Utah State University

DigitalCommons@USU
Reports Utah Water Research Laboratory

1-1-1990

Pipe Network Simulation Analysis Computer

Program - NETWK
Roland W. Jeppson

Follow this and additional works at: http://digitalcommons.usu.edu/water_rep

Part of the Civil and Environmental Engineering Commons, and the Water Resource
Management Commons

Recommended Citation
Jeppson, Roland W., "Pipe Network Simulation Analysis Computer Program - NETWK" (1990). Reports. Paper 298.
http://digitalcommons.usu.edu/water_rep/298

This Report is brought to you for free and open access by the Utah Water
Research Laboratory at DigitalCommons@USU. It has been accepted for
inclusion in Reports by an authorized administrator of
DigitalCommons@USU. For more information, please contact
[email protected].
 USERS MANUAL

PIPE NETWORK
SIMULATION ANALYSIS
COMPUTER PROGRAM - NETWK

Civil and Environmental Engineering

Utah Water Research Laboratory
College of Engineering
Utah State University
Logan, Utah 84322

By
Roland W. Jeppson,
Professor
. -~
~ March 1990
'j;
USERS MANUAL FOR PIPE NETWORK SIMULATION
ANALYSIS COMPUTER PROGRAM USU-NETWRK

by

Roland W. Jeppson

Civil and Environmental Engineering

and
Utah Water Research Laboratory
Utah State University
Logan, UT 84322

The software described in this document is furnish under a license

and may only be used in accordance with the terms of the license.

Copyright C 1990 by Roland W. Jeppson

Department of Civil Engineering
Utah State University

Previous editions; April 1977, September 1978.

