P1 P3 Outlet
I1
I3

Schematic for Problem 6

Rainfall Data for Problem 6
Duration (min) Rainfall Intensity (mm/h)
5 Years 10 Years 25 Years
5 165 181 205
10 142 156 178
15 123 135 154
30 91 103 120
60 61 70 80

Inlet Information for Problem 6

Inlet Ground Impervious Pervious Time of
Elevation Area (ha) Area (ha) Concentration
(m) (min)
I-1 17.9 0.13 0.32 6.0
I-2 18.0 0.15 0.58 5.0
I-3 17.6 0.08 0.36 5.0

Pipe Information for Problem 6

Pipe Upstream Downstream Diameter Length
Invert (m) Invert (m) (mm) (m)
P-1 16.7 16.15 300 56
P-2 16.8 16.1 375 46
P-3 16.1 15.3 375 54

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 105

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

7. You have been asked by the lead project engineer for a water supply util-
ity to design the stormwater collection system for the proposed ground
storage tank and pump station facility shown in the layout. Pipe lengths
for P-1, P-2, P-3, and P-4 are 88, 92, 185, and 46 ft, respectively. See the
CAD drawing and data for the system layout. Assume C = 0.3 for pervi-
ous areas and C = 0.9 for impervious areas.
(a) Using the StormCAD program’s Automatic Design feature, size
the system using the following design data. Use concrete pipe
(n = 0.013) and the 25-year intensity–duration–frequency data
provided in Problem 6 (Hint: StormCAD can mix SI and U.S.
customary units). The top of bank elevation at the outfall ditch is
846.1 ft. The outfall pipe invert must be located at or above eleva-
tion 838.0 ft. Assume that the water surface elevation at the outfall
is 842.0 ft, and that the pipes should have matching soffit (crown)
elevations at every structure. Present your design in tabular form and
provide a profile plot of your design.
(b) During agency review, the county engineer requests that the water
utility and the county work cooperatively to accommodate the
planned construction of an elementary school nearby by increas-
ing the size of the proposed storm system so that it can handle the
design runoff from the school. The county engineer performs his
own calculations and asks that you increase the size of pipes P-3
and P-4 to handle an external CA of 9.5 acres with a time of con-
centration of 12 min. Using StormCAD, introduce the additional
flow at inlet I-3 and revise the facility design using the Automatic
Design functionality of the program. Are all the design constraints
met? What can you say about the flow conditions in pipe P-3 and
pipe P-4? In the rest of the pipes?
(c) If necessary, manually fine-tune and revise the design to meet all
design criteria. Document your design as in part (a).

Water
storage
tank

I-1 P-1
I-2
P-2

I-4
I-3 P-3
P-
4

Outlet

106 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

CAD Drawing for Problem 7

Inlet Information for Problem 7
Inlet Ground Impervious Pervious Time of
Elevation Area (ac) Area (ac) Concentration
(ft) (min)
I-1 848.9 0.25 0.25 6.0
I-2 848.3 0.09 – 5.0
I-3 847.0 0.20 0.49 5.0
I-4 846.5 0.18 0.5 7.5

Design Constraints for Problem 7

Constraint Minimum Maximum
Velocity 2 ft/s 15 ft/s
Cover 4 ft* –
Pipe Slope 0.005 ft/ft 0.100 ft/ft
*Cover constraints may be relaxed at the outfall, where invert elevation
requirements will typically govern.

8. An inlet with a ground elevation of 260 m is connected to an outlet with

a ground elevation of 259 m and a tailwater elevation of 256 m. The pipe
connecting the inlet and outlet is a 20-m-long, concrete, circular pipe
with a diameter of 525 mm and an upstream invert/sump elevation of
257 m. Print and compare pipe slope, average velocity, pipe flow time,
capacity, energy slope, hydraulic grade line upstream, and flow profiles
for each of the following conditions.
(a) Scenario 1:
Inlet inflow = 0.3 m3/s
Downstream invert/sump elevation = 256.9 m
(b) Scenario 2:
Inlet inflow = 0.5 m3/s
Downstream invert/sump elevation = 256.9 m
(c) Scenario 3:
Inlet inflow = 0.5 m3/s
Downstream invert/sump elevation = 256.7 m
(d) Scenario 4:
Inlet inflow = 0.5 m3/s
Downstream invert/sump elevation = 256.2 m
What are the differences between the four scenarios, and why do they occur?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 107

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

9. The system described below drains into a river. Because the water surface
elevation of the river varies, analyze the following system with different
tailwater elevations. Use the Darcy–Weisbach equation because it is more
appropriate for pressure flow than Manning’s equation. All pipes are con-
crete (roughness = 0.122 mm), circular pipes. The pipe and inlet data
follow. The ground elevation at the outlet is 312.0 m.
What is the hydraulic grade at the entrance of inlet I-2 under the follow-
ing tailwater conditions for the 10-year storm event?
(a) Tailwater elevation = 310.0 m
(b) Tailwater elevation = 311.0 m
(c) Tailwater elevation = 313.0 m
I-3
I-1

P-2

P-1

I-2
P-3

I-4
4
P-

Outlet

Schematic for Problem 9

Rainfall Data for Problem 9
Duration (min) Rainfall Intensity (mm/h)
2 years 10 years 100 years
5 80 130 210
15 65 97 145
30 45 68 100
60 25 37.5 55

108 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

Pipe Information for Problem 9

Pipe Length Diameter Upstream Downstream
(m) (mm) Invert Invert Elevation
Elevation (m) (m)
P-1 23 300 312.4 311.5
P-2 25 300 311.5 311.0
P-3 32 525 311.0 310.5
P-4 17 525 310.5 310.4

Inlet Information for Problem 9

Inlet Area1 C1 Area2 C2 tc Ground Head Loss Coefficient Sump
(ha) (ha) (min) Elev. (m) (Standard Method) Elev. (m)
I-1 0.2 0.7 0.35 0.6 5.0 315.0 0.5 312.4
I-2 0.3 0.8 0.18 0.6 5.0 314.5 0.7 311.0
I-3 0.8 0.4 0.3 0.7 6.0 314.3 0.5 311.5
I-4 0.9 0.4 0.3 0.5 8.0 313.0 0.5 310.5

10. A schematic of an existing storm system for a residential subdivision

is shown below. The rainfall, pipe, inlet, and hydrologic information
are provided in the data tables. All pipes are circular, PVC pipes with
Manning’s n = 0.010.

I-1
P-1

P-2 I-2

I-3

P-3

Outlet

Schematic for Problem 10

(a) Analyze the system for a 25-year storm event. Assume a free outfall (criti-
cal depth) condition. Provide output tables summarizing pipe flow, veloc-
ity, and hydraulic grade at the upstream end of each pipe. How is this
system performing?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 109

 3 Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

(b) If the existing system were constructed using CMP (n = 0.024) instead
of PVC, how would the system perform in a 25-year storm event?
(c) If the outfall were discharging to a pond that had a water surface elevation
equal to that of the ground (using the original PVC pipes), how would
the system perform in a 25-year storm? A 50-year storm?
(d) Using the tailwater conditions specified in part (a), if the park were paved
(with a C = 0.9 and time of concentration = 8 min), how would the
system perform in a 25-year storm? In a 50-year storm?

Rainfall Data for Problem 10

Duration (min) Rainfall Intensity (in./h)
25 years 50 years
5 4.00 4.60
15 2.80 3.20
30 1.80 2.00
60 0.75 0.90
100 0.55 0.65

Pipe Information for Problem 10

Pipe Length Diameter Upstream Downstream
(ft) (in) Invert Invert
Elevation (ft) Elevation (ft)
Pipe-1 70 18 166.70 165.30
Pipe-2 40 18 165.20 164.80
Pipe-3 53 24 164.80 164.40

Inlet Information for Problem 10

Inlet Ground Rim Sump Head Loss Coefficient
Elevation Elevation Elevation (Standard Method)
(ft) (ft) (ft)
1 172.00 172.00 166.50 0.5
2 172.00 172.00 165.00 0.5
3 168.00 168.00 164.50 0.5
Outfall 166.50 166.50 164.40 –

110 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Inlets, Gravity Piping Systems, and Storm Sewer Design
Problems

Inlet Hydrologic Information for Problem 10

Inlet Area (ac) Area Type Description C tc (min)
1 8 Residential Single Family 0.5 12
1 Pavement Asphalt/Concrete 0.7
0.04 Misc. Playground 0.2
2 1.2 Pavement Asphalt/Concrete 0.7 5
3 10 Misc. Park 0.2 15

11. Use the same network and data provided in part (a) of Problem 10.
During a 25-year storm, it was determined that inlet 1 is not capable of
capturing all of the flow produced by the runoff for its catch basin. A
flow of 5 cfs is bypassed to inlet 2, which is in sag (all gutter flow is cap-
tured). Hint: You can apply negative and positive additional carryovers
to inlets I-1 and I-2 under the Catchment tab.
(a) What is the gutter flow captured by inlet 1?
(b) What is the gutter flow captured by inlet 2?
(c) How is the system performing?
(d) A flow monitoring study was performed and the following data was
collected:

Pipe 25-Year Storm 50-Year Storm

(cfs) (cfs)
Pipe-1 10 11.5
Pipe-2 17.3 19.9
Pipe-3 21.0 25.0

Analyze the system using these known flow values for both the 25- and 50-year
storms. How is the system performing? Report the flows, upstream velocities,
and hydraulic grade lines for each pipe.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 111

 4 Culvert Hydraulics

4.1 Culvert Systems

Culverts are commonplace in practical hydraulic design, including appli-
cations such as roadway crossings and detention pond outlets. A roadway
cross-drainage culvert is typically designed to carry flows from one side of the
road the other, without allowing the headwater (HW) (water surface elevation
just upstream of the culvert) to exceed safe levels. When engineers analyze
these culvert systems, they are usually trying to solve for one or more of the
following:
• the size, shape, and number of new or additional culverts required to pass
a design discharge;
• the hydraulic capacity of an existing culvert system under an allowable HW
constraint;
• the upstream flood level at an existing culvert system resulting from a dis-
charge rate of special interest; and
• hydraulic performance curves for a culvert system, which are used to assess
hydraulic risk at a crossing or as input for another hydraulic or hydrologic
model.
Similar to a storm sewer system, a culvert system consists of a hydrologic com-
ponent, a culvert component, and a tailwater component. Although the water-
shed hydrology and tailwater conditions are almost identical to those that you
would normally use in analyzing a storm sewer or other open channel transport
system, there are additional hydraulic computations typically reserved for cul-
vert analysis only.
In this chapter, procedures for designing culverts using CulvertMaster are
presented. The use of nomographs as aids to hand calculations is described
in detail in the Hydraulic Design of Highway Culverts manual (Normann
et al., 2001, and Schall et al., 2012). Further discussion on hydraulic theory,
end treatments, alignments, and roadway overtopping is presented by Haestad
Methods and R. Durrans (2003)

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 113

 4 Culvert Hydraulics
Outlet Control Hydraulics

Culvert Hydraulics
Obtaining accurate solutions in culvert hydraulics can be a formidable compu-
tational task. Culverts act as a significant constriction to flow and are subject to
a range of flow types, including both gradually varied and rapidly varied flows.
It is this mix of flow conditions and the highly transitional nature of culvert
hydraulics that makes the hydraulic solutions so difficult. For this reason, the
documented approach is to simplify the hydraulics problem and analyze the
culvert system using two different assumptions of flow control:
• Outlet control assumption—Computes the upstream HW depth using
conventional hydraulic methodologies that consider the predominant
losses due to the culvert barrel friction, as well as the minor entrance and
exit losses. The tailwater condition during the design storm has an impor-
tant effect on the culvert system.
• Inlet control assumption—Computes the upstream HW depth resulting
from the constriction at the culvert entrance, while neglecting the culvert
barrel friction, tailwater elevation, and other minor losses.
The controlling HW depth is the larger value of the computed inlet control and
outlet control HW depths. Because the culvert system may operate under inlet
control conditions for a range of flow rates and under outlet control conditions
for another range, calculations must be performed for both control conditions.

4.2 Outlet Control Hydraulics

Calculations for outlet control HW depths are similar to those performed for
storm sewer analysis. The HW depth is found by summing the tailwater depth,
entrance minor loss, exit minor loss, and friction losses along the culvert barrel.
The energy basis for solving the outlet control HW for a culvert is given by the
basic energy equation and is presented graphically in Figure 4.1.
Full flow conditions (Cases A, B, and C in Figure 4.1) result in the same velocity
(and therefore the same velocity head) throughout the length of the culvert; there-
fore the energy grade line and hydraulic grade line are parallel throughout the length
of the barrel. At the culvert entrance, there is a slight increase in velocity and dip in
the Hydraulic Grade Line (HGL) caused by the flow contraction that occurs there.
The energy equation can be rewritten specifically for culvert terms, which
results in the following form:

V2 V2
HWO + 2ug = TW + 2dg + HL
where
HWO = headwater depth above the outlet invert (m, ft);
Vu = approach velocity (m/s, ft/s);
TW = tailwater depth above the outlet invert (m, ft);

114 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Outlet Control Hydraulics

Water (A)
surface
H
W.S.
HW

(B)

H
W.S. W.S.
HW

(C)

W.S.
H
HW
W.S.

(D)

W.S. E.G.
H
HW
W.S.

(E)

E.G. H
W.S.
HW W.S.

Figure 4.1: Outlet control flow conditions.

Vd = exit velocity (m/s, ft/s); and

HL = sum of all losses, including the entrance minor loss (HE), barrel friction
losses (HF), the exit loss (HO), and other losses (m, ft).
Culverts often connect ponds or other bodies of water with negligible veloc-
ities, so the approach velocity and the velocity downstream of the culvert are
often neglected, resulting in the following equation:

HWO = TW + HL
where
HWO = headwater depth above the outlet invert (m, ft),
TW = tailwater depth above the outlet invert (m, ft), and
HL = sum of all losses as listed above (m, ft).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 115

 4 Culvert Hydraulics
Outlet Control Hydraulics

Friction Losses
Culverts are frequently hydraulically short (entrance and exit losses exceed pipe
friction losses), and uniform flow depths are not always achieved. For this rea-
son, gradually varied flow methods are well suited to the analysis. For a more
detailed description of gradually varied flow, see Chapters 3.6 and 9.4

Entrance Minor Loss

The entrance loss is caused by the contraction of flow as it enters the cul-
vert and is a function of the barrel velocity head just inside the entrance. It is
expressed by the following equation:

⎛V 2 ⎞
He = ke ⎜⎜⎜ ⎟⎟⎟
⎜⎝ 2 g ⎟⎠

where
He = entrance loss (m, ft),
ke = entrance loss coefficient,
V = velocity just inside the barrel entrance (m/s, ft/s), and
g = gravitational acceleration (m/s2, ft/s2).
The entrance loss coefficient varies depending on the type of inlet that is pres-
ent. The smoother the transition from the channel or pond into the culvert,
the lower the loss of coefficient is. Values for the coefficients are presented in
Table 4.1.

Table 4.1: Entrance Loss Coefficients (Normann et al., 2001)

Culvert Type Entrance Type and Description Entrance Loss Coefficient, ke

Projecting from fill, socket end (groove end) 0.2
Projecting from fill, square cut end 0.5
Headwall or headwall with wingwalls
Socket end of pipe (groove end) 0.2
Square edge 0.5
Pipe, concrete
Rounded (radius = 1/12D) 0.2
Mitered to conform to fill slope 0.7
End-section conforming to fill slope* 0.5
Beveled edges, 33.7° or 45° bevels 0.2
Side- or slope-tapered inlet 0.2

116 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Outlet Control Hydraulics

Culvert Type Entrance Type and Description Entrance Loss Coefficient, ke

Projecting from fill (no headwall) 0.9
Headwall or headwall and wingwalls square edge 0.5
Pipe or Pipe Arch, Mitered to conform to fill slope, paved or unpaved slope 0.7
Corrugated Metal End-section conforming to fill slope* 0.5
Beveled edges, 33.7° or 45° bevels 0.2
Side- or slope-tapered inlet 0.2
Headwall parallel to embankment (no wingwalls)
Square-edged on 3 edges 0.5
Rounded on 3 edges to radius of 1/12 bbl dimension 0.2
Or beveled edges on 3 sides
Wingwalls at 30°–75° to barrel
Square-edged at crown 0.4
Box, Reinforced
Crown edge rounded to radius of 1/12 bbl dimension 0.2
Concrete
Or beveled top edge
Wingwalls at 10°–25° to barrel
Square-edged at crown 0.5
Wingwalls parallel (extension of sides)
Square-edged at crown 0.7
Side- or slope-tapered inlet 0.2

*
Note: “End-section conforming to fill slope,” made of either metal or concrete, are the sections commonly available from
manufacturers. From limited hydraulic tests, they are equivalent in operation to a headwall in both inlet and outlet
control. Some end sections with a closed taper in their design have superior hydraulic performance.

Exit Minor Loss

The exit loss is an expansion loss, which is a function of the change in velocity
head that occurs at the discharge end of the culvert. In culvert hydraulics, this
sudden expansion loss is expressed as:

⎡ 2 Vd 2 ⎤
HO = 1.0 ⎢V − 2g ⎥
⎢⎣ 2 g ⎦⎥

where
Vd = velocity of the outfall channel (m/s, ft/s),
V = velocity just inside the end of the culvert barrel (m/s, ft/s), and
g = gravitational acceleration (m/s2, ft/s2).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 117

 4 Culvert Hydraulics
Inlet Control Hydraulics

When the discharge velocity is negligible (as for a pond or slow-moving

channel), the exit loss is equal to the barrel velocity head.

Gradually Varied Flow Analysis

Culverts operating under outlet control with unsubmerged outlets (Cases D
and E in Figure 4.1) require an analysis of the hydraulic profile in order to
determine the HW elevation. In Case D, the water surface reaches the crown of
the culvert at some point upstream of the barrel. In both cases, the flow passes
through critical depth at the outlet end of the barrel. The flow is represented
by the M2 profile (see Figure 3.8).

4.3 Inlet Control Hydraulics

When a culvert is operating under inlet control conditions (as in Figure 4.2),
the hydraulic control section is the culvert entrance itself. This means that the
friction and minor losses within the culvert are not as significant as the losses
caused by the entrance constriction.
Because the control section for culverts operating under inlet control condi-
tions is at the upstream end of the culvert barrel, critical depth generally occurs
at or near the inlet, and flows downstream of the inlet are supercritical. The
hydraulic profile and outlet velocities are determined using frontwater gradu-
ally varied flow techniques.
Three types of inlet control hydraulics are in effect over a range of culvert
discharges:
• Unsubmerged (A and B)—For low discharge conditions, the culvert
entrance acts as a weir. The hydraulics of weir flow are governed by empiri-
cal working equations developed as a result of model tests.
• Submerged (C)—When the culvert entrance is fully submerged, it is
assumed to be operating as an orifice.
• Transitional (D)—This flow type occurs in the poorly defined region just
above the unsubmerged zone and below the fully submerged zone.

Unsubmerged Flow
There are two equations for unsubmerged (weir) flow. The first is based on the
specific head at critical depth (with correction factors), whereas the second
equation is more closely related to a weir equation. Although either equation
will produce adequate results, the second equation is more commonly used
during hand calculations because it is easier to apply.
Note that these equations were developed by the U.S. FHWA and as such are
only intended for use in U.S. customary units. SI units should be converted
before applying these equations.

118 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Inlet Control Hydraulics

(A)

HW Water surface

W.S.

Inlet and outlet unsubmerged

(B)

Water surface
HW W.S.

Inlet unsubmerged
outlet submerged

(C)

Water surface
HW

W.S.
Inlet submerged
outlet unsubmerged

(D)
Median drain

HW W.S.

Water surface

Inlet and outlet submerged

Figure 4.2: Inlet control flow conditions.

Unsubmerged, Form 1

HWi H ⎡ Q ⎤M
= c +K ⎢ AD0.5 ⎥ − 0.5S
D D ⎢⎣ ⎥⎦

Unsubmerged, Form 2

HWi ⎡ Q ⎤M
=K⎢ ⎥
D ⎢⎣ AD0.5 ⎥⎦

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 119

 4 Culvert Hydraulics
Inlet Control Hydraulics

where
HWi = headwater depth above the control section invert (ft);
D = interior height of the culvert barrel (ft);
Hc = specific head at critical depth, yc + Vc2/2g (ft), where yc = critical depth
(ft) and Vc = velocity at critical depth (ft/s);
Q = culvert discharge (ft3/s);
A = full cross-sectional area of the culvert barrel (ft2);
S = culvert barrel slope (ft/ft); and
K, M = constants from Table 4.2.
These equations are applicable up to approximately Q/AD0.5 = 3.5. When
using the first equation with mitered inlets, use a slope correction factor (last
term of Form 1 equation) of +0.7S instead of −0.5S.

Submerged Flow
The equation for submerged (orifice) flow is:

HWi ⎡ Q ⎤2
= c⎢ ⎥ + Y − 0.5S
D ⎢⎣ AD0.5 ⎥⎦
where
HWi = headwater depth above the control section invert (ft),
D = interior height of the culvert barrel (ft),
Hc = specific head at critical depth, yc + Vc2/2g (ft),
Q = culvert discharge (ft3/s),
A = full cross-sectional area of the culvert barrel (ft2),
S = culvert barrel slope (ft/ft), and
c, Y = constants from Table 4.2.
This equation for submerged flow is applicable above approximately Q/AD0.5 =
4.0. When using this equation with mitered inlets, use a slope correction factor
(last term of equation) of +0.7S instead of −0.5S.

Table 4.2: Coefficients for Inlet Control Design Equations

Shape and Equation Unsubmerged Submerged

Material Inlet-Edge Description Form K M c Y
Square-edge w/headwall 0.0098 2.0 0.0398 0.67
Circular
Groove end w/headwall 1 0.0018 2.0 0.0292 0.74
Concrete
Groove end projecting 0.0045 2.0 0.0317 0.69

120 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Inlet Control Hydraulics

Shape and Equation Unsubmerged Submerged

Material Inlet-Edge Description Form K M c Y
Headwall 0.0078 2.0 0.0379 0.69
Circular
Mitered to slope 1 0.0210 1.33 0.0463 0.75
CMP
Projecting 0.0340 1.50 0.0553 0.54
Beveled ring, 45° bevels 0.0018 2.50 0.0300 0.74
Circular 1
Beveled ring, 33.7° bevels 0.0018 2.50 0.0243 0.83
30°–75° wingwall flares 0.026 1.0 0.0347 0.81
Rectangular
90° and 15° wingwall flares 1 0.061 0.75 0.0400 0.80
Box
0° wingwall flares 0.061 0.75 0.0423 0.82
Rectangular 45° wingwall flare d = 0.043D 0.510 0.667 0.0309 0.80
2
Box 18°–33.7° wingwall flare d = 0.083D 0.486 0.667 0.0249 0.83
90° headwall w/3/4″ chamfers 0.515 0.667 0.0375 0.79
Rectangular
90° headwall w/45° bevels 2 0.495 0.667 0.0314 0.82
Box
90° headwall w/33.7° bevels 0.486 0.667 0.0252 0.865
3/4″ chamfers; 45° skewed headwall 0.545 0.667 0.0505 0.73
Rectangular 3/″ chamfers; 30° skewed headwall 0.533 0.667 0.0425 0.705
2
Box 3/4″ chamfers; 15° skewed headwall 0.522 0.667 0.0402 0.68
45° bevels; 10°–45° skewed headwall 0.498 0.667 0.0327 0.75
45° nonoffset wingwall flares 0.497 0.667 0.0339 0.803
Rectangular
18.4° nonoffset wingwall flares 0.493 0.667 0.0361 0.806
Box 2
18.4° nonoffset wingwall flares
3/4″ Chamfers
30° skewed barrel 0.495 0.667 0.0386 0.71
Rectangular 45° wingwall flares—offset 0.497 0.667 0.0302 0.835
Box 33.7° wingwall flares—offset 2 0.495 0.667 0.0252 0.881
Top Bevels 18.4° wingwall flares—offset 0.493 0.667 0.0227 0.887
90° headwall 0.0083 2.0 0.0379 0.69
CM Boxes Thick wall projecting 1 0.0145 1.75 0.0419 0.64
Thin wall projecting 0.0340 1.5 0.0496 0.57
Horizontal Square edge w/headwall 0.0100 2.0 0.0398 0.67
Ellipse Groove end w/headwall 1 0.0018 2.5 0.0292 0.74
Concrete Groove end projecting 0.0045 2.0 0.0317 0.69

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 121

 4 Culvert Hydraulics
CulvertMaster and Tutorials

Shape and Equation Unsubmerged Submerged

Material Inlet-Edge Description Form K M c Y
Vertical Square edge w/headwall 0.0100 2.0 0.0398 0.67
Ellipse Groove end w/headwall 1 0.0018 2.5 0.0292 0.74

Concrete Groove end projecting 0.0095 2.0 0.0317 0.69

Pipe Arch 90° headwall 0.0083 2.0 0.0379 0.69

18″ Corner Mitered to slope 1 0.0300 1.0 0.0463 0.75
Radius CM Projecting 0.0340 1.5 0.0496 0.57
Pipe Arch Projecting 0.0300 1.5 0.0496 0.57
18″ Corner No bevels 1 0.0088 2.0 0.0368 0.68

Radius CM 33.7° bevels 0.0030 2.0 0.0269 0.77

Pipe Arch Projecting 0.0300 1.5 0.0496 0.57

31″ Corner No bevels 1 0.0088 2.0 0.0368 0.68

Radius CM 33.7° bevels 0.0030 2.0 0.0269 0.77

90° headwall 0.0083 2.0 0.0379 0.69

Arch CM Mitered to slope 1 0.0300 1.0 0.0463 0.75
Thin wall projecting 0.0340 1.5 0.0496 0.57
Smooth tapered inlet throat 0.534 0.555 0.0196 0.90
Circular 2
Rough tapered inlet throat 0.519 0.64 0.0210 0.90
Tapered inlet-beveled edges 0.536 0.622 0.0368 0.83
Elliptical
Tapered inlet-square edges 2 0.5035 0.719 0.0478 0.80
Inlet Face
Tapered inlet-thin edge projecting 0.547 0.80 0.0598 0.75
Rectangular Tapered inlet throat 2 0.475 0.667 0.0179 0.97
Rectangular Side tapered—less favorable edges 0.56 0.667 0.0446 0.85
2
Concrete Side tapered—more favorable edges 0.56 0.667 0.0378 0.87
Rectangular Slope tapered—less favorable edges 0.50 0.667 0.0446 0.65
2
Concrete Slope tapered—more favorable edges 0.50 0.667 0.0378 0.71

CMP is Corrugated Metal Pipe; CM is Corrugated Metal.

4.4 CulvertMaster and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of

122 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Problems

hydraulic systems without needing to spend a lot of time inputting data. As

new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software.

What Does CulvertMaster Do?

CulvertMaster is a program that helps civil engineers design and analyze cul-
vert hydraulics. Just click a button to create a new worksheet, enter data in the
clearly labeled fields, and click to calculate. You can solve for most hydraulic var-
iables, including culvert size, flow, and HW. It also allows you to generate and
plot rating tables and graphical output showing computed flow characteristics.
CulvertMaster contains options to automatically generate peak discharges,
including the Rational Method and SCS Graphical Peak Discharge Method.
You have a choice of culvert barrel shapes, including circular pipes, arches, boxes,
and more. Calculations handle free surface, pressure, and varied flow situations,
including backwater and drawdown curves. CulvertMaster’s flexible reporting
features allow you to print the results in report format or as a graphical plot.
The theory and background used in CulvertMaster are described in detail in
this chapter, as well as in the CulvertMaster online help system.

How Can You Use CulvertMaster?

CulvertMaster can design or analyze culverts and compute HW for a range of
flow rates. For a typical CulvertMaster project, an engineer may be interested
in several culvert locations and try several designs for each location. You can
use CulvertMaster to
• size culverts,
• compute and plot rating tables and curves,
• view output in both English and SI units, and
• generate professional-looking reports and graphs.

4.5 Problems
Solve the following problems using CulvertMaster’s Quick Culvert Calculator.
1. A culvert is 11-m long and has upstream and downstream inverts of
263.4 and 263.1 m, respectively. The downstream tailwater elevation is
below the downstream pipe invert.
For a Ke of 0.5 and a Manning’s n value of 0.013, what minimum diam-
eter concrete circular culvert (in mm) is required to pass 1.4 m3/s under a
roadway with a maximum allowable headwater elevation of 265.2 m?
What is the headwater elevation for the selected culvert?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 123

 4 Culvert Hydraulics
Problems

2. An existing 9.73-m-long, 560×420-mm steel and aluminum var CR

arch (n = 0.025) has a 90-degree headwall. The inverts are 33.11 and
33.09 m. Assuming no tailwater effects, what is the maximum discharge
that can pass through this culvert before the maximum allowable water
surface elevation of 34.25 m is exceeded?
3. Twin 1,220×910-mm box culverts (n = 0.013, entrance has 90-degree
and 15-degree wingwall flares) carry 8.5 m3/s along a 31-m length of pipe
constructed at a 1.0% slope. The tailwater depth is 0.61 m.
(a) What is the headwater depth?
(b) Are the culverts flowing under inlet or outlet control conditions?
(c) What would the headwater depth and flow regime for a flow rate of
380 cfs?
4. A 100-ft horizontal concrete ellipse pipe (n = 0.013, Ke = 0.5) on a 3.5%
slope is required to carry 65 cfs. Assume there is a free outfall. The maxi-
mum allowable headwater is 4.7 ft.
(a) What is the minimum pipe size (in in.) required?
(b) What would be the minimum size for a vertical concrete ellipse
pipe?
5. A 12.2-m-long, 920 × 570-mm concrete arch pipe (n = 0.013, groove
end with headwall entrance) constructed at a 0.8% slope carries 1.84 m3/s.
(a) If there is a constant tailwater depth of 0.3 m, what is the headwater
depth for both inlet and outlet control conditions?
(b) Is the culvert flowing under inlet or outlet control conditions?
(c) What would be the result if the tailwater were 0.5 m deeper?
6. Triple 3,050 × 1,830-mm, 200-m-long, concrete box culverts (n =
0.013, Ke = 0.5) carry 110 m3/min. The culverts are constructed on
a 1.4% slope and discharge into a pond with a depth 0.8 m above the
downstream culvert invert. What is the exit velocity of the culverts?
7. Twin culverts are proposed to discharge 6.5 m3/s. The culverts will be
36.6-m long and have inverts of 20.1 and 19.8 m. The design engineer
analyzed the three culvert systems described below. The tailwater eleva-
tion is free outfall. Which of the following proposed culverts will result in
the highest headwater elevation? The lowest?
(a) 1,200-mm circular concrete pipes (n = 0.013 and Ke = 0.5)
(b) 1,220 × 910-mm concrete box culverts (n = 0.013, 90° and 15°
wingwall flares at entrance)
(c) 1,630 × 1,120-mm steel and aluminum var CR arches (n = 0.025
and Ke = 0.5)

124 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Culvert Hydraulics
Problems

8. A 40-ft-long elliptical pipe (n = 0.013 and Ke = 0.5) will be constructed

to carry 80 cfs with inverts of 22.6 and 22.1 ft. The tailwater is constant at
an elevation of 24.0 ft. Which pipe will provide a lower headwater eleva-
tion: a 38 × 60-in. horizontal ellipse or a 60 × 38-in. vertical ellipse?
9. A circular concrete culvert has a free outfall. The culvert is 60-ft long on a
2% slope and is 30 in. in diameter. The culvert entrance will project from
the embankment. Create a rating table that correlates the culvert dis-
charge to the depth of the headwater, from 0 to 6 ft in 0.5-ft increments.
10. A detention pond drains through a circular concrete culvert that has a
square-edge inlet with a headwall (n = 0.013, Ke = 0.5). The peak dis-
charge from the pond is 4 m3/s, and the headwater elevation is 15.5 m.
The culvert is 10-m long on a 5% slope. The depth of the headwater is
4 m, and the tailwater elevation is below the elevation of the downstream
invert. What is the elevation of the upstream invert? What size culvert is
required?
11. A stream flows under a road through a 48-in. concrete circular culvert
that has a square-edge inlet with a headwall (n = 0.013, Ke = 0.5). The
culvert is on a 5% slope with a length of 31.2 ft, a downstream invert
elevation of 92.5 ft, and a free tailwater outfall. The surface of the road is
at an elevation of 108.3 ft, and the stream has a base flow rate of 100 cfs.
(a) If the runoff from a rainstorm adds another 200 cfs of flow to the
stream, will the road be flooded?
(b) What is the maximum flow rate that will not flood the road?
12. A 15-m-long culvert that has a projecting groove-end inlet is needed on
a 2% slope with a free outfall. The elevation of the upstream invert is
100 m, and the maximum headwater depth is 5 m.
(a) What size concrete arch culvert (n = 0.013, Ke = 0.5) is needed to
carry 6 m3/s of water?
(b) Is the culvert operating under inlet or outlet control?
(c) If the tailwater depth changes to 0.3 m, what is the flow control?
(d) If the tailwater depth changes to 3.3 m, what is the flow control?
(e) Why does the flow control change or not change?
13. A 15-m-long concrete vertical ellipse culvert that has a square-edge inlet
with a headwall (860 × 550 mm, n = 0.013, Ke = 0.5) has upstream
and downstream invert elevations of 2.00 and 1.46 m, respectively. If the
tailwater elevation is 3.02 m, at what flow rate (greater than 0.1 m3/s) are
inlet and outlet control headwater elevations equal? (Hint: Use rating
curves.) Is this free-surface flow or pressurized flow?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 125

 5 Detention Pond Design

5.1 Overview of Stormwater Detention

Stormwater detention is the temporary storage of runoff in basins, under-
ground containers, or on rooftops. Wet or dry detention basins (also called
detention ponds) are the most commonly used structures and are the subject
of this chapter. Once runoff has been collected in a detention basin, it can be
released to downstream properties or conveyances in a controlled manner such
that downstream flooding and other adverse impacts are prevented or allevi-
ated. The construction of stormwater detention facilities is typically associated
with new developments, as these usually result in increased stormwater runoff
rates and volumes.
Figure 5.1 demonstrates the basic principles of detention pond analysis. An
inflow runoff hydrograph (see Chapter 2) from one or more contributing
drainage areas is directed to a storage facility (in this case, a graded detention
pond). Runoff is then released from the facility at a controlled rate through a
properly sized outlet structure such as a culvert. The result is a pond outflow
hydrograph that is substantially flatter (i.e., has a lower peak flow) than the
inflow hydrograph.
This chapter provides basic information on stormwater detention design and
analysis concepts. The topics covered include basic detention facility types,
pond volume calculations, outlet structure types, and the storage-indication
method of pond routing. For more in-depth coverage of stormwater detention,
the reader is referred to Stormwater Conveyance Modeling and Design (Haestad
Methods and R. Durrans, 2003)

5.2 Basic Design Considerations

Design criteria for detention ponds vary widely. In the United States, deten-
tion regulations are typically adopted at the municipal or county level, so local
regulations and guidelines must be consulted before the stormwater detention

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 127

 5 Detention Pond Design
Basic Design Considerations

6,970

72

68
Outlet

74
structure

64

62

66
0
66

6,96
68
Outflow
Inflow hydrograph
hydrograph

Spillway

6,9
72 70

Storage

Figure 5.1: Conceptual drawing of a detention pond.

facility can be designed. Despite the wide variation in design practices, the
underlying philosophy of most regulations is that peak runoff rates discharged
from the undeveloped site for one or more design storm events should not be
exceeded after the site development.
Several basic categories of detention facilities exist. Ponds may be wet or dry
between storm events; detention may occur on the surface or underground;
and a facility may lie in a main drainage path or be removed from it. The type
of facility may be dictated by regulations or by the needs and limitations of the
individual site or drainage area.

Predevelopment versus Postdevelopment Criteria

and Recurrence Frequency
As a watershed is developed, the peak discharge rate and the amount of runoff
will tend to increase, whereas the time to peak will typically decrease. The
intent of stormwater detention is typically to limit postdevelopment peak
discharges from a site to discharge rates no greater than those that occurred
prior to development. Thus the conveyance capacities of existing downstream
structures will still be sufficient if these structures were adequately sized prior
to development.
Sometimes allowable detention facility discharges may be based on criteria
other than predevelopment flows. For instance, if a structure downstream of
the property is undersized, the developer may be required to actually lower
peak discharge rates below predevelopment values so as not to exceed its capac-
ity. It is important to note, however, that even if peak discharge objectives are
met, detention ponds typically do not mitigate the increase in runoff volume

128 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Basic Design Considerations

caused by land development. Although the peak flow rate from a site for a
given storm event may be the same as it was prior to development, the duration
of the discharge can be substantially longer.
The criteria described previously must be met for one or more specified design
storm events. For most localities, these design events will be specified in their
regulations. For example, the regulations may specify that postdevelopment
peak flows be less than predevelopment peaks for 24-h, synthetic storms hav-
ing return frequencies of 2, 5, and 10 years. (It is advisable, however, to check
the design for storms of different durations.) The regulations may further
state that an emergency overflow structure capable of handling flows from the
100-year storm be constructed. The magnitude of the largest storm that the
facility must be designed to accommodate sometimes depends on factors such
as site area and land use.

Types and Configurations of Stormwater

Detention Facilities
A dry pond has an outlet that is positioned at or below the lowest elevation in
the pond, such that the pond drains completely between storm events. A wet
pond, on the other hand, is one, the outlet of which is at an elevation above the
pond bottom. Water remains in the pond during dry-weather periods between
storm events and this water can only be depleted by infiltration into the soils
and by evaporation from the water surface. A retention pond is a type of wet
pond in which no outlet structure exists; infiltration and evaporation are the
only means by which water can escape.
Detention ponds can be designed in an almost unlimited number of shapes
and sizes. An aboveground pond usually consists of a depressed or excavated
area with an earthen berm. In cases where there is not ample surface area avail-
able to meet storage requirements, underground detention may be necessary.
Underground detention may consist of a series of large pipes or prefabricated
custom chambers manufactured specifically for underground detention.
Online and off-line ponds are distinguished from one another based on whether
they are positioned along the alignment of the main stormwater drainage path.
An online pond, as its name suggests, is positioned along the pathway and all
runoff must pass through it. An off-line pond is located outside of the main
drainage pathway, and only a portion of the total contributing stormwater
runoff is diverted through it.
Multiple ponds may be necessary to achieve detention requirements under
certain site conditions. Ponds can be connected in series such that the out-
flow hydrograph from one pond is also the inflow hydrograph for the next
pond downstream. In some cases, particularly where the topography is flatter,
the ponds may be close enough in elevation that the downstream pond stage
has tailwater effects on the rating curve for the upstream pond. This type of

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 129

 5 Detention Pond Design
Detention Pond Modeling Concepts

calculation is called interconnected pond modeling and is beyond the scope of

this chapter.

5.3 Detention Pond Modeling Concepts

Stormwater detention design requires estimation of a complete runoff hydro-
graph, as opposed to the single peak discharge typically used in storm sewer
design. Procedures for inflow hydrograph estimation were presented in Chapter 2.
This section provides a conceptual review of stormwater runoff hydrographs
and discusses how they are shaped and altered by added detention facilities.
Figure 5.2 serves as a definition sketch for the basic hydrograph concepts
associated with detention. The inflow hydrograph represents the runoff from
a watershed. Its principal attributes are the peak discharge rate (in ft3/s or
m3/s) and the time from the beginning of the storm to the peak discharge.
The area under this curve represents the total volume of runoff resulting from
this storm event.
The outflow hydrograph represents the discharge from a detention facility. The
outflow hydrograph is obtained by performing a routing analysis (discussed
later in this chapter) for a given detention pond, outlet configuration, and
inflow hydrograph. The reduction in the peak discharge seen in the outflow
hydrograph is referred to as attenuation, and the increase in the time to peak
discharge is referred to as lag.
If the infiltration and evaporation occurring in the pond during the storm
event are negligible, the areas under the two curves (i.e., the runoff volumes)
must be equal. The area between the inflow and outflow hydrographs repre-
sents the maximum volume of water present in the pond during the storm

Inflow
hydrograph

Attenuation
Discharge

Change in
storage volume
Outflow
hydrograph

Lag

Time

Figure 5.2: Detention pond inflow and outflow hydrographs.

130 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Components of Detention Facilities

event. If this volume is greater than the storage capacity of the pond, the pond
will overflow.
The following three scenarios summarize the relationship of a pond’s inflow
rate, outflow rate, storage, and water surface elevation for any given time inter-
val within a stormwater runoff event:
• If the average inflow rate is greater than average outflow rate during a time
interval, the volume of stored water in the pond increases during the time
interval, and the water surface elevation in the pond increases.
• If the average inflow rate is equal to the average outflow rate during a time
interval, the volume of stored water does not change during the interval,
and the water surface elevation remains constant.
• If the average outflow rate is greater than the average inflow rate during a
time interval, the volume of stored water decreases during the time interval,
and the water surface elevation in the pond decreases.
Most detention ponds have uncontrolled outlets (i.e., their outlets have no
valves or gates that can be opened or closed). For a pond with an uncontrolled
outlet into a free outfall, the peak of the outflow hydrograph from the pond
will occur at the point where the outflow hydrograph intersects the receding
limb of the inflow hydrograph (see Figure 5.2).
This concept can be used as a check on the validity of a set of routing calcu-
lations and can be reasoned as follows. Prior to the time of the intersection
point, the inflow rate is larger than the outflow rate and the volume of water
in the pond is increasing. After that intersection time, the volume is decreasing
because the outflow rate is greater than the inflow rate. Thus the maximum
stored volume of water in the pond occurs at the time where the two hydro-
graphs intersect. Because the outflow rate from a pond increases as the head-
water depth in the pond increases, it follows that the maximum outflow rate
must occur at the maximum depth, which also corresponds to the maximum
storage during the event. Finally, at the point of intersection, the outflow is
equal to the inflow rate, resulting in a zero rate of change of stored volume at
the moment in time when the water surface crests in the pond.

5.4 Components of Detention Facilities

Detention facilities usually consist of an aboveground pond that is excavated
and graded into the land surface, or they may consist of shallow collection areas
in parking lots. In areas where available space is at a premium, it may be eco-
nomically advantageous to use underground storage consisting of pipes and/
or vaults. Thus the physical components of individual detention facilities can
be quite variable, but some general observations can be made. The following
subsections provide descriptions of the most common elements of detention
facilities.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 131

 5 Detention Pond Design
Components of Detention Facilities

Dam Embankments
In many cases, detention facilities are formed by excavating and grading a
depression in the land surface, and the resulting pond has no dam embank-
ment per se. In other cases, a stream or localized depression is dammed to
create a detention pond. Most typically, a combination of these two approaches
is used in which an existing depression is enlarged and dammed. The design
of the dam should conform to accepted geotechnical engineering practices.
Governmental dam safety and permitting requirements may be applicable if
the embankment height or storage exceeds a regulatory limit.

Freeboard
Freeboard for a storm event is defined as the vertical distance from the maxi-
mum water surface elevation (stage) to the top of the dam. The storage availa-
ble within the freeboard elevation range adds a “safety” volume to the required
detention storage volumes that exist below this elevation. Some jurisdictions
and reviewing agencies mandate freeboard requirements for the maximum
design event.

Pond Bottoms and Side Slopes

Side slopes of ponds, especially for wet ponds that become attractions for peo-
ple, should be gentle. Barrier vegetation, such as cattails, can also be installed
along the perimeters of wet ponds to limit access to the water’s edge.
Detention pond bottoms should have a slope sufficient to ensure complete
drainage between storm events in dry ponds, and to facilitate pond drainage
for maintenance in the case of wet ponds. For dry ponds, the engineer should
consider providing a riprap or concrete low-flow channel from the inlet to the
outlet structure to assist in draining the pond bottom.

Outlet Structures
The detention pond outlet structure allows flow to discharge from the pond at
a controlled rate. Early detention ponds were usually designed to control only a
single runoff event having a specified recurrence interval; thus these ponds had
very simple outlet structures. This practice provided little to no benefit with
smaller or larger storm events.
Modern detention design typically focuses on control of multiple storm events
and often requires several outlet openings at different pond stages. These mul-
tiple openings can consist of multiple pipe outlets with differing diameters or
invert elevations, orifices at various levels, overflow weirs, or any combination
of these elements. Figure 5.3 depicts a profile view of an outlet structure con-
figuration consisting of several structural components.

132 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Standpipe or Multiple riser

inlet box side openings Spillway

Culvert

Pond invert

Figure 5.3: Detention pond inflow and outflow hydrographs.

Overflow Spillways
All detention ponds, regardless of the type, should be fitted with overflow
spillway facilities to convey flows in excess of those for which the detention
facility was designed. The overflow spillway may be an extension of a nor-
mal weir outlet, or it may be completely separate. To prevent undue damages
downstream, consideration must be given to the paths that pond overflows will
take. These outflow pathways are part of the major drainage system referred to
in previous chapters.

5.5 Routing Data: Storage and Hydraulic Relationships

Hydrograph routing through a detention facility is the process of computing
an outflow hydrograph and water surface elevation data based on a specified
inflow hydrograph. Pond routing requires two key types of data in order to
describe the pond’s storage and hydraulic characteristics:
• stage versus storage volume data to describe the pond’s shape and size
• stage versus discharge data to describe the pond outlet hydraulics.
These relationships are discussed in the following subsections.

Stage versus Storage Volume

The stage versus storage relationship for a pond relates pond water surface
elevation (the stage) or water depth (h) to the volume of water stored (S). The
relationship may be given in a table or graphed as in Figure 5.4. Precision of
the curve or table increases as the incremental elevation between the points
decreases. The designer must judge the appropriate degree of storage approxi-
mation when setting the elevation increment size for the rating curve.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 133

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Figure 5.4 demonstrates how stage–storage data is used in routing to deter-

mine the relative rise or fall in water surface for a given change in volume.
In this example, h1 and S1 are, respectively, the elevation and volume at the
beginning of the time step, and ΔS is the change in volume during the time
step calculated from the difference between the integrated inflow and outflow
hydrographs. At the end of the time step, volume S2 = S1 + ΔS. Elevation h2
at the end of the time step can be found from the curve based on S2. In this
example, the net change in storage is positive; thus the change in water surface
elevation is also positive (the water is rising).
If a detention facility has a regular geometric shape (e.g., it is a cylindrical
tank), a geometric equation can be applied to describe the stage–storage vol-
ume relationship. In most cases, however, the engineer must first determine
a detention basin’s surface area (A) at incremental elevations or depths and
then use this information to develop stage–storage data. Two methods for
determining the stage–storage relationship—the average-end-area method
and the conic method—are presented in this section. Numerical integration
may also be used in some cases, although this technique is beyond the scope
of this text.

Average-End-Area Method
The average-end-area method uses the same concept frequently seen in
computations for estimating earthwork volumes. In this method, the incre-
mental storage volume available between two water surface elevations is
computed as:

70

68
Elevation, ft

66

64
h2
h1 62
Positive change in volume causes rise
in water surface.
60

58
0 1 2 3 4 5 6 7
ΔS
Volume, ac ft
S1 S2

Figure 5.4: Relationship of stage–storage data to pond routing.

134 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

A 1+ A 2
ΔS = (h 2− h 1 )
2

where
h1 = water surface elevation 1
h2 = water surface elevation 2 (h2 > h1)
A1 = available water surface area corresponding to stage h1
A2 = available water surface area corresponding to stage h2
ΔS = ncremental storage volume available between h1 and h2.
Computation of incremental storage amounts for a series of known stages and
surface areas, followed by summing of the incremental storages, yields the
desired relationship between h and S.
One limitation of the average-end-area method is that it applies a linear aver-
aging technique to describe area, whereas the change in area is actually a sec-
ond-order function of height for a pond with sloped sides. However, the error
associated with numerical volume integration is well within the error bounds of
other key design assumptions, such as the selected design storm and watershed
homogeneity.

Example 5.1: Computing Storage Volume Using the Average-End-Area

Method
An underground storage vault is to be constructed to serve as a detention pond
for a heavily urbanized area. The walls of the vault will be vertical, but the floor
will be sloped along its length to promote drainage. In plan view, the vault is
rectangular with a length of 75 ft and a width of 40 ft. The slope of the floor,
along the 75-ft dimension, is 2%. The invert of the lower end of the vault is at
an elevation of 123.00 ft.
Determine the elevation versus surface area relationship for the vault and use
this information to calculate and graph the elevation–storage relationship. Use
the average-end-area method.

Solution
The difference in elevation along the length of the floor of the vault is
75(0.02) = 1.50 ft

Because the floor elevation at its lower end is at an elevation (stage) of 123.00 ft,
the elevation at its high end is 124.50 ft. For any stage between these elevations,
the rectangular water surface area is equal to the width of the vault (W = 40 ft)
times the length of the water surface area. That length depends on the depth of
water in the vault and can be expressed in terms of the elevation, h, as

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 135

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

(h − 123.00)
L=
0.02

The corresponding water surface area is

40(h − 123.00)
As = WL =
0.02

These expressions apply to values of h ranging from 123.00 to 124.50 ft. For
stages above 124.50 ft, the surface area is constant because of the vertical walls
and is equal to

As = 40(75) = 3,000 ft2

The first two columns of Table 5.1 give the elevation versus surface area rela-
tionship for elevations from 123.00 to 130.00 ft.
To compute the elevation–storage relationship using the average-end-area
method, adjacent area values in Column 2 are added and then divided by the
stage increment for those areas. The steps are described next, and results are
reported in Table 5.1. The relationship is plotted in Figure 5.5.
• Column 3 = [Column 2 (current row)]+[Column 2 (previous row)]
• Column 4 = [Column 1 (current row)]−[Column 1 (previous row)]
• Column 5 = [Column 3 (current row)/2]×[Column 4 (current row)]
• Column 6 = [Column 5 (current row)]+[Column 6 (previous row)]

Table 5.1: Calculation of Detention Volume using the Average-End-Area Method

(1) (2) (3) (4) (5) (6)
Elevation Area As(h1)+As(h2) h1−h2 ΔS Cumulative S
(ft) (ft2) (ft2) (ft) (ft3) (ft3)
123.00 0 0 0.00 0.0 0.0
123.25 500 500 0.25 62.5 62.5
123.50 1,000 1,500 0.25 187.5 250.0
123.75 1,500 2,500 0.25 312.5 562.5
124.00 2,000 3,500 0.25 437.5 1,000.0
124.25 2,500 4,500 0.25 562.5 1,562.0
124.50 3,000 5,500 0.25 687.5 2,250.0
125.00 3,000 6,000 0.50 1,500 3,750.0
130.00 3,000 6,000 5.00 15,000 18,750.0

136 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

131
130
129
128

Elevation, ft
127
126
125
124
123
122
0 5,000 10,000 15,000 20,000
Storage, cu ft

Figure 5.5: Plot of elevation–storage relationship for Example 5.1.

Conic Method
This method applies a conic geometry to each stage increment to approximate
the nonlinear relationship between stage and water surface area, as illustrated
in Figure 5.6. The incremental storage volume between two stages h1 and h2
(h2>h1) is defined as

⎛ h 2− h 1 ⎞⎟
ΔS = ⎜⎜ ⎟( A + A2 + A 1 A 2 )
⎜⎝ 3 ⎟⎠ 1

Example 5.2: Computing Storage Volume Using the Conic

Method
Use the conic method to calculate and graph the elevation versus storage rela-
tionship for a pond with an elevation versus surface area relationship as given
in Table 5.2.

Solution
Table 5.3 is used to determine the cumulative volume at each elevation. Each
column represents a step in the calculations used for the conic method. A plot
of the elevation–storage relationship is given in Figure 5.7.
• Columns 1 and 2 are the elevation versus surface area relationship and are
taken from Table 5.2.
• Column 3 = {[Column 1 (current row)] − [Column 1 (previous row)]}/3.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 137

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Area A2
h2

h2–h1

Area A1
h1

Figure 5.6: Illustration of conic method for computing pond volume.

Table 5.2: Elevation-Area Data for Example 5.2
Elevation (ft) Area (ac)
100.00 0.158
100.50 0.170
101.00 0.182
101.50 0.194
102.00 0.207
102.50 0.221
103.00 0.234
103.50 0.248
104.00 0.263
104.50 0.278
105.00 0.293
105.50 0.309
106.00 0.325
106.50 0.342
107.00 0.359

• Column 4 = [(Column 2 (current row)] + [Column 2 (previous row)] +

{[Column 2 (current row)] × [Column 2 (previous row)]}0.5.
• Column 5 = (Column 3) × (Column 4).
• Column 6 = [Column 6 (previous row)] + [Column 5 (current row)].

138 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Table 5.3: Calculations for Example 5.2

(1) (2) (3) (4) (5) (6)

Elevation (ft) Area (ac) (h2−h1)/3 (ft) A1 + A 2 + A 1 A 2 (ac) ΔS (ac ft) Cumulative S (ac ft)
100.00 0.158 0.000 0.000 0.000 0.000
100.50 0.170 0.167 0.492 0.082 0.082
101.00 0.182 0.167 0.528 0.088 0.170
101.50 0.194 0.167 0.564 0.094 0.264
102.00 0.207 0.167 0.602 0.100 0.364
102.50 0.221 0.167 0.641 0.107 0.471
103.00 0.234 0.167 0.682 0.114 0.585
103.50 0.248 0.167 0.723 0.121 0.705
104.00 0.263 0.167 0.766 0.128 0.833
104.50 0.278 0.167 0.811 0.135 0.968
105.00 0.293 0.167 0.856 0.143 1.111
105.50 0.309 0.167 0.903 0.150 1.262
106.00 0.325 0.167 0.951 0.158 1.420
106.50 0.342 0.167 1.000 0.167 1.587
107.00 0.359 0.167 1.051 0.175 1.762

108
107
106
105
Elevation ,ft

104
103
102
101
100
99
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Storage, ac ft

Figure 5.7: Stage–storage relationship for Example 5.2.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 139

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Stage versus Discharge

The discharge Q from a detention facility depends on the stage h of the pond,
on the geometric and hydraulic characteristics of the outlet(s), and possibly
on tailwater effects that may influence the outlet hydraulics. In cases where
interconnected ponds are to be routed, the discharge from one pond to another
depends on the stages in the ponds, and on the hydraulic characteristics of the
channel or conduit connecting them.
The hydraulic characteristics for a pond outlet structure are described by a
stage versus discharge rating table, which is applied during routing to deter-
mine the relative increase or decrease in outflow, given a change in the stage
(water surface elevation). Figure 5.8 demonstrates how elevation versus dis-
charge data is related to routing. The rating curve is calculated by computing
the flow for various stage elevations. Precision of the curve increases as the
incremental elevation between points decreases. The designer must judge the
appropriate degree of flow approximation when setting the elevation incre-
ment size for the rating curve.
In Figure 5.8, h1 and Q1 are the elevation and outflow at the beginning of the
time step, respectively. Elevation h2 at the end of the time step is determined
from the stage versus storage curve based on the change in volume. From h2,
Q2 can be found using the stage versus discharge curve. Thus for this example
the net change in elevation is positive (rising), so the change in outflow is
positive (increasing).
Most dry detention facilities have both low- and high-level outlets. The low-
level outlet may consist of a pipe or culvert, that is, culvert one or more orifices

70

68

66
Elevation, ft

64
h2
h1 62
Rise in water surface
increases outflow
60

58
0 20 40 60 80 100 120 140 160 180 200

Q1 Q2 Discharge, cfs

Figure 5.8: Relationship of elevation–discharge data to detention routing.

140 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

or a weir positioned at or near one or more orifices, or a weir. The high-level

outlet is usually an overflow spillway and is expected to function only during
extreme runoff events with recurrence intervals that exceed the pond’s design
constraints. Where ponds are designed to control runoff rates for several dif-
ferent recurrence intervals, intermediate outlets are usually also provided. They
are similar to low-level outlets in practice and design but are installed at higher
elevations.
The stage versus discharge relationship for a pond is developed using hydraulic
relationships depending on the number and types of individual outlets. The
relationships for various types of pond outlets are discussed next.

Pipe or Culvert Outlets

The stage versus discharge relationship for a pipe or culvert outlet can be devel-
oped using culvert design procedures described in Chapter 4. The outlet may
perform under either inlet or outlet control, depending on downstream effects.
A culvert performance curve delineating discharge as a function of the headwa-
ter depth (or stage) is the desired stage versus discharge relationship.
For small diameter pipes (D<12 in.), culvert hydraulics can be adequately rep-
resented using the orifice equation (discussed in Section 1.3). This assumption
holds true if the ratio of headwater depth to pipe diameter is at least 1.5, and if
tailwater (outlet control) effects are negligible (Brown et al., 2009).

Orifices and Weirs

The hydraulic properties of orifices and weirs are discussed in Section 1.3.

Overflow Spillways
An overflow spillway has the purpose of safely conveying discharges when the
design storage capacity of a detention pond is exceeded during large storm
events. Overflow spillways are commonly referred to as emergency spillways
because they are designed to operate under extreme rainfall conditions to pre-
vent dam overtopping.
The storm recurrence interval for which an overflow spillway should be
designed is an issue that requires exercise of professional judgment, as well
as consideration of the adverse effects of downstream flooding or damages.
In areas where the major drainage system is evaluated for a designated storm
recurrence interval, the same recurrence interval might be used for overflow
spillway design. On the other hand, if failure of a dam embankment could
cause considerable damages or loss of life downstream, regulatory requirements
pertaining to dams may be appropriate.
Detailed coverage of spillway hydraulics is beyond the scope of this book. For
more advanced discussion on spillway and outlet design, refer to Stormwater
Conveyance Modeling and Design (Haestad Methods and R. Durrans, 2003).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 141

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Composite Stage versus Discharge Relationships

The overall, or composite, stage versus discharge relationship for a detention
pond represents the total discharge from a pond, possibly occurring through
several outlets, as a function of the water surface elevation (or depth). Assembly
of a composite relationship involves computation and summing of the dis-
charges through each outlet for each stage of interest. Example 5.3 demon-
strates the development of the stage–discharge relationship.

Structures in Parallel
The most basic type of composite outlet structure is one that has different
outlet devices working independently of one another. In this situation, the
hydraulic rating characteristics for each device are simply summed at each ele-
vation to compute the total outflow. Figure 5.9 displays a composite set of
outlet structures working in parallel, and Figure 5.10 shows the rating curve
resulting from this outlet configuration.

Example 5.3: Stage versus Discharge for a Composite Structure

A simple composite outlet structure is being designed with an orifice and a weir
operating in parallel. The orifice is a 6-in. diameter orifice plate with an invert
elevation of 100.00 ft and an orifice coefficient of 0.60. The orifice discharges
into the top section of a large culvert; under maximum head conditions in the
pond, the water within the culvert will not rise above the orifice invert (i.e.,
the orifice controls throughout). The concrete rectangular suppressed weir has
a weir coefficient approximated as 2.6 for the entire range of headwaters, a weir
length of 15 ft, and a crest elevation at 105.00 ft. It operates under free outfall
conditions.

A
Top of dam
Weir spillway

Culvert

A
Front view Section A–A

Figure 5.9: Composite outlet with outlet structures discharging in parallel.

142 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

10
way
spill
rt +
C ulve
Spillway crest

Pond depth, ft
5
rt
lve
Cu

0
0 50 100 150 200
Discharge, cfs

Figure 5.10: Depth versus discharge for a composite outlet with outlet
structures discharging in parallel.

Calculate and graph the elevation versus discharge rating curve from elevations
100.00 to 107.00 ft for each 0.50-ft increment. (Note that a smaller increment
would yield a smoother, more precise rating curve.)

Solution
Table 5.4 shows the stage–storage calculation results described by the following
steps. Figure 5.11 is a plot of the composite structure rating curve.
• Orifice centroid elevation = 100.00 ft + (1/2 × 0.5 ft) = 100.25 ft.
• Orifice area = πD2/4 = 0.196 ft2.
• Column 1 is the range of rating table elevations (100.0 –107.0 ft, in
0.5-ft increments).
• Column 2 is the head on the orifice = (Column 1)−(orifice centroid
elevation).
• Column 3 is the orifice discharge = solution to orifice equation for C =
0.60, A = 0.196, h = Column 2 (the orifice equation is Q = CA(2gH )0.5,
as introduced in Chapter 1.).
• Column 4 is the head on the weir = (Column 1)−(weir crest elevation).
• Column 5 is the weir discharge = solution to rectangular weir equation for
C = 2.6, L = 15.0, h = Column 4 (the weir equation is Q = CLH 3/2, as
introduced in Chapter 1.).
• Column 6 is the composite structure discharge = (Column 3) + (Column 5).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 143

 5 Detention Pond Design
Routing Data: Storage and Hydraulic Relationships

Table 5.4: Stage–Discharge Calculations for Example 5.3

(1) (2) (3) (4) (5) (6)
Elevation Centroid = 100.25 ft Crest = 105.00 ft
Total Flow (cfs)
(ft) Head (ft) Flow (cfs) Head (ft) Flow (cfs)
100.00 0.00 0.00 0.00 0.00 0.00
100.50 0.25 0.47 0.00 0.00 0.47
101.00 0.75 0.82 0.00 0.00 0.82
101.50 1.25 1.06 0.00 0.00 1.06
102.00 1.75 1.25 0.00 0.00 1.25
102.50 2.25 1.42 0.00 0.00 1.42
103.00 2.75 1.57 0.00 0.00 1.57
103.50 3.25 1.70 0.00 0.00 1.70
104.00 3.75 1.83 0.00 0.00 1.83
104.50 4.25 1.95 0.00 0.00 1.95
105.00 4.75 2.06 0.00 0.00 2.06
105.50 5.25 2.16 0.50 13.79 15.95
106.00 5.75 2.26 1.00 39.00 41.27
106.50 6.25 2.36 1.50 71.65 74.01
107.00 6.75 2.45 2.00 110.31 112.76

108
107
106
Elevation, ft

105
104
103
102
101
100
99
0 20 40 60 80 100 120
Flow, cfs
Figure 5.11: Rating curve for outlet structure in Example 5.3.

144 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Storage Indication Method

Top of dam
A
Weir spillway

V-notch weir Inlet box

Culvert

Culvert

Front view Section A–A

Figure 5.12: Composite outlet with outlet structures discharging in

parallel and in series.

Structures in Series
When outlet structures are in series, it is possible that backwater effects caused
by hydraulic controls at downstream structures will affect the discharges
through connecting upstream structures. In these cases, hydraulic grade line
calculations are required to determine the interactions between individual
structure components.
Figure 5.12 illustrates a pond outlet with structures in series. In this example,
as the stage rises, the culvert eventually throttles (controls) the combined flow
through the v-notch weir and the inlet box, causing the inlet box riser to be
fully submerged. This phenomenon is shown graphically near the center por-
tion of the stage versus discharge relationship in Figure 5.13.
This text provides only an introduction to basic outlet hydraulics. For more
detailed explanations on analyzing complex outlet structures such as outlets in
series, refer to Stormwater Conveyance Modeling and Design (Haestad Methods
and R. Durrans, 2003).

5.6 Storage Indication Method

Routing of a detention pond is the process by which a given inflow hydrograph
to the pond is transformed into the corresponding outflow hydrograph. In
addition to the inflow hydrograph, the outflow hydrograph also depends on
the storage and hydraulic relationships of the pond (described in the previous

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 145

 5 Detention Pond Design
Storage Indication Method

10
way
spill
ert +
Culv
Spillway crest

Pond depth, ft otch

+ V-n
5 Inlet Inlet
crest
ch t

rt
V-no

lve
Cu
0
0 50 100 150 200
Flow, cfs

Figure 5.13: Rating curve for a composite outlet with outlet structures
discharging in parallel and in series.

sections) and on the initial water surface elevation in the pond when the inflow
hydrograph begins.
Fundamentally, routing of a pond amounts to a solution of the conservation
of mass equation:

dS
= I (t ) − O(t )
dt

where
S = the volume of water in the pond (ft3, m3);
I(t) = specified inflow at time t (cfs, m3/s); and
O(t) = the outflow at time t (cfs, m3/s).
This equation states that the change in storage during a time period is equal to
the difference between the inflow and the outflow. There are two unknowns in
this equation (S and O); thus information on both storage characteristics and
hydraulic characteristics is required to solve it. Various routing methods are
available, but the most commonly used method in simple stormwater deten-
tion is the storage indication method, which is based on a finite difference
approximation of the conservation of mass equation.
Figure 5.14 shows the inflow and outflow hydrographs for a pond. The graph
magnifies a time slice (step) for which the inflow is greater than the outflow. The
x-axis of the slice has units of time (t), and the y-axis of the slice, which represents

146 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Storage Indication Method

Average inflow – average outflow

Δ Volume

Inflow Storage

Pond crests
Flow

Outflow

Δ Time Time, h

Figure 5.14: Change in pond volume for a routing time step.

flow rate, has units of volume (L3) per time, or L3/t. Thus the bounded area of
the slice has units of T × (L3/t) = L3, or volume. Because the inflow is greater
than the outflow for this time slice, the change in volume is positive, which
means the pond water surface (stage) is rising during this time step.
At the point where the inflow equals the outflow, the change in storage is zero,
the water surface elevation (stage) is constant, and the pond is cresting. After the
pond has crested, the outflow rate is greater than the inflow rate, resulting in a net
loss in volume that causes the water surface (stage) to fall during the time step.
Because the shaded area of Figure 5.14 represents the bounded area for which
inflow is greater than outflow before the pond crests, this area represents the
total volume stored while the pond is filling. As stated previously, the shaded
area represents the maximum storage volume.
Figure 5.15 shows a portion of the inflow and outflow hydrographs for a pond.
The graph illustrates a time period in which the inflow rate is greater than
the outflow rate, but the relationships that follow are applicable for any time
period, regardless of the relative magnitudes of the inflow and outflow rates.
Over a time interval Δt = tn+1 − tn, the inflow rate varies from In to In+1, and

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 147

 5 Detention Pond Design
Storage Indication Method

In+1 Inflow

In

Discharge, cfs

Outflow
On+1

On

Dt

tn Time tn+1

Figure 5.15: Definition sketch for terms in finite difference approximation.

therefore the average inflow rate over the time interval can be approximated
as (In + In + 1)/2. Over the same time interval, the average outflow rate can be
approximated as (On + On + 1)/2. The change in storage over the time interval
is ΔS = Sn + 1 − Sn, whereas the time rate of change of storage may be approx-
imated as ΔS/Δt = (Sn + 1 − Sn)/Δt. Substitution of the average inflow and
outflow rates and the approximate rate of change of storage into the conserva-
tion of mass equation yields

Sn +1 − Sn I + In +1 On + On +1
= n −
Δt 2 2

Because the inflow hydrograph is a known quantity in detention pond rout-

ing, the terms In and In+1 are known for every time interval in the routing
process. For the first time interval, at the beginning of a pond inflow event,
the terms On and Sn are also initially known. Note that On and Sn are related
as for each pond stage hn there is a corresponding Sn and a corresponding On.
Rearranging the finite difference approximation with unknown quantities
on the left and known quantities on the right yields the storage indication
equation:

⎛ 2Sn +1 ⎞⎟ ⎛ 2Sn ⎞⎟
⎜⎜ ⎜
⎜⎝ Δt + On +1 ⎟⎟⎠ = ⎜⎜⎝ Δt − On ⎟⎟⎠ + ( In + In +1 )

148 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Stormwater Detention Analysis Procedure

Thus knowledge of the quantities Sn, On, and In at the beginning of a time
interval (at time tn), as well as knowledge of the inflow rate In + 1 at the end
of the time interval, can be used to compute the value of the left side of the
equation. However, there are two unknowns on the left side (On + 1 and Sn + 1).
Additional information must be established to expose the relationship between
On + 1 and Sn + 1 so that these two values can be determined.
A storage indication curve is a graph delineating the relationship between O
and (2S/Δt + O). Such a graph can be constructed by realizing that for any
stage h in a pond, there are corresponding values of O and S; hence, there are
corresponding values of O and (2S/Δt + O).
In performing storage indication calculations for a specified time interval n+1,
the storage indication equation is applied to compute the numerical value of
the quantity (2Sn + 1/Δt + On + 1). Next, the value of (2Sn + 1/Δt − On + 1)
for the time interval n + 1 must be found. Because the value of (2Sn + 1/Δt +
On + 1) is known, and this value can be used with the storage indication curve
to find On+1, (2Sn + 1/Δt − On + 1) can be computed as

⎛ 2Sn +1 ⎞⎟ ⎛ 2Sn +1 ⎞⎟
⎜⎜ ⎜
⎜⎝ Δt − On +1 ⎟⎟⎠ = ⎜⎜⎝ Δt + On +1 ⎟⎟⎠ − 2On +1

The term on the left side of the equation can now be used in the next time
step (n + 2) to compute (2Sn + 2/Δt + On + 2), and the process repeats through
every time step.
Section 5.7 describes the full procedure used in storage indication routing and
presents an example problem.

5.7 Stormwater Detention Analysis Procedure

This chapter introduced each of the basic components required for modeling
how a pond functions during a rainfall event. In order to apply pond routing
techniques to model a pond, the various components of hydrographs, storage
volume, and outlet structures are used in the following steps:
1. Calculate the postdeveloped inflow hydrograph discharging into the
pond. Methods for calculating hydrographs are covered in Chapter 2.
2. Calculate the storage volume rating curve (stage versus storage) for the
given pond design. The rating curve should start at the dry invert of the
pond and extend to the top of the dam.
3. Calculate the outflow rating curve (stage versus outflow) for the given
pond outlet structure. The rating curve should match the range of eleva-
tions of the physical pond design, that is, match the storage rating curve
range of elevations. This holds true even if lower elevations will have zero
discharge within a wet pond.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 149

 5 Detention Pond Design
Stormwater Detention Analysis Procedure

4. Establish the relationship between stage, storage, and discharge by creat-

ing a storage indication curve for (2S/Δt+O) versus outflow.
5. Use the hydrograph from Step 1 and the storage indication curve from
Step 4 to route the inflow hydrograph through the storage and discharge
relationship to obtain a pond outflow hydrograph.

Example 5.4: Detention Pond Routing Example

Route the inflow hydrograph given in Table 5.5 through the pond described by
the stage versus storage data computed in Example 5.2 and the stage versus dis-
charge data computed in Example 5.3. Determine the maximum water surface
elevation for the detention basin, the maximum storage volume, maximum
outflow rate, and peak lag resulting from detention.

Table 5.5: Inflow Hydrograph for Example 5.4

Time (min) Inflow (cfs)
0 0.00
10 0.58
20 5.94
30 15.44
40 20.75
50 20.27
60 17.44
70 14.19
80 11.42
90 9.44
100 8.12
110 7.27
120 6.61
130 5.49
140 3.32
150 1.52
160 0.66
170 0.29
180 0.12
190 0.05
200 0.02
210 0.00

150 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Stormwater Detention Analysis Procedure

Solution
The results from Examples 5.2 and 5.3 are combined in Table 5.6. Note the
importance of calculating the same elevations for storage and discharge in
order to directly correlate flow and storage at each elevation.
Follow the five steps outlined for storage indication pond routing:
1. Compute the postdeveloped pond inflow hydrograph using the proce-
dures presented in Chapter 2 or another method. In this example, the
final inflow hydrograph is given.
2. Calculate the stage versus storage curve from the pond grading plan or
storage structure geometry. This information was calculated in Example
5.2 (no calculations are necessary for this example).
3. Calculate the stage versus outflow curve for the pond outlet structure.
This information was calculated in Example 5.3.
4. Calculate and graph the storage indication curve that correlates 2S/Δt + O
and outflow. To calculate this curve, the rating elevations for the stage
versus storage curve must be identical to those used for the stage versus
outflow curve. Use a 10-min time step in this example.
Step 4 results for 2S/Δt + O are given in Table 5.7. Values given in
Column 4 were computed using

Table 5.6: Elevation–Volume–Flow Data for Example 5.4

Elevation (ft) Total Volume (ac-ft) Total Flow (cfs)
100.00 0.000 0.00
100.50 0.082 0.47
101.00 0.170 0.82
101.50 0.264 1.06
102.00 0.364 1.25
102.50 0.471 1.42
103.00 0.585 1.57
103.50 0.705 1.70
104.00 0.833 1.83
104.50 0.968 1.95
105.00 1.111 2.06
105.50 1.262 15.95
106.00 1.420 41.27
106.50 1.587 74.01
107.00 1.762 112.76

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 151

 5 Detention Pond Design
Stormwater Detention Analysis Procedure

Column 4 = {2.0 × [(Column 2) × 43,560 ft2/ac]/(10 min × 60 s/min)}

+ (Column 3)
Because the maximum inflow is about 20 cfs, we already know that the max-
imum outflow value should be equal to or less than 20 cfs. The storage indi-
cation curve shown in Figure 5.16 is an expanded view for the outflow range
below 20 cfs, which makes it easier to read the smaller flow values during the
routing tabulation.
5. Use the inflow hydrograph from Step 1 and the storage indication
curve from Step 4 to route the hydrograph using a systematic worksheet
approach as shown in Table 5.8.
• Columns 1 and 2 represent the given inflow hydrograph.
• Under initial conditions, Columns 3–6 have values of 0. Because the
pond is initially empty, the initial elevation (Column 6) is 100.00 ft.
• For subsequent time steps,
• [Column 3 (current row)] = [Column 2 (previous row)] + [Column
2 (current row)]

Table 5.7: Relationship between Elevation and 2S/Δt + O for Example 5.4

(1) (2) (3) (4)

Elevation (ft) Volume (ac ft) Flow (cfs) 2S/Δt + O (cfs)
100.00 0.000 0.00 0.00
100.50 0.082 0.47 12.38
101.00 0.170 0.82 25.50
101.50 0.264 1.06 39.39
102.00 0.364 1.25 54.10
102.50 0.471 1.42 69.81
103.00 0.585 1.57 86.51
103.50 0.705 1.70 104.07
104.00 0.833 1.83 122.78
104.50 0.968 1.95 142.50
105.00 1.111 2.06 163.38
105.50 1.262 15.95 199.19
106.00 1.420 41.27 247.45
106.50 1.587 74.01 304.44
107.00 1.762 112.76 368.60

152 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Stormwater Detention Analysis Procedure

400
350
300

(2S/t) + O
250
200
150
100
50
0
0 5 10 15 20
Flow, cfs

Figure 5.16: Storage indication curve for Example 5.4.