June 1980, April 1982, Jan. 1985

March 1990

Price $25

-i-
=

-ii-
 USU-NElWK ON IBM-PC COMPUTERS
= The next couple of pages are applicable compressed form on them. In order to get these
if you have obtained a version of USU-NETWK programs off the diskette and in executable form
for use on a PC. Read them carefully before it is necessary that you properly install and
installing USU-NETWK. decompress them by completing the following
The diskette that you have received steps (These steps assume your PC is operating
contains the executible element of the network under DOS and that the DOS commands are
analysis and design program that has been accessible):
developed at Utah State University. It also
contains the executible element of a pre-processor 1. Place the distribution diskette labelled
program PIPEINPT to assist in preparing input USU-NETWK in one of your diskette drives. (If
data files, and a post-processor program PLTNET the distribution diskette is a 5 1/4 inch, high
that will display the network and its solution on density, diskette then this must be a 1.2 meg
the PC monitor provided that the PC contains a drive.) Make sure that the default drive is where
graphics card. (All IBM compatibles will not you want the program to go, i.e. when installing
allow the graphics even if they have a graphics USU-NETWK on a hard disk, the default drive
card.) The diskette also contains a couple of should be C:, D: etc.
example data files for small networks These few
pages are intended to get you started and provide 2. Type A:INST_NET (or B:INST_NET
instructions to get the program up on your Pc. if the diskette is in drive B:)
To effectively use the program you will [Note: _ is the underscore not the minus
need to read this manual This manual is written sign -, and is typed by pressing
to also serve the needs of users of USU-NETWK this key while simultaneously holding
on large computers. Most of the input described down the Shift key]
in the manual applies to this PC version of the
program. The manual indicates which options and 3. Respond to the prompts. You will be
commands are acceptable only to the larger ask to provide your license number. This number
program. will be in the license agreement with the
Unless you have indicated otherwise, the distribution diskette. You will be ask give the
program you have received requires that the PC drive that the distribution diskette is in. Type
contain co-math processor such as an 8087 and either 1 or 2 as appropriate. You will be ask to
will require a PC that has 512 kbytes of RAM. If give a subdirectory that the files will be stored on.
the PC you plan to use USU-NETWK on does A good choice for a subdirectory is NETWK.
not have a co-math processor, you should request To use USU-NETWK thereafter you
that a diskette be sent that does not require it. should first make the subdirectory into which the
Also if you wish to use USU-NETWK on a PC programs are installed your default directory by
that has only 256 kbytes of memory you should typing CD\NETWK (or whatever the
indicate this. The program for a PC with only 256 subdirectory name is). This subdirectory should
kbytcs will accommodate a pipe network with 300 then be used for input and output files as well.
pipes.
Ignore any message such as "File not
found" that you get during the installation of
Installing program on your PC USU-NETWK. If you get the message that
directory NETWK (or the name you gave) cannot
The following instructions assume that be created, it simply means that this directory
your PC has a hard disk. If it does not then you already exists, and the installation will overwrite
will need to use one diskette for the program and any files with the same name in this directory.
another for the input and output data files (see
the next section). It is important that you follow Should you have ordered the PC version
these instructions carefully. of USU-NETWK that also has the capability to do
time-dependent problems, and have requested that
The diskette(s) that contains the USU- you also receive the steady state version of USU-
NETWK package will have the programs in NETWK, then this steady state version will be
under a directory \STEADY\ if the distribution directly from the diskelle by placing this diskette
diskette is a 1.2 mega 5 1/4 inch (high density) in a floppy drive, make this the default drive, and
diskette, or on a 1.4 mega 3 1/2 inche diskette. type TUTOR. Should TUTOR be on the same
You can determine whether this subdirectory exist distribution diskette with the other USU-NETWK
by giving the command DIR A: (or DIR B: if programs, then it is necessary to put TUTOR on
the distribution diskette is in drive B:). The your hard disk or another floppy diskette. If you
steady-state version of USU-NETWK will be wish to install TUTOR on your hard disk, then
under the name NETWKS.EXE. The added S place the distribution diskette that contains
stands for steady state, and to execute it rather TUTOR in a floppy drive compatible with the
than the simulation version type its name. You diskette. If a subdirectory TUTOR exists on this
can install the steady state version of USU- diskette, then make this the default directory of
NETWK in the same subdirectory on a hard disk the floppy by typing A: (or B: if the distribution
as the simulation version, or a different diskette is in drive B) followed by CD\TUTOR.
subdirectory. In either case you need to complete Next make the hard disk the default drive again by
the following steps to also install the steady-state typing C: (or the drive you want TUTOR installed
version of USU-NETWK. on). With the hard disk your default drive type
1. Make the subdirectory STEADY on A:INSTALL ( or B:INSTALL). (Note the name
the floppy drive where the distribution diskette is is now INSTALL and not INST_NET as above.)
the default drive by typing: A: (or B:) then type
CD\STEADY.
2. Make the hard disk the default drive EXECUTING USU-NETWK
again by typing C: (or D: if you which this
installed on drive D:, etc.) To execute the program after it is
3. Type A:INST_NET (or B:INST_NET installed give the command, NETWK. When the
if the distribution diskette is in drive B) program asks: WHAT FILE CONTAINS INPUT
4. Respond to the prompts as above DATA? give the name of the file that you have
selecting 5 (to install everything). Generally only prepared with an editor or the pre-processor
the element NETWKS.EXE will be installed. program PIPEINPT, to define the network
However, another version of PLTNET.COM might problem you want solved. If you respond with
also be installed. This latter version of PLTNET TTY or REMOTE then USU-NETWK expects to
with the extension .COM utilizes only CGA receive input data for the problem directly from
graphics capabilities, whereas PLTNET.EXE the keyboard. Such use is practical only for very
utilizes EGA, or VGA graphics capabilities. small networks. When the program asks: WHAT
Should PLTNET.COM be installed on the same FILE NAME SHOULD CONTAIN OUTPUT?
directory as PLTNET.EXE, and the PC has EGA give the file to which the output is to be written.
or VGA graphics, you should either delete You can display this output file on your monitor
PLTNET.COM or rename PLTNET.COM to afterward with the command TYPE followed by
PLTNETC.COM or some other name or else it the same name. If you want the output to be
will be executed when you give the command displayed to the monitor directly given TTY or
PLTNET, rather than PLTNET.EXE. On the REMOTE for this file name. The length of the
other hand if the PC only has CGA graphics, then file names that USU-NETWK accepts is 12
delete PLTNET.EXE. characters or less including the extension and
drive deSignation if it precedes the name.
NOTE: If you use a word processor such
INSTALLATION OF TUTOR FOR USU- as WordPerfect or WordStar, it is necessary that
NETWK the file be written to disk as an ASCII (or TEXT
files) for NETWK to read them. For example
With the distribution of USU-NETWK with WordPerfect, you will need to use the test
you will likely have also received a TUTOR IN/OUT key when writing the data to disk.
program to help you get started. This TUTOR
may eome on a separate diskette, or be in a
subdirectory TUTOR on the distribution diskette EXAMPLE PROBLEMS
if this distribution diskette is 3 1/2 inch in size or
is a high density 5 1/4 diskette. If the TUTOR is Several small example pipe network
on a separate diskette then you can use TUTOR problems will be included with the distribution of