• [Column 5 (current row)] = [Column 3 (current row)] + [Column 4

(previous row)].
• Determine values in Column 6 by using the values in Column 5 and solv-
ing for the associated outflow from the storage indication curve calculated
in Step 4.
• Solve Column 7 (current row) using stage versus outflow curve created in
Step 3.
• Finally, to complete the current row, solve
• [Column 4 (current row)] = [Column 5 (current row)] − 2 × [Column
6 (current row)].
• The tabulated results shown stop at t = 300 min into the routing. To com-
plete the routing, the calculations need to be continued until the pond is
fully drained (the routed outflow in Column 6 recedes to zero). The graph
for the inflow and outflow hydrographs displays the entire routing time.
From the routing table, one can find:
• Maximum stage = 105.30 ft, which is above spillway crest.
• Maximum storage = 1.2 ac ft (determined from the stage versus storage
data for an elevation of 105.30 ft).
• Maximum flow = 10.38 cfs.
• Outflow peak time = 90 min.
• Peak flow was delayed 50 min from inflow peak time of 40 min.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 153

 5 Detention Pond Design
Stormwater Detention Analysis Procedure

Table 5.8: Routing Results from Example 5.4

(1) (2) (3) (4) (5) (6) (7)

Time (min) Inflow (cfs) I1 + I2 (cfs) 2S/Δt − O (cfs) 2S/Δt + O (cfs) Outflow (cfs) Elevation (ft)
0 0 0 0.00 0.00 0.00 100.00
10 0.58 0.58 0.54 0.58 0.02 100.02
20 5.94 6.52 6.52 7.06 0.27 100.28
30 15.44 21.38 26.18 27.90 0.86 101.09
40 20.75 36.19 59.69 62.37 1.34 102.26
50 20.27 41.02 97.36 100.71 1.68 103.40
60 17.44 37.71 131.26 135.07 1.90 104.31
70 14.19 31.63 158.77 162.89 2.06 104.99
80 11.42 25.61 163.97 184.38 10.21 105.29
90 9.44 20.86 164.07 184.83 10.38 105.30
100 8.12 17.56 163.35 181.63 9.14 105.25
110 7.27 15.39 162.70 178.74 8.02 105.21
120 6.61 13.88 162.22 176.58 7.18 105.18
130 5.49 12.1 161.71 174.32 6.30 105.15
140 3.32 8.81 160.86 170.52 4.83 105.10
150 1.52 4.84 159.78 165.70 2.96 105.03
160 0.66 2.18 157.86 161.96 2.05 104.97
170 0.29 0.95 154.73 158.81 2.04 104.89
180 0.12 0.41 151.11 155.14 2.02 104.80
190 0.05 0.17 147.29 151.28 2.00 104.71
200 0.02 0.07 143.41 147.36 1.98 104.62
210 0 0.02 139.52 143.43 1.95 104.52
220 0 0 135.65 139.52 1.93 104.42
230 0 0 131.84 135.65 1.91 104.33
240 0 0 128.07 131.84 1.89 104.23
250 0 0 124.34 128.07 1.86 104.13
260 0 0 120.66 124.34 1.84 104.04
270 0 0 117.03 120.66 1.82 103.94
280 0 0 113.45 117.03 1.79 103.85
290 0 0 109.92 113.45 1.77 103.75
300 0 0 106.44 109.92 1.74 103.66

154 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
PondPack and Tutorials

25

20
Inflow

Flow, cfs
15

10
Outflow
5

0
0 50 100 150 200 250 300 350
Time, min

Figure 5.17: Pond inflow and outflow hydrographs.

5.8 PondPack and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software.

What Is PondPack?
PondPack is a stormwater modeling program capable of analyzing a wide
range of situations, from simple sites to complex networked watersheds. The
program analyzes pre- and postdeveloped watershed conditions and sizes
ponds.
PondPack also computes outlet structure rating curves with tailwater effects,
pond infiltration, pond detention times, and channel routing effects. It
includes the capability to model interconnected ponds with divergent (mul-
tiple) outfalls.
The program can use any rainfall duration or distribution to compute hydro-
graphs for multiple events. Hydrographs can be added at junctions and routed
through multiple reaches and ponds. PondPack helps automate almost every
aspect of drainage design computations.
PondPack graphically displays such items as watershed diagrams, hydrographs,
rainfall curves, I–D–F curves, outlet rating curves, volume curves, time ver-
sus elevation curves, time versus volume curves, cross sections, channel rating
curves, and a wide variety of other output diagrams.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 155

 5 Detention Pond Design
Problems

PondPack builds customized reports organized by categories and automatically

creates section and page numbers, tables of contents, and indexes. You can dis-
play a short results summary for an entire watershed or build a comprehensive
drainage report showing any or all report items.

How Can You Use PondPack for Windows?

You can apply this software to all phases of a stormwater detention project, from
feasibility studies to analysis of the final design. This product can be used to
• create full drainage reports with table of contents and index,
• compute pre- and postdeveloped flows using popular methods such as
Technical Release-55 or Rational Method,
• estimate storage requirements,
• size ponds,
• model multistage outlets with tailwater considerations,
• analyze culverts with inlet/outlet control checking,
• model simple pond sites,
• solve complex watershed networks with interconnected ponds,
• model unlimited diversions,
• review tidal outfall effects on pond outlets,
• analyze channel capacity,
• check outflow rates against predeveloped conditions,
• compute water quality parameters such as minimum drain time and
detention time,
• model any rainfall duration or time distribution, and
• simulate gauged rainfall events and synthetic rainfall distribution events.

5.9 Problems
The following problems can be solved with the PondPack computer program.
1. A detention pond has the shape of a trapezoidal basin. The specifications are:

Top elevation 120 ft

Top length 400 ft
Top width 300 ft
Bottom elevation 100 ft
Bottom length 280 ft
Bottom width 180 ft

156 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Problems

Construct an elevation versus volume graph for the basin. What is the volume
at an elevation of 120 ft?
2. Given the elevation-area data for a detention pond, construct an elevation
versus volume graph for the pond. What is the volume of the pond at an
elevation of 205 m?

Elevation—Volume Data for Problem 2

Elevation (m) Area (m2)
200.0 1,700
200.5 2,500
201.0 2,900
201.5 3,700
202.0 4,300
202.5 4,800
203.0 5,500
203.5 6,900
204.0 8,100
204.5 12,300
205.0 15,000

3. An outlet structure for a pond consists of a riser pipe with two orifices.
The riser is a 3-ft diameter pipe with a rim elevation of 105 ft. A 1-ft
diameter orifice has an invert of 101 ft and a 0.5-ft orifice has an invert
of 100 ft. Use 3.33 for the weir coefficient and 0.614 for the orifice coef-
ficient. Construct an elevation versus flow diagram for a headwater range
of 100–120 ft, with an increment of 1 ft. What is the discharge at a head-
water elevation of 120 ft?
4. A composite outlet has a 4-m broad crested weir at an elevation of
203.5 m and a 50-cm diameter orifice at an elevation of 200 m. The weir
coefficient is 1.84 and the orifice coefficient is 0.614. Construct an ele-
vation versus discharge curve for the structure over a headwater range
of 200–205 m with a step of 1 m. What is the discharge at a headwater
elevation of 205 m?
5. Use the detention pond described in Problem 5.2 and the outlet struc-
ture described in Problem 5.4 to route the following hydrograph. What
is the peak discharge from the pond? What is the maximum elevation in
the pond?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 157

 5 Detention Pond Design
Problems

Inflow Hydrograph Data for Problem 5

Time (h) Flow (m3/s) Time (h) Flow (m3/s)
0.0 0.00 3.6 3.12
0.2 0.24 3.8 2.64
0.4 0.80 4.0 2.24
0.6 1.52 4.2 1.84
0.8 2.48 4.4 1.66
1.0 3.76 4.6 1.44
1.2 5.28 4.8 1.18
1.4 6.56 5.0 0.90
1.6 7.44 5.2 0.86
1.8 7.92 5.4 0.70
2.0 8.00 5.6 0.62
2.2 7.92 5.8 0.53
2.4 7.44 6.0 0.44
2.6 6.88 6.2 0.26
2.8 6.24 6.4 0.14
3.0 5.44 6.6 0.08
3.2 4.48 6.8 0.00
3.4 3.68 7.0 0.00

6. A proposed development will consist of single-family homes and a small

park on a 25-ac site. Local regulations require that on-site detention be
used to maintain peak runoff rates at predevelopment levels for a 25-year
storm. Design a detention facility to meet this requirement.
Use the SCS 24-h Type II storm with a total depth of 6.5 in. and the SCS
Unit Hydrograph Method to compute the runoff hydrograph. The pond is
to be a trapezoidal basin with a total depth of 5 ft and 3H:1V side slopes. For
the outlet structure, use a 1-ft diameter orifice with an invert at the bottom of
the basin together with a rectangular weir. Size and place the weir such that the
predevelopment peak flow is not exceeded and a minimum of 1 ft of freeboard
is maintained. Document your design by filling in the table.

158 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Detention Pond Design
Problems

Land Cover Data for Problem 6

Stage Land Cover Area (ac) CN tc (min)
Wooded 15 58
Predevelopment 45
Meadow 10 65
0.5-ac residential lots 21 75
Postdevelopment Roads 0.3 98 30
Park 3.7 79

Proposed Design for Problem 6

Predevelopment peak runoff flow rate/allowable outflow rate
Estimated storage volume required
Dimensions of bottom of pond
Dimensions of top of pond
Computed pond volume below freeboard elevation
Weir length
Height of weir above bottom of pond
Maximum water surface elevation
Peak pond outflow rate

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 159

 6 Pressure Piping Systems
and Water Quality Analysis

6.1 Pressure Systems

Pressure piping network analysis has many applications, including well pump-
ing systems and heating and cooling systems. This chapter deals primarily with
the topic of pressure piping as it relates to potable water distribution systems.
The main purpose of a water distribution system is to meet demands for pota-
ble water. People use water for drinking, cleaning, gardening, and any number
of other uses, and this water needs to be delivered in some fashion. A secondary
purpose of many distribution systems is to provide water for fire protection.
If designed correctly, the network of interconnected pipes, storage tanks,
pumps, and regulating valves provides adequate pressure, adequate supply, and
good water quality throughout the system. If incorrectly designed, some areas
may have low pressures, poor fire protection, and even present health risks.

Water Demands
Just as storm sewer analysis is driven by the watershed runoff flow rate, water
distribution system analysis is driven by customer demand. Water usage rates
and patterns vary greatly from system to system and are highly dependent on
climate, culture, and local industry. Every system is different, so the best source
of information for estimating demands is directly recorded system data.

Metered Demand
Metered demands are often a modeler’s best tool and can be used to calculate
average demands, minimum demands, peak demands, and so forth. This data
can also be compiled into daily, weekly, monthly, and annual reports that show
how the demands are influenced by weather, special events, and other factors.
Unfortunately, many systems still do not have complete system metering. For
these systems, the modeler is often forced to use other estimation tools (including
good engineering judgment) to obtain realistic demands.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 161

 6 Pressure Piping Systems and Water Quality Analysis
Pressure Systems

Demand Patterns
A pattern is a function relating water use to time of day. Patterns allow the
user to apply automatic time-variable changes within the system. Different
categories of users, such as residential or industrial customers, will typically be
assigned different patterns to accurately reflect their particular demand varia-
tions. A diurnal curve is a type of pattern that describes changes in demand
over the course of a daily cycle, reflecting times when people are using more or
less water than average. Most patterns are based on a multiplication factor ver-
sus time relationship, whereby a multiplication factor of 1.0 represents the base
value (often the average value). In equation form, this relationship is written as:

Q t = A t× Qbase

where
Qt = demand at time t,
At = multiplier for time t, and
Qbase = baseline demand.
Using a representative diurnal curve for a residence (Figure 6.1), we see that
there is a peak in the diurnal curve in the morning as people take showers and
prepare breakfast, another slight peak around noon, and a third peak in the
evening as people arrive home from work and prepare dinner. Throughout the
night, the pattern reflects the relative inactivity of the system, with very low
flows compared to the average. (Note that this curve is conceptual and should
not be construed as representative of any particular network.)
There are two basic forms for representing a pattern: stepwise and continuous.
A stepwise pattern is one that assumes a constant level of usage over a period
of time and then jumps instantaneously to another level where it again remains

2.0
Continuous
Stepwise
Multiplication factor

1.5

Average
1.0

0.5

0.0
0 6 12 18 24
Time of day

Figure 6.1: Typical Diurnal Curve.

162 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Pressure Systems

steady until the next jump. A continuous pattern is one for which several
points in the pattern are known and sections in between are transitional, result-
ing in a smoother pattern. Notice that, for the continuous pattern in Figure 6.1,
the magnitude and slope of the pattern at the start and end times are the
same—a continuity that is recommended for patterns that repeat.
Because of the finite time steps used in the calculations, most computer pro-
grams convert continuous patterns into stepwise patterns for use by the algo-
rithms, with the duration of each step equal to the time step of the analysis.

Nonrevenue Water and Leaks

Not all the water that is produced by a treatment facility is delivered to the
users. A portion of that water does not make it to the customers due to leaks or
is used as unmetered flow. This nonrevenue water flow includes (1) what is lost
in the system due to leaks and breaks in pipes and joints, (2) unmetered uses
such as fire protection and maintenance and meter inaccuracy, (3) water theft,
and (4) a variety of other minor water losses. Generally in all cases, more water
will be produced and enter the piping network than is delivered to consumers.
Many times, the unaccounted-for water is used to gauge the performance of a
pipe network. On an annual basis, water suppliers try to maintain nonrevenue
water to less than 10% of the total water production. However, for older sys-
tems, the nonrevenue water can be much higher due to the aging pipe network,
which can have a significant amount of leakage and other losses.
A hydraulic model can be used to evaluate leakage in a water distribution sys-
tem. Pipe leakage is a pressure-dependent water loss (demand) that can be dis-
tributed throughout the pipe network, placed with a few large leaks at specific
locations, or a combination of the two. Without complete knowledge of the
type, size, and location of all leaks, some assumptions will need to be made
when using a hydraulic model to assess and analyze leaks.
The first step is to estimate the amount of leakage. This is done by subtracting
the metered flow for consumers from the total flow entering a water system.
To estimate the total flow for the entire water system, the metered production
water from a treatment plant can be used, or for a subsection of the water
system, the metered flow at points where there is a measured flow such as a
pumping station can be used. The difference between the total flow and the
metered flow to consumers is the amount of unmetered flow. A portion of the
unmetered flow would be due to leaks. If field tests are conducted during times
when there is no unmetered flow for fire protection or maintenance, such as
late at night without a fire event, it is likely that much of the calculated unme-
tered flow would be due to leaks.
Q leaks = Q total − Q meatered − Q fire and maintenance
The calculated flow rate due to leaks is then distributed throughout the water
system as many small leaks and/or a few large leaks at system nodes. Various
methods are used to distribute small leaks throughout the pipe network.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 163

 6 Pressure Piping Systems and Water Quality Analysis
Energy Losses

Conventional method—The simplest method to distribute the leakage flow

would be to divide the calculated leakage flow among the nodes in the water
system as an added demand. The baseline leakage rate is typically set to be
proportional to the metered demand. This method does not take into account
the pressure-dependent flow rate of the leak.
Uniform emitter coefficient—If the actual locations of the leaks are unknown,
a uniform emitter coefficient (K) can be used for each node to distribute the
leakage flow. The emitter coefficient is estimated so that the total leakage flow
rate is distributed based on the pressure head at each node. This ensures that
the leakage flow rate is pressure-dependent.

Qleaks = K ∑ i =1 Pin
N

where
N = total number of demand nodes,
Pi = Pressure head at node i, and
n = exponent.
Proportional methods—There are other various methods that use an emit-
ter coefficient, the node pressure head, and proportional values for the pipe
length, and metered consumption. These methods use weighted factors for
pipe length, and metered consumption in addition to the pressure-dependent
leakage flow rate at each node.

6.2 Energy Losses

Friction Losses
The hydraulic theory behind friction losses is the same for pressure piping as it
is for open-channel hydraulics. The most commonly used methods for deter-
mining head losses in pressure piping systems are the Hazen–Williams equa-
tion and the Darcy–Weisbach equation, both discussed in Chapter 1. Many of
the general friction loss equations can be simplified and revised because of the
following assumptions that can be made for a pressure pipe system:
• Pressure piping is almost always circular, so the flow area, wetted perimeter,
and hydraulic radius can be directly related to diameter.
• Pressure systems flow full (by definition) throughout the length of a given
pipe, so the friction slope is constant for a given flow rate. This means that
the energy grade and hydraulic grade drop linearly in the direction of flow.
• Because the flow rate and cross-sectional area are constant, the velocity must
also be constant. By definition, then, the energy grade line and hydraulic
grade line (HGL) are parallel, separated by the constant velocity head.
These simplifications allow for pressure pipe networks to be analyzed much
more quickly than systems of open channels or partially full gravity piping.

164 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Energy Losses

Several hydraulic components that are unique to pressure piping systems, such
as regulating valves and pumps, add complexity to the analysis.

Minor Losses
Localized areas of increased turbulence cause energy losses within a pipe, creat-
ing a drop in the energy and hydraulic grades at that point in the system. These
disruptions are often caused by valves, meters, or fittings (such as the pipe
entrance in Figure 6.2), and are generally called minor losses. These minor
losses are often negligible relative to friction losses and may be ignored during
analysis.
Although the term “minor” is a reasonable generalization for most large-scale
water distribution models, these losses may not always be as minor as the name
implies. In piping systems that contain numerous fittings relative to the total
length of pipe, such as heating or cooling systems, the minor losses may actu-
ally have a significant impact on the energy loss.
The equation most commonly used for determining the loss in a fitting, valve,
meter, or other localized component is:

V2
Hm = K
2g

where
Hm = minor loss (m, ft),
K = minor loss coefficient for the specific fitting,
V = velocity (m/s, ft/s), and
g = gravitational acceleration (m/s2, ft/s2).
Typical values for the fitting loss coefficient are included in Table 6.1. As can be
seen with similar fitting types, the K-value is highly dependent on bend radius,
contraction ratios, and so forth. Gradual transitions create smoother flow lines
and smaller head losses than sharp transitions because of the increased turbu-
lence and eddies that form near a sharp change in the flow pattern. Figure 6.2
shows flow lines for a pipe entrance with and without rounding.

Table 6.1: Typical Fitting K Coefficients

Fitting K-Value Fitting K-Value

Pipe Entrance 90° Smooth bend
Bellmouth 0.03–0.05 Bend radius/D = 4 0.16 – 0.18
Rounded 0.12–0.25 Bend radius/D = 2 0.19 – 0.25
Sharp edged 0.50 Bend radius/D = 1 0.35 – 0.40
Projecting 0.80 Mitered bend

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 165

 6 Pressure Piping Systems and Water Quality Analysis
Energy Gains—Pumps

Table 6.1: (Continued)

Fitting K-Value Fitting K-Value

Contraction—sudden θ = 15° 0.05
D2/D1 = 0.80 0.18 θ = 30° 0.10
D2/D1 = 0.50 0.37 θ = 45° 0.20
D2/D1 = 0.20 0.49 θ = 60° 0.35
Contraction—conical θ = 90° 0.80
D2/D1 = 0.80 0.05 Tee
D2/D1 = 0.50 0.07 Line flow 0.30 – 0.40
D2/D1 = 0.20 0.08 Branch flow 0.75 – 1.80
Expansion—sudden Cross
D2/D1 = 0.80 0.16 Line flow 0.50
D2/D1 = 0.50 0.57 Branch flow 0.75
D2/D1 = 0.20 0.92 45° Wye
Expansion—conical Line flow 0.30
D2/D1 = 0.80 0.03 Branch flow 0.50
D2/D1 = 0.50 0.08
D2/D1 = 0.20 0.13

Sharp entrance: K = 0.50 Rounded entrance: K = 0.03

Figure 6.2: Flow Lines in Minor Losses.

6.3 Energy Gains—Pumps

Pumps are an integral part of many pressure systems and are an important
part of modeling head change in a network. Pumps add energy (head gains)
to the flow to counteract head losses and hydraulic grade differentials within
the system. There are several types of pumps that are used for various purposes;
pressurized water systems typically have centrifugal pumps.

166 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Energy Gains—Pumps

A centrifugal pump is defined by its characteristic curve, which relates the

pump head (head added to the system) to the flow rate. To model the behavior
of the pump system, additional information is needed to ascertain the actual
point at which the pump will be operating.
The system operating point is the point at which the pump curve crosses the
system curve—the curve representing the static lift (Hs) and head losses (HL)
due to friction and minor losses. When these curves are superimposed (as in
Figure 6.3), the operating point is easily located.
As water surface elevations and demands throughout the system change, the
static head (HS) and head losses (HL) vary. These changes cause the system
curve to move around, whereas the pump characteristic curve remains constant.
These shifts in the system curve result in a shifting operating point over time.

Pum
pc
urv
e
Head

Operating point

urve
em c
Syst HL

Hs

Flow rate

Figure 6.3: System Operating Point.

Variable-Speed Pumps
A centrifugal pump’s characteristic curve is fixed for a given motor speed and
impeller diameter but can be determined for any speed and any diameter by
applying the affinity laws. For variable-speed pumps, these affinity laws are
presented as:
2
Q1 n1 H1 ⎛⎜ n1 ⎞⎟
= and = ⎜ ⎟⎟
Q2 n2 H2 ⎜⎝ n2 ⎟⎠

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 167

 6 Pressure Piping Systems and Water Quality Analysis
Energy Gains—Pumps

180

160 n = 0.95
n = 0.90
140
n = 0.85
120 n = 0.80

Head, ft 100

80

60
Best
efficiency
40
point

20 Speed
n=
Full speed
0
0 100 200 300 400
Flow, gpm

Figure 6.4: Relative Speed Factors for Variable-Speed Pumps.

where
Q = pump flow rate (m3/s, ft3/s),
H = pump head (m, ft), and
n = pump speed (rpm).
Thus pump discharge rate is proportional to pump speed, and the pump dis-
charge head is proportional to the square of the speed. Using this relationship,
once the pump curve is known, the curve at another speed can be predicted.
Figure 6.4 illustrates the affinity laws applied to a variable-speed pump. The
line labeled “Best efficiency point” indicates how the best efficiency point
changes at various speeds.

Constant Horsepower Pumps

During preliminary studies, the exact characteristics of the pump may not be
known. In these cases, the assumption is often made that the pump is adding
energy to the water at a constant rate. Horsepower is input as the actual power
added to the system, and not the rated horsepower of the motor (because there
is a loss of efficiency in the motor, and motors usually run at less than their
rated capacity). Specifying a pump as a constant horsepower pump means that
the pump will add the same power to the water at any flow rate. Although
this assumption is useful for some applications, a constant horsepower pump
should only be used for preliminary studies.

168 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Control Valves

6.4 Control Valves

There are several types of valves that may be present in a typical pressurized
pipe system. These valves have different behaviors and different applications,
but all valves are used to automatically control parts of the system, opening,
closing, or throttling to achieve the desired result.

Check Valves (CVs)

Check valves (CVs) are used to maintain flow in one direction only by closing
when the flow begins to reverse. When the flow is in the same direction as the
specified direction of the CV, the valve is considered to be fully open.

Flow Control Valves (FCVs)

A flow control valve (FCV) limits the flow rate through the valve to a specified
value in a specified direction. A flow rate is used to control the operation of
an FCV. These valves are commonly found in areas where a water district has
contracted with another district or a private developer to limit the maximum
demand to a value that will not adversely affect the provider’s system.

Pressure-Reducing Valves (PRVs)

Pressure-reducing valves (PRVs) are often used to separate pressure zones in
water distribution networks. These valves prevent the pressure downstream
from exceeding a specified level, in order to avoid pressures and flows that
could otherwise have undesirable effects on the system. A pressure or a hydrau-
lic grade is used to control the operation of a PRV.

Pressure-Sustaining Valves (PSVs)

Pressure-sustaining valves (PSVs) maintain a specified pressure upstream of
the valve. Similar to the other regulating valves, PSVs are often used to ensure
that pressures in the system (upstream, in this case) will not drop to unacceptable
levels. A pressure or a hydraulic grade is used to control the operation of a PSV.

Pressure Breaker Valves (PBVs)

Pressure breaker valves create a specified head loss across the valve and are
often used to model components that cannot be easily modeled using standard
minor loss elements.

Throttle Control Valves (TCVs)

Throttle control valves (TCVs) simulate minor loss elements, the head loss
characteristics of which change over time. With a TCV, the minor loss K is
adjusted based on some other system flow or head.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 169

 6 Pressure Piping Systems and Water Quality Analysis
Pipe Networks

6.5 Pipe Networks

In practice, pipe networks consist not only of pipes but also of miscellaneous
fittings, services, storage tanks, reservoirs, meters, regulating valves, pumps,
and electronic and mechanical controls. For modeling purposes, these system
elements can be organized into four fundamental categories:
• Junction nodes: Junctions are specific points (nodes) in the system where
an event of interest is occurring. Junctions include points where pipes inter-
sect, points where major demands on the system (such as a large industry,
a cluster of houses, or a fire hydrant) are located, or critical points in the
system where pressures are important for analysis purposes.
• Boundary nodes: Boundaries are nodes in the system where the hydraulic
grade is known, and they define the initial hydraulic grades for any com-
putational cycle. They set the HGL used to determine the condition of all
other nodes during system operation. Boundary nodes are elements such as
tanks, reservoirs, and pressure sources. A model must contain at least one
boundary node for the HGLs and pressures to be calculated.
• Links: Links are system components such as pipes that connect to junc-
tions or boundaries and control the flow rates and energy losses (or gains)
between nodes.
• Pumps and valves: Pumps and valves are similar to nodes in that they
occupy a single point in space, but they also have link properties because
head changes occur across them.
An event or condition at one point in the system can affect all other locations in
the system. Although this fact complicates the approach that the engineer must
take to find a solution, there are some governing principles that drive the behavior
of the network, such as the conservation of mass and the conservation of energy.

Conservation of Mass—Flows and Demands

This principle is a simple one. At any node in the system under incompressible
flow conditions, the total volumetric or mass flow entering must equal the
mass flow leaving (plus the change in storage).
Separating the total volumetric flow into flows from connecting pipes,
demands, and storage, we obtain the following equation:

∑QinΔt = ∑Qout Δt + Δ∀S

where
ΣQin = total flow into the node,
ΣQout = total flow out of the node,
Δ∀S = change in storage volume, and
Δt = change in time.

170 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Network Analysis

A HL3 C Path from A to C

HL3 = HL1 + HL2

Path from A to B

HL1 = HL3 − HL2

HL1 HL2
Loop from A to A

0 = HL1 + HL2 − HL3

Figure 6.5: Conservation of Energy.

Conservation of Energy
Chapter 1 introduced the application of the energy equation to hydraulic
analysis. The principle of conservation of energy dictates that the head losses
through the system must balance at each point (Figure 6.5). For pressure net-
works, this means that the total head loss between any two nodes in the system
must be the same regardless of the path taken between the two points. The
head loss must be “sign consistent” with the assumed flow direction (that is,
head loss occurs in the direction of flow, and head gain occurs in the direction
opposite that of the flow).
Although the equality can become more complicated with minor losses
and controlling valves, the same basic principle can be applied to any path
between two points. As shown in Figure 6.5, the combined head loss around
a loop must equal zero in order to compute the same hydraulic grade for a
given point.

6.6 Network Analysis

Steady-State Network Hydraulics
Steady-state analysis is used to determine the operating behavior of a system
at a specific point in time, or under steady-state conditions. This type of anal-
ysis can be useful in discovering the short-term effect of fire flows or average
demand conditions on the system.
For this type of analysis, the network equations are determined and solved with
tanks being treated as fixed grade boundaries. The results that are obtained from
this type of analysis are instantaneous values and may not be representative of
the values of the system a few hours—or even a few minutes—later in time.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 171

 6 Pressure Piping Systems and Water Quality Analysis
Water Quality Analysis

Extended-Period Simulation
An extended-period simulation is used to determine the behavior of the
system over time. This type of analysis allows the user to model tanks fill-
ing and draining, regulating valves opening and closing, and pressures
and flow rates changing throughout the system in response to varying
demand conditions and automatic control strategies formulated by the
modeler.
While a steady-state model may tell the user whether the system has the capa-
bility to meet a specific demand, an extended-period simulation indicates
whether the system has the ability to provide acceptable levels of service over
a period of minutes, hours, or days. Extended-period simulations can also
be used for energy consumption and cost studies, as well as for water quality
modeling.
Data requirements for an extended-period simulation go beyond what is
needed for a steady-state analysis. The user must determine water usage pat-
terns, provide more detailed tank information, and enter operational rules for
pumps and valves.

6.7 Water Quality Analysis

In the past, water distribution systems were designed and operated with little
consideration of water quality, due in part to the difficulty and expense of
analyzing a dynamic system. The cost of extensive sampling and the complex
interaction between fluids and constituents makes numeric modeling the ideal
method for predicting water quality.
To predict water quality parameters, an assumption is made that there is com-
plete mixing across finite distances, such as at a junction node or in a short
segment of pipe. Complete mixing is essentially a mass balance given by:

ΣQiCi
Ca =
ΣQi
where
Ca = average (mixed) constituent concentration,
Q i = inflow rates, and
Ci = constituent concentrations of the inflows.

Age
Water age provides a general indication of the overall water quality at any given
point in the system. Age is typically measured from the time that the water
enters the system from a tank or reservoir until it reaches a junction.

172 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Water Quality Analysis

Along a given link, water age is computed as:

x
Aj = Aj −1 +
V

where
Aj = age of water at jth node,
x = distance from node j − 1 to node j, and
V = velocity from node j − 1 to node j.
If there are several paths for water to travel to the jth node, the water age is
computed as a weighted average using the equation:

AAj =
∑Q i ⎡⎣⎢ AA i + ( x / V )i ⎤⎦⎥
∑Q i
where
AAj = average age at the node immediately upstream of node j and
Qi = flow rate to the jth node from the ith node.

Trace
Identifying the origin of flow at a point in the system is referred to as flow
tracking or trace modeling. In systems that receive water from more than one
source, trace studies can be used to determine the percentage of flow from each
source at each point in the system. These studies can be very useful in delineat-
ing the area influenced by an individual source, observing the degree of mixing
of water from several sources, and viewing changes in origins over time.

Constituents
Reactions can occur within pipes that cause the concentration of substances to
change as the water travels through the system. Based on conservation of mass
for a substance within a link (for extended-period simulations only):

∂c ∂c
=V + θ(c )
∂t ∂x
where
c = substance concentration as a function of distance and time,
t = time increment,
V = velocity,
x = distance along the link, and
θ (c) = substance rate of reaction within the link.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 173

 6 Pressure Piping Systems and Water Quality Analysis
Water Quality Analysis

In some applications, there is an additional term for dispersion, but this term
is usually negligible (plug flow is assumed through the system).
Assuming that complete and instantaneous mixing occurs at all junction
nodes, additional equations can be written for each junction node with the
following conservation of mass equation:

Ck x =0 =
∑Q jC j x =L + QeCe
∑Q j + Qe
where
Ck = concentration at node k,
j = pipe flowing into node k,
L = length of pipe j,
Qj = flow in pipe j,
Cj = concentration in pipe j,
Qe = external source flow into node k, and
Ce = external source concentration into node k.
Once the hydraulic model has solved the network, the velocities and the mix-
ing at the nodes are known. Using this information, the water quality behavior
can be derived using a numerical method.

Initial Conditions
Just as a hydraulic simulation starts with some amount of water in each storage
tank, initial conditions must be set for a water age, trace, or constituent con-
centration analysis. These initial water quality conditions are usually unknown,
so the modeler must estimate these values from field data, a previous water
quality model, or some other source of information.
To overcome the problem of unknown initial conditions at the vast majority
of locations within the water distribution model, the duration of the analysis
must be long enough for the system to reach equilibrium conditions. Note that
a constant value does not have to be reached for equilibrium to be achieved;
rather, equilibrium conditions are reached when a repeating pattern in age,
trace, or constituent concentration is established.
Pipes usually reach equilibrium conditions in a short time, but storage tanks are
much slower to show a repeating pattern. For this reason, extra care must be taken
when setting a tank’s initial conditions, in order to ensure the model’s accuracy.

Numerical Methods
There are several theoretical approaches available for solving water qual-
ity models. These methods can generally be grouped as either Eulerian or
Lagrangian in nature, depending on the volumetric control approach that is
taken. Eulerian models divide the system into fixed pipe segments, and then

174 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Water Quality Analysis

track the changes that occur as water flows through these segments. Lagrangian
models also break the system into control volumes but then track these water
volumes as they travel through the system. This chapter presents two alterna-
tive approaches for performing water quality constituent analyses.

Discrete Volume Method

The Discrete Volume Method (DVM) is an Eulerian approach that divides
each pipe into equal segments with completely mixed volumes (Figure 6.6).
Reactions are calculated within each segment, and the constituents are then
transferred to the adjacent downstream segment. At nodes, mass and flow
entering from all connecting pipes are combined (assuming total mixing).
The resulting concentration is then transported to all adjacent downstream
pipe segments. This process is repeated for each water quality time step
until a different hydraulic condition is encountered. When this occurs, the
pipes are divided again under the new hydraulic conditions, and the process
continues.

Time-Driven Method
The Time-Driven Method (TDM) is an example of a Lagrangian approach
(Figure 6.7). This method also breaks the system into segments, but rather
than using fixed control volumes as in Eulerian methods, the concentration
and size of water parcels are tracked as they travel through the pipes. With
each time step, the farthest upstream parcel of each pipe elongates as water
travels into the pipe, and the farthest downstream parcel shortens as water
exits the pipe.
Similar to the DVM, the reactions of a constituent within each parcel are cal-
culated, and the mass and flow entering each node are summed to determine

Original concentrations

Reacted concentrations

Transport into nodes

Transport into links

Figure 6.6: Eulerian DVM.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 175

 6 Pressure Piping Systems and Water Quality Analysis
Automated Optimization

Original concentrations

Reacted concentrations

Transport through system

Figure 6.7: Lagrangian TDM.

the resulting concentration. If the resulting nodal concentration is significantly

different from the concentration of a downstream parcel, a new parcel will be
created rather than elongating the existing one. These calculations are repeated
for each water quality time step until the next hydraulic change is encountered
and the procedure begins again.

6.8 Automated Optimization

WaterGEMS has the capability to optimize a model based on field data or
design criteria. Many times, water utility managers will use a model to make
design decisions or gather field data to calibrate a model. This process is typ-
ically a trial-and-error approach where the modeler will modify a few param-
eters in a model to either compare design solutions based on cost or benefit
or have the model better predict the real conditions. Because this can be very
time-consuming, WaterGEMS has the capability to create many potential
solutions and provide a measure of which solution is the “better” solution
based on specific boundary conditions and input criteria.
WaterGEMS employs a genetic algorithm search method to find “better” solu-
tions based on the principles of natural selection and biological reproduction.
This genetic algorithm program first creates a population of trial solutions based
on modeled parameters. The hydraulic solver then simulates each trial solution
to predict the HGL and flow rates within the network and compares them to
any input criteria. Based on this comparison, a goodness-to-fit value is assigned.
This information is now used to create a new population of trial solutions which
again are used to find new solutions. The program compares these solutions to
the specific boundary conditions and input criteria until the goodness-to-fit
value is optimized. In other words no better solution can be generated.

Model Calibration
Model calibration is the process of modifying parameters or values in a model
so it better matches what is happening in the real system. The calibration of

176 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Automated Optimization

water distribution models is very complicated. There are many unknown val-
ues and parameters that are needed at any one time to reduce the discrepancy
between the model and real system. Many times the pipe roughness value is
adjusted to have the model results match the measured or expected values in
the real system. However, there are many other parameters that could influ-
ence the modeled results. The water demand at junctions and the status of
pipes and valves in the system could also be adjusted when calibrating a model.
Calibrating a model relies on accurate field measurement data. Field measure-
ments of pressures in the system, pipe flow rates, water levels in tanks, valve status,
and pump operating status and speed are all used to calibrate models. Critical to
all these measurements is the time for which the measurements are made. The
time of these measurements must all be synchronized to the time frame of the
model. In addition, because the conditions within a real system changes through-
out the day or year, field data should be collected for many different conditions
and times. The calibration process is used to adjust the model to simulate multiple
demand loading and operational boundary conditions. Only then can the mod-
eler be confident that the model is valid for many different conditions.
WaterGEMS uses Darwin Calibrator to assist in optimizing the model to
match field measurement data. The Darwin Calibrator allows the modeler
to input field data then request the software to determine the optimal solu-
tion of pipe roughness values, junction demands, or status (on/off). Pipes
that have the same hydraulic characteristics where one roughness values is
assigned to all pipes can be grouped together. Junctions can also be grouped
based on the demand pattern and location. Caution should be used group-
ing pipes and junctions since this could greatly affect the model calibration
accuracy.

System Design
The goal of water distribution system design is to maximize the benefits of
the system while minimizing the cost. The optimal solution is a design that
meets all the needs of the system at minimal cost. Some planning is needed to
account for additional future needs of the system, including potential growth
of the system in terms of demand and its location. The modeler must work
with the system owner and planning groups to account for both the current
and future needs.
WaterGEMS uses Darwin Designer to assist engineers to plan and design water
distribution networks. The Darwin Designer can be used to size new pipe and/
or rehabilitate old pipes to minimize cost, maximize benefit, or create a sce-
nario to trade-off cost and benefit. The least cost optimization is used to deter-
mine the pipe material and size to satisfy the needed design requirements. The
maximum benefit optimization is used to determine the most beneficial solu-
tion based on a known budget. The Darwin Designer will generate a number
of solutions that meet the design requirements at minimal cost or maximum

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 177

 6 Pressure Piping Systems and Water Quality Analysis
WaterGEMS and Tutorials

benefit. In either case, the best solution of new pipe or rehabilitation of old
pipe will be based upon the following input hydraulic criteria:
• minimum and maximum allowable pressures;
• minimum and maximum allowable pipe flow velocity;
• additional demand requirements; and
• pipe, pump, tank, valve, etc. status change requirements.
Critical to creating an accurate designed system is time and peak demand require-
ments. The peak demand and fire flow conditions are used to size pipes since the
pipe network must work for all conditions. Using average demand values to size
pipe without accurately accounting for peaking factors can create networks that
are either undersized and will not deliver the required water needs, or oversized
and much more expensive than need be. The daily and seasonal variations can
also greatly affect the final design. Demand variations need to be synchronized in
the model to accurately reflect what could happen in the real system.

6.9 WaterGEMS and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software.

What Does WaterGEMS Do?

WaterGEMS is a powerful, easy-to-use program that helps civil engineers
design and analyze water distribution systems. It is also used by water util-
ity managers to as a tool for the efficient operation of distribution systems.
WaterGEMS provides intuitive access to the tools you need to model complex
hydraulic situations. WaterGEMS’s sophisticated modeling capabilities can
• perform steady-state, extended-period, and water quality simulations;
• analyze multiple time-variable demands at any junction node;
• model FCVs, PRVs, PSVs, pressure-breaking valves, and TCVs;
• model cylindrical and noncylindrical tanks and constant hydraulic grade
source nodes
• track conservative and nonconservative chemical constituents;
• determine water source and age at any element in the system;
• estimate construction costs;
• simulate the operating cycles of constant or variable speed pumps;

178 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Problems

• estimate the cost of pumping over any time period;

• use field measurement data to calibrate a model; and
• provide design solutions to maximize benefits at minimal cost.
WaterGEMS can be used as a stand-alone program, integrated with AutoCAD,
or linked to a Geographical Information System via the GEMS component.
The theory and background used in WaterGEMS are presented in this chapter
and in the WaterGEMS online help system.

How Can You Use WaterGEMS?

WaterGEMS can analyze complex distribution systems under a variety of con-
ditions. For a typical WaterGEMS project, you may be interested in determin-
ing system pressures and flow rates under average loading, peak loading, or
fire flow conditions. Extended-period analysis tools also allow you to model
the system’s response to varying supply and demand schedules over a period of
time; you can even track chlorine residuals or determine the source of the water
at any point in the distribution system.
In summary, you can use WaterGEMS forpipe sizing
• pump sizing
• master planning
• operational studies
• rehabilitation studies
• vulnerability studies
• water quality studies.
WaterGEMS is a state-of-the-art software tool primarily for use in the mod-
eling and analysis of water distribution systems. Although the emphasis is on
water distribution systems, the methodology is applicable to any fluid system
with the following characteristics:
• steady or slowly changing turbulent flow
• incompressible, Newtonian, single-phase fluid
• full, closed conduits (pressure system).
Examples of systems with these characteristics include potable water sys-
tems, sewage force mains, fire protection systems, well pumps, and raw water
pumping.

6.10 Problems
Solve the following problems using the WaterGEMS computer program.
1. The ductile iron pipe network shown below carries water at 20°C.
Assume that the junctions all have an elevation of 0 m and the reservoir

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 179

 6 Pressure Piping Systems and Water Quality Analysis
Problems

is at 30 m. Use the Hazen–Williams formula (C = 130) and the pipe and

demand data below to perform a steady-state analysis and answer the
following questions:
(a) Which pipe has the lowest discharge? What is the discharge (in
l/min)?
(b) Which pipe has the highest velocity? What is the velocity (in m/s)?
(c) Calculate the problem using the Darcy–Weisbach equation
(k = 0.26 mm) and compare the results.
(d) What effect would raising the reservoir by 20 m have on the pipe
flow rates? What effect would it have on the hydraulic grade lines at
the junctions?

R-1 P-5 J-1 P-1 J-2

P-
P-4

2
P-3
J-4 J-3

Schematic for Problem 1

Pipe and Junction Information for Problem 1

Pipe Diameter (mm) Length (m) Junction Demand (l/min)

P-1 150 50 J-1 570
P-2 100 25 J-2 660
P-3 100 60 J-3 550
P-4 100 20 J-4 550
P-5 250 760

2. A pressure gage reading of 288 kPa was taken at J-5 in the pipe network
shown below. Assuming a reservoir elevation of 100 m, find the appro-
priate Darcy–Weisbach roughness height (to the hundredths place) to
bring the model into agreement with these field records. Use the same
roughness value for all pipes.

180 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Problems

The pipe and junction data are shown below.

(a) What roughness factor yields the best results?
(b) What is the calculated pressure at J-5 using this factor?
(c) Other than the pipe roughness, what other factors could cause the
model to disagree with field-recorded values for flow and pressure?

J-5 P-5 J-4 P-4 J-3

P-3
P-6

R-1 P-1 P-2

J-1 J-2

Schematic for Problem 2

Pipe and Junction Information for Problem 2

Pipe Diameter (mm) Length (m) Junction Elevation (m) Demand (l/min)
P-1 250 1,525 J-1 55 950
P-2 150 300 J-2 49 1,060
P-3 150 240 J-3 58 1,440
P-4 150 275 J-4 46 1,175
P-5 150 245 J-5 44 980
P-6 200 230

3. A distribution system is needed to supply water to a resort development

for normal usage and emergency purposes (such as fighting a fire). The
proposed system layout is shown in the following figure:

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 181

 6 Pressure Piping Systems and Water Quality Analysis
Problems

Well and Tank

R-1 T-1
P-1
Pro Shop

P-15
PMP-1 P-2
J-1 P-3 J-2 P-4 J-3

Hotel

P-11

P-12
J-11

P-5
10 J-9
P-

P-13 P-14
J-8
P- P-7 P-6
9 J-5
8 J-6 J-4
P-
J-7 J-10
Entrance Administration
and Maintenance

Proposed Network for Problem 3

The source of water for the system is a pumped well. The water is treated and
placed in a ground-level tank (shown above as a reservoir because of its plen-
tiful supply), which is maintained at a water surface elevation of 210 ft. The
water is then pumped from this tank into the rest of the system.
The well system alone cannot efficiently provide the amount of water needed
for fire protection, so an elevated storage tank is also needed. The bottom of
the tank is at 376 ft (high enough to produce 35 psi at the highest node), and
the top is approximately 20-ft higher. To avoid the cost of an elevated tank,
this 80-ft diameter tank is located on a hillside, 2,000 ft away from the main
system. Assume that the tank starts with a water surface elevation of 380 ft.
The pump was originally sized to deliver 300 gpm with enough head to pump
against the tank when it is full. Three defining points on the pump curve are
as follows: 0 gpm at 200 ft of head, 300 gpm at 180 ft of head, and 600 gpm at
150 ft of head. The pump elevation is assumed to be the same as the elevation
at J-1, although the precise pump elevation is not crucial to the analysis.
The system is to be analyzed under several demand conditions with minimum
and maximum pressure constraints. During normal operations, the junction
pressures should be between 35 and 80 psi. Under fire flow conditions, how-
ever, the minimum pressure is allowed to drop to 20 psi. Fire protection is
being considered both with and without a sprinkler system.
Demand alternatives: WaterGEMS enables you to store multiple demand
alternatives corresponding to various conditions (such as average day and peak
hour). This feature allows you run different scenarios that incorporate various
demand conditions within a single project file without losing any input data.

182 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Problems

For an introduction and more information about scenarios and alternatives,

see WaterGEMS’s online help system and Appendices A and B.
Junction Information for Problem 3

Fire with Fire without

Average Day Peak Hour Minimum
Junction Elevation (ft) Sprinkler Sprinkler
(gpm) (gpm) Hour (gpm)
(gpm) (gpm)
J-1 250 0 0 0 0 0
J-2 260 0 0 0 0 0
J-3 262 20 50 2 520 800
J-4 262 20 50 2 520 800
J-5 270 0 0 0 0 800
J-6 280 0 0 0 0 800
J-7 295 40 100 2 40 40
J-8 290 40 100 2 40 40
J-9 285 0 0 0 0 0
J-10 280 0 0 0 360 160
J-11 270 160 400 30 160 160

Pipe network: The pipe network consists of the pipes listed in the following
tables. The diameters shown are based on the preliminary design and may not
be adequate for the final design. For all pipes, use ductile iron as the material
and a Hazen–Williams C-factor of 130.

Pipe Information for Problem 3

Diameter Length Diameter

Pipe Pipe Length (ft)
(in.) (ft) (in.)
P-1 8 20 P-9 6 400
P-2 8 300 P-10 6 200
P-3 8 600 P-11 6 500
P-4 6 450 P-12 8 500
P-5 6 500 P-13 6 400
P-6 6 300 P-14 6 200
P-7 8 250 P-15 10 2,000
P-8 6 400

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 183

 6 Pressure Piping Systems and Water Quality Analysis
Problems

To help keep track of important system characteristics (such maximum velocity

and lowest pressure), you may find it helpful to keep a table such as the following:

Results Summary for Problem 3

Fire with Fire without

Variable Average Day Peak Hour Minimum Hour
Sprinkler Sprinkler
Node w/low pressure
Low pressure (psi)
Node w/high pressure
High pressure (psi)
Pipe w/max. velocity
Max. velocity (ft/s)
Tank in/out flow (gpm)
Pump discharge (gpm)

Another way to quickly determine the performance of the system is to color-

code the pipes according to some indicator. In hydraulic design, a good
performance indicator is often the velocity in the pipes. Pipes consistently
flowing below 0.5 ft/s may be oversized. Pipes with velocities over 5 ft/s are
fairly heavily stressed, and those with velocities above 8 ft/s are usually bottle-
necks in the system under that flow pattern. Color-code the system using the
ranges in the table below. After you define the color-coding, place a legend
in the drawing.

Color-Coding Range for Problem 3

Max. Velocity (ft/s) Color

0.5 Magenta
2.5 Blue
5.0 Green
8.0 Yellow
20.0 Red

(a) Fill in or reproduce the Results Summary table after each run to get
a feel for some of the key indicators during various scenarios.
(b) For the average day run, what is the pump discharge?
(c) If the pump has a best efficiency point at 300 gpm, what can you say
about its performance on an average day?

184 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Problems

(d) For the peak hour run, the velocities are fairly low. Does this mean
you have oversized the pipes? Explain.
(e) For the minimum hour run, what was the highest pressure in the
system? Why would you expect the highest pressure to occur during
the minimum hour demand?
(f) Was the system (as currently designed) acceptable for the fire
flow case with the sprinkled building? On what did you base this
decision?
(g) Was the system (as currently designed) acceptable for the fire
flow case with all the flow provided by hose streams (no sprin-
klers)? If not, how would you modify the system so that it will
work?
4. A ductile iron pipe network (C = 130) is shown below. Use the Hazen–
Williams equation to calculate friction losses in the system. The junctions
and pump are at an elevation of 5 ft and all pipes are 6 in. in diameter.
(Note: Use a standard, three-point pump curve. The data for the pump,
junctions, and pipes are in the tables below.) The water surface of the
reservoir is at an elevation of 30 ft.

R-1 P-6 P-5 J-1 P-1 J-2

PMP-1

P-
P-4

2
P-3

J-4 J-3

Schematic for Problem 4

Pump Information for Problem 4
Head (ft) Flow (gpm)
200 0
175 1,000
100 2,000

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 185

 6 Pressure Piping Systems and Water Quality Analysis
Problems

Junction Information for Problem 4

Junction Label Demand (gpm)
J-1 400
J-2 550
J-3 550
J-4 350

Pipe Information for Problem 4

Pipe Label Length (ft)

P-1 78
P-2 40
P-3 90
P-4 39
P-5 10
P-6 10

(a) What are the resulting flows and velocities in the pipes?
(b) What are the resulting pressures at the junction nodes?
(c) Place a check valve on pipe P-3 such that the valve only allows flow
from J-3 to J-4. What happens to the flow in pipe P-3? Why does
this occur?
(d) When the check valve is placed on pipe P-3, what happens to the
pressures throughout the system?
(e) Remove the check valve on pipe P-3. Place a 6-in. flow control valve
node at an elevation of 5 ft on pipe P-3. The FCV should be set so
that it only allows a flow of 100 gpm from J-4 to J-3 (Hint: a check
valve is a pipe property). What is the resulting difference in flows in
the network? How are the pressures affected?
(f) Why doesn’t the pressure at J-1 change when the FCV is added?
(g) What happens if you increase the FCV’s allowable flow to
2,000 gpm? What happens if you reduce the allowable flow to zero?
5. A local country club has hired you to design a sprinkler system that will
water the greens of their nine-hole golf course. The system must be able to
water all nine holes at once. The water supply has a water surface elevation
of 10 ft. All pipes are PVC (C = 150, use the Hazen–Williams equation
to determine friction losses). Use a standard, three-point pump curve for
the pump, which is at an elevation of 5 ft. The flow at the sprinkler is
modeled using an emitter coefficient. The data for the junctions, pipes,

186 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Pressure Piping Systems and Water Quality Analysis
Problems

and pump curve are given in the tables that follow. The initial network
layout is shown below.

J-9 P-10 J-8 J-4

P-
5
P-9
J-6

J-3
J-7

P-7

P-4
R-1 P-1 PMP-1
J-10 J-5

P-8
6

P-
P- J-2
P-2

11
P-3
J-1

Schematic for Problem 5

Junction and Pipe Information for Problem 5

Emitter Coefficient
Junction Label Elevation (ft) Pipe Label Diameter (in.) Length (ft)
(gpm/psi0.5)
J-1 – 10 P-1 4 10
Hole 1 8 7 P-2 4 1,000
Hole 2 10 7 P-3 4 800
Hole 3 15 40 P-4 3 750
Hole 4 12 5 P-5 3 500
Hole 5 8 5 P-6 3 700
Hole 6 8 15 P-7 2 400
Hole 7 10 20 P-8 4 800
Hole 8 15 10 P-9 3 500
Hole 9 8 12 P-10 2 400
P-11 2 500

Pump Information for Problem 5

Head (ft) Flow (gpm)
170 0
135 300
100 450

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 187

 6 Pressure Piping Systems and Water Quality Analysis
Problems

(a) Determine the discharge at each hole.

(b) What is the operating point of the pump?
6. A subdivision of 36 homes is being constructed in a new area of town.
Each home will require 1.7 l/s during peak periods. All junction nodes are
192 m in elevation. All pipes are ductile iron (C = 130, use the Hazen–
Williams equation to determine the friction losses in the pipe). The cur-
rent lot and network layout is shown below.

Elm St.

J-8

J-2
J-7

P-1
P-8
P-9
J-3 J-1
Street D

Street E
Town

P-6

Street C

P-7

P-5
Street B
h
ighwa

Common land
open
y #64

P-4
P-3 J-4
P-2 J-5
J-9 J-6 Street A

Schematic for Problem 6

Junction and Pipe Information for Problem 6

Junction Label Number of Lots Serviced Pipe Label Length (m) Diameter (mm)
J-1 5 P-1 60.0 150
J-2 4 P-2 60.0 150
J-3 4 P-3 110.5 150
J-4 5 P-4 164.0 150
J-5 6 P-5 152.5 150
J-6 6 P-6 204.0 100
J-7 6 P-7 148.0 150
P-8 61.0 100
P-9 194.0 150

188 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Problems

Currently, a model of the entire water system does not exist. However, hydrant
tests were conducted using hydrants located on two water mains, one in Town
Highway 64 and the other in Elm Street. The following data were obtained:
Town Highway 64 Hydrant Test
Static pressure 310.3 kPa
Residual pressure 98.5 kPa at 32 l/s
Elevation of pressure gauge 190 m
Elm Street Hydrant Test
Static pressure 413.7 kPa
Residual pressure 319.3 kPa at 40 l/s
Elevation of pressure gauge 191.5 m
The subdivision will connect to existing system mains in these streets at nodes
J-8 and J-9. (Hint: Model the connection to an existing water main with a
reservoir and a pump.)
(a) What are the demands at each of the junction nodes? What is the
total demand?
(b) Does the present water distribution system have enough capacity to
supply the new subdivision?
(c) Which connection to the existing main is supplying more water to
the subdivision? Why?
(d) Are the proposed pipe sizes adequate to maintain velocities between
0.15 and 2.44 m/s, and pressures of at least 140 kPa?
(e) Would the subdivision have enough water if only one connection
were used? If so, which one?
(f) What do you think are some possible pitfalls of modeling two con-
nections to existing mains within the same system, as opposed to
modeling back to the water source?
7) Use the pipe sizes given in the table below for the subdivision in
Problem 6.
Pipe Information for Problem 7

Pipe Label Diameter (mm)

P-1 200
P-2 150
P-3 150
P-4 150
P-5 150
P-6 150
P-7 150
P-8 150
P-9 150

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 189

 6 Pressure Piping Systems and Water Quality Analysis
Problems

City ordinances require the following:

The pressure at the fire flow discharge and at other points in the distribution
system cannot fall below 125 kPa during a fire flow of 34 l/s. (Hint: The total
flow at the fire flow node does not need to include the baseline demand.)
(a) If a residential fire occurs at J-7, would the current system be able to
meet the fire flow requirements set by the city?
(b) If not, what can be done to increase the available flow to provide
adequate fire flow to that hydrant?
(c) If a fire flow is placed at J-4, does the system meet the require-
ments with the proposed improvements? Without the proposed
improvements?
8. A local water company is concerned with the water quality within its
water distribution network. They want to determine the age and the chlo-
rine concentration of the water as it exits the system at different junctions.
The water surface at the reservoir is 70 m.
Chlorine is injected into the system at the source of flow, R-1, at a concentra-
tion of 1 mg/l. It has been determined through a series of bottle tests that the
average bulk reaction rate of the chlorine in the system (including all pipes and
tanks) is approximately −0.5/day.

R-1 PMP-1
P-7
P-

3 J-9
P-1
8

J-4 P-4 J-5 P-5 J-6 P-6

J-7
P-1

P-11
P-10

P-12

J-1
P-2
P-14
J-8
J-2 P-3 J-3
P-9

T-1

Schematic for Problem 8

The cylindrical tank has a diameter of 15 m. The base and minimum eleva-
tions are 99 m. The maximum elevation is 104 m, and the initial elevation is
103.4 m.

190 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Problems

Pump Information for Problem 8

Head (m) Discharge (l/min) Controls
40 0 Off if node T-1 above 103.5 m
35 3,000
24 6,000 On if node T-1 below 100.5 m

Stepwise Demand Pattern Data for Problem 8

Time from Start (h) Multiplier Time from Start (h) Multiplier
0 0.80 13 1.30
1 0.60 14 1.40
2 0.50 15 1.50
3 0.50 16 1.60
4 0.55 17 1.80
5 0.60 18 1.80
6 0.80 19 1.40
7 1.10 20 1.20
8 1.50 21 1.00
9 1.40 22 0.90
10 1.30 23 0.80
11 1.40 24 0.80
12 1.40

Junction Data for Problem 8

Junction Elevation (m) Demand (l/min)
J-1 73 151
J-2 67 227
J-3 81 229
J-4 56 219
J-5 67 215
J-6 73 219
J-7 55 215
J-8 84 180
J-9 88 151

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 191

 6 Pressure Piping Systems and Water Quality Analysis
Problems

Pipe Data for Problem 8

Pipe Length (m) Diameter (mm) Roughness
P-1 300 200 130
P-2 305 200 130
P-3 300 200 130
P-4 200 200 130
P-5 300 300 130
P-6 200 200 130
P-7 1 300 130
P-8 5,000 300 130
P-9 300 300 130
P-10 500 200 130
P-11 500 200 130
P-12 500 200 130
P-13 150 150 130
P-14 200 200 130

(a) Perform an age analysis on the system using a duration of 300 h and
a time step of 2 h. Fill in the results table, indicating the maximum
water age at each junction and tank after the system reaches equi-
librium (a pattern of average water age vs. time becomes evident).
What point in the system generally has the oldest water? Explain
why the water is oldest at this location.
(b) Perform a constituent analysis using the same duration and time step
as in part (a). Fill in the results table, indicating the minimum chlo-
rine concentration for each junction and tank after the system has
reached equilibrium (a pattern of concentration versus time becomes
evident). What point in the system has the lowest chlorine concen-
tration? Explain why the chlorine residual is lowest at this location.

Results Table for Problem 8

Junction J-1 J-2 J-3 J-4 J-5 J-6 J-7 J-8 J-9 T-1
Age (h)
Chlorine
Concentration
(mg/l)

192 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Pressure Piping Systems and Water Quality Analysis
Problems

(c) From the above table and graphs of demand, age, and concentration
versus time generated within WaterGEMS, determine the following
correlations:
(1) age and chlorine concentration
(2) demand and chlorine concentration at a junction
(3) demand and water age at a junction
(d) Why is it necessary to run the model for such a long time? Do you
feel that 300 h is too long or too short a time period for testing the
model? Why?
9. A planning commission has indicated a new industry may be connected
to the water system described in Problem 8. You are to determine the pipe
diameters in the network to minimize the installation cost assuming all
the pipes are ductile iron. Use the Darwin Designer to determine the total
cost and size each pipe for each of the following conditions. Use the pipe
cost information from Tutorial 4 for the ductile iron pipe.
(a) Size the pipes using a demand multiplier (peaking factor) of 3.2.
The pressure must remain between 170 and 550 kPa during peak
demand. Exclude pipes P-7 and P-8 in your analysis when deter-
mining the pipe sizes. Hint: you will need to specify an additional
demand of zero (0 l/min) with the default pressure constraints at the
junctions or a fatal error will occur.
(b) It is expected that a new industry with an expected additional
demand of 2,000 l/min with a required minimum pressure of
260 kPa will be added to the system. It could be tapped into the
network at either junction 6, 7, or 8. Size the pipes for the condi-
tions in part (a) above along with the industry added to all proposed
junctions. You will need to analyze the network three times, once
for the industry at J-6, again with the industry at J-7, then finally
with the industry at J-8.
a. Indicate which option(s) would work.
b. Which junction should the industry be tapped into to be the
least costly and what is the expected cost?
c. What is the size of each pipe for the best solution for the least
costly option with the industry added?
d. What is the calculated minimum pressure at the industry for
the best solution?

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 193

 7 Hydraulic Transients in Distribution
Systems

7.1 Hydraulic Transients

A hydraulic transient occurs when there is a sudden change in flow. Hydraulic
transient flow occurs between the initial steady-state condition and the final
steady-state condition after the change has occurred. The most common
cause of transient flow is when the fluid velocity suddenly changes due to
a change to a system component such as valve closure, pump shut down, or
pump failure. The result of transient flow is a series of pressure waves trav-
eling at approximately the speed of sound through the distribution system
that can temporarily increase or decrease the system pressures to values that
are much larger or much smaller than pressures expected during steady-state
conditions. A hydraulic transient is sometimes also referred to as a surge or
water hammer.
Hydraulic transients are important in water distribution systems because
they can cause pipe failure (both rupture and collapse), vibrations, damage
to pipe support structures and joints, vapor cavities, and negative (vacuum)
pressure regions. In regions with a negative pressure, nontreated water from
outside the pipe can be pulled into it and contaminate the water distribution
system.
Hydraulic transients can also occur in pressurized wastewater force mains, sur-
charging sewer systems, hydropower piping systems, slurry or oil pumping,
and industrial systems such as heaters, boilers, steam, coolers, and other cogen-
eration systems.

Transient Flow
Transient flow is the transition from an initial flow condition to the final flow
condition due to a change in the system. A simple illustration of this is when a
valve suddenly closes in a single pipe fed by a tank (Figure 7.1). When the valve
in the pipe instantaneously closes, the fluid velocity suddenly drops behind the
valve. A positive pressure wave is formed and travels up the pipe toward the tank.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 195

 7 Hydraulic Transients in Distribution Systems
Hydraulic Transients

100 m
Globe valve
Q (closes in 20 s)
a = = 330
0
1,0 l
00 /s
m/
s

L = 20,000 m D = 500 mm Datum

300
t = 24.2 s
Packing
220
Head, m

140
ΔH0
60

–20

–100
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance, m (×10,000)

300
t = 30.2 s Packing
220
Head, m

140
ΔH0
60

–20

–100
0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Distance, m (×10,000)

Figure 7.1: Pressure wave traveling up a pipe after valve closure.

As the pressure wave in the pipe moves toward the tank, the velocity behind
the wave is zero and the velocity in front of the wave is approximately equal to
the initial velocity. When the pressure wave reaches the tank, water from the
pipe reenters the tank causing a small increase of the water level in the tank. At
that instant, the pressure wave reaches the tank and the pipe velocity is zero,
but the head in the pipe is higher than in the tank. This head imbalance causes
flow to reverse in the pipe and the pressure in the pipeline to drop, creating a
“pressure reduction wave” that moves back toward the valve. The pressure wave
will oscillate back and forth in the pipe until it is dampened by the system and
reaches the final steady-state condition (Figure 7.2).

196 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Hydraulic Transients

p(t)

pmax

Final steady-state flow condition

p2

p1

pmin
Valve closure
TT x

TM
t
Start of valve closure

Figure 7.2: Oscillating pressure wave near a recently closed valve.

On the downstream side of the valve, the water will initially move away from
the closed valve causing a region where the pressure is suddenly reduced. There
is the potential for the pressure to drop to a level where dissolved gasses can
come out of the water, and also below the vapor pressure of the fluid itself to
create vaporous cavitation (column separation). If the flow (liquid columns)
reverses, these pockets of gasses or vapor could suddenly collapse creating
a strong transient with the potential to damage the pipe lining, or rupture
the pipe.

Common Causes
The majority of hydraulic transients are the result of changes at system bound-
aries, typically at the upstream or downstream ends of the systems, or at high
points (locations of high elevation) in the pipe network. Three common causes
of transients in hydraulic systems are described below.
Pumps—A pump delivers energy from the pump impeller to the fluid as the
fluid passes from the suction side to the discharge side of the pump. Pumps
force water through the piping network and components on the discharge side
of the pump. When the pump starts, it can greatly increase the pressure rap-
idly, or when a pump sags or fails, a sudden drop on pressure can propagate
downstream.
Turbines—A turbine is located on the downstream end of a penstock (the
feeding conduit). As water flows through the penstock and enters the turbine,
the water energy rotates a turbine impeller, which creates an electrical current

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 197

 7 Hydraulic Transients in Distribution Systems
Hydraulic Transient Theory

from the generator. If electricity from the turbine is no longer needed, the
flow through the penstock is stopped, which can create a pressure increase that
propagates upstream into the penstock.
Valves—A valve can change, start, or stop flow suddenly. The rate of closing
or opening of the valve can allow for a gradual increase or decrease of flow
to reduce the magnitude of the transient pressure wave. However, if a valve
suddenly closes (valve slam), there is a sudden high-pressure pulse that prop-
agates upstream of the valve and a sudden low-pressure pulse that propagates
downstream. If the low-pressure pulse is severe enough, gas or vapor pockets
may form, or external water could be pulled into the pipe network. Also, there
is the potential for the pipe to rupture or collapse.
Sudden changes in flow (fluid momentum) create a pressure pulse that prop-
agates away from the disruption in every direction and throughout the entire
pipe network. If no other transient event occurs, than the pressure wave will
continue to propagate as unsteady flow until the transient is dampened by fric-
tion. After the transient is dampened, the system reaches the final steady-state
condition.

7.2 Hydraulic Transient Theory

Hydraulic transient flow is classified as unsteady flow. When the flow rate
changes rapidly in a hydraulic system the flow momentum changes as the fluid
accelerates or decelerates. A hydraulic transient is analyzed mathematically by
solving for the fluid velocity and pressure, and both are a function of position
and time. The continuity equation and momentum equations are needed to
determine the fluid velocity and pressure. Solving these two equations simulta-
neously has shown to reflect actual field measurements as long as the data and
assumptions used are valid. Hydraulic transients are analyzed using one of two
types of models.
Rigid-Column Model—The unsteady flow is modeled as an incompressible
fluid within a rigid system. When the continuity equation is used with this
model, the fluid flow changes instantaneously throughout the system and trav-
els as a single mass inside a nonflexible pipe. This model is limited because it
does not accurately describe the pressure wave propagation in a pipe and does
not apply to very rapid changes in flow. This model is used for slow-flow tran-
sients known as surge or mass oscillation.
Elastic Model—The unsteady flow is modeled as a compressible fluid within
a flexible system. Since the fluid is compressible, it can experience density
changes and the propagation of the pressure wave in the fluid column can be
described. The pressure wave velocity (celerity) is dependent upon the elasticity
of the fluid and pipeline (pipe material and type of joint). This model is more
commonly used to describe hydraulic transient flow for rapid operational sys-
tem changes.

198 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Hydraulic Transient Theory

The derivation of the complete equations for transient analysis is beyond the
scope of this book but can be found in Almeida and Koelle (1992) and Wylie
and Streeter (1993). Also, for the remainder of this book, only the Elastic
Model will be used to describe hydraulic transient flow.

General Modeling Equations

The continuity equation for a fluid is based on the principle of conservation of
mass. The simplified general form of the equation for unsteady one-dimensional
pipe flow that is used in transient flow models is as follows:

∂H a2 ∂Q
+ =0
∂t gA ∂x

where
∂H / ∂t = change in hydraulic head (m, ft) with time (s),
∂Q / ∂x = change in flow (m3/s, cfs) with position (m, ft),
a = pressure wave celerity in the fluid (m/s, ft/s),
A = cross-sectional area of the fluid (m2, ft2), and
g = gravity constant (m/s2, ft/s2).
The momentum equation for a fluid can be derived from considering the forces
acting upon a small fluid volume (control volume), including the shear stresses
generated by the fluid motion and viscosity. The three-dimensional momen-
tum equations for a fluid are also known as the Navier–Stokes equations. The
simplified general form of the equation for unsteady one-dimensional flow that
is used in transient flow models is as follows:

∂Q ∂H fQ Q
+ gA + =0
∂t ∂x 2 DA

where
∂Q / ∂t = change in flow (m3/s, cfs) with time (s),
∂H / ∂x = change in hydraulic head (m, ft) with position (m, ft),
Q = fluid flow rate (m3/s, cfs),
D = inside pipe diameter (m, ft), and
f = Darcy–Weisbach friction coefficient.
The last term in the momentum equation represents the friction losses in the
direction of fluid flow.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 199

 7 Hydraulic Transients in Distribution Systems
Hydraulic Transient Theory

Hydraulic transient modeling consists of solving these equations for a wide

variety of boundary conditions and system topography. The equations cannot
be analytically solved so various approximate methods are used. Except for the
simplest forms of these equations, the solutions are found by using computers.
Bentley’s HAMMER software uses the method of characteristics to convert the
two partial differential equations into four ordinary differential equations. It
does this by determining the characteristic curves when the partial differential
equations become ordinary differential equations. The ordinary differential
equations are solved along the characteristic curve and then transformed into a
solution for the original partial differential equations.

Speed and Pressure of Transients

The pressure wave that is formed during a hydraulic transient will propagate
through the pipe network at a speed up to the speed of sound in the fluid. The
speed of sound in fresh water is approximately 1,447 m/s (4,748 fps). The pressure
wave celerity (a) is less than the speed of sound since the pipe material and joints
are flexible, which dampens the pressure wave. A more flexible pipe material or
expanding joints will have a much slower hydraulic transient pressure wave speed.
In 1878, Diederik Korteweg developed an equation that relates the pressure
wave celerity to the properties of the pipe material and fluid. The pressure wave
celerity used in the Elastic Model is determined from this equation.

E f ρf
a=
1 + (DE f eE )ψ

where
a = pressure wave celerity (m/s, fps),
Ef = modulus of elasticity for the fluid (Pa, lbf/ft2),
ρf = fluid density (kg/m3, slugs/ft3),
D = inside pipe diameter (mm, in.),
e = pipe wall thickness (mm, in.),
E = modulus of elasticity for the pipe material (Pa, lbf/ft2), and
Ψ = pipeline support factor.
The pipeline support factor (Ψ) is determined from the Poisson’s ratio of the
pipe material. For thin-walled pipes, if the pipeline has functioning expand-
ing joints, Ψ = 1; if the pipeline anchored against axial (longitudinally) move-
ment, Ψ = 1−μ2; if the pipeline is anchored at one end only, Ψ = 5/4−μ
(see Advanced Water Distribution Modeling and Management. Heastad Press,
2003). Properties of pipe materials and common fluids can be found in
Tables 7.1 and 7.2.

200 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Hydraulic Transient Theory

In a single pipeline, the pressure wave will reach the other end of the pipe in a
time interval equal to L/a seconds where L in the pipe length (m, ft). For the
pressure wave to return to the original location that caused the transient, the time
interval will be 2L/a seconds. The quantity 2L/a is known as the characteristic
time for the pipeline. Flow control operations that occur faster than the charac-
teristic time are considered “rapid” and have a high risk to create very dangerous
transients, flow control operations that are slower than the characteristic time
are considered “gradual,” and operations with a time much greater than 2L/a
are considered “slow.” The characteristic time can be used to determine which
hydraulic transient flow model to use. For “slow” operations, it is possible to use
the Rigid-Column Model, and for all other conditions, the Elastic Model is used.

Table 7.1 Physical Properties of Some Common Pipe Materials

Modulus of Elasticity
Material Poisson’s Ratio, μ
109 lbf/ft2 GPa
Steel 4.32 207 0.30
Cast iron 1.67–3.55 80–170 0.25–0.27
Ductile iron 3.59 172 0.28
Concrete 0.42–0.63 20–30 0.10–0.15
Reinforced concrete 0.63–1.25 30–60 0.25
Asbestos cement (AC) 0.50 24 0.30
Polyvinylchloride (PVC) 0.069 3.3 0.45
Polyethylene (PE) 0.017 0.8 0.46
Polystyrene (PS) 0.10 5.0 0.40

Table 7.2 Physical Properties of Some Common Fluids

Modulus of Elasticity Density
Material 6 2
10 lbf/ft GPa Slugs/ft3 kg/m3
Fresh water 43.7 2.15 1.94 998
Salt water 47.4 2.27 1.99 1,025
Mineral oils 31.0–40.0 1.5–1.9 1.67–1.73 860–890
SAE 30 oil 31.0 1.5 1.77 912

In 1897, Joukowsky demonstrated that the momentum equation can be used

to correctly predict the maximum pressure created by a hydraulic transient in
the Moscow, Russia water system (Rouse and Ince, 1957). This equation can
be used to estimate the pressure for a hydraulic transient in a single pipeline
connected to a reservoir.

ΔP = aρ f Δv

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 201

 7 Hydraulic Transients in Distribution Systems
Hydraulic Transient Theory

where
ΔP = pressure change due to hydraulic transient (Pa, psf ) and
Δv = change in fluid velocity (m/s, fps).
For the case where the fluid velocity goes to zero rapidly due to the fluid stop-
ping suddenly, the maximum pressure near the control that stopped the flow
can be estimated:

where
ΔPmax = maximum pressure change (Pa, psf ) and
vi = initial fluid velocity (m/s, fps).

Example 7.1: Estimating Maximum Hydraulic Transient Pressure

Estimate the maximum internal pressure for a 24-in. ductile iron pipe
(I.D. = 24.86 in.; e = 0.47 in.) with expanding joints after a valve suddenly
closes 2,500 ft downstream from a reservoir. The regular operating pressure is
95 psi and velocity is 6.5 fps at the valve location.

Solution
Determine the pressure wave celerity using the fluid and pipe material prop-
erties. Since the joints are allowed to expand, the pipeline support factor (Ψ)
is equal to 1.

E f ρf 43, 700, 000 psf 1.94 slugs/ft3

a= = = 3, 701.8 fps
1 + (DE f eE ) ψ 24.86 in. × 43, 700, 000 psf
1+ ×1
0.47 in. × 3, 590, 000, 000 psf

For the fluid velocity going to zero, vi = 6.5 fps the maximum pressure due to
the hydraulic transient can be determined.

3,701.8fps × 1.94 slugs / ft3 × 6.5fps

ΔPmax = aρ f vi = = 324 psi
144 in.2 / ft2

The total pressure in the pipeline would be the operating pressure plus
the pressure increase caused by the hydraulic transient.

Ptotal = Pop + ΔPmax = 95 psi + 324 psi = 419 psi

The characteristic time for this pipeline at the valve can be determined.
Characteristic time = 2L/a = 2 × 2,500 ft/3,830.7 fps = 1.35 s
Since this valve closed suddenly, this would be considered a “rapid” flow oper-
ational control.

202 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Minimizing Transients in Systems

7.3 Minimizing Transients in Systems

Network systems should be designed and operations set to minimize the like-
lihood of hydraulic transient events. However, transients will still occur and
methods must be evaluated to minimize damage caused by transients. There
are two general strategies used to control transients. First is to minimize the
possibility of transient conditions during project design and system flow con-
trol operations. Emergency backup operational systems are placed to avoid
unusual system conditions. The second is to strategically place transient con-
trol devices.

Design Strategies
When designing pipe network systems, the engineer will need to consider the
technical requirements and demands on the system. Sufficient safety factors,
such as the pipe’s wall thickness, sufficient storage, and emergency operations
need to be evaluated to minimize transient conditions. A long pipe with a
high flow velocity and minimal branching is most vulnerable to hydraulic tran-
sients. Emergency flow control scenarios should be analyzed and tested during
the design phase to ensure that the proper pipe diameter and wall thickness,
and system components are specified.
The topographic conditions of the pipe network need to be considered. It is
common to have local high points along the pipe network. It may be more
desirable to reroute the pipeline to avoid high points where low-pressure con-
ditions exist that are prone to the low-pressure pulses caused by transients. The
presence of high points could lead to pipe collapse or column separation within
the pipe if the pressure pulse is extreme.
Selecting larger diameter pipe will lower the fluid velocity within the pipe.
This could reduce the risk of damage caused by transients; however, the eco-
nomic cost would increase, and water age within the pipe would be higher
which could reduce the water quality and disinfectant residual. For long pipe-
line systems, the diameter should be selected to optimize the construction and
operational cost. In most cases, long pipelines will require transient protection
devices.
Steel, plastic, and thin-walled ductile iron pipes are most prone to damage
caused by hydraulic transients. The pipe will be weakened by repeated exposure
to transients either due to the pressure wave itself or by vibrations caused by the
transient. A pipe weakened by age or corrosion would also be vulnerable. The
engineer may specify a more expensive, thick-walled pipe in areas that are most
vulnerable to transients. Also, the type of joint specified could be influenced
by the possibility of transients.
With the use of a transient model, the engineer can test different valve and
pump operations to minimize the effect of hydraulic transients. The most severe
transients with pumps typically occur with a power outage when the pump

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 203

 7 Hydraulic Transients in Distribution Systems
Minimizing Transients in Systems

suddenly stops. For normal pump operations, a slow opening/closing valve will
gradually change the fluid velocity, which reduces the pressure pulse created by
transients. Different ramping speeds with soft starts, or using variable speed
drive motors on pumps, can be tested to minimize transients, although this
is typically accomplished much more economically with slow opening/clos-
ing valves near the pump. The control valve after the pump should be opened
slowly after the pump started and closed slowly before the pump shuts down.