iv-
 USU-NETWK. These are in the compressed file of NETWK terminate the first lines of the general
INPUT. ZIP, and will be decompressed during the graphics file with /,s rather than 999 either replace
above "install" procedure if you select to have this the / with a 999 or attempting to read the graphics
done when prompted i.e. by selection menu item file with PROFILM. On the other hand when
5 (to have everything installed). The example using OFPLOT either replace the 999's with a lor
problem on the file EXAMPLT.DAT is designed place a I in front of the 999's with an editor.
to write files that allow you to use the graphics The program PROFILM will ask:
post processors PLTNET, PROFILM, OFPLOT, SELECT THE NO. FOR TYPE OF PLOT YOU
etc. The x and y coordinates of the nodes are DESIRE.
given in this file, which is necessary to do any For the first plot select 2 (for "Pressure bars at
graphics. Also to display networks on the monitor nodes of network"). Thereafter you will be ask to
of your PC it must contain a graphics adaptor. A GIVE FOLLOWING: 1 - type of bar, 2 - scale
good way to get started in using PLTNET is to factor, 3 - width and 4 - units. The numbers listed
use the example network, EXAMPLT. The file after these items are values that PROFILM
that PLTNET utilizes is written by NETWK if the suggests. For your first display type in these
option NETPLT=1 is in the $SPECIF list of values, i.e. type in 2 .3 10 0 After the plot is
options. The more general purpose graphics file, displayed take a carefully look at the pipe
that PROFILM and OFPLOT use, is written by . numbers and the network's layout. To reactivate
NETWK if the option NETPLT=3. PROFILM from just displaying the plot press the
It is suggested that you complete the enter key.
following exercises using the example problem Have PROFILM make a second display,
EXAMPLT.DAT: but this time when ask: SELECT THE NO. FOR
1. Execute USU-NETWK by typing TYPE OF PLOT YOU DESIRE give 1 for "Head
NETWK, and when ask for the input file give the profiles over selected sequences of pipes. Place
name EXAMPLT.DAT, and give TrY for the three such profile on the display by giving 3 when
output file, so that the solution will be displayed ask, and then give the follow three sequences of
on the screen. Upon completion of the solution pipe numbers:
USU-NETWK will have written a file 65421/
PLTNET.DAT for PLTNET to use. 6 5 10 11 16 27 28 31/
2. Execute the graphics program 3029 17 27/
PLTNET by typing PLTNET. When ask if you Next have PROFILM make a contour plot
wish to change options type an N for no the first by responding with 3 ("Contour map of heads or
time, also when ask WHAT DO YOU WANT ON pressures") for the TYPE OF PLOT YOU
PLOT? you can select "All setting OK" by typing DESIRE. You will be ask to give 5 values for the
10. You will see this network displayed on your following 5 items: (1) No. of Contours, (2)
PC's monitor. Press the return to activate Value of smallest contour, (3) Contour interval,
PLTNET again. Reading the instructions related (4) Frequency of heavy line, & (5) maintain x and
to the use of PLTNET will inform you about what y scales or fill screen. Appropriate values might
this processor can do. be:
3. Use an editor to change the option 32400540
NETPLT in the file EXAMPLT.DAT to The values give by PROFILM, associated with the
NETPLT=3, save this altered file, and execute request, help you decide what to give for these 5
NETWK again. Setting option NETPLT=3 cause items.
the general purpose graphics file to be written by With this preliminary introduction with
USU-NETWK. If you are ask during this PLTNET and PROFILM you might wish to play
execution to give the file name for the graphics with the options, changing colors, character styles
data type PLOTlO.DAT. and sizes until you get what is most pleasing to
4. Execute the graphics program you.
PROFILM by typing PROFILM. This program is
written in PASCAL and it requires strings of three
nines to terminate reading of the first lines of PC's with two floppy diskette drives but
input from the graphics file PLOTlO.DAT. no hard disk
Programs, such as OFPLOT, written in
FORTRAN require that a / terminate reading If your PC does not have a hard disk,
these lines rather than 999. Should your version