Protection Devices
When possible, design strategies should be used to minimize transients.
However, if transients cannot be avoided, then specific protection devises to
control transients may be needed.
Surge tanks—Surge tanks work by allowing water to bleed out of the system
during high pressures and add water to the system during low pressures. A
surge tank is a small tank that has a water height equal to the hydraulic grade
line (HGL) and feeds the protected pipe by gravity (see Figure 7.3). The surge
tank is sized to provide adequate hydraulic capacity during low-pressure con-
ditions, and the connecting valve is sized to prevent the tank from filling or
draining too slowly or quickly. The transient is dissipated as water enters or
leaves the tank. These devices should be equipped to allow for water to circu-
late into the tank to keep water fresh and if necessary, these devices should be
protected from freezing in colder climates.
A one-way tank is a type of surge tank. It will store water under atmospheric
conditions, and through a check valve allow water to enter the protected pipe
during low-pressure transients. This will reduce the low-pressure transient.

Sight glass Compressed air

Water level

Bypass

Air compressor
x
(A) Air chamber (B) Surge tank

Figure 7.3: Examples of transient protection devices.

204 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Minimizing Transients in Systems

After the transient event, a float valve opens and allows for the gradual refilling
of the one-way tank. The advantage of using the one-way tank is the water
level in the one-way tank can be much lower than the HGL. It is common to
place one-way tanks on the discharge pipe after a pump in locations that are
vulnerable to low-pressure transients after the pump shuts off or experiences a
sudden failure (e.g., power failure).
Hydropneumatic tanks—A hydropneumatic tank is also known as an air
chamber or gas vessel. A hydropneumatic tank is a closed tank that con-
tains compressed air, or nitrogen, above a volume of water (see Figure 7.3).
The gas or air pressure in the chamber can be preset to gauge zero or higher.
During conditions of low pressure in the protected pipe, the compressed air
will expand and force water from the tank into the pipe, which could avoid
or reduce column separation and surges. During conditions of high pressure
in the protected pipe, the air or gas in the tank acts as a “cushion” against the
transient. These devices should be protected from freezing in colder climates.
There are three types of hydropneumatic tanks in HAMMER. A sealed tank
is simply a closed pressure vessel allowing water to move between the tank
and protected pipe based on the pressure within the pipe. A hydropneumatic
tank can also be vented with an air valve to allow air to enter the tank during
conditions of low pressure and effectively become a one-way surge tank. They
can also be vented with an air valve and dipping (ventilation) tube inside. If
the HGL drops and lowers the tank water level below the dipping tube, the air
valve opens and the tank effectively becomes a surge tank open to the atmos-
phere. When the HGL later increases, the air is expelled from the tank through
the air valve and the water level will go above the bottom of the dipping tube,
and the tank will act as a sealed pressure tank again.
Surge relief valves—In the case of a high-pressure transient, a surge relief (pres-
sure relief ) valve can be placed to allow the transient pressure to be relieved.
These valves are set to open and allow water to discharge when the pressure in
the protected pipe exceeds a predetermined value. This controls the maximum
pressure within the protected pipe. After the high-pressure transient is relieved,
the valve will close slowly to avoid causing another transient. Similar to a surge
relief valve is a surge-anticipator relief valve. This valve is controlled by the sys-
tem emergency operations and will start to open immediately in anticipation
of a high transient, usually because of a power failure.
Air inlet valves—Air inlet valves are also called vacuum breakers. These valves
are installed in high points of the pipe network to control vacuum conditions.
During low-pressure transients, these valves will allow air to enter the system.
This will control the minimum pressure within the protected pipe to avoid col-
umn separation or pipe collapse. An adequate period of time should be allowed
for the air that entered the system to be expelled before normal operations are
allowed to restart.
Rupture disk—A rupture disk are also called bursting disk or burst diaphragm.
A rupture disk is a one-time use diaphragm or membrane that will rupture

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 205

 7 Hydraulic Transients in Distribution Systems
Modeling Hydraulic Transients

when a predetermined high or low pressure ruptures the disk. This will create a
“designed” opening in the protected pipe that can be later be easily repaired by
replacing the ruptured disk with a new one. Many times a rupture disk is used
as a backup protection device to a surge relief or air inlet valve.

7.4 Modeling Hydraulic Transients

The primary objectives of transient modeling are to determine the extreme
pressure values that can result from sudden changes to the flow in pipe net-
works and analyze transient control alternatives. This can be due to opera-
tional changes within the system or due to unforeseen events such as power
failure or pipe breaks. Hydraulic transients are pressure waves caused during
the transition from an initial steady-state condition to a final steady-state
condition. Because of the complexity of the equations needed to model tran-
sients, numerical computer models are used to analyze transient flow hydrau-
lics. A well-constructed model is needed for the engineer to have confidence
with identifying and analyzing alternative scenarios for controlling hydraulic
transients. These models are used to establish design specifications for pipes
and system components, as well test alternative normal and emergency oper-
ational controls to minimize the impact caused by hydraulic transients.
When the steady-state flow conditions are altered, the flow characteristics
(velocity and pressure) at positions along the pipe network are altered with time
until the final steady-state conditions are established. The time interval can be
altered to provide enough modeling points to effectively observe the effect of the
hydraulic transient. The time frame for a hydraulic transient can be a few seconds
to a couple minutes in length depending on the size of the pipe network. The
results of a transient model are the time-dependent change in pressure at differ-
ent locations within the pipe network. The modeled maximum and minimum
pressures are evaluated by the engineer, and operational controls or protection
devices are tested to determine alternatives to minimize the hydraulic transient.
Normally, the entire pipe network does not need to be evaluated for hydrau-
lic transients. Transient analysis should be performed on large, high-velocity
pipelines, especially near pumping stations. Normal flow conditions, along
with emergency operations and uncommon flow conditions should be evalu-
ated. Since most hydraulic models used to analyze water distribution systems
such as WaterGEMS do not consider momentum in their analysis, a hydraulic
transient analysis model is required after the piping network is skeletonized.
Skeletonization is the process of selecting only parts of the piping network that have
a significant impact of the modeled behavior. Capturing every feature of a piping
network in a transient model would require a tremendous amount of data and
computational power. Also, the impact of including small pipes to the modeled
results would be minimal since it is relatively easy to identify the most vulnerable
locations to hydraulic transients. Much of a pipe network would have small pipes
or pipes with a relatively low velocity which would not likely be damaged by a

206 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
HAMMER and Tutorials

hydraulic transient. Skeletonization is a more practical approach to modeling tran-

sients that allows the modeler to focus on the piping that is more likely to be dam-
aged by a transient without investing unnecessary time and resources. For a more
detailed explanation and examples of skeletonization, see Heastad Press (2003).
The validation and calibration of hydraulic transient models depend on adequate
field instrumentation and dynamic measurements. Instead of conventional pres-
sure gauges and SCADA systems, high-speed datalogging equipment is needed to
accurately track a hydraulic transient pressure wave. A pressure sensor will need
to be very sensitive over a large pressure range, have high resolution, and be con-
nected to a high speed data acquisition unit. With the measured data, the pres-
sure wave celerity and extreme pressure values can be recorded. It is common to
have at least two locations established in the system for measurements. Also, the
operational conditions such as value open/closed positions, pump impeller rota-
tional speed, and water levels in transient protection devices must be measured.

7.5 HAMMER and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software.

What Does HAMMER Do?

HAMMER is a powerful, easy-to-use program that helps engineers design and
analyze hydraulic transients in pumping systems and water distribution networks
as they transition from one steady state to another. Since the hydraulic transient
can last for a few seconds to a few minutes, the time frame for analysis is very
short. But because of the extreme pressures, both positive and negative, a hydrau-
lic transient can cause significant damage to a pumping system or pipe network.
HAMMER is also used by water utility managers to as a tool to analyze
operation strategies or protection devices to minimize hydraulic transients.
HAMMER’s sophisticated modeling capabilities can be used to
• analyze operation strategies and protection devises to reduce the risk of
transient-related damage to piping networks;
• set operations to reduce wear and tear on pumps and pumping systems;
• prepare operations checklists for use in emergency failures such as power
outages, pipe breakage, and component (valve, pump), and/or control
failures;

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 207

 7 Hydraulic Transients in Distribution Systems
Problems

• reduce the risk of external water entering a piping network that can con-
taminate distribution water; and
• identify locations or pipes that are of high risk to damage caused by a
hydraulic transient.
HAMMER can be used as a stand-alone program, or interface with
MicroStation, AutoCAD or ArcGIS, HAMMER shares the same file format
of WaterCAD and WaterGEMS so models can be easily shared between these
applications. It will also import files from EPANET.

7.6 Problems
Solve the following problems using the HAMMER computer program.
1. A reservoir discharges to atmosphere through 5,000 ft of 12-in. pipeline,
as shown in the following diagram.

Reservoir
elevation = 350 ft
Valve
elevation = 100 ft
Typical pressure drop = 100 psi
Typical flow = 3,000 gpm
Discharge element type = valve
Initial status = open

5,000 ft of 12-in. diameter ductile iron pipe Expansion joints throughout Discharge to
atmosphere
Hazen–Williams C = 120 Wall thickness = 0.28 in.

Schematic for Problem 1

Use HAMMER’s “Discharge to Atmosphere” element to represent the valve in
your model. Assume the fluid is water at 39°F. Also set the Transient Friction
Method to Unsteady, Run Duration Type to Time, Run Duration to 600 s, and
use a Vapor Pressure of −14.2 psi.
(a) What are the static, nontransient pressure and hydraulic grade line (HGL)
at the valve when the valve is closed?
(b) What are the maximum and minimum pressures and HGLs at the valve
for closing times of 2, 10, 30, and 60 s?

Valve P Max P Min HGL Max HGL Min

Closing Time (s) (psi) (psi) (ft) (ft)
2
10
30
60

208 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Hydraulic Transients in Distribution Systems
Problems

2. A pump draws water from reservoir R-1 and delivers it to reservoir R-2
through 10,500 ft of 12-in. pipeline. The pipeline passes over a high point
at Junction J-1, as shown in the following diagram.

Reservoir R-2
Junction J-1 Elevation = 1,000 ft
Elevation = 900 ft

Pipe data

ft
3,

00
12-in. ductile iron 00

2,0
Hazen–Williams C = 120 0
ft
Wall thickness = 0.28 in.
Expansion joints throughout
ft
00
5,0

Junction J-2
Elevation = 400 ft

Reservoir R-1
Elevation = 500 ft

500 ft

Pump
Elevation = 490 ft

Schematic for Problem 2

Pump definition data follows:

Flow (gpm) TDH (ft)

0 840
2,500 630 (BEP)
5,000 0

Additional pump information: Pump efficiency at BEP = 65%

RPM = 1,800 rpm
Specific speed = 980 (US)
Reverse spin allowed
Pump control valve = 8-in. diameter
Valve closure on power failure = 10 s

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 209

 7 Hydraulic Transients in Distribution Systems
Problems

Motor = constant speed

Motor efficiency = 90%

Assume the fluid is water at 68°F. Set the Transient Friction Method to
Unsteady, Run Duration Type to Time, Run Duration to 240 s, and use a
Vapor Pressure of −14.2 psi.
(a) What are the steady-state pressure and hydraulic grade line (HGL) at
Junction J-1 with the pump running?
(b) What are the maximum pressure and maximum HGL at Junction J-1
following a power failure?
(c) If Junction J-1 is replaced with a combination air/vacuum valve (4-in.
air inlet diameter, 1/4-in. air outlet diameter), what are the maximum
pressure and maximum HGL at the air/vacuum valve following a power
failure?

210 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 8 Sanitary Sewer Design

8.1 Sanitary Sewer Systems

Sanitary sewers collect wastewater from its sources and transport it to a treat-
ment facility. Sanitary sewers are intended to convey only wastewater. However,
their design must allow for the infiltration of groundwater that invariably
occurs in sewers. Older systems that convey both sanitary wastes and stormwa-
ter are referred to as combined sewers.
Wherever practicable, sewers are gravity flow systems. When the natural slopes
are not sufficient to convey flow, a combination of gravity and pressure flow
systems is used. The gravity sewer transports flow to a collection point such as
a wet well. The wastewater is pumped from the wet well through a force main
over some obstruction or hill to another gravity sewer, or directly to a waste-
water treatment facility.
The hydraulics in the gravity portion of a sanitary sewer are analyzed with the
same techniques used in storm sewers hydraulics, as discussed in Section 1.6
and in Chapter 3. When the flow depth is constant, Manning’s equation is
used. When obstructions or changes in pipe slope exist, the gradually varied
flow analysis procedure presented in Section 3.6 is applied.
The pressure portion of the sanitary sewer is analyzed in the same manner as a
water distribution system, as discussed in Chapter 6. The primary difference is
that flow is generally withdrawn from a water distribution network, whereas in
a sewer system, flow is injected into force mains.
In this chapter, a brief introduction to sanitary sewer design is presented. The
use of SewerCAD to assist in the design process is demonstrated. More com-
prehensive reviews of sewer design may be found elsewhere (Tchobanoglous,
1981, or ASCE and WEF, 2007). Additional discussion on the development
and application of hydraulic sewer models is presented in Wastewater Collection
System Modeling and Design (Haestad Press, 2004).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 211

 8 Sanitary Sewer Design
Loading

Common Sanitary Sewer Elements

Although all of the various types of sewer appurtenances and structures are too
numerous to mention in this text, the most commonly modeled sanitary sewer
elements include the following:
• Manholes: Manholes provide access for maintenance of a sanitary sewer.
They are generally located at regular intervals (as dictated by local ordi-
nances), at places where pipe characteristics such as diameter and slope
change, and at places where several pipelines join. In sanitary sewer models,
such as SewerCAD, manholes also represent points where loads enter the
system.
• Junction chambers: Junction chambers are underground structures built
at points in the sewer where multiple pipes join. Junction chambers are
rarely used in sanitary sewers because their underground construction
makes access and maintenance difficult.
• Wet wells: Wet wells are structures that collect and store wastewater until it
is pumped into force mains for transport to another gravity system or to a
wastewater treatment facility. Wet wells are generally designed to minimize
wastewater retention time in order to prevent odors and the buildup of
gases, while allowing the pumps to cycle in an efficient manner.
• Pumps and pump stations: Pumps, as discussed in Chapter 6, add head to
flow as it passes through the sanitary sewer system.
• Pressure junctions: Pressure junctions represent points in the pressure sys-
tem where pressure pipes are joined together. They also represent points
where loads directly enter a force main, or points where the modeler wishes
to know a pressure.
• Pipes: Pipes are conduits through which flow is transferred either by the
influence of gravity or by energy supplied from pumping stations. The
most common pipe section shapes in a gravity system are circular, box
(rectangular), arched, elliptical, and egg-shaped. In pressure systems, pipes
are usually circular. Pressurized sewer pipes are commonly referred to as
force mains.
• Regulators: When stormwater loads cause flows in sewers to exceed capac-
ity, control devices are used to divert flows. These devices are called regu-
lators. They split the flow between an interceptor sewer and an overflow.
Possible destinations for controlled releases include wet wells, temporary
storage basins, treatment facilities, parallel relief sewers, or receiving waters.

8.2 Loading
Sanitary sewer loads are divided into two categories: sanitary loads and
wet-weather loads.

212 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Loading

• Sanitary loads (dry-weather loads) result from human activity and are not
weather dependent. Common sources of sanitary loads are various residen-
tial, commercial, recreational, and industrial uses.
• Wet-weather loads are related to rainfall activity such as groundwater
infiltration (water leaking into a pipe through cracks, joints, and defects)
and structure inflow (surface water entering a structure through openings
around the cover, or due to a missing cover).
Time-based loads can be classified as either wet-weather or sanitary loads, and
they may have hydrograph- or pattern-based formats.

Common Load Types

Two loading types that can be applied as either wet-weather or sanitary loads are
• hydrographs and
• pattern loads.
Hydrographs
Flow versus time data, also known as a hydrograph, can be entered as a load.
The hydrograph will then be directly added to any other loads coming to that
point, and then routed downstream.
In the case of wet-weather loading, the hydrograph may reflect infiltration or
inflow resulting from an actual or design storm event. Sanitary load hydro-
graphs show the variation in customer loading over time, such as over the
course of a typical day. An example of a hydrograph is shown in Figure 8.1.
Hydrographs may be obtained from measurements collected in a combined
sewer during a rainfall event or they may be developed using the hydrologic
procedures described in Chapter 2.

5.5
5.0
4.5
4.0
3.5
Flow, cfs

3.0
2.5
2.0
1.5
1.0
0.5
0.0
0 5 10 15 20 25 30 35
Time, h

Figure 8.1: A hydrograph.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 213

 8 Sanitary Sewer Design
Loading

During a steady-state analysis, a hydrograph load must be converted into an

instantaneous load. There are several ways of choosing an appropriate instan-
taneous load, and the method chosen depends on the purpose of the analysis.
For instance, the hydrograph peak flow, average flow, or minimum flow may
be used. If other data are used to generate steady-state loading, the hydrograph
may be disregarded entirely.

Pattern Loads
A pattern load comprises a base load and an associated loading pattern. The
pattern is a series of multipliers relating sanitary sewer loading to time, and
it allows the user to automatically apply time-variable changes within the
system.
Different categories of users, such as residential or industrial customers, may
be assigned different patterns to accurately reflect their particular load varia-
tions. A diurnal curve is a type of pattern that describes change in loading
over the course of a daily cycle. A dry-weather diurnal curve for sanitary load-
ing often resembles the diurnal curve for water use at the same location (see
Section 6.1), reflecting times when people are using more and less water than
average.
Patterns are typically based on a multiplication factor versus time relationship,
in which a multiplication factor of 1.0 represents the base value (usually the
average value). This relationship can be expressed:

Qt = At × QBase

where
Q t = load at time t,
At = multiplier for time t, and
Q Base = baseline load.

For the residential usage, diurnal curve in Figure 8.2, we see that there is a peak
in the diurnal curve in the morning, another slight peak around noon, and a
third peak in the evening. During the nighttime hours, the pattern reflects the
relative inactivity of the system. (Note that this curve is conceptual and should
not be construed as representative of any particular network.)
As explained in Chapter 6, there are two basic forms for representing a pattern:
stepwise and continuous. A stepwise pattern is one that assumes a constant
level of usage over a period of time and then jumps instantaneously to another
level where it again remains steady until the next jump. A continuous pat-
tern is one for which several points in the pattern are known and sections in
between are transitional, resulting in a smoother pattern. For the continuous
pattern in Figure 8.2, the magnitude and slope of the pattern at the start and
end times are the same—a continuity that is recommended for patterns that
repeat.

214 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Loading

2.0
1.8
1.6
1.4
1.2

Multiplier
1.0
0.8
0.6
0.4
0.2
0.0
0 5 10 15 20 25
Time, h

Figure 8.2: An example of a diurnal pattern for sanitary loading.

Sanitary Unit Loading

Ideally, sanitary loads are based on data measured at individual loading points
throughout the sewer system. In other words, the load from each source is
modeled as a direct flow into the sewer system. However, data with this level
of detail are not typically available. Instead, wastewater sources are grouped
together, and average loads are computed based on land usage or the activities
of the population generating the load.
For example, consider a small, middle-class subdivision that houses approxi-
mately 50 people. Studies have shown that, on average, a resident of middle
class housing generates 280 l/day (74 gpd) of wastewater. Thus the average load
entering the sewer system at this subdivision is 14,000 l/day (3,700 gpd).
Sanitary loads are typically estimated based on a number of contributing
units, with a specified average load per unit (such as 260 l/day per apartment
resident). The unit is typically a measure of population such as residents or
employees; however, loads can also be based on other criteria such as contrib-
uting area or user-defined counts of items indicative of loading behavior. See
Table 8.1 for a list of standard loading sources and their associated average
base loads.

Extreme Flow Factors

Sewer design and analysis generally consider a variety of loading conditions,
such as minimum, average, and peak conditions. Base (average) sanitary loads
are transformed into minimum or peak loads using an extreme flow factor
(EFF) method (Figure 8.3).

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 215

 8 Sanitary Sewer Design
Loading

Table 8.1 Typical Unit Sanitary Loads from Different Sources

(Tchobanoglous, 1981)

Base Load
Unit Sanitary Load Loading Unit
(l/day per Loading Unit)

Airport Passenger 10

Apartment Resident 260

Automobile service Employee 50

Bar Customer 8

Cabin resort Guest 160

Campground Guest 120

Coffee shop Guest 20

Department store Toilet room 2,000

Dormitory Guest 150

Home (average) Resident 280

Home (better) Resident 310

Home (luxury) Resident 380

Hospital (medical) Bed 650

Hotel Employee 40

Prison Inmate 450

Restaurant Meal 10

School (large) Student 80

School (medium) Student 60

School (small) Student 40

Swimming pool Employee 40

Shopping center Parking space 4

The most common type of EFF is the variable peaking factor (PF). In this
method, a computed EFF is multiplied by the specified base flow.

Q peak = Q base × EFF

216 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Loading

where
Qpeak = transformed flow (l/day, gpd),
Qbase = base flow (l/day, gpd), and
EFF = extreme flow factor (unitless).

Common Variable Peaking Factors (PF)

Some of the most common variable PF calculation methods are as follows:

• Babbitt (1958)

5.0
PF =
(P 1, 000)
0.2

where P = contributing population.

• Harmon (1918)

14.0
PF = 1.0 +
4.0 + ( P 1, 000)
0.5

• 10 States Standard (Great Lakes Upper Mississippi River Board, 1997)

P
18 +
PF = 1000
P
4+
1000

• Federov (Jakovlev et al., 1975)

2.69
PF =
Q 0.121

where Q = base sanitary flow (l/s).

For population-based PF methods, the PF decreases as the population

increases. A larger population means that peak loads from different sources are
likely to occur in a more staggered manner. Thus the peak loads are likely to
be less pronounced compared to the average loading rate. In systems servicing
smaller populations, peak loads from different sources are more likely to coin-
cide, causing more pronounced differences in average loading rates. The graph
in Figure 8.3 shows how the EFF decreases with increasing population when
applying the 10 States Standard.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 217

 8 Sanitary Sewer Design
Loading

4.0
3.5

Extreme flow factor

3.0
2.5
2.0
1.5
1.0
0.5
0.0
0.0 1.0 2.0 3.0 4.0 5.0
Contributing population in millions

Figure 8.3: EFF versus contributing population, from the 10 States

Standard.

Example 8.1: Determining Peak Sanitary Loads

A shopping center and a residential subdivision of average and luxury homes
load a gravity sanitary sewer. The shopping center has 100 parking spaces.
Eighty people live in the “average” housing, whereas 30 people live in the “lux-
ury” housing. What is the average base load contributing to the sanitary sewer
at this point? What would be an approximate peak flow applied to the sys-
tem? The local ordinance dictates that the Harmon’s method must be applied
to peak all population-based unit sanitary loads, whereas a constant PF of 2
should be applied to all nonpopulation-based loads.

Solution
First, to get the average base load, find the amount of load per unit for the
three loading categories mentioned in Table 8.1. Then, multiply this unit flow
by the number of units.

Shopping center 100 Parking spaces × 4 l/(day-space) = 400 l/day

Average housing 80 Residents × 280 l/(day-resident) = 22,400 l/day

Luxury housing 30 Residents × 380 l/(day-resident) = 11,400 l/day
Total base flow = 34,200 l/day

Apply Harmon’s method to find the PF for the residential population of 80

residents of “average” housing and 30 residents of “luxury” housing.
• Assuming that both population-based loads have similar peaking behavior,
sum the two populations:
80 Residents of average housing + 30 residents of luxury housing = 110
Residents

218 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Loading

• Apply the Harmon’s method using a contributing population of 110:

14
PF = 1.0 + = 4.23
4 + (110 1, 000)
0.5

Apply this PF to the base loads calculated from the subdivision. Then, apply
the given PF of 2 to the load derived from the shopping center to compute the
approximate peak flow.
4.23 × (22,400 l/day + 11,400 l/day) + 2 × 400 l/day = 144,000 l/day

Wet-Weather Loading
Wet-weather loading represents the intrusion of ground and surface water into
the sewer system and is divided into two components: infiltration and inflow.
Infiltration represents groundwater entering pipes through cracks, holes,
and other defects. Inflow represents wet-weather flow that enters the sewer
through manhole covers, roof drains, catch basins, sump pumps, and so forth.
Collectively, these flows are referred to as I/I. It is important to minimize intru-
sion of rain water into the sewer system because of the cost of treating the added
flow. Also, wet-weather loads increase the potential for overloading the system,
which causes flow to back up into houses or overflow into the environment.
Infiltration causes noticeable increases in sewer flow, especially after rainstorms
and subsequent rises in the groundwater table. Sewer pipes placed in low areas,
such as near creek beds, are more subject to infiltration than those placed in
higher ground because they are more likely to be located below the water table.
Older sanitary sewers are more susceptible to infiltration than newer sewer
systems because modern techniques and materials used in newer systems can
generally prevent or greatly reduce infiltration.
Infiltration at an individual pipe is generally hard to quantify. Instead, infiltra-
tion is estimated for the whole system or a portion of the system. This global
infiltration is then distributed to individual pipes using one of the following
methods:
• proportional to pipe length
• proportional to pipe diameter/length
• proportional to pipe surface area
• proportional to some user-defined count (e.g., number of defects in
the pipe)
Inflows are unwanted flows entering from the surface and are modeled based
on estimates or requirements from local or state regulatory agencies. Like infil-
tration, inflows can be quite significant and can vary considerably depending
on the depth of runoff, the size of the manhole cover, and the number of
openings in the manhole cover.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 219

 8 Sanitary Sewer Design
Extended-Period Simulations

8.3 Extended-Period Simulations

Overview
An extended-period simulation (EPS) models how a sewer network will
behave over time. This type of analysis allows the user to model how wet
wells fill and drain; pumps toggle on and off; and pressures, hydraulic
grades, and flow rates change in response to variable loading conditions
and automatic control strategies formulated by the modeler. EPS is a useful
tool for assessing the hydraulic performance of alternative pump and wet
well sizes.
The SewerCAD algorithm proceeds in a general downstream direction toward
the outfall, the whole procedure of this is described here:
1. The analysis begins in the gravity portion of the network. The hydro-
graphs enter the gravity system and are successively routed and summed
as the flows approach the downstream wet well or outfall. Ultimately, the
total inflow hydrograph to the wet well is determined.
2. Knowing the inflow into the wet well, the pressure calculations for the
force main system bounded by the wet well are performed. In addition
to flow velocities and pressures, the levels in the wet well over time are
determined.
3. The calculation then returns to the gravity portion of the network discussed
in step 1. The hydraulics and hydraulic grade line (HGL) profiles are cal-
culated through out the gravity system for each time step using the known
level of the wet well as the boundary condition for the backwater analysis.
The process repeats, continuing through the system downstream of the pres-
sure network until an outlet is reached.

Routing Overview
As a hydrograph is routed through a conduit, it undergoes changes in shape
and temporal distribution caused by translation and storage effects.
SewerCAD uses one of two methods to determine the shape and distribution
of a hydrograph routed through a gravity pipe.
• convex routing
• weighted translation routing
Convex Routing
The underlying assumptions of the convex routing method are that the routed
outflow for a time step is based on the inflow and outflow for the previous
time step, and that the flow does not back up in the pipe (i.e., no reverse
flow or reduced flow due to tailwater effects exists). Each outflow ordinate is
calculated as:

220 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Extended-Period Simulations

Ot +Δt = cIt + (1 − c )Ot

where
Ot + Δt = outflow at time t + Δt;
t = current time (s, min),
Δt = hydrologic time step (s, min),
c = convex routing coefficient,
It = inflow at time t(l/s, gpm), and
Ot = outflow at time t(l/s, gpm).

The convex routing coefficient is essentially a ratio of the hydrologic time

step and representative flow travel time through the pipe and is calculated as
follows:

V Δt
c = Δt =
L tt

where
Δt = hydrologic time step (s),
tt = travel time (s),
V = velocity established for representative flow (m/s, ft/s), and
L = length of pipe (m, ft).

In SewerCAD, the velocity used to calculate the coefficient is either the normal
velocity or full velocity generated for a user-specified percentage of the peak
of the inflow hydrograph. In other words, if the percentage of the peak flow
is greater than the capacity of the pipe, the full flow velocity is used. If the
percentage of the peak flow is less than the pipe capacity, the flow velocity for
normal depth is used.
The higher the percentage of flow, the faster the velocity used to calculate the
convex routing coefficient, and the closer the routed hydrograph will be to a
pure translation of the inflow hydrograph.
The user-specified percentage can be modified in the calculation options. A
typical value is around 75%, but this value may be modified for oddly shaped
hydrographs with sharp, uncharacteristic peaks or for calibration purposes.

Weighted Translation Routing

The convex routing method is only valid when the convex routing coefficient, c,
is less than 1 or when the hydrologic time step is less than the calculated travel
time. For certain cases in which the travel time exceeds the hydrologic time
step, SewerCAD automatically uses an alternative method of routing called
weighted translation routing.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 221

 8 Sanitary Sewer Design
SewerCAD, SewerGEMS and Tutorials

Each ordinate of the outflow hydrograph is derived from a weighted average of

the ordinates for the current and previous time steps of the inflow hydrograph.
The weights are calculated based on the convex routing coefficient.
Each ordinate of the outflow hydrograph is calculated as follows:

1 ⎛ 1⎞
Ot = It −Δt + ⎜⎜⎜1 − ⎟⎟⎟ It
c ⎝ c⎠

where
Ot = outflow at current time step (l/s, gpm),
c = convex routing coefficient
It − Δt = inflow at previous time step (l/s, gpm), and
It = inflow at current time step (l/s, gpm).

Hydrologic and Hydraulic Time Steps

SewerCAD uses two distinct time steps when running an EPS.
• Hydrologic time step: This time step is used to calculate the routed hydro-
graphs and represents the time increment of all hydrographs generated dur-
ing the analysis. The hydrologic time step is also used as the calculation
increment for the pressure calculations.
• Hydraulic time step: This time step represents how often the hydrau-
lic calculations are performed for gravity flow. Flows are interpolated
from the previously generated hydrographs using the hydraulic time step
and are then used to perform the gradually varied flow analysis for that
time step.
The hydrologic time step should be less than or equal to the hydraulic time
step, and the hydraulic time step should be a multiple of the hydrologic
time step.

8.4 SewerCAD, SewerGEMS and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software can
be found at aka.bentley.com/CAiHE.supplement. If you have trouble access-
ing the site, please send a message to [email protected]. The tutorials are
designed for student users to become familiar with the design and analysis of
hydraulic systems without needing to spend a lot of time inputting data. As
new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software. Information
about SewerGEMS can be found in Chapter 9.

222 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
SewerCAD, SewerGEMS and Tutorials

What Does SewerCAD Do?

SewerCAD is a powerful, easy-to-use program for the design and analysis of
wastewater collection systems. SewerCAD’s intuitive graphical interface and
powerful data exchange capabilities make it easy to develop and load complex
models of combined gravity and pressure networks. Using SewerCAD, you
can design and analyze the gravity portion of the system according to either a
gradually varied flow calculation or a standard capacity analysis.
The powerful automatic design capabilities of SewerCAD allow you to design
the system based on user-defined constraints for velocity, cover, and slope. You
can also design for partially full pipes, multiple parallel sections, maximum sec-
tion size, invert/crown-matching criteria, and the allowance of drop structures.
You can disable the design feature on a pipe-by-pipe basis, to design all or only
a portion of your system.
SewerCAD’s powerful loading model provides complete support for all types
of sanitary and wet-weather loads. The unit sanitary load library is completely
user-customizable and supports population, area, discharge, and count-based
sanitary loads, as well as hydrographs and load patterns. Wet-weather inflow can
also be added to the model as instantaneous loads, hydrographs, or base loads
with associated patterns. Infiltration can be computed based on pipe length, pipe
diameter/length, pipe surface area, or unit count. Loads are peaked according
to a PF or EFF equations that you select from a user-customizable EFF library.

How Can You Use SewerCAD?

You can use SewerCAD to design new systems and analyze the performance of
existing systems. SewerCAD’s intuitive graphical editor and powerful scenario
management capabilities facilitate the process of analyzing a large number of
design alternatives and finding potential problems in an existing system. In
summary, SewerCAD’s sophisticated modeling techniques enable you to
• design and analyze multiple sanitary sewer networks in a single project;
• examine your system using SewerCAD’s gradually varied flow algorithms
or a standard capacity analysis;
• design the system using SewerCAD’s automatic, constraint-based design;
• load your model based on contributing population, service area, total sani-
tary flow, or your own loading type;
• peak your loads using Babbitt’s, Harmon’s, 10 States, and Federov’s equa-
tions, or use your own formulas or tables;
• calculate infiltration based on pipe length, diameter, surface area, length/
diameter, or user-defined data;
• generate plan and profile plots of a network;

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 223

 8 Sanitary Sewer Design
Problems

• perform EPSs that include time-variable loads and hydrologic routing; and
• animate plans and profiles showing sanitary sewer system performance
over time.
The theory and background used in SewerCAD are presented in this chapter
and in the SewerCAD online help system.
SewerCAD can analyze complex sewage collection systems under a variety of
loading conditions. SewerCAD’s powerful loading capabilities can be used to
track system response to an unlimited range of sanitary and wet-weather load-
ing combinations. It can be used for
• pipe sizing
• pump sizing
• master planning
• operational studies
• rehabilitation studies

8.5 Problems
1. In the sanitary sewer represented by the following diagram, a load from
a development of high-rise apartments (at wet well WW-1) with 10,000
residents is pumped to the top of a hill, where it is then transported
via a circular gravity line to a treatment plant (represented by O-1). At
manhole MH-1 is a resort apartment with 300 guests. The load gen-
erated here flows down a circular gravity pipe to the same gravity line
mentioned above. A bar that serves, on average, 50 people per day and 2
large cafeterias with 20 employees each are located near manhole MH-2.

MH-1
P-1

-2 MH-2 P-2
FM

PMP-1
-1
FM
O-1

WW-1

224 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Problems

Schematic for Problem 1

Pipe Data for Problem 1
Diameter Length Upstream Downstream
Pipe Pipe Type Material
(mm) (m) Invert (m) Invert (m)
P-1 Gravity PVC n = 0.010 200 100 98.90 97.62
P-2 Gravity PVC n = 0.010 375 200 97.62 96.62
FM-1 Pressure Ductile iron C = 130 300 1 70.00 70.00
FM-2 Pressure Ductile iron C = 130 350 1,000 70.00 97.62

Node Data for Problem 1

Ground Rim Sump Structure
Node
Elevation (m) Elevation (m) Elevation (m) Diameter (m)
MH-1 100.30 100.30 98.90 1.00
MH-2 99.10 99.10 97.62 1.00
O-1 100.00 100.00 96.62 N/A
WW-1 74.00 – – –
The base and minimum elevations for the wet well are both 70 m. The initial
elevation is 73 m, and the maximum elevation is 73.5 m. The diameter of the
circular wet well is 3 m.
The tailwater elevation at outlet O-1 is 98 m.
The three defining points of the pump curve are 0 m3/min at 33.33 m, 5.00 m3/
min at 25.00 m, and 10 m3/min at 0 m. The pump’s elevation is 70 m.
(a) What is the total peak sanitary outflow if no peaking factor method
is applied to the four unit sanitary loads mentioned above? If
Babbitt’s peaking factor method is applied? If Harmon is applied?
Which peaking factor method is the most conservative?
(b) With the Harmon peaking factor applied to each of the four unit
dry loads, what is the hydraulic grade at MH-2? How does this
peaking factor change the hydraulic load and the flow velocity of
pipe P-1 from when no peaking factor was applied?
(c) Identify and describe any problems for each of the three scenarios
from part (a).
(d) The estimated infiltration rate for each of the concrete gravity pipes
in the proposed sanitary sewer system represented below is 20 l/day/
mm km. The estimated inflow into each of the manholes during a
5-year storm event is approximately 75 l/day. The ground elevations
for MH-1, MH-2, MH-3, JC-1, and O-1 are 12.10, 12.10, 11.20,
11.80, and 10.25 m, respectively. The top of junction chamber JC-1
is 11.8 m. The tailwater condition is a free outfall.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 225

 8 Sanitary Sewer Design
Problems

Use SewerCAD’s automatic design feature to design the inverts and

sizes of concrete pipes in the proposed sanitary sewer. The design
constraints and pipe lengths are given in the tables below. Apply
both the wet-weather (given above) and dry-weather sanitary loads
(given in the table below). Use Federov’s equation to calculate the
peaking factor for each of the unit sanitary loads.

MH-1 MH-2

2
P-

P-
1

MH-3
JC-1

P-
3 4
P-

O-1

Schematic for Problem 2

Pipe and Constraint Data for Problem 2
Pipe Length (m) Constraint Minimum Maximum
P-1 200 Velocity (m/s) 0.5 3.0
P-2 150 Cover (m) 1.0 3.0
P-3 200 Slope (m/m) 0.005 0.1
P-4 300

Node Data for Problem 2

Node Unit Sanitary Load Loading Unit Loading Unit Count
Hospital (medical) per bed Bed 400
MH-1 Cafeteria per employee Employee 20
Apartment resort Guest 100
Apartment Resident 400
Shopping center per
MH-2 Employee 100
employee
Laundromat per wash Wash 200
MH-3 School (boarding) Student 500

226 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Problems

(a) List the diameter and slope for each newly designed pipe. Are there
any problems with the designed system?
(b) What percentage of the total flow is wet-weather flow? Which pipe
has the most infiltration?
(c) For a larger magnitude storm, the inflow rate into each manhole
is estimated at 100 l/day. Analyze the model using the previously
designed system and apply the larger wet-weather loading. What
percentage of the total flow is wet-weather loading?
(d) Do you consider the amount of wet-weather flow into the system
significant? What are some methods for alleviating infiltration?
3. The following network is an initial design for a system of force mains in
a sanitary sewer. All pipes are ductile iron (Hazen–Williams’ C = 130).
The pump, PMP-1, is at an elevation of 691 m. Enter the system data
given in the tables below and answer the following questions. The ground
elevation at the outlet is 715 m and the sump elevation is 712 m.
Hints: Make sure that the Fixed Level in Steady State box is checked under
the “Section” tab of the Wet Well dialog. In addition, before running the
model, make the following modification to the calculation options: within the
GO dialog, click the Options button, scroll to the right, and in the Pressure
Hydraulics tab, check the Use Pump Loads box.

WW-1 FM-1 PMP-1 FM-2 J-1 FM-3 J-2 FM-4 J-3

FM-5
FM-7

FM-8
J-4
J-5 FM-6

O-1

Schematic for Problem 3

Pump and Wet Well Data for Problem 3
Pump Curve Head Flow Ground elevation (m) 694.7
Point (m) (m3/min) Maximum elevation (m) 694.0
Shutoff 30 0 Initial elevation (m) 691.5.
Design 20 4 Minimum elevation (m) 688.5
Max
5 6 Base elevation (m) 687.0
operating
Diameter (m) 3.0

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 227

 8 Sanitary Sewer Design
Problems

Junction Data for Problem 3

Pressure Junction Elevation (m)
J-1 698
J-2 701
J-3 703
J-4 705
J-5 703

Pipe Data for Problem 3

Diameter Length Upstream Downstream
Pipe
(mm) (m) Invert (m) Invert (m)
FM-1 400 2 687.00 691.00
FM-2 250 58 691.00 698.00
FM-3 350 45 698.00 701.00
FM-4 350 88 701.00 703.00
FM-5 350 71 703.00 705.00
FM-6 450 67 705.00 712.00
FM-7 450 39 701.00 703.00
FM-8 450 68 703.00 705.00

(a) What is the head loss across the entire system?

(b) Why is there more flow in FM-7 than FM-4?
(c) Fill in the following table.

Answer Table for Part (c) of Problem 3

Pipe Flow (m3/s) Velocity (m/s) Head Loss (m)
FM-2
FM-3
FM-4
FM-5
FM-6
FM-7
FM-8

228 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Problems

(d) If the minimum velocity required in the force main to keep particles
from settling is 0.6 m/s, which areas are going to have problems?
(e) What are some possible changes to the design to fix the problem
portions of the system?
(f) Implement a solution suggested in part (e). Describe the fix(es) and
fill in the following chart.

Answer Table for Part (f) of Problem 3

Pipe Diameter (mm) Flow (m3/s) Velocity (m/s) Head Loss (m)
FM-2
FM-3
FM-4
FM-5
FM-6
FM-7
FM-8

4. A major interceptor along a river collects laterals from subdivisions. The

lower residential area loads are collected in a wet well and pumped to the
major interceptor on the other side of a hill. The layout of the system
is shown below. All pipes shown as double lines are circular, concrete,
gravity-flow pipes (n = 0.013). The two pressure pipes (FM-1 and FM-2)
are ductile iron (Hazen–Williams’ C = 130). There is an overflow diver-
sion at MH-5. All input data is given below.
To determine the performance of the system, set up and run three scenarios:
1. a steady-state analysis of the average (base) sanitary loading only,
2. an extended-period analysis of the sanitary loading only, and
3. an extended-period analysis of both the sanitary and wet-weather loading.
For the extended-period analyses, use a 24-h duration with a 1-h hydraulic
time step and a 0.1-h hydrologic time step.
Hint: For this problem, it is only necessary to create one Sanitary Loading
Alternative that will contain the base and pattern loads. During a steady-state
analysis, SewerCAD will ignore the time-based pattern. However, it will be
necessary to create two infiltration and inflow loading alternatives (one with-
out the wet-weather loads and the other with the wet-weather loads) because
Scenario 2 should not consider the wet-weather loads. Define the wet-weather
loading pattern in the manhole prototype before laying out the network.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 229

 8 Sanitary Sewer Design
Problems

MH-6
MH-10

P-6
P-10
MH-11 MH-7 MH-1

P-11

P-7

P-1
P-8
MH-8
MH-9 MH-2

P-9

P-2
O-1 P-5 P-4 MH-4 P-3 MH-3

MH-5
-2
FM

1
P-12 FM- PMP-1

MH-12 WW-1

Schematic for Problem 4

Pipe Data for Problem 4
Section Upstream Invert Downstream Invert Length
Pipe
Size (in) Elevation (ft) Elevation (ft) (ft)
P-1 18 146.50 141.50 400
P-2 18 141.50 136.00 400
P-3 24 136.00 119.00 400
P-4 36 119.00 114.00 400
P-5 36 114.00 94.00 400
P-6 12 157.00 146.50 400
P-7 18 146.50 136.50 400
P-8 18 136.50 126.00 400
P-9 24 126.00 114.00 400
P-10 12 147.00 136.50 400
P-11 18 136.50 126.00 400
P-12 18 74.50 45.00 400
FM-1 24 45.00 45.00 2
FM-2 24 45.00 136.00 60

230 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Problems

Node Data for Problem 4

Ground Sump
Node
Elevationn (ft) Elevation (ft)
MH-1 150.00 146.50
MH-2 145.00 141.50
MH-3 140.00 136.00
MH-4 129.00 119.00
MH-5 124.00 114.00
MH-6 160.00 157.00
MH-7 150.00 146.50
MH-8 140.00 136.50
MH-9 130.00 126.00
MH-10 150.00 147.00
MH-11 140.00 136.50
MH-12 80.00 74.50
Outlet 100.00 94.00
PMP-1 45.00 N/A
WW-1 60.00 N/A

Overflow Data for MH-5 in Problem 4

System Flow (gpm) Divertedw Flow (gpm)
0 0
20,000 0
30,000 10,000

All manholes have a sanitary base load of 100 gpm and a continuous diurnal
pattern applied to them as defined below. Each manhole also has wet-weather
loading given by the following hydrograph.

Domestic Pattern Data for Problem 4

Time(h) Multiplier
Starting 0.4
3 1.0
6 1.4
9 1.2
12 1.4
15 0.9
21 0.6
24 0.4

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 231

 8 Sanitary Sewer Design
Problems

Wet-Weather Loading Data for Problem 4

Time (h) Flow (gpm)
0.00 0
3.00 0
4.00 100
5.00 300
6.00 900
7.00 1,500
8.00 1,800
9.00 1,600
10.00 1,000
11.00 600
12.00 300
13.00 100
14.00 0

Wet Well Data for Problem 4

Ground Base Minimum Initial Alarm Maximum Wet Well
Wet Well Elevation Elevation Elevation Elevation Elevation Elevation Diameter
(ft) (ft) (ft) (ft) (ft) (ft) (ft)
WW-1 60.00 45.00 45.00 55.00 59.00 60.00 20.00

Pump Data for Problem 4

Pump Curve Point Head (ft) Flow (gpm)
Shutoff 100 0
Design 80 4,000
Max Operating 40 8,000
The pump turns on when the elevation in the wet well rises to 57.0 ft and shuts
off when the elevation drops to 45.0 ft.
To verify important information such as minimum and maximum velocities, it
will be helpful to keep a table like the following.

Variable Steady State EPS Sanitary Only EPS Dry and Wet
Maximum flow at outlet (gpm)
Time of max. flow at outlet (h) N/A
Maximum velocity in system at
this time (ft/s)

232 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Sanitary Sewer Design
Problems

Pipe with max. velocity at this

time
Minimum flow at outlet (gpm) N/A
Time of min. flow at outlet (h) N/A
Min. velocity in system at this
time (ft/s)
Pipe with min. velocity at this
time
Maximum diverted flow (gpm)
At what hour?

(a) Fill in the table for each scenario.

(b) Is there surcharging or flooding in the system? Explain the difference
between surcharging and flooding.
(c) Does the diversion divert any flow? If so, for which scenario?
(d) If the pump has a best efficiency point at 4,000 gpm, what can you
say about its performance for the dry- and wet-weather scenarios?
(e) Plot profiles from MH-1 to the outlet for all three scenarios and
compare the HGLs.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 233

 9 Low-Impact Development
and Dynamic Modeling

9.1 Introduction
It was a typical summer, but one accentuated with periods of extreme heat
and intense rainfall. One area in the city was nearing the completion of a
2-year beautification and utility project that had local residents and mer-
chants convinced would solve a history of minor, inconvenience flooding.
Unfortunately, one of the thunderstorms resulted in damaging flooding at
multiple locations. The headlines in the local paper the next morning read
“2.7 Inches of Rain in 70 Minutes: $50 Million to End Flooding.” The State
Climatologist reported the event return period was between 5 and 10 years.
Certainly, a modern urban drainage system should accommodate runoff
from such an event, particularly since local regulations require sewer systems
designed to accommodate peak flows from a 25-year event. What happened?
Why was there so much flooding?
When they designed the local drainage system, engineers used traditional meth-
ods to size the storm sewer pipes, culverts, and open channels to accommodate
estimated peak flows. They did not consider the impact of limited storage in
conduits, and the additive effects and timing of local inflows. Nonuniform
flow profiles were determined to delineate the floodplain. Otherwise, down-
stream boundary conditions and interactions among the different drainage
system components were not considered. All calculations were performed
assuming steady flows.
The crux of the problem was that the steady flow methods did not account for
the dynamic response of the watershed and its internal drainage system. No true
picture, that is, understanding of the drainage system performance was estab-
lished. The results were insufficient and potentially misleading. Consequently,
there was a lack of adequate planning and flood prevention measures.
To overcome these deficiencies, the city undertook a study to identify the
causes of flooding and possible solutions. That study included a hydraulic
analysis of the drainage system using a dynamic model. The results provided
a clear understanding of the performance of individual components and their
interactions at a system wide scale.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 235

 9 Low-Impact Development and Dynamic Modeling
Why to Use an Unsteady Flow Model?

The purposes for this chapter are to introduce dynamic modeling and demon-
strate its application to the analysis and design of hydraulic systems using the
SewerGEMS computer software. The option to include low-impact develop-
ment (LID) to reduce the overall surface water runoff due to land planning and
green infrastructure will be discussed. Example applications include stormwa-
ter conveyance systems that include green infrastructure design to size storm-
water sewer networks.

9.2 Why to Use an Unsteady Flow Model?

Although unsteady flow models are computationally more sophisticated than
steady flow models, they represent the physical processes influencing flow and
its properties better and provide a more detailed understanding of real flow
events. Also in many scenarios, LID structures (that is, green roof, rain garden,
and parking lot with porous pavement) are relatively small, could be numer-
ous in an urban design, and have dynamically different impacts on the runoff
hydrograph from a wide range of storm events. An unsteady flow model always
should be used when hydrograph attenuation and flow restrictions are con-
sidered. Storage within a reach can be significant, which leads to substantial
attenuation. Flow restrictions increase storage, thereby amplifying attenuation.
These effects have further implications at downstream confluences and struc-
tures. An unsteady flow model accounts for these storage effects and thereby
predicts total system performance more accurately.
A steady flow model projects the peak flow entering one branch of a convey-
ance system and translates it downstream, where it is added to the peak flows
from other branches. As such, the peak flow in the downstream links likely is
overpredicted. This approach is conservative, but not realistic, particularly for
input to planning for system operation and maintenance, and potential expan-
sion. In other words, an unrealistic picture of system performance can amount
to substantial time and capital improvement costs.
Typically, the steady flow approach is conservative in its determination of water
levels in that, other things being equal, steady flow analysis tends to overes-
timate flow and thus depth. This is due to the assumption flow is constant
within a reach; as a consequence, the effects of in-stream storage on the hydro-
graph shape, peak, and timing are ignored.
There are instances, however, in which a steady flow model may not be con-
servative in its estimation of water levels. For cases in which flow reversal may
occur, the designer must be aware of the implications of his or her choice of
flow modeling techniques. Examples of these situations include stream junc-
tions where flow may reverse such as the confluence of a tributary and a tidal
river where the flood tide can generate a wave as it moves upstream and the
waterway narrows.
Dynamic modeling opens a whole new realm of design alternatives and solu-
tions than were previously possible with steady flow modeling. It is more

236 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Low-Impact Development

accurate because it improves the hydraulic wave effects observed in the field
during periods when there are significant changes in loading patterns, such as
weekends and holidays when water consumption and wastewater production
patterns change, sometimes dramatically.
This discussion has identified some of the reasons and applications for unsteady
modeling. Although steady flow equations are significantly simpler and easier
to use, their solutions are accurate only for those situations in which the chan-
nel and flow conditions do not vary greatly. When faced with more complex
modeling problems, the unsteady flow equations provide far more appropriate
solutions.

9.3 Low-Impact Development

The urbanization of land has a direct impact on the volume of runoff and
the peak flow rate. Urban areas tend to reduce the amount of water that
would infiltrate into the subsurface, eliminate vegetation that would inter-
cept rainfall, and increase the velocity that the runoff travels over the devel-
oped surfaces (that is, roads, rooftops, and parking lots). This will increase
the volume of runoff, and since the runoff travel time to a discharge location
is reduced, the peak flow rate will be greatly increased. Traditional strategies
for stormwater management focused on reducing the peak flow rate from
urban runoff by storing runoff (that is, detention ponds), and channeling
the stormwater out of the urbanized areas using engineered channels and
piping systems. However, there is a growing awareness of constructing engi-
neered systems to mimic natural processes in urban areas can better manage
stormwater runoff.
LIDs are techniques to mimic the natural hydrology that existed before urban-
ization. They tend to be small-scale systems that are located near the origin of
the runoff to store stormwater and slow its travel time. This would not only
decrease the peak flow rate but also change the timing of runoff from urban
areas that utilize these strategies. A further benefit is the potential to allow
for stormwater to infiltrate then replenish groundwater systems, have more
“green” space in urban areas, and can improve the water quality leaving the
urban environment.
There are several LID techniques that can be utilized in urban design. Most LID
techniques are considered to be structural best management strategies (BMPs)
used to control and treat runoff close to the origin of the runoff. They tend to be
smaller than traditional stormwater structures and spread out throughout the
urban area. The Water Environment & Reuse Foundation maintains a database
on the effectiveness of these LID techniques (www.bmpdatabase.org). Table 9.1
compares some LID techniques to traditional stormwater management struc-
tures. These LID structures will reduce the volume of runoff, increase the time
of concentration (travel time), trap or reduce pollutants, and allow for increased
infiltration and evapotranspiration.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 237

 9 Low-Impact Development and Dynamic Modeling
Dynamic Modeling

Table 9.1: Comparison of Traditional Stormwater Structures and Low-

Impact Development Structures.

Traditional Structure Low-Impact Development Structure

Vegetated swales, infiltration trench,
Catch basins and pipes
constructed runoff filters
Planted basins, bioretention cells, rain bar-
Stormwater detention basins
rels, subsurface infiltration beds, wet ponds
Large centralized detention Several small detention basins, rain gar-
basin dens, constructed wetlands
Pervious pavement, green roofs, landscape
Impervious surfaces
restoration, soil restoration

The design objectives for LID in urban areas is to manage the peak flow rates
and maximum water level of stored water, while simultaneously protecting
the urban infrastructure (buildings, roads, and other structures), human lives,
and surrounding environment. The use of LIDs, since they tend to be smaller
and decentralized, requires the hydraulic modeler to carefully examine the pre-
dicted runoff peak flows rates, volumes, and resulting hydrographs throughout
the system design since potentially many LID structures are in use at one time.
As the smaller LID structures fill and overflow into other traditional structures
used to manage the runoff, a dynamic model with detail of how all these struc-
tures are hydraulically connected is needed.
A challenge to add LIDs to existing urban areas is to find locations to place pro-
posed LIDs. The acquisition of land to build these structures is limited in many
urbanized areas and if a site is found, it may not be optimally placed. Also, the
soil for many urban areas tends to have a low infiltration rate or additional
water in the urban subsurface could have undesirable consequences, which
would make LIDs that use infiltration impractical to implement. Regardless of
the challenges, the use of LIDs for managing stormwater is gaining popularity,
and it is common to evaluate many design alternatives when these structures
are considered. A detailed hydraulic model of the stormwater system can be
used to evaluate how LIDs should be placed and sized to manage the runoff
and improve the water quality leaving urban areas.

9.4 Dynamic Modeling

Simply stated, dynamic means change. Flow in sewers and open channels
changes in time and space. Flow that changes with time is unsteady; flow that
changes in space is nonuniform (see Chapter 1). Dynamic models account for
unsteady, nonuniform flow.
In unsteady open-channel flow, the velocity and depth at any location (cross
section) continually change due to changes in loading patterns and bound-
ary conditions. Unsteady open channel flow by nature is also nonuniform.
Consequently, dynamic models are also known as unsteady flow models.

238 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Low-Impact Development and Dynamic Modeling
Dynamic Modeling

St. Venant Equations

Flows in stormwater conveyance systems and sewers are usually free surface
open-channel flows. The governing equations are the St. Venant equations,
based on the principles for the conservation of mass and linear momentum.

∂Q ∂( A + Ao )
+ −q =0
∂x ∂t

∂Q ∂(βQ 2 / A ) ⎛ ∂y ⎞⎟
+ + gA ⎜⎜⎜ + So + S f + Si ⎟⎟ + L = 0
∂t ∂x ⎝ ∂x ⎠

where
Q = discharge (cfs, m3/s)

x = distance along the longitudinal axis of the channel (ft, m)

A = active cross-sectional area of flow (ft2, m2)
Ao = inactive (off-channel storage) cross-sectional area of flow (ft, m)
t = time (s)
q = lateral inflow or outflow (ft2/s, m2/s)
β = coefficient for nonuniform velocity distribution in cross section
g = gravitational acceleration constant (ft/s2, m/s2)
y = flow depth (m, ft)
So = slope of channel bed in longitudinal direction
Sf = friction slope
Si = slope due to severe local expansion and contraction effects
L = momentum effect of lateral flow (ft3/s2, m3/s2)

The first equation is the conservation of mass or continuity equation. It relates

the change in the cross-sectional area of flow at a point over time to the change
in flow over space.
The second equation is the conservation of momentum equation based on
Newton’s second law, which states if a net force is acting on a body, the momen-
tum of that body will change, where momentum is mass times acceleration.
This equation was derived by considering how forces applied to a control vol-
ume affect the movement of water through the control volume. For example,
if the driving forces on the control volume (due to gravity and water surface
slope) are greater than the resistive forces, the flow will accelerate. The acceler-
ation can be an increase in velocity at one point over time (local or temporal

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 239

 9 Low-Impact Development and Dynamic Modeling
Dynamic Modeling

acceleration), an increase in velocity over space (convective or spatial accelera-

tion), or both. Conversely, if the resistive forces are greater, the flow will deceler-
ate, meaning the flow slows and the depth increases, which frequently happens
upstream of undersized road cross-drains during periods of stormwater runoff.
The local and convective acceleration terms (respectively, the first two terms in
the momentum equation) are inertial terms. The other terms are the various
applied forces (pressure imbalance, gravity, friction, and expansion/conver-
gence effects), expressed as slope terms. The friction slope generally is evalu-
ated using an empirical equation in the same way as the loss term in the energy
equation when performing flow profile analyses. The Manning equation is
commonly used, but other similar equations may also be used. This loss term
encompasses not only the effects of boundary friction but also all processes
creating flow resistance, notably turbulence and internal shear.
The St. Venant equations are also known as the shallow water or dynamic
wave equations. Two assumptions implicit in the derivation of these equations
are: (a) the flow is one-dimensional and (b) the flow is gradually varied. The
first assumption means it is only necessary to consider velocities in the down-
stream direction and not in the transverse or vertical directions. This reduces
the cross-sectional properties to single parameters, that is, cross-sectional area
and average velocity. The second implies the streamlines are nearly parallel.
This means the vertical pressure distribution is essentially hydrostatic, that is,
increases linearly with depth, and the friction slope term can be approximated
with one of the uniform flow formulae.

Numerical Solution
There is no known analytical solution to the St. Venant equations; therefore
they are solved numerically. The solution method in SewerGEMS uses a finite
difference scheme that converts the system of partial differential equations to
a system of algebraic equations involving the unknowns at discrete points in
time and space.
Specifically, a weighted four-point implicit scheme is used. An implicit scheme
is preferred over explicit since these schemes have the advantage of maintaining
numerical stability for large computational time steps and exhibit robustness in
modeling systems that integrate the complex hydraulic interactions encountered
in gravity sewer systems. The scheme was chosen because it handles unequal dis-
tance steps, its stability–convergence properties can be conveniently modified,
and the internal and external boundary conditions can be easily applied.
The numerical analog uses the following four-point finite-difference quotients
to approximate the first- and zero-order derivative terms:

∂f f j +1 + fi +j +1 1 − fi j − fi +j 1
= i
∂t Δt

240 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Low-Impact Development and Dynamic Modeling
Dynamic Modeling

∂f f j +1 + fi +j +1 1 − fi j − fi +j 1
= i
∂t Δt

θ( fi j +1 + fi +j +1 1 ) + (1 − θ )( fi j + fi j+1 )
f =
2

in which f represents the unknown variables, Q and A, and θ is a weighting

factor. The weighted four-point implicit scheme is unconditionally stable for θ
> 0.5. The optimal range of θ for maintaining stability and accuracy for large
computational time steps is 0.6–0.8.
The finite difference equations are solved with a Newton–Raphson iteration
routine that includes an algorithm that iterates banded matrixes and is com-
putationally efficiency.

Initial and Boundary Conditions

Solution of the dynamic wave equations requires initial and boundary condi-
tions. Initial conditions prescribe the state of the system for the first time step
and usually are specified as dry channel or baseflow. Boundary conditions are
functional relationships between depth (or area) and flow at the upstream and
downstream ends of the reach of interest. (Note: Reach of interest indicates
the length of channel or sewer through which a hydrograph is being routed.)
Effectively, boundary conditions are constraints the numerical solution must
meet, particularly if a stable and accurate solution is to be obtained. Boundary
conditions usually consist of the inflow hydrograph and a tailwater condition
at the outlet.
The numerical solution may use one or both sets of boundary conditions,
depending on flow classification. This is explained with the concept of charac-
teristics, which are mathematical expressions that describe the speed and path
by which a physical disturbance, or its influence, propagates though the x,t
solution domain. A physical disturbance is any change in the rate and/or depth
of flow due to influences such as an inflow hydrograph and flow constriction.
There are two characteristics: forward and backward, indicated symbolically as
C+ and C− and expressed mathematically as

dx
C+ : =V −c
dt

and

dx
C− : =V −c
dt

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 241

 9 Low-Impact Development and Dynamic Modeling
Hydrograph Routing Effects

where x,t are longitudinal distance and time, v is cross-sectional average veloc-
ity, and c is the speed at which the wave effect created by the disturbance trav-
els. This wave speed is called celerity and is evaluated as

c= gD

where g is gravity and D is the hydraulic depth of water (see Chapter 1).
For flow in a rectangular channel, the Froude number is expressed in terms of
wave celerity as

V V
F= =
gD c

For subcritical flow, F < 1, meaning c > V. The backward characteristic is

nonzero and negatively sloped, which means there is a mathematical basis
for information to travel upstream. That is, the influence of a downstream
disturbance will propagate upstream and affect the flow conditions there. A
well-known example is the backwater curve that forms upstream of an under-
sized culvert. This causes an increased depth of flow at the culvert inlet as the
headwater increases to a higher state of specific energy to overcome the flow
resistance. This influence propagates upstream, forming the backwater curve.
Determining the upstream conditions depends greatly on the downstream
boundary condition. In other words, accurate solution of the St. Venant equa-
tions for subcritical flow requires properly defined upstream and downstream
boundary conditions.
For supercritical flow, F>1, meaning V>c. Both the forward and backward
characteristics are nonzero and positively sloped. Therefore all physical dis-
turbances are swept downstream. As such, determining flow conditions in the
reach of interest depends only on the upstream boundary condition.
For flows near critical depth, the location of the boundary conditions is chang-
ing and the numerical solution may become unstable. Numerical instability
when the flow is near critical depth is one of the greatest problems encountered
when modeling unsteady flow.

9.5 Hydrograph Routing Effects

The dynamic equations are solved to route a hydrograph through a link from
the upstream or inflow node to the downstream or outflow node. Routing
is a mathematical procedure to describe the movement and change in shape
of a stormwater hydrograph as it travels through a channel, storm sewer, or
pond, and of a wastewater hydrograph as it passes through a pump station,
force main, or gravity sewer. Routing involves the processes of translation and
attenuation, where translation typically is measured by the time-shift between

242 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Simplified Routing Models

the occurrence of the inflow and outflow hydrograph peaks or the centers of
mass. Attenuation is the reduction in peak flow rate resulting from the spread
of the hydrograph shape. Spreading occurs due to diffusion and dispersion
caused by varying conduit conditions that alternately accelerate and decelerate
the flow, for the cases of channel and sewer routing, and hydrograph storage
and release, for the cases of pond and pump station routing. Spreading results
in elongation of the time base, reduction of the peak flow rate, and skewing
of the hydrograph shape. This leads to the development of a “tail” to the
recession limb. Diffusion is spreading that does not involve mass (or volume)
shift, whereas dispersion involves mass transfer from the rising limb to the
recession limb.

9.6 Simplified Routing Models

The dynamic wave equations are the standard for routing hydrographs through
channels and sewers. Historically, two perceived drawbacks to their wholesale
general use have been the large data requirements and the necessity for sophis-
ticated numerical integration. Very often, based on channel conditions and
hydrograph characteristics, it is possible to make simplifying assumptions that
allow one to use approximations to these equations. When this is possible,
advantages in terms of ease of solution and reduced data requirements are often
realized, but at the cost of potential loss of solution accuracy. Currently, many
widely accepted stormwater conveyance models incorporate such approxima-
tions. Some of the approximate methods are discussed in this section.
The range of models available for evaluating unsteady flow in conveyance
systems can be classified into two categories: hydraulic and hydrologic. The
dynamic, diffusion, and kinematic wave models are hydraulic routing models.
They are based on the conservation equations for mass and momentum, or
simplified forms of these equations, and analyze both flow and hydraulic grade
through the system. They also are referred to as distributed parameter models
because they explicitly account for spatial variations in channel properties. As
such, they calculate flow and depth at several cross sections along the reach of
interest simultaneously.
Hydrologic models, such as the Muskingum and Muskingum–Cunge, are
based on continuity and an empirical relationship between channel storage and
discharge; therefore they are not as rigorous as the hydraulic models and do not
accommodate spatial variations in channel properties. They are termed lumped
parameter models in that the channel properties are assumed constant along
a reach. They calculate flow at one location—the downstream end of a reach.

Diffusion Wave
Two approximations to the dynamic wave model are the diffusion and kine-
matic wave models. The diffusion wave model that assumes the inertia terms

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 243

 9 Low-Impact Development and Dynamic Modeling
Simplified Routing Models

in the momentum equation are negligible compared with the force terms for
pressure, friction, and gravity. The diffusion wave equations are continuity and
the following reduced form of momentum.

∂y
= So − S f
∂x

For prismatic channels, the diffusion wave equations can be combined into the
single equation

∂Q ∂Q ∂Q
2
+c =D
∂t ∂x ∂x
2

where c is wave celerity and D is a hydrograph diffusion term. This equation is

in the classic form of the advective–diffusive transport equation widely used for
water quality modeling. This equivalence indicates the diffusion wave model
transports mass—the hydrograph—and accounts somewhat for the diffusion
of the hydrograph shape. This model applies to situations where the accelera-
tion effects are minimal and there are minor backwater effects.
One of the challenging features associated with flow in stormwater conveyance
systems is interchanging regimes between subcritical and supercritical flows.
This is due to the fact that an urban conveyance system may experience a wide
range of conduit slopes and it is common to have significant slope changes at
pipe junctions. A good numerical model must be able to handle the mixed
flow regimes and interchanges with robustness. For certain mixed flow applica-
tions, the dynamic routing technique using the four-point implicit numerical
scheme tends to be less numerically stable than diffusion routing. This is attrib-
uted to the fact that the diffusion wave model neglects the acceleration terms.
To take advantage of the diffusion model’s stability and to retain the accuracy
of the dynamic wave model, the solution technique in SewerGEMS employs a
numerical filter on the acceleration terms.

Kinematic Wave
The kinematic wave model neglects all dynamic terms, which includes the
derivative terms for local and convective acceleration, and pressure imbalance.
This reduces the momentum equation to

S f ≈ So

which means it can be approximated with one of the uniform flow formulae.
The kinematic wave model accounts only for hydrograph translation and not
for any diffusion or dispersion. The most common application for kinematic
wave modeling is overland flow during periods of stormwater runoff.

244 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Low-Impact Development and Dynamic Modeling
Simplified Routing Models

Muskingum Routing
The Muskingum routing method was developed for use in a flood control
study in the Muskingum River basin in Ohio, hence the origin of its name.
This method uses a lumped parameter model that treats the channel reach of
interest as a single spatial element, that is, the reach is not divided into shorter
segments, and the channel cross-sectional geometry, slope, and roughness are
assumed constant. Model input is the inflow hydrograph at the upstream end
and model output is the outflow hydrograph from the downstream end. This
method does not compute flows at intermediate locations along the reach as do
hydraulic models.
The model governing equations are the hydrologic form of continuity

dS
I−Q =
dt

and a conceptual model for reach storage

S = K [θ I + (1 − θ )Q ]

I is inflow, Q is outflow, S is reach storage, K is a timing parameter interpreted

as travel time, and θ is a weighting coefficient.
These equations are solved using a simple finite difference technique that yields
the following routing equation:

−K θ + 0.5Δt K θ + 0.5Δt K (1 − θ) − 0..5Δt

Q2 = I2 + I1 + Q1
K (1 − θ ) + 0.5Δt K (1 − θ ) + 0.5Δt K (1 − θ ) + 0.5Δt

The subscripts refer to inflow and outflow values at the beginning and end of a
computational time step, Δt.
The Muskingum method traditionally has been applied to channels where
there are recorded inflow and outflow hydrographs that are used to determine
the model parameters that give a best fit between observed and simulated
outflow hydrographs. As such, the model is calibrated to the channel. It then
is applied to other inflow hydrographs to predict the outflow hydrographs.
While the Muskingum method has been used for decades, values for the coeffi-
cients, K and θ, are not readily available or easily obtained for measured hydro-
graphs. A general rule of thumb is that K can be estimated by the travel time
through the reach and θ lies between 0.2 and 0.4. The upper limit for θ is
0.5. Values less than 0.5 define dispersing hydrograph shapes. θ equal to 0.5
defines a hydrograph that does not change shape and only exhibits translation
effects. A value of 0.0 is the practical lower limit for θ and suggests inflow has
little effect. Some users have reported values for θ from 0.4 to 0.5 for natural

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 245

 9 Low-Impact Development and Dynamic Modeling
SewerGEMS and Tutorials

streams, while others have suggested 0.1–0.3. Unfortunately, these guidelines

are imprecise and somewhat conflicting.

Muskingum–Cunge Method
To overcome the limitation with parameter estimation and extend the method
to ungaged channels, Cunge (1969) developed hydraulic-based equations to
estimate the routing parameters by matching coefficients of the advective–
diffusive transport form of the Muskingum routing model

∂Q Δx ∂Q ⎡ 1 Δx ⎤ ∂ Q
2 2
+ = ⎢Δx (1 − θ )ck − ⎥
∂t K ∂x ⎢⎣ 2 K ⎥⎦ ∂x 2

to those of the diffusion wave model.

∂Q ∂Q QR ∂ Q
2
+ cK =
∂t ∂x 2TSo ∂x 2

In these equations, cK is kinematic celerity (= (5QR / 3 A ), T is channel top

width, and QR is a reference discharge value, typically chosen as two-thirds of
the peak flow rate.
Cunge obtained the following expressions to estimate K and θ.

Δx
K =
cK

1⎡ QR ⎤
θ= ⎢1 − ⎥
2 ⎢⎣ TSo cK Δx ⎥⎦

When these equations are used to estimate the routing parameters, the method
is known as the Muskingum–Cunge routing method. Numerical experience
that has shown negative values for θ can sometimes occur but do not cause
numerical problems.
An important feature and a perceived advantage of the Muskingum–Cunge model
is that it is diffusive for θ<0.5 and offers the same advantages of the diffusion wave
model. This method is included in many contemporary stormwater models.

9.7 SewerGEMS and Tutorials

Tutorials and additional problems using the Bentley OpenFlows software
can be found at aka.bentley.com/CAiHE.supplement. If you have trouble

246 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Low-Impact Development and Dynamic Modeling
SewerGEMS and Tutorials

accessing the site, please send a message to [email protected]. The tutorials

are designed for student users to become familiar with the design and analysis
of hydraulic systems without needing to spend a lot of time inputting data.
As new features are added to the software and as the software is upgraded, the
tutorials will be updated too. There will also be additional problems added to
challenge the students and encourage further use of the software. Information
about SewerGEMS can be found in Chapter 9.

What Does SewerGEMS Do?

SewerGEMS is a fully dynamic, multiplatform program for the analysis and
design of urban stormwater conveyance systems, and wastewater collection
and pump station systems. SewerGEMS is a dynamic unsteady flow model
rather than a steady state or standard step model. It is therefore capable of
delivering results far more accurately and closer to real life than a steady
flow model. Dynamic models allow the effects of storage and backwater in
channels and sewers, and the timing of the hydrographs, to yield a true rep-
resentation of the hydraulic grade line (HGL) at any point in space and
time. It provides a fast solution for analyzing the performance of complex
conveyance systems subject to time varying loading patterns and dynamic
boundary conditions.
SewerGEMS includes several LID structures. These include a Bioretention
Cell, Green Roof, Infiltration Trench, Porous Pavement, Rain Barrels, Rain
Gardens, Rooftop Disconnection, and Vegetated Swales. These structures can
be modeled for a wide range of storm events and design scenarios.
Some of the other features of SewerGEMS include geographic information
systems (GIS) integration, results animation, and scenario management. The
GIS integration allows the user to build complete data bases using shapefiles,
coverages, and geodatabases. The results animation allows one to review and
present model results in customizable animations that offer an unparalleled
visualization of model results. Scenario management provides an easy way to
examine “what if ” scenarios within a single project without the need of adding
redundant data.
SewerGEMS’s implicit fully dynamic engine provides numerical solutions for
the toughest interconnected pipe, pond, and open-channel networks, with
stunning performance and numerical robustness. SewerGEMS will account
for storage within structures and quantify overflows should they occur. The
dynamic engine will precisely calculate flow splits at manholes with multiple
outgoing pipes and can handle complex, looped systems.
SewerGEMS provides a broad range of functions for importing and allocating
sewer loads in the form of hydrographs, patterned loads, and unit loads.
SewerGEMS can model pump stations with both constant-speed and
variable-speed pumps. It uses state-of-the-practice methods to estimate storm-
water runoff.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 247

 9 Low-Impact Development and Dynamic Modeling
Problems

How Can You Use SewerGEMS?

You can use SewerGEMS to accurately model every aspect of your stormwater
system, including rainfall, runoff, inlet capture and bypass, gravity and pres-
sure piping, detention ponds, outlet structures, open channels, and culverts.
You can use SewerGEMS for analysis and design purposes, to generate plan-
ning input, and as an education and research tool. With SewerGEMS, you can
analyze all sanitary and combined sewer system elements in one package and
have the option of performing the analyses with the SWMM algorithm or the
four-point implicit solution of the full St. Venant equations, which is faster
and more stable.
Enjoy the easy-to-use stand-alone interface, or choose to work directly inside
AutoCAD, ArcGIS, or MicroStation. As one of the four modeling environ-
ments in SewerGEMS, the ArcGIS interface let you leverage ESRI’s geoda-
tabase architecture to guarantee a single dataset for modeling and GIS. With
SewerGEMS, you can build your sewer model from scratch with simple
drag-and-drop layout tools or by leveraging existing data sources to drasti-
cally streamline the model development process. You can utilize an unlimited
number of background images (DXF, SHP, TIF, SID, BMP, JPG, and others)
to simplify model development and improve presentation. You can create cus-
tomized output tables and graphs.
You can use the scenario manager with a single model to evaluate an unlimited
number of design, operational, loading, and even network topology scenarios
for better decision-making support. You can use this feature to answer a long
list of the proverbial “what if ” questions.
You can use the result’s animation to view the real-time performance selected
portions of the system and the impact of their interaction. This feature is
important when analyzing system performance, particularly for forensic stud-
ies into why a system is not performing as expected.

9.8 Problems
The following diagram shows the stormwater conveyance system for an area
that underwent staged development from agricultural to suburban land use.
CM-1 was developed first, prior to any stormwater regulations. Following
standard practice at the time, stormwater runoff was collected in the sewer
system and diverted to an outfall into the creek at cross section of CS-2.
CM-2 was developed after stormwater regulations were in place. To control
the runoff peak from a 10-year design rainfall at or below the predevelopment
rate, a detention pond was constructed between the outlet from CM-2 and
the point where local runoff enters the sewer system at MH-3. During the
development period, CM-3 was built out for mixed land use—multifamily
residential, commercial, and a park. Runoff from CM-3 flows directly to the
existing creek.

248 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Low-Impact Development and Dynamic Modeling
Problems

Tables are included that provide data for the catchments, channel, cross sections,
conduits, manholes, and pond. The pond outflow structure is a two-stage riser
with lower stage orifice (two 6-in. diameter openings at elevation 533.25 ft MSL,
orifice coefficient = 0.60) and an upper stage rectangular, sharp crested weir
(length = 1 ft, weir coefficient = 3.10, and crest elevation at 536.75 ft MSL).
Downstream of the confluence of the storm sewer and creek, there is a road
crossing with a culvert cross-drain. Conduit CO-9 is the culvert. Note that it
connects CS-3 and CS-4. This system discharges to a free outfall.
Following build-out of CM-1, CM-2, and CM-3, the road crossing began
to flood (overtop) more frequently. Local residents noted the sewer sys-
tem between manholes MH-4 and MH-6 was surcharging with occasional
aboveground flooding during the events when the crossing overtopped. This
resulted in minor system flooding. It also was noted during these events the
water depth in the creek was above the crown of the outfall sewer. The residents
quickly concluded that the minor system flooding they were experiencing was
due to an undersized culvert at the road crossing and demanded that the city
make appropriate repairs. The city engineer analyzed the system and deter-
mined the inconvenience flooding they were experiencing was not due to an
undersized culvert, but probably to an excessive pond release rate.
Apply SewerGEMS to analyze this system. Use the default curvilinear NRCS
unit hydrograph, NRCS Type II rainfall distribution, and a 24-h rainfall depth
of 5.15 in.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 249

 9 Low-Impact Development and Dynamic Modeling
Problems

Were the citizens correct about the cause of flooding in the sewer system or was
the city engineer correct?
Were both parties wrong? Was the flooding due to other conditions?
Would it be a more appropriate solution—not necessarily cheaper in terms of
material costs—to replace one or more links of sewer with a larger pipe? If so,
which one(s)?
Hint: Run animations for a profile between the detention pond and the system
outfall.

Catchment Data for Problem 1

Catchment Area (ac) Curve Number Time of Concentration (min)
CM-1 3.30 82 15
CM-2 4.20 85 15
CM-3 15.0 78 26

Channel Data for Problem 1

Downstream
Upstream Invert
Channel Invert Elevation Length (ft)
Elevation (ft MSL)
(ft MSL)
CH-1 527.00 526.00 2,000
CH-2 526.00 524.80 2,000
CH-3 524.20 524.00 1,000

Cross-Sectional Data for Problem 1

Bottom Manning Invert Elevation
Cross Section Geometry Left Side Z Left Side Z
Width (ft) n-Value (ft MSL)
CS-1 Trapezoid 5.0 1 0 0.030 527.00
CS-2 Trapezoid 7.0 1 1 0.050 526.00
CS-3 Trapezoid 7.0 1 1 0.050 524.80
CS-4 Trapezoid 10.0 1 1 0.500 524.20

Conduit Data for Problem 1

Upstream Invert Downstream Invert
Conduit Diameter (in.) Length (ft) Manning n-Value
Elevation (ft MSL) Elevation (ft MSL)
CO-1 24 325 533.80 533.00 0.013
CO-2 24 330 533.00 532.20 0.013
(Continued )

250 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Problems

Upstream Invert Downstream Invert

Conduit Diameter (in.) Length (ft) Manning n-Value
Elevation (ft MSL) Elevation (ft MSL)
CO-3 24 278 532.20 530.90 0.013
CO-4 24 290 530.90 530.10 0.013
CO-5 24 305 530.10 529.40 0.013
CO-6 24 345 529.40 529.00 0.013
CO-7 24 490 529.00 526.00 0.013
CO-8 24 88 533.25 532.20 0.013
C0-9 18 66 524.80 524.20 0.013

Manhole Data for Problem 1

Manhole Diameter (ft) Head Loss Method Ground Elevation (ft MSL) Invert Elevation (ft MSL)
MH-1 3 HEC-22 537.00 533.80
MH-2 3 HEC-22 536.80 533.00
MH-3 3 HEC-22 535.50 532.00
MH-4 3 HEC-22 535.00 530.90
MH-5 3 HEC-22 534.00 530.10
MH-6 3 HEC-22 532.00 529.40
MH-7 3 HEC-22 531.80 529.00

Pond Data for Problem 1

Elevation (ft MSL) Surface Area (ac)
533.25 0.23
534.00 0.26
535.00 0.29
536.00 0.32
537.00 0.35

This figure shows a stormwater drainage system at an apartment complex

located on a barrier island. Runoff from the apartment buildings flows across
adjacent, contiguous paved parking lots to the stormwater collection system.
The parking lots are graded and curbed to create surface storage along the
low-lying, back edge. This design was chosen because the tidal control on the
sewer outflow potentially can surcharge the sewer to the point of surface over-
flow. As such, the low-lying portion of the parking lot becomes a surcharge
detention storage system. The stormwater inlets comprised a series of combi-
nation inlets located in sag meaning each inlet has 100% capture.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 251

 9 Low-Impact Development and Dynamic Modeling
Problems

The accompanying tables provide data about the system components and tidal
data from a nearby U.S. Geological Survey gaging station. The system has a
free outfall into a tidal creek, that is, it has a tidal boundary condition. The
outfall ground elevation is 5.00 ft MSL with invert elevation of 3.80 ft MSL.
Apply SewerGEMS to analyze this system. Use the default curvilinear NRCS
unit hydrograph, NRCS Type III rainfall distribution, and a 24-h rainfall
depth of 5.30 in.
Part of the design is setting curb height to contain any surcharge overflow. Is a
curb height of 4 in. sufficient?
Does the flow in the sewer reverse during high tide? In other words does water
flow from the tidal creek into the sewer? Or does any system surcharge result
due to a “backwater effect?”
What is the duration of any about ground surcharge during the design event?
What signage should be placed along the lower portion of the parking lot?
What would be the effect of upsizing one or more of the sewer pipes? Explain
your answer in terms of the effects of in-line storage. Does this suggest in-line
storage can be a solution at other, noncoastal locations?

252 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Problems

Tidal Curve Data for Problem 2

Hours Elevation (ft) Hours Elevation (ft) Hours Elevation (ft)
0 4.90 9 2.73 18 1.33
1 5.08 10 3.78 19 0.97
2 4.74 11 4.70 20 1.23
3 3.96 12 5.33 21 1.93
4 2.99 13 5.55 22 2.84
5 2.03 14 5.22 23 3.76
6 1.35 15 4.35 24 4.53
7 1.26 16 3.24
8 1.81 17 2.18

Catchment Data for Problem 2

Catchment Area (ac) Curve Number Time of Concentration (min)
CM-1 1.00 84 9.6
CM-2 0.67 85 6.0
CM-3 1.10 88 15.0

Channel Data for Problem 2

Upstream Downstream
Channel Invert Elevation Invert Elevation Length (ft)
(ft MSL) (ft MSL)
CH-1 4.00 3.90 500
CH-2 3.90 3.80 500

Cross-Sectional Data for Problem 2

Bottom
Cross Left Side Left Side Manning
Geometry Width Invert Elevation (ft MSL)
Section Z Z n-Value
(ft)
CS-1 Trapezoid 2.0 0.5 0.5 0.045 4.00
CS-2 Trapezoid 6.0 0.5 0.5 0.045 3.90

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 253

 9 Low-Impact Development and Dynamic Modeling
Problems

Conduit Data for Problem 2

Upstream Invert Downstream Invert Manning
Conduit Diameter (in.) Length (ft)
Elevation (ft MSL) Elevation (ft MSL) n-Value
CO-1 18 325 5.50 5.45 0.013
CO-2 18 330 5.45 5.40 0.013
C0-3 18 66 5.40 4.00 0.013

Manhole Data for Problem 2

Ground Invert
Head Loss
Manhole Diameter (ft) Elevation Elevation
Method
(ft MSL) (ft MSL)
MH-1 3 HEC-22 7.50 5.50
MH-2 3 HEC-22 7.45 5.45
MH-7 3 HEC-22 7.50 5.40

A nearby town proposes to add a new school onto its wastewater collection
system. The new school will discharge into MH-4 and connect with the system
trunk line at MH-3. From there, the wastewater flows to the local treatment
plant located at OF-1. There are concerns about system capacity, particularly
since this addition is immediately downstream of the location where a man-
ifold pump station system discharges into the trunk line. Specific concerns
include the potential for system surcharge with surface overflow at MH-3.
The figure and accompanying tables summarize pertinent system data. The
system has a free outfall at the head of the wastewater treatment plant with
ground elevation of 263.70 ft MSL and invert elevation of 260.00 ft MSL.
Use SewerGEMS to analyze this system. Perform simulations for a full diurnal
loading pattern at MH-1, WW-1, WW-2, and MH-4. Both pump stations are
duplex stations, so use two pumps at each station, operating in lead–lag status.
It is highly recommended that each station has two identical pumps. Data are
provided for two recommended pumps; evaluate system performance for both
pumps at each pump station.
Will MH-3 surcharge? If so, will it surcharge aboveground? If so, is the flood-
ing at MH-3 sensitive to which pump is used at the two pump stations?
What are the maximum and minimum pressures at PJ-2?
Local regulations stipulate the minimum pump runtime must be 2 min or
longer, the average number of pump cycles per hour should be in the range of
4–6, and wet well detention time should not exceed 30 min. Does each pump
station meet regulations?
What are the minimum and maximum HGL values at MH-3?

254 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Problems

Average Daily Wastewater Loading

Data for Problem 3
Node Flow (gpm)
WW-1 40
WW-2 50
MH-1 1,000
MH-4 1,125

Dimensionless Diurnal Wastewater Loading Curve Data for Problem 3

Hours Multiplier Hours Multiplier Hours Multiplier Hours Multiplier
1 0.49 7 0.46 13 2.50 19 1.60
2 0.46 8 0.51 14 2.20 20 1.42
3 0.44 9 0.63 15 1.68 21 0.94
4 0.42 10 0.76 16 1.41 22 0.76
5 0.40 11 1.08 17 1.34 23 0.63
6 0.43 12 1.51 18 1.39 24 0.54

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 255

 9 Low-Impact Development and Dynamic Modeling
Problems

Pump 1 Data for Problem 3

Head (ft) Flow (gpm)
86.0 0
66.0 200
44.0 300

Pump 2 Data for Problem 3

Head (ft) Flow (gpm)
78.0 0
72.0 450
42.0 800

Pump Station Data for Problem 3

Pump Lead-On Lag-On Pumps-Off Ground Cross-Sectional
Station Elevation (ft) Elevation (ft) Elevation (ft) Elevation (ft) Area (sq ft)
WW-1 196.50 197.25 195.00 202.00 35
WW-2 250.50 251.50 248.75 255.00 30

Pressure Pipe (Force Main) Data for Problem 3

Minor Loss Hazen-Williams
Pipe Diameter (in.) Length (ft)
Coefficient Coefficient
PP-1 6 860 2 130
PP-2 6 900 2 130
PP-3 6 1000 0 130
PP-4 6 800 0 130

Pressure Junction Data for Problem 3

Junction Ground Elevation (ft MSL) Node Elevation (ft MSL)
PJ-1 222.00 219.33
PJ-2 264.00 260.00

Conduit Data for Problem 3

Upstream Invert Downstream Invert Manning
Conduit Diameter (in.) Length (ft)
Elevation (ft MSL) Elevation (ft MSL) n-Value
CO-1 18 500 271.00 267.00 0.013
CO-2 18 400 267.20 263.00 0.013
(Continued )

256 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Problems

Upstream Invert Downstream Invert Manning

Conduit Diameter (in.) Length (ft)
Elevation (ft MSL) Elevation (ft MSL) n-Value
CO-3 18 700 271.00 260.00 0.013
C0-4 12 500 264.60 263.00 0.013

Manhole Data for Problem 3

Head Loss Ground Elevation Invert Elevation
Manhole Diameter (ft)
Method (ft MSL) (ft MSL)
MH-1 3 Standard 275.00 271.00
MH-2 3 HEC-22 270.00 267.00
MH-3 3 HEC-22 266.00 263.00
MH-4 3 HEC-22 268.00 264.60

The following diagram shows the layout for a stormwater conveyance system
proposed for a 400-ft segment of a local street. Runoff from the street and adja-
cent areas (indicated as catchments CM-1, CM-2, and CM-3) flows into the
gutter and enters the underground sewer system through inlets at three catch
basins. The gutter is a triangular channel formed by the intersection of a 6-in.
barrier curb and the transversely sloped (H:V = 40:1) street surface. Pertinent
data about system components are given in the tables below.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 257

 9 Low-Impact Development and Dynamic Modeling
Problems

The system has a free outfall with ground and invert elevations of 97.50 and
93.00 ft MSL, respectively.
Apply SewerGEMS to analyze the system. Use the default curvilinear NRCS
unit hydrograph, NRCS Type III rainfall distribution, and a 24-h rainfall
depth of 5.10 in. Use the HEC-22 Headloss Method to compute head losses in
the catch basins and manholes.
Will the system surcharge, that is, have HGL above the crown of the pipe, at
one or more locations if the tailwater at the free outfall is at elevation 95 ft?
If so, which pipes, manholes, and catch basins surcharge? Will the roadway
flood? If so, where?
If the system will surcharge, can the problem be resolved by upsizing one
link of pipe? If so, which link should be upsized? What is the minimum
upsized pipe diameter? What are the implications about in-line detention
storage?
Are any of the proposed catch basins undersized?

Catchment Data for Problem 4

Time of
Catchment Area (ac) Curve Number
Concentration (min)
CM-1 1.00 90 7
CM-2 1.10 92 8
CM-3 0.8 91 5

Gutter Data for Problem 4

Gutter Length (ft) Constructed Slope (ft/ft) Manning n-Value

GU-1 200 0.003 0.013
GU-2 200 0.003 0.013

Catch Basin Data for Problem 4

Ground ELEV Invert ELEV Max Inflow or
Catch Basin Length (ft) Width (ft) Inlet Type
(ft MSL) (ft MSL) Capture Efficiency
Maximum
CB-1 3.0 3.0 5.50 5.45 1.50 cfs
capacity
Maximum
CB-2 3.0 3.0 5.45 5.40 2.00 cfs
capacity
CB-3 3.0 3.0 5.40 4.00 Full capture 100%

258 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Low-Impact Development and Dynamic Modeling
Problems

Manhole Data for Problem 4

Ground Elevation Invert Elevation
Manhole Diameter (ft)
(ft MSL) (ft MSL)
MH-1 3 100.00 96.00
MH-2 3 99.50 95.50
MH-7 3 99.00 95.00

Conduit Data for Problem 4

Downstream
Upstream Invert Manning
Conduit Diameter (in.) Length (ft) Invert Elevation
Elevation (ft MSL) n-Value
(ft MSL)
CO-1 18 200 96.00 95.50 0.013
CO-2 18 200 95.50 95.00 0.013
C0-3 18 200 95.00 93.00 0.013
CO-4 18 20 97.00 96.00 0.013
CO-5 18 20 96.50 95.50 0.013
CO-65 18 20 96.00 95.00 0.013

Regulations governing stormwater ponds vary from the classic statement “the
post development peak discharge cannot exceed the pre-development peak
discharge for a given design storm event” to the 10% and no net increase
rules. Under the 10% rule, the engineer must demonstrate no increase in peak
flow over predevelopment conditions at a downstream location where the land
use change area is no more than 10% of the total contributing area at that
point. At watersheds where there is substantial hydrograph attenuation due
to in-stream storage and routing effects or the presence of a pond would delay
the timing of peak outflow from a downstream pond such that it more closely
aligned with the peak runoff from upstream areas, thereby creating a greater
peak than would have occurred naturally, this criterion may actually be used
to show no pond is required on-site. Calculations to demonstrate this clearly
must include unsteady pond and channel routing.
The no net increase rule extends the requirements for stormwater ponds beyond
peak runoff control and also requires no net increase in runoff volume. Some
interpretations also hold the outflow hydrograph shape and timing must not be
altered substantially. Here again, unsteady pond and channel routing is required.
Given a 100-ac watershed with the following land use conditions: 20 ac in for-
est (good condition) over HSG C soils; 40 ac in open space (good condition)
over HSG B soils; 35 ac in single-family residential over HSG B soils; and 5 ac

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 259

 9 Low-Impact Development and Dynamic Modeling
Problems

in commercial land use over HSG B soils. The watershed hydraulic length is
3,750 ft and the area weighted average watershed slope is 1.6%.
At the upper end of this watershed, it is proposed to develop 10 ac of the open
space land use to single-family residential land use. Runoff from this develop-
ment will be controlled by a stormwater pond (inverted quadrilateral frustum
with 3:1 side slopes and length to width ratio of 2:1), which is to be designed
according to controlling regulations. The 10-ac site currently has an average
slope of 1.8%, which after development will be reduced to 1.5%. The hydraulic
length is 1,280 ft.
The channel connecting outflow from the 10-ac development to the outlet
of the 100-ac watershed (10 ac or 10% plus 90 ac or 90%) is 3,200 ft long
with an approximate trapezoidal cross section with the following parameters:
base width = 8 ft; side slope Hor:Vert = 2.5:1.0; longitudinal slope = 0.3%;
Manning n-value = 0.065.
The design rainfall is a 24-h event with 4.80-in. accumulation. Use the NRCS
synthetic unit hydrograph.
Apply SewerGEMS to design a stormwater pond at the outlet of the 10-ac
development site such that (a) the peak flow at the outlet of the 100-ac basin
(downstream point to check the 10% rule for the 10-ac site) does not exceed the
predevelopment rate for the design storm event and (b) the peak and volume at
the outlet of the 100 ac do not violate the “no net increase rule.”

260 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

Bibliography

AASHTO Task Force on Hydrology and Hydraulics, 1991, Model Drainage

Manual, American Association of State Highway and Transportation
Officials, Washington, DC.
Almeida A.B. and E. Koelle, 1992, Fluid Transients in Pipe Networks,
International Series on Computational Engineering. Computational
Mechanics Publications.
American Society of Civil Engineers, 1963, Friction Factors in Open Channels:
Progress Report of the Task Force on Friction Factors in Open Channels of the
Committee on Hydrodynamics of the Hydraulics Division, American Society
of Civil Engineers, New York, NY.
American Society of Civil Engineers and the Water Environment Federation,
1992, Design and Construction of Urban Stormwater Management Systems,
American Society of Civil Engineers, New York, NY.
American Society of Civil Engineers and the Water Environment Federation,
2007, Gravity Sanitary Sewer Design and Construction, American Society of
Civil Engineers, New York, NY.
Babbitt, H.E. and E.R. Baumann, 1958, Sewerage and Sewage Treatment, John
Wiley & Sons, New York, NY.
Benedict, R.P., 1980, Fundamentals of Pipe Flow, John Wiley & Sons, New
York, NY.
Bodhaine, G.L., 1968, Measurement of Peak Discharge at Culverts by
Indirect Methods, In U.S. Geological Survey Techniques of Water-Resources
Investigations, U. S. Department of the Interior, Washington, DC, Book 3,
Chapter A3.
Brater, E.F. and H.W. King, 1976, Handbook of Hydraulics, McGraw-Hill,
New York, NY.
Brown, S.A., J.D. Schall, J.L. Morris, C.L. Doherty, S.M. Stein, and J.C.
Warner, 2009, Urban Drainage Design Manual Hydraulic Engineering
Circular No. 22, 3rd ed., U.S. Department of Transportation, Federal
Highway Administration, Washington, DC.
Cesario, A.L., 1995, Modeling, Analysis, and Design of Water Distribution
Systems, American Water Works Association, Denver, CO.
Chow, V.T., 1959, Open-Channel Hydraulics, McGraw-Hill, New York, NY.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 261

 Bibliography

Clark, R.M., W.M. Grayman, R.M. Males, and A.F. Hess. 1993. Modeling
Contaminant Propagation in Drinking Water Distribution Systems.
Journal of Environmental Engineering, 119(2), 349.
Cunge, J.A. 1969. On the Subject of a Flood Propagation Computation
Method (Muskingum Method). Journal of Hydraulic Research, 7(2), 205.
Debo, T.N. and A.J. Reese, 1995, Municipal Storm Water Management, CRC
Press, Boca Raton, FL.
Featherstone, R.E. and C. Nalluri, 1982, Civil Engineering Hydraulics,
Granada, New York, NY.
Federal Aviation Administration (FAA), 1970, Report A/C 150-5320-
58, Advisory Circular on Airport Drainage, U.S. Department of
Transportation, Washington, DC.
Federation, 1982, Gravity Sanitary Sewer Design and Construction, American
Society of Civil Engineers, New York, NY, ASCE Manual of Practice,
No. 60.
Frederick, R.H., V.A. Meyers, and E.P. Auciello, 1977, Five to 60-minute
Precipitation Frequency for the Eastern and Central United States, In
NOAA Technical Memorandum NEW Hydro-35. www.nws.noaa.gov/oh/
hdsc/studies/prcpfreq.html (accessed June 2004).
French, R.H., 1985, Open-Channel Hydraulics, McGraw-Hill, New York, NY.
Great Lakes-Upper Mississippi River Board of Sanitary Engineers, 1997,
Recommended Standards for Wastewater Facilities, Health Research, Inc.,
Albany, NY.
Hathaway, G.A. 1945. Design of Drainage Facilities. Transactions, American
Society of Civil Engineers, 110, 697.
Harmon, W.G. 1918. Forecasting Sewage at Toledo Under Dry Weather
Conditions. Engineering News-Record, 80.
Haestad Methods, 2000, CulvertMaster: Culvert Design & Analysis Software,
Bentley Systems.
Haestad Methods, 2000, FlowMaster: Open Channel, Pipe & Inlet Software,
Bentley Systems.
Haestad Methods, 2000, StormCAD: Storm Sewer Design & Analysis Software,
Bentley Systems.
Haestad Methods, 2001, SewerCAD: Sanitary Sewer Modeling Software,
Bentley Systems.
Haestad Methods, 2002, PondPack: Urban Hydrology & Detention Pond
Modeling Software, Bentley Systems.

262 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Bibliography

Haestad Methods, 2002, WaterGEMS: Geospatial Water Distribution Software

for GIS, Bentley Systems.
Haestad Methods, 2002, Proceedings of the Water Security Summit, Haestad
Press, Waterbury, CT, December 3–4, 2001.
Haestad Methods, 2007, HAMMER V8 XM Edition User’s Manual, Bentley
Systems.
Haestad Methods and R. Durrans, 2003, Stormwater Conveyance Modeling and
Design, Haestad Press, Waterbury, CT.
Haestad Methods, G. Dyhouse, J. Hatchett, and J. Benn, 2003, Floodplain
Modeling Using HEC-RAS, Haestad Press, Waterbury, CT.
Haestad Methods, T.M. Walski, T.E. Barnard, L.B. Merritt, E. Harold,
N. Walker, and B. Whitman, 2004, Wastewater Collection System Modeling
and Design, Haestad Press, Waterbury, CT.
Haestad Methods, T.M. Walski, D.V. Chase, D.A. Savic, W.A. Grayman,
S. Beckwith, and E. Koelle, 2003, Advanced Water Distribution Modeling
and Management, Haestad Press, Waterbury, CT.
Henderson, F.M., 1966, Open Channel Flow, Macmillan Publishing Co. Inc.,
New York, NY.
Henderson, F.M. and R.A. Wooding. 1964. Overland Flow and Groundwater
Flow from a Steady Rain of Finite Duration. Journal of Geophysical Research
69(8), 1531.
Hersfield, D.M., 1961, Rainfall Frequency Atlas of the United States for
Durations from 30 Minutes to 24 Hours and Return Periods from 1 to
100 Years, In Technical Paper 40, U.S. Department of Commerce, Weather
Bureau. www.nws.noaa.gov/oh/hdsc/studies/prcpfreq.html.
Horton, R.E. 1939. Analysis of Runoff Experiments with Varying Infiltration
Capacity. Transactions of the American Geophysical Union 20, 693-711.
Huff, F.A. and J.R. Angle, 1992, Rainfall Frequency Atlas of the Midwest, In
Bulletin 71, Illinois State Water Survey, Champaign, IL.
Hwang, N.H.C. and C.E. Hita, 1987, Hydraulic Engineering Systems,
Prentice-Hall, Englewood Cliffs, NJ.
Hydraulics Research Station, 1951, Velocity Equations for Hydraulic Design of
Pipes, Metric ed., Her Majesty’s Stationary Office, London.
Izzard, C.F. 1946. Hydraulics of Runoff from Developed Surfaces. Proceedings,
Highway Research Board 26, 129.
Jakovlev, S.V., J.A.A. Karelm, A.I. Zukov, S. Kolobanov, 1975, Kanalizacja,
Moscow: Stroizdat.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 263

 Bibliography

Johnstone, D. and W.P. Cross, 1949, Elements of Applied Hydrology, Ronald

Press, New York, NY.
Joint Task Force of the American Society of Civil Engineers and the Water
Pollution Control.
Kerby, W.S. 1959. Time of Concentration Studies. Civil Engineering 60, 174.
Kirpich, Z.P. 1940. Time of Concentration of Small Agricultural Watersheds.
Civil Engineering 10(6).
Masch, F.D., 1984, Hydrology, Washington, DC: Federal Highway
Administration, HEC No. 19.
Males R.M., W.M. Grayman, and R.M. Lark. 1988. Modeling Water
Quality in Distribution System. Journal of Water Resources Planning and
Management 114, 197-209.
Mays, L.W., 2001, Stormwater Collection Systems Design Handbook,
McGraw-Hill, New York, NY.
McKay, M. and D.S. Wilkes, 1995, Atlas of Short-Duration Precipitation
Extremes for the Northeastern United States and Canada, Northeast Regional
Climate Center Research Series Publication No. RR95-1, Ithaca, NY.
Metcalf & Eddy, 1981, Wastewater Engineering: Collection and Pumping of
Wastewater, McGraw-Hill, New York, NY.
Miller, J.F., R.H. Frederick, and R.J. Tracey, 1973, Precipitation-Frequency
Atlas of the Coterminous Western United States, Silver Spring, MD, National
Weather Service, NOAA Atlas 2, 11 vols.
National Resources Conservation Service, 2001, National Engineering
Handbook Part 630 – Hydrology, U.S. Department of Agriculture,
Washington, DC. http://www.wcc.nrcs.usda.gov/water/quality/common/
neh630/4content.html (accessed June 2004).
Normann, J.M., R.J. Houghtalen, and W.J. Johnson, 2001, Hydraulic Design of
Highway Culverts, Hydraulic Design Series No. 5, 2nd ed., Federal Highway
Administration, Washington, DC.
Pilgrim, D.H., 1987, Australian Rainfall and Runoff, The Institute of Engineers,
Barton, Australia.
Rawls, W., P. Yates, and L. Asmussen, 1976, Calibration of Selected Infiltration
Equations for the Georgia Coastal Plain, U.S. Department of Agriculture,
Agricultural Research Service, Washington, DC, Technical Report
ARS-S-113.
Roberson, J.A., and C.T. Crowe, 1990, Engineering Fluid Mechanics, 4th ed.,
Houghton Mifflin Company, Boston, MA.
Roberson, J.A., J.J. Cassidy, and H.M. Chaudhry, 1988, Hydraulic Engineering,
Houghton Mifflin Company, Boston, MA.

264 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Bibliography

Rossman, L.A. and P.F. Boulos. 1996. Numerical Methods for Modeling Water
Quality in Distribution Systems: A Comparison. Journal of Water Resources
Planning and Management 122(2), 137.
Rossman, L.A., 1993, EPANet User’s Manual (AWWA Workshop Edition), Risk
Reduction Engineering Laboratory, Office of Research and Development,
US Environmental Protection Agency, Cincinnati, OH.
Rossman, L.A., R.M. Clark, and W.M. Grayman. 1994. Modeling
Chlorine Residuals in Drinking-Water Distribution Systems. Journal of
Environmental Engineering 120(4), 803.
Rouse, H., and S. Ince. 1957. History of Hydraulics. Iowa City, Iowa. Iowa
Institute of Hydraulic Research.
Sanks, R.L., 1998, Pumping Station Design, 2nd ed., Butterworth-Heinemann,
Inc., Stoneham, MA.
Schaake, J.C. Jr., J.C. Geyer, and J.W. Knapp. 1967. Experimental Examination
of the Rational Method. Journal of the Hydraulics Division 93(HY6).
Schall, J.D., P.L. Thompson, S.M. Zerges, R.T. Kilgore, and J.L. Morris, 2012,
Hydraulic Design Of Highway Culverts, Hydraulic Design Series No. 5, 2nd
ed., Federal Highway Administration, Washington, DC.
Simon, A.L., 1976, Practical Hydraulics, John Wiley & Sons, New York, NY.
Singh, V.P., 1988, Hydrologic Systems, Volume I: Rainfall-Runoff Modeling,
Prentice Hall, Englewood Cliffs, NJ.
Singh, V.P., 1992, Elementary Hydrology, Prentice Hall, Englewood Cliffs, NJ.
Soil Conservation Service, 1969, National Engineering Handbook, Section 4,
Chapter 16, National Resources Conservation Service, Washington, DC.
www.wcc.nrcs.usda.gov/water/quality/common/neh630/4content.html
(accessed June 2004).
Soil Conservation Service, 1986, TR-55, Urban Hydrology for Small Watersheds,
Natural Resources Conservation Service, Washington, DC. http://www.
wcc.nrcs.usda.gov/water/quality/common/tr55/tr55.pdf (accessed June
2004).
Strafaci, A.M., 1998, Essential Hydraulics and Hydrology, Haestad Press,
Waterbury, CT.
Streeter, V.L. and E.B. Wylie, 1985, Fluid Mechanics, McGraw-Hill, New
York, NY.
Tchobanoglous, G., 1981, Wastewater Engineering: Collection and Pumping of
Wastewater, Metcalf & Eddy, Inc. McGraw-Hill, Inc.
Van Havern, B.P., 1986, Water Resources Measurements, American Water Works
Association, Denver, CO.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 265

 Bibliography

Wanielista, M.P., 1990, Hydrology and Water Quantity Control, John Wiley &
Sons, New York, NY.
Walski, T.M., 1984, Analysis of Water Distribution Systems, Van
Nostrand-Reinhold, New York, NY.
Walski, T.M., 1993, Water System Modeling Using CYBERNET®, Haestad
Methods, Waterbury, CT.
Williams, G.B. 1922. Flood Discharges and the Dimensions of Spillways in
India. Engineering (London) 134, 321.
Wylie, E.B. and V.L. Streeter, 1993, Fluid Transients in Systems, Prentice-Hall.
Zipparro, V.J. and H. Hasen, 1993, Davis’ Handbook of Applied Hydraulics,
McGraw-Hill, New York, NY.

266 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 A Bentley Systems Software

A.1 Software Products

Purchasing this book entitles the reader to use student versions of several
Bentley OpenFlows products. These provide all software features available in
the professional version. However, the licenses for programs incorporating
networked hydraulic systems (i.e., StormCAD, WaterGEMS, HAMMER,
SewerGEMS, SewerCAD, and PondPack) limit the number of network ele-
ments that may be included in a single project. Further, programs with aca-
demic licenses may be used only for academic purposes such as completing the
exercises in this book. Professional or commercial use is prohibited.
Table A 1 describes the capabilities of the included licenses associated with the
book, as well as the additional options that are available with the purchase of a
professional license or upgraded academic license. (Academic license upgrades
are available only to academic institutions.) For more information on these
packages, visit www.bentley.com.
The software provided is intended to be used with the standalone version of the
product. If the user wants to use the AutoCAD or ArcGIS platform versions, it
is the user’s responsibility to install and integrate those products.

A.2 Software Installation and Licensing

Readers of this book can access the OpenFlows software several ways:
1. If they own commercial versions of the software through their employ-
ment or other sources, those versions can be used, and there is no need to
install student versions.
2. If their school owns academic licenses of the software, there is no need to
install additional software.
3. If they don’t have access to either source, they can install and license their
software using the steps described in the “Downloading and Licensing”
in the article: aka.bentley.com/CAiHE.supplement.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 267

 A Bentley Systems Software
Software Installation and Licensing

There is a need to supply a unique registration code to install the software,

and obtaining the code varies depending on whether a paper or eBook is
being used.
a. Paper book—the unique registration code is included on an inserted
sheet in the book. If you do not have the insert, please contact bipress@
bentley.com
b. eBook—contact [email protected] for a registration code

With this code, it is possible to download a license file and install the software
according to the above article.

Table A.1: Software access provided with this book

Program Description Academic License Professional License

Options
FlowMaster Hydraulic solver for pressure All features present Same features as academic
pipes, open channels, weirs, version but licensed for
orifices and inlet structures professional use
Culvert Design and analysis tool for All features present Same features as academic
Master culvert hydraulics version but licensed for
professional use
StormCAD Storm sewer design and 10 inlets; Stand-Alone inter- Additional inlets;
modeling package face AutoCAD integration
WaterGEMS Water distribution design 25 pipe segments; Stand- Additional pipe
and modeling package Alone, segments
AutoCAD, ArcGIS, and
MicroStation interfaces
SewerCAD Sanitary sewer design and 25 pipe segments; Stand- Additional pipe segments;
modeling package Alone interface AutoCAD integration
PondPack Stormwater detention limited to 1 pond Additional detention
design package basins; interconnected pond
modeling
Sewer Dynamic sanitary sewer and 25 links; Stand- Alone, Auto- Additional link
GEMS stormwater analysis CAD, ArcGIS, and segments
Micro-Station interfaces
HAMMER Hydraulic transient analysis 5 pipe segments; Additional pipe
software Stand-Alone, AutoCAD, segments
ArcGIS and MicroStation
interfaces

268 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Bentley Systems Software
General Tips and Common Tools

Installation for the MicroStation, AutoCAD,

or ArcGIS Platform
It is recommended that beginners use the stand-alone versions of the prod-
ucts. Some of the products are available with MicroStation, AutoCAD, and/or
ArcGIS integration. (The user must already have one of these programs to use
the integrated version. Currently AutoCAD integration is available only with
WaterGEMS, SewerGEMS and HAMMER).
After completing the installation instructions previously described, you should
see shortcuts for applicable integrated versions on your Desktop and in the
product’s Start menu folder. For example “WaterGEMS for AutoCAD 2020”.
If you do not see the shortcut, check the Readme.pdf file included with the
installation to check if you have a compatible version of the integrated appli-
cation. If the integrated application (such as AutoCAD) was installed after
the Bentley software, you will need to run the Integrator tool, located under
Start Menu

A.3. General Tips and Common Tools

Bentley OpenFlows’ software programs share many common features and
interface tools. This section summarizes the basic characteristics of some of
these shared features. You are encouraged to “play” with the software to become
more familiar with them. These features are quite intuitive and can help to
increase the efficiency of your water resources modeling activities.

Online Help
Electronic documentation can be kept up-to-date more easily than paper doc-
umentation. It is accessible from within the program, and it is environmen-
tally friendly. For these reasons, Bentley OpenFlows’ programs have extensive
online Help systems to guide the user through the analysis and design process.
The Help system is used most often to:
• Obtain context-sensitive information about an area of the program
• Search for topics such as background theory and equations
• Use “How Do I” to get detailed step-by-step instructions for completing
a task

Context-Sensitive Help
The quickest way to access the Help topic related to your current task (i.e.,
context-sensitive help) is to press the F1 key, click a Help button, or right-click
the mouse and select Help from the pop-up menu. Context-sensitive help pro-
vides instant access to the help topics related to the current dialog.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 269

 A Bentley Systems Software
General Tips and Common Tools

How Do I?
The How Do I? section of the Help system has step-by-step instructions for
common applications of the model. You may find it beneficial to read over
these topics before using the software or at least to familiarize yourself with
them. When a related question arises, you can quickly find the answer.

Pop-ups and Jumps

Certain words within many Help topics are underlined and displayed in a dif-
ferent color. These words are known as pop-ups and jumps. When you select
a pop-up (dotted underline), a definition of the word or phrase appears on the
screen. To close the pop-up window, click anywhere outside the pop-up dialog.
When you select a jump (solid underline), a related Help topic is displayed.

Graphical Editor
Network models such as StormCAD, WaterGEMS, SewerCAD, and
PondPack have a graphical editor to aid the engineer in laying out the system.
The Bentley OpenFlows graphical editor is an “intelligent” drafting environ-
ment, meaning that it recognizes the characteristics and behavior of different
modeling elements and maintains connectivity when the user drags and drops
elements on-screen.
This intelligent graphical editor prevents problems such as incorrect pipe con-
nections and “hanging” pipes (pipes with a node defined at only one end).
Models that do not have this type of behavior can be easily “broken,” and it
may take hours (or days) to locate the source of the problem.

DXF Files
A drawing exchange file (DXF) is the standard format used for translating from
one CAD system, such as AutoCAD®, to another, such as MicroStation™.
DXF files can also be imported into StormCAD, WaterGEMS, and SewerCAD
for use as background drawings. For example, a roadway network plan can
be imported and used as a background map for laying out and designing a
proposed storm sewer system, sanitary sewer system, or water distribution
network.
If desired, drawing elements from DXF files can be automatically converted
to network components using the polyline-to-pipe import feature. DXF lines
and polylines can be converted to pipes, and DXF blocks can be converted to
node elements.
To learn more about importing and using DXF backgrounds, look in the
online help system of StormCAD, WaterGEMS, or SewerCAD. You can find
topics in the How Do I? section or perform a search with a keyword such as
“DXF” or “import.”

270 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Bentley Systems Software
General Tips and Common Tools

Scaled or Schematic?
In StormCAD, WaterGEMS, and SewerCAD, you may choose to either create
a schematic drawing or define the horizontal and vertical distance scale. The
scaled mode is typically used when a DXF or CAD background file is present.
The scale is set so that one drawing unit is equal to a unit of measure (e.g., 1
DXF unit equals 1 foot). In a scaled plan, you can also set the drawing scale
that you want to use when printing or plotting.
A schematic plan is not drawn at any defined scale, so there is no connection
between dimensions shown on the plan and real-world dimensions. There may
not even be a correlation between individual dimensions shown on a schematic
plan. For example, in a schematic plan, the length representing a pipe with a
length of 50 m may be less than, equal to, or greater than, the line representing
a pipe with a length of 20 m.
To get to the drawing scale options, choose Options from the Tools menu and
go to the Drawing tab. The available options are as follows:
• Schematic—You enter the pipe length in the Pipe Properties dialog for
each pipe. Absolute positioning in the drawing editor is not used.
• Scaled—You enter the drawing unit scale to be used to determine the
length of the drawn pipes. The pipe length will be calculated based on the
starting and ending coordinates. The length in the drawing editor corre-
sponds to a real-world length. You can also enter the horizontal and vertical
drawing scales at which output should be plotted.

Color Coding
One way to determine the performance of a system is to color code the sys-
tem according to some indicator. For hydraulic design work, pipeline veloc-
ity is often a useful indicator. For example, pipes consistently flowing below
0.15 m/s may be oversized. Pipes with velocities over 1.5 m/s are fairly heavily
stressed, and those above 2.4 m/s are often bottlenecks in the system.

Table Manager and Table Customization

StormCAD, WaterGEMS, SewerCAD, and FlowMaster have tabular data sheets
called FlexTables. These tables aid the modeler in entering and editing data, as
well as in organizing and presenting the results of the model. FlexTables have
several built-in features that are discussed in detail in the online help system. For
information about customizing FlexTables, search the online help for topics such
as “table” or “customize.” The following are a few of the topics related to table
preferences.
• Changing the Table Title—When you choose to print a table, the table
name is used as the title in the printed report. You can change the report
title by using the Table Manager to rename the table.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 271

 A Bentley Systems Software
General Tips and Common Tools

• Adding and Removing Columns—You can add and remove column

headings by using the table setup options from the table manager.
• Drag-and-Drop Column and Row Placement—Select the column or
row that you would like to move by holding down the left mouse button
on its heading and simply dragging it to the desired location.
• Resizing Columns—In the column heading section of the table, place
your mouse over the vertical separator between columns. The cursor shape
changes to indicate that you are over the separator. Hold down the left
mouse button and drag the mouse to the left or right to set the new column
width.
• Changing Column Properties—The display units for various data types
can be changed in the tables. To view or edit the properties of any numeric
column, right-click the heading area of the column and choose Properties
from the pop-up menu. The unit properties will be displayed in the Set
Field Options dialog.
• Changing Column Labels—To change the label of any column,
right-click the column heading and choose Edit Column Label from the
context menu.
• Using Local Units—Local units allow data columns to have fixed units
and display precision, regardless of the current unit settings of the model.
This option can be used to create standardized reports or to present the
same data side by side in different units (especially handy for projects that
use both English and SI units).
• In addition to changing the previous table properties, you can also per-
form data management operations such as sorting, filtering, and global
editing.
• Sorting—A table can be sorted in ascending or descending order according
to any variable. Right-click on the column heading, select Sort, and then
select Ascending or Descending. In this way, the elements can be quickly
reorganized based on the alphanumeric labels, pipe flow rates, pipe diame-
ters, hydraulic grade lines, and so on.
• Filtering—Occasions often arise when a modeler wants to view only a sub-
set of the entire system. For example, the modeler may be concerned with
all pipes carrying more than 3 m3/s in a storm or sanitary sewer, or there
may be problems with junction nodes having a pressure below 150 kPa in
a water distribution network. The ability to show only some of a system’s
elements is called filtering, and it is accomplished by specifying the desired
filter criteria in the table’s Filter dialog box. Each element in the filter list
requires three items:
• Column—The table column containing the element property by which
to filter.

272 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Bentley Systems Software
General Tips and Common Tools

• Operator—The operator used to compare the specified filter value to the

values in the specified column. Operators include: =, >, ≥, <, ≤, and ≠.
• Value—The comparison value for the filter.
Any number of criteria can be added to a filter. Multiple criteria are implicitly
joined with a logical AND statement. So, when multiple criteria are defined,
only rows that meet all of the specified criteria will be displayed.
The status pane at the bottom of the Tabular window shows the number of
elements displayed and the total number of rows available (e.g., 10 of 20 ele-
ments displayed). When a filter is active, the element labels will turn blue.
Table filtering allows you to perform global editing (see next bullet) on a subset
of elements that you want to change.
• Global Editing—Factors and constants can be applied to every value of
any editable variable within a FlexTable. Global editing operations include
multiplication, division, addition, subtraction, or directly setting all values,
depending on the modeler’s needs. Only the elements displayed in the table
will be changed, so global editing is often used in conjunction with filter-
ing to obtain the desired results. To access global editing, right-click the
column heading of the column that you want to edit and select the Global
Edit option from the pop-up menu.

FlexUnits
You can set field options—the properties of a numeric value—from almost
any field in any of Bentley OpenFlows’ programs. These properties include the
units, decimal precision, scientific notation display option, and the allowable
range of values. This feature is referred to as FlexUnits.
To set these options for a particular field type, simply right-click the field and
select Properties from the context menu. This action opens the Set Field
Properties dialog where you can select the options that you want.
If you would like to set up the units for many data types from the same loca-
tion in the program, choose FlexUnits from the Tools pull-down menu to
open the FlexUnits dialog.

Units
FlexUnit offers a wide variety of possible units for any field. In the Set Field
Options dialog, click the Units drop-down menu and select the desired unit
from the list. The value in the field will be automatically converted to the new
unit using the appropriate conversion factor. The FlexUnit feature allows you
to mix and match any units, even if you are using a combination of English
and SI units.
Once you begin changing units, you will probably notice that some units have
multiple representations (e.g., psi and lbs/in2, cfs and ft3/s). Computationally,

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 273

 A Bentley Systems Software
General Tips and Common Tools

these units are the same; however, the additional choices allow you to control
the way the units are presented on-screen and in program output.

Display Precision
Display precision can be used to control the number of digits displayed after
a decimal point or the rounding of a numeric value. When the precision is
greater than or equal to zero, it specifies the number of digits that are displayed
after the decimal point. For example, π (3.14159265…) with a decimal preci-
sion of 4 would be presented as 3.1416.
Specifying a negative number for display precision results in a displayed value
that is rounded to the nearest power of 10. −1 rounds to 10, −2 rounds to 100,
−3 rounds to 1000, and so on. For example, the number 1,234 with a display
precision of −2 would be displayed as 1,200.
Note: The display precision affects only the way the numbers are displayed, not
their actual values. The internal values stored by the software are still carried
out to their maximum decimal precision—they are just displayed differently.
This concept is important concept to keep in mind, especially when check-
ing calculations by hand or working with values that are lower than the usual
decimal precision. For example, be aware that a 0.75-in diameter pipe with a
display precision of 0 will result in a displayed value of 1 in, even though the
calculations will be performed based on the true 0.75-in diameter.

Scientific Notation
Scientific notation displays a value as a real number multiplied by some power
of 10. It is displayed as an integer or real value followed by the letter “e” and a
positive or negative integer.
For example, 12,345 could be written in scientific notation as 1.2345 e4, and
0.12345 could be written as 1.2345 e–1.
Scientific notation follows the same display precision rules outlined previously.
To turn scientific notation on or off, just check or uncheck the box labeled
Scientific notation.

Minimum and Maximum Allowable Values

These options are available for selected input data and control the range of
input values that the program will accept. For example, the user may want
to set minimum or maximum values for roughness coefficients, slope, and so
forth in order to limit these values to a realistic range, thus helping to prevent
typographic errors during data entry.

274 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 B Scenario Management

B.1 Overview
The Scenario Manager is included in many of the Bentley Systems software. This
feature can dramatically increase your productivity in the “What If?” areas of
modeling, including calibration, operations analysis, and planning. By invest-
ing a little time to understand scenario management, you can avoid unneces-
sary editing and data duplication. Tutorial 2 with Chapter 3 demonstrates how
to manage and compare scenarios using SewerCAD and SewerGEMS, which
can also be used for WaterGEMS, HAMMER, and PondPack.
In contrast to the old methods of scenario management (editing or copying
data), automated scenario management using inheritance gives you these
advantages:
• A single project file makes it possible to generate an unlimited number
of “What If?” conditions without becoming overwhelmed with numerous
modeling files and separate results.
• Because the software maintains the data for all the scenarios in a single pro-
ject, it can provide you with powerful automated tools for directly compar-
ing scenario results. Any set of results is immediately available at any time.
• The Scenario/Alternative relationship empowers you to mix and match
groups of data from existing scenarios without having to declare any data
again.
• With inheritance, you do not have to re-enter data if it remains unchanged
in a new alternative or scenario, avoiding redundant copies of the same
data. Inheritance also enables you to correct a data input error in a par-
ent scenario and automatically update the corrected attribute in all child
scenarios.
These advantages, while obvious, may not seem compelling for small projects.
It is as projects grow to hundreds or thousands of network elements that the
advantages of true scenario inheritance become clear. On a large project, being
able to maintain a collection of base and modified alternatives accurately and

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 275

 B Scenario Management
Developing Scenarios

efficiently can be the difference between evaluating optional improvements

and being forced to ignore them.

B.2 About This Appendix

The depth of scenario management as implemented by Bentley Systems is
probably far beyond what you have ever seen before. With that in mind, this
appendix is intended as an introduction to the philosophy and terminology
upon which scenario management is based.
This appendix is not intended as a step-by-step guide to using the software.
If you are a moderately experienced Windows software user, you should have
no difficulty learning and exploring the scenario management interface.
In addition to the tutorials with this book, excellent tutorials and
context-sensitive online help are also available within the software itself. These
learning tools will help you with all aspects of the software, and should cer-
tainly not be ignored if you are having difficulty. For more information, just
click the Help button, which is available from anywhere within the program.
In addition, you can contact Bentley Systems or visit www.bentley.com for the
schedule of training seminars held around the world.

B.3 Developing Scenarios

Let us begin by understanding the approaches that have historically been used
to attempt “What If?” analyses. Traditionally, there have only been two possi-
ble ways of analyzing the effects of change on a software model:
• Change the model, recalculate, and review the results
• Create a copy of the model, edit that copy, calculate, and review the results
Although either of these methods may be adequate for a relatively small system,
the data duplication, editing, and re-editing becomes very time-consuming
and error-prone as the size of the system — and the number of possible con-
ditions — increase. Additionally, comparing conditions requires manual data
manipulation, because all output must be stored in physically separate data
files.
The scenario-management feature developed by Bentley Systems logically
organizes scenarios to meet the following objectives:
• Minimize the number of project files the modeler needs to maintain (one,
ideally).
• Maximize the usefulness of scenarios through easy access to things such as
input and output data, and direct comparisons.

276 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
The Scenario Cycle

• Maximize the number of scenarios you can simulate by mixing and match-
ing data from existing scenarios (data reuse).
• Minimize the amount of data that needs to be duplicated to consider con-
ditions that have a lot in common.
A single project file enables you to generate an unlimited number of “What If?”
conditions, edit only the data that needs to be changed, and quickly generate
direct comparisons of input and results for desired scenarios.

B.4 The Scenario Cycle

The process of working with scenarios is similar to the process of manually
copying and editing data, but without the disadvantages of data duplication
and troublesome file management. This process allows you to cycle through
any number of changes to the model, without fear of overwriting critical data
or duplicating important information (see Figure B.1). Of course, it is possible
to directly change data for any scenario, but an “audit trail” of scenarios can be
useful for retracing the steps of a calibration series or for understanding a group
of master plan updates.

Figure B.1: Self-Contained Scenarios

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 277

 B Scenario Management
The Scenario Cycle

Scenario Anatomy: Attributes and Alternatives

Before exploring scenario management further, a few key terms should be
defined:
• An attribute is a fundamental property of an object, often a single numeric
quantity. For example, the attributes of a pipe include diameter, length,
and roughness.
• An alternative holds a family of related attributes so pieces of data that you
are most likely to change together are grouped for easy referencing and edit-
ing. For example, a physical properties alternative groups physical data for
the network’s elements, such as elevations, sizes, and roughness coefficients.
• A scenario has a list of referenced alternatives (which hold the attributes),
and combines these alternatives to form an overall set of system conditions
that can be analyzed. This referencing of alternatives enables you to easily
generate system conditions that mix and match groups of data that have
been previously created. Note that scenarios do not actually hold any attrib-
ute data — the referenced alternatives do.

A Familiar Parallel
Although the structure of scenarios may seem a bit difficult at first, anyone
who has eaten at a restaurant should be able to relate fairly easily. A meal
(scenario) is comprised of several courses (alternatives), which might include a
salad, an entrée, and a dessert. Each course has its own attributes. For example,
the entrée may have a meat, a vegetable, and a starch. Examining the choices,
we could present a menu as in Figure B.2.
The restaurant does not have to create a new recipe for every possible meal
(combination of courses) that could be ordered. They can just assemble any
meal based on what the customer orders for each alternative course. Salad 1,
Entrée 1, and Dessert 2 might then be combined to define a complete meal.

Figure B.2: A Restaurant Meal “Scenario”

278 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
Scenario Behavior: Inheritance

Generalizing this concept, we see in Figure B.3 that any scenario simply refer-
ences one alternative from each category to create a “big picture” that can be
analyzed. Note that different types of alternatives may have different numbers
and types of attributes, and any category can have an unlimited number of
alternatives to choose from.

Figure B.3: Generic Scenario Anatomy

B.5 Scenario Behavior: Inheritance

The separation of scenarios into distinct alternatives (groups of data) meets
one of the basic goals of scenario management: maximizing the number of
scenarios you can develop by mixing and matching existing alternatives. Two
other primary goals have also been addressed: a single project file is used, and
easy access to input data and calculated results is provided in numerous formats
through the intuitive graphical interface.
But what about the other objective: minimizing the amount of data that needs to
be duplicated to consider conditions that have a lot of common input? Surely an
entire set of pipe diameters should not be re-entered if only one or two change?
The solution is a familiar concept to most people: inheritance.
In the natural world, a child inherits characteristics from a parent. This may
include such traits as eye color, hair color, and bone structure. There are two
significant differences between the genetic inheritance that most of us know
and the way inheritance is implemented in software:
• Overriding inheritance
• Dynamic inheritance

Overriding Inheritance
Overriding inheritance is the software equivalent of cosmetics. A child can
override inherited characteristics at any time by specifying a new value for that

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 279

 B Scenario Management
Scenario Behavior: Inheritance

characteristic. These overriding values do not affect the parent, and are there-
fore considered “local” to the child. Local values can also be removed at any
time, reverting the characteristic to its inherited state. The child has no choice
in the value of his inherited attributes, only in local attributes.
For example, suppose a child has inherited the attribute of blue eyes from his
parent. Now the child puts on a pair of greentinted contact lenses to hide his
natural eye color. When the contact lenses are on, we say his natural eye color
is “overridden” locally, and his eye color is green. When the child removes
the tinted lenses, his eye color instantly reverts to blue, as inherited from his
parent.

Dynamic Inheritance
Dynamic inheritance does not have a parallel in the genetic world. When
a parent’s characteristic is changed, existing children also reflect the change.
Using the eye-color example, this would be the equivalent of the parent chang-
ing eye color from blue to brown, and the children’s eyes instantly inheriting
the brown color also. Of course, if the child has already overridden a character-
istic locally, as with the green lenses, his eyes will remain green until the lenses
are removed. At this point, his eye color will revert to the inherited color, now
brown.
This dynamic inheritance has remarkable benefits for applying wide-scale
changes to a model, fixing an error, and so on. If rippling changes are not
desired, the child can override all of the parent’s values, or a copy of the parent
can be made instead of a child.

When Are Values Local, and When Are They

Inherited?
Any changes that are made to the model belong to the currently active scenario
and the alternatives that it references. If the alternatives happen to have chil-
dren, those children will also inherit the changes unless they have specifically
overridden that attribute. Figure B.4 demonstrates the effects of a change to a
mid-level alternative. Inherited values are shown as gray text, local values are
shown as black text.

Figure B.4: A Mid-level Hierarchy Alternative Change

280 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
Scenario Behavior: Inheritance

Minimizing Effort through Attribute Inheritance

About Scenario Management
Inheritance has an application every time you hear the phrase “just like x
except for y”. Rather than specifying all of the data from x again to form this
new condition, we can simply create a child from x and change y appropriately.
Now we have both conditions, with no duplicated effort.
We can even apply this inheritance to our restaurant analogy as follows.
Inherited values are shown as gray text, local values are shown as black text.

• “Salad 2 is just like Salad 1, except for the dressing.”

• “Salad 3 is just like Salad 1, except for the dressing.”
Note: Salad 3 could inherit from Salad 2, if we prefer: “Salad 3 is just like Salad
2, except for the dressing.”

• “Entrée 2 is just like Entrée 1, except for the meat and the starch.”
• “Entrée 3 is just like Entrée 2, except for the meat.”
Note: If the vegetable of the day changes (say from green beans to peas), only
Entrée 1 needs to be updated, and the other entrées will automatically inherit
the vegetable attribute of “Peas” instead of “Green Beans”.

• “Dessert 2 is just like Dessert 1, except for the topping.”

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 281

 B Scenario Management
A Water Distribution Example

Note: Dessert 3 is a parent alternative and has nothing in common with the
other desserts, so it can be created as a “root” or “base” alternative. It does not
inherit its attribute data from any other alternative.

Minimizing Effort through Scenario Inheritance

Just as a child alternative can inherit attributes from its parent, a child scenario
can inherit which alternatives it references from its parent. This is essentially
still the phrase “just like x except for y”, but on a larger scale.
Carrying through on the meal example, consider a situation where you go out
to dinner with three friends. The first friend places his order, and the second
friend orders the same thing except for the dessert. The third friend orders
something totally different, and you order the same meal as hers except for
the salad.
The four meal “scenarios” could then be presented as follows (inherited values
are shown as gray text, local values are shown as black text):

• “Meal 2 is just like Meal 1, except for the dessert.” The salad and entrée
alternatives are inherited from Meal 1.
• “Meal 3 is nothing like Meal 1 or Meal 2.” A totally new “base” or “root”
is created.
• “Meal 4 is just like Meal 3, except for the salad.” The entrée and dessert
alternatives are inherited from Meal 3.

B.6 A Water Distribution Example

This section presents a fairly simple water distribution system: a single reservoir
supplies water by gravity to three junction nodes (see Figure B.5).
Although true water distribution scenarios include such alternative categories
as initial settings, operational controls, water quality, and fire flow, we are going
to focus on the two most commonly changed sets of alternatives: demands and
physical properties. Within these alternatives, we are going to concentrate on
junction baseline demands and pipe diameters.

282 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
A Water Distribution Example

Figure B.5: Example Water Distribution System

Building the Model (Average Day Conditions)

During model construction, probably only one alternative from each category
is going to be considered. This model is built with average demand calcula-
tions and preliminary pipe diameter estimates. At this point we can name our
scenario and alternatives, and the hierarchies look like the following (showing
only the items of interest):

Analyzing Different Demands

(Maximum Day Conditions)
In our example, the local planning board also requires analysis of maximum
day demands, so a new demand alternative is required. No variation in demand
is expected at J-2, which is an industrial site. As a result, the new demand
alternative can inherit J-2’s demand from “Average Day” while the other two
demands are overridden.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 283

 B Scenario Management
A Water Distribution Example

Now we can create a child scenario from “Average Day” that inherits the phys-
ical alternative, but overrides the selected demand alternative. As a result, we
get the following scenario hierarchy:

Because no physical data (pipe diameters) have been changed, the physical
alternative hierarchy remains the same as before.

Another Set of Demands (Peak Hour Conditions)

Based on pressure requirements, the system is adequate to supply maximum
day demands. Another local regulation requires analysis of peak hour demands,
with slightly lower allowable pressures. Because the peak hour demands also
share the industrial load from the “Average Day” condition, “Peak Hour” can
be inherited from “Average Day.” In this instance, “Peak Hour” could inherit
just as easily from “Maximum Day.”

Another scenario is also created to reference these new demands, as shown here:

Note again that the physical data was not changed, so the physical alternatives
remain the same.

Correcting an Error
This analysis results in acceptable pressures, until it is discovered that the
industrial demand is not actually 500 gpm — it is 1,500 gpm! Because of the
inheritance within the demand alternatives, however, only the “Average Day”

284 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
A Water Distribution Example

demand for J-2 needs to be updated. The changes will ripple through to the
children. After the single change is made, the demand hierarchy is as follows:

Notice that no changes need to be made to the scenarios to reflect these cor-
rections. The three scenarios can now be calculated as a batch to update the
results.
When these results are reviewed, it is determined that the system does not have
the ability to adequately supply the system as it was originally thought. The
pressure at J-2 is too low under peak hour demand conditions.

Analyzing Improvement Suggestions

To counter the head loss from the increased demand load, two possible
improvements are suggested:
• A much larger diameter is proposed for P-1 (the pipe from the reservoir).
This physical alternative is created as a child of the “Preliminary Pipes”
alternative, inheriting all the diameters except P-1’s, which is overridden.
• Slightly larger diameters are proposed for all pipes. Because there are no
commonalities between this recommendation and either of the other phys-
ical alternatives, this can be created as a base (root) alternative.
These changes are then incorporated to arrive at the following hierarchies:

This time, the demand alternative hierarchy remains the same since no
demands were changed. The two new scenarios (“Peak, Big P-1,” “Peak, All Big
Pipes”) can be batch run to provide results for these proposed improvements.

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 285

 B Scenario Management
A Water Distribution Example

Now features like Scenario Comparison Annotation (from the Scenario

Manager) and comparison Graphs (for extended period simulations, from
the element editor dialogs) can be used to directly determine which proposal
results in the most improved pressures.

Finalizing the Project

It is decided that enlarging P-1 is the optimum solution, so new scenarios
are created to check the results for average day and maximum day demands.
Notice that this step does not require handling any new data. All of the infor-
mation we want to model is present in the alternatives we already have!

Also note that it would be equally effective in this case to inherit the “Avg.
Day, Big P-1” scenario from “Avg. Day” (changing the physical alternative) or
to inherit from “Peak, Big P-1” (changing the demand alternative). Likewise,
“Max. Day, Big P-1” could inherit from either “Max. Day” or “Peak, Big P-1.”
Neither the demand nor physical alternative hierarchies were changed in order
to run the last set of scenarios, so they remain as they were.

286 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 Scenario Management
A Water Distribution Example

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 287

Index

A Colebrook–White equation. See D

Aboveground pond, 129 Darcy–Weisbach equation Dam embankments, 132
Abstractions, 39, 40 Combination inlets on grade, 87–88 Darcy–Weisbach equation, 11–12,
depression storage, 39 Combined sewers, 211 13, 24, 164
infiltration, 39–40 Composite stage vs. discharge Darwin Designer, 177–178
initial rainfall, 45 relationships, 142 Demand patterns, 162–163
interception, 39 structures in parallel, 142–145 Demands, 170
Adverse slope, 93 structures in series, 145 Depression storage, 39
Affinity laws, 167 Computer applications, 23–24 Design storm hyetograph from SCS
Air chamber. See Conic method, 137–139 distributions, 35–38
Hydropneumatic tank Conservation Detention basins, 127
Air inlet valves, 205 of energy, 171 Detention pond design
Allowable spread length, 81 of mass, 170 basic design considerations,
Antecedent moisture condition (AMC of mass equation, 174 127–130
condition), 48 Constant horsepower pumps, 168 components of detention facilities,
Area, 2–3 Continuity equation, 198 131–133
Attenuation, 130 Continuous pattern, 163, 214 modeling concepts, 130–131
Automated optimization, 176 Contracted weir, 20 PondPack and tutorials, 155–156
model calibration, 176–177 Contributing area, 215 postdevelopment criteria, 128–129
system design, 177–178 Control valves, 169 predevelopment criteria, 128–129
Average-end-area method, 134–136 Conventional method, 164 recurrence frequency, 128–129
Convex routing method, 220–221 storage and hydraulic relationships,
Cover considerations, 42–43 133–145
B Crest, 19 storage indication method,
Babbitt method, 217 Critical depth, 15–17, 98 145–149
Backwater analysis, 93 Critical slope, 92 stormwater detention analysis
Base flow, 65 Crown elevation, 98 procedure, 127, 149–155
Best management strategies Culvert hydraulics, 113 types and configurations of
(BMPs), 237 CulvertMaster and tutorials, stormwater detention facilities,
Broad-crested weir, 20 122–123 129–130
Burst diaphragm. See Rupture disk culvert systems, 113–114 Detention ponds. See Detention
Bursting disk. See Rupture disk inlet control hydraulics, 114, basins
118–122 Diffusion wave model, 243–244
outlet control hydraulics, 114–116 Dimensionless unit hydrograph, 70
C CulvertMaster and tutorials, Direct-step method, 95
C coefficients, 57 122–123 Direct runoff hydrograph, 65
Celerity, 242 Culvert outlets, 141 Discrete convolution, 74–76
Centrifugal pump, 167 Culvert systems, 113–114 Discrete volume method (DVM), 175
Change in energy (ΔE), 6 Curb inlets on grade, 85–87 Distributed parameter models, 243
Characteristic curve, 167 Curve number Diurnal curve, 162, 214
Check valves (CVs), 169 method, 45 Drawdown analysis, 93
Chézy’s equation, 10 tables, 48–51 Dry-weather loads. See Sanitary loads

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 289

 E

Dry detention basins, 127 rapidly varied, 98 Hydraulic(s), 211

Dry pond, 129 rate, 1–2 grade, 7–8
Dynamic modeling, 236–237, 238 tracking, 173 jumps, 99
initial and boundary conditions, zone classification, 93 model, 163
241–242 Free-surface flow. See radius, 2–3
numerical solution, 240–241 Open-channel flow routing model, 243
problems, 248–260 Freeboard, 132 theory, 164
St. Venant equations, 239–240 Frequency, 29–30 time steps, 222
Friction losses, 9, 116, 164–165 Hydraulic grade line (HGL), 4,
Chézy’s equation, 10 8, 164
E Darcy–Weisbach equation, 11–12 Hydraulic transient, 195
Effective precipitation, 39 Hazen–Williams equation, 10–11 design strategies, 203–204
Effective rainfall hyetographs. See Manning’s equation, 9–10 general modeling equations,
Runoff hyetographs typical roughness factors, 12 199–200
Elastic model, 198–199 Friction slope, 8 HAMMER and tutorials, 207–208
Emergency spillways. See Overflow Frontwater analysis, 93 minimizing transients in
spillways Froude number, 15–16, 242 systems, 203
Employees, 215 Full-flow conditions, 3 modeling, 206–207
Energy balance, 94–95 problems, 208–210
Energy equation, 7, 114 protection devices, 204–206
Energy gains, 8–9, 166–168 G speed and pressure of transients,
Energy grade, 8 Gas vessel. See Hydropneumatic tank 200–202
Energy grade line (EGL), 4, 8 Gradually varied flow, 91–92 theory, 198–199
Energy losses, 8–9, 164 analysis, 118 transient flow, 195–198
friction losses, 164–165 energy balance, 94–95 HYDRO-35, 32
minor losses, 165–166 flow classification, 92–94 Hydrograph(s), 64, 220
Energy principle, 6–9 Gutter, 81 discrete convolution, 74–76
Entrance minor loss, 116 sections on grade, 82–83 load, 213–214
Eulerian DVM, 175 runoff hyetographs, 65–68, 76–78
Eulerian models, 174 unit hydrographs, 68–73
Exit minor loss, 117–118 H Hydrograph routing, 133
Extended-period simulations (EPS), HAMMER software, 200 effects, 242–243
172, 220 and tutorials, 207–208 Hydrologic routing model, 243
routing, 220–222 Harmon method, 217 Hydrologic time steps, 222
Extreme flow factors (EFF), 215–217 Hazen–Williams equation, 10–11, Hydrology, 29
13, 24, 164 hydrographs, 64–76
Head gains, 7 model, 97
F Head losses, 7 peak runoff flow rate, 51–63
Federal Highway Administration Headwater depth (HW depth), rainfall, 29–41
(FHWA), 82 113–114 unit hydrographs, 68–73
Federov method, 217 HEC-22 Urban Drainage Design Hydropneumatic tank, 205
FlexTables, 24 Manual, 82
Flow Control Valves (FCVs), 169 High-level outlets, 140
FlowMaster, 24 Horizontal slope, 93 I
Flows, 170 Horton equation, 38, 66 Infiltration, 39–40, 219
accumulation, 97 Horton infiltration equation, 41–42 Inflow, 219
area, 2 infiltration rate, 43 hydrograph, 130
characteristics, 5 land use and cover Initial rainfall abstractions, 45
classification, 92–95 considerations, 42–43 Inlet control hydraulics, 114,
conveyance, 3 runoff estimation, 43 118–122

290 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 J

Inlet design overview, 81, 90–91 Muskingum–Cunge method, 246 rational method, 56–60
inlet and gutter network problems Muskingum routing method, time of concentration, 51–55
using StormCAD, 91 245–246 Pipe leakage, 163
Inlets in sag, 82, 88–89 Pipe networks, 170–171
Inlets on grade, 82, 84–85 Pipe outlets, 141
combination, 87–88 N Pond bottoms, 132
curb, 85–87 Nappe, 20 PondPack
Intensity–duration–frequency curves Navier–Stokes equations, 199 and tutorials, 155–156
(IDF curves), 30–32, 98 Network analysis, 171–172 for windows, 156
Interception, 39 Nonrevenue water and leaks, Precipitation, 40
Interconnected pond modeling, 130 163–164 Pressure-Reducing Valves (PRVs), 169
International System of Units (SI), 1 Nonuniform flow, 4 Pressure-Sustaining Valves
Normal depth, 4, 98 (PSVs), 169
NRCS curve number equation, 38 Pressure Breaker Valves (PBVs), 169
J NRCS (SCS) curve number Pressure flow, 3, 12, 14
Jet, 17 method, 46–47 Pressure of transients, 200–202
antecedent moisture condition, Pressure pipe friction losses, 12
48 Pressure piping systems, 161
K cover type, 48 automated optimization, 176–178
Kinematic wave model, 244 curve number tables, 48–51 control valves, 169
Kutter’s equation. See Chézy’s hydrologic condition, 48 energy gains, 166–168
equation soil group, 47 energy losses, 164–166
NRCS (SCS) peak flow network analysis, 171–172
estimation, 60–63 pipe networks, 170–171
L NRCS synthetic unit problems, 179–193
Lag, 130 hydrographs, 70–73 water demands, 161–164
Lagrangian TDM, 175 WaterGEMS and tutorials,
Laminar flow, 4–6 178–179
Land use, 42–43 O water quality analysis, 172–176
Loading, 212 Off-line pond, 129 Pressure wave celerity, 200
EFF, 215–217 Online pond, 129 Prismatic channel, 2
hydrographs, 213–214 Open-channel flow, 3, 14 Profile
load types, 213 critical flow, 14–17 classifications, 93–94
pattern loads, 214–215 specific energy, 14–17 in StormCAD, 98–99
sanitary unit loading, 215 Orifices, 17–19, 141 Proportional methods, 164
variable PF, 217–219 Oscillating pressure wave, 197 Pumps, 166
Low-Impact Development (LID), Outflow hydrograph, 130 constant horsepower pumps, 168
236–238 Outlet control hydraulics, 114–116 variable-speed pumps, 167–168
Low-level outlets, 140 Outlet structures, 132
Overflow spillways, 133, 141
R
M Rainfall, 29
Manning’s equation, 9–10, 24, 55 P abstractions, 39, 40
Mass conservation, 45 Pattern loads, 214–215 characteristics, 29
Metered demand, 161 Patterns, 162, 214 data types, 30–38
Mild slope, 93 Peaking factor (PF), 216 Horton infiltration
Minor losses, 165–166 variable, 217–219 equation, 41–43
Mixed flow profiles, 95–96 Peak runoff flow rate, 51 hyetograph, 32
Model calibration, 176–177 NRCS (SCS) peak flow NRCS (SCS) curve number
Momentum equations, 198–199, 201 estimation, 60–63 method, 46–51

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 291

 S

Rainfall (continued ) Simplified routing models, 243 Supercritical flow, 15

return period and frequency, 29–30 diffusion wave model, 243–244 Suppressed weir, 20
runoff coefficient, 44–45 kinematic wave model, 244 Surge. See Hydraulic transients
runoff volume, 38–41 Muskingum routing method, Surge relief valves, 205
temporal distribution, 33 245–246 Surge tanks, 204–205
watershed area, 38–39 Skeletonization, 206–207 Synthetic temporal distributions of
Rational method, 56–60 Slope classification, 92–93 rainfall, 33
Recurrence interval, 30 Soil group, 47 System curve, 167
Residents, 215 Specific energy, 14–15 System design, 177–178
Retention, 45 Speed of transients, 200–202 System operating point, 167
pond, 129 Spread, 81
Return period, 29–30 Stage vs. discharge, 140
Reynolds number, 4–6 orifices and weirs, 141 T
Rigid-column model, 198 overflow spillways, 141 10 States Standard for PF, 217–219
Routing, 220, 242 pipe or culvert outlets, 141 Throttle Control Valves (TCVs), 169
analysis, 130 rating table, 140 Time-driven method (TDM), 175
convex routing, 220–221 Stage vs. storage volume, 133 Time of concentration, 51–55
of detention pond, 145–146 average-end-area method, 134–136 Time of concentration, 33
hydrologic and hydraulic time conic method, 137–139 Total hydrograph, 65
steps, 222 Standard-step method, 95 Trace modeling, 173
weighted translation routing, Steady-state network hydraulics, 171 Transient flow, 195–198
221–222 Steady flow, 3 models, 199
Runoff coefficient, 38, 44–45 model, 236 Transient protection devices, 198
Runoff hydrograph, 51, Steep slope, 93 Transitional flow, 118
65–68, 76–78 Stepwise pattern, 162–163, 214 Turbulent flow, 4–6
Runoff volume, 38–43, 128 Storage and hydraulic Typical roughness coefficients, 13
Rupture disk, 205–206 relationships, 133 Typical roughness factors, 12
composite stage vs. discharge
relationships, 142–145
S stage vs. discharge, 140–141 U
Sanitary loads, 213 stage vs. storage volume, 133–139 Uncontrolled outlets, 131
hydrographs, 213 Storage indication Underground detention, 129
Sanitary sewer design, 211–212 curve, 149 Uniform emitter coefficient, 164
elements, 212 method, 145–149 Uniform flow, 4
extended-period simulations, StormCAD, 91, 98–99 Uniform open-channel flow, 16–17
220–222 analysis and design, 99 Unit hydrographs, 68
loading, 212–219 profiles, 99–100 estimation methods, 64
problems, 224–233 Storm sewer analysis, 161 NRCS synthetic, 70–73
SewerCAD, 222–224 Storm sewer applications, 96–97 visualizing and developing, 68–70
Sanitary unit loading, 215 flow accumulation, 97 Unsteady flow, 3
unit sanitary loads from different hydrology model, 97 model, 236–237
sources, 216 sources of water, 97–98 Unsubmerged flow, 118–120
Sealing conditions, 96 Stormwater detention, 127 Urbanization of land, 237
SewerCAD, 211, 222–224 analysis procedure, 149–155 User-specified tailwater, 98
algorithm, 220–221 facilities, 129–130
SewerGEMS software, 98, 236, Stormwater management, 237
240–241, 244 Streamlines, 4 V
and tutorials, 246–280 St. Venant equations, 239–240 V-notch weir coefficients of
Sharp-crested weir, 20 Subcritical flow, 15 discharge, 21
Side slopes, 132 Submerged flow, 118, 120–122 Variable-speed pumps, 167–168

292 $PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition

 W

Variable PF, 217–219 Water Environment & Reuse TDM, 175–176

Velocity, 1–2 Foundation, 237 trace, 172
Vena contracta, 17, 19, 20 WaterGEMS, 176–177, 206 Watershed area, 38–39, 57
Volumetric flow rate, 1 and tutorials, 178–179 Water sources, 97–98
Water hammer. See Hydraulic Wave celerity, 15
transients Weighted runoff coefficient, 45
W Water quality analysis, 172 Weighted translation routing method,
“Water-hammer” effect, 24 age, 172–173 221–222
Water demands, 161 constituents, 173–174 Weirs, 19–23, 141
demand patterns, 162–163 DVM, 175 Wet-weather loading, 213, 219
metered demand, 161 initial conditions, 174 Wet detention basins, 127
nonrevenue water and leaks, numerical methods, Wet pond, 129
163–164 174–175 Wetted perimeter, 2–3

$PNQVUFS
"QQMJDBUJPOT in Hydraulic Engineering - 9th Edition 293