-v-
then it will be necessary to keep your input and Give the command: NETWK and then enter the
output data files on another diskette, and run data as shown below.
NETWK from the program diskette in the other
drive. The following description assumes that the
program diskette will be in drive A: and the data WHAT FILE CONTAINS INPUT DATA?
diskette in drive B:. To install the program
initially do the following: TTY
1. After turning the PC on place the
DOS diskette in drive A and wait for it to boot. WHAT FILE NAME SHOULD CONTAIN
2. Place an unused diskette in drive B, OUTPUT?
and give the command FORMAT B: and follow
the instructions. TTY
3. Copy the EDLIN editor on the
diskette in drive B with the command COPY THREE RESERVOIR PROBLEM
EDLIN. COM B: (if you want to use another 1*
editor, or word processor delete this step.)
4. Give the command B: to make B your TITLE GIVEN TO NETWRK
default drive. THREE RESERVOIR PROBLEM
5. Replace the system diskette with the ALL DEMAND FLOWS ARE MULTIPLIED
diskette included herein and give the command BY 1.000
A:INSTALL. When asked for a file where the
output to the example problem should go give $SPECIF NPRINT=-I,COEFRO=.005 $END
TTY, that indicates the ouput will go to the PIPES
monitor directly. (If you give another name then 10 11500 8
that filename will be established, and you can see 20110006
the solution output, thereafter, by either giving the 3 10 1200
command TYPE filename, or TYPE NODES
TYPEFILENAME>PRN: to have it printed. 10350
Every time you want to use a new data RESER
diskette for input and output data for network 1500
problems it is necessary to follow this same 2450
procedure. Subsequently you will want to issue 3300
the command DEL *.DAT to delete the data files RUN
for the example networks, however.
Output from NETWK

Examples: PIPES 3
NODES 1
In the following example USU-NETWK is SOURCE PUMPS 0
used to solve the classical 3 reservoir problem that BOOSTER PUMPS 0
can be found in most fluid mechanics books. RESERVOIRS 3
MINOR LOSSES 0
PRVS 0
NOZZLES 0
CHECK VALVE 0
BACK PRES. V. 0
t-::~-; 300' DIE HEAD DEV. 0
all roushneslil coef. '" 0.006" RESERVOIR(NOZZLE) PIPES AND THEIR
ELEVATIONS ARE
1 500.0 2 450.00 -3 300.00
FLOW FROM PUMPS AND RESERVOIRS
EQUALS .000

ITERATION 1 SUM .125E+Ol

ITERATION 2 SUM .579E+00

-vi-
ITERATION 3 SUM .172E+OO
ITERATION 4 SUM .166E-Ol
ITERATION 5 SUM .157E-03
UNITS OF SOLUTION ARE
DIAMETERS - INCH
LENGTH - FEET
ELEVATIONS - FEET
PRESSURES (PSI)
FLOWRATES -(CFS)

DARCYWEISBACH FORMULA USED FOR

COMPUTING HEAD LOSS
PIPE DATA
PIPE NODES HEAD HLOSS
NO. FROH TO LENGTH DIAM COEF FLOW RATE VELOCITY LOSS /1000
1 a 1 1500 B.O .005000 3.05 8.74 49.17 32.76
.. 2 1 o 1000 6.0 .005000 .21 1.07 .06 .86
.. 3 1 o 1200 6.0 .005000 2.84 14.46 150.06 125.72
NODE DATA:
NODE DEMAND HGL
NO. (CFS) (GPM) ELEV HEAD PRESSURE ELEV
1 .000 .00 350. 100.66 43.71 450.86
5 1 4650
60 4700
A more usual mode of solving network PIPES
problems is to prepare the input data on a file 1 7 1 1000 10
first with an editor as illustrated below for the 21220008
small network shown in the sketch. In this 3 2 3 1700 6
example EDLIN will be used to prepare the input 4 4 3 1800
data. 5 1 4 1500 8

6 5 2 1500
~m
I" 7 5 3 15006
.& IU
II .. 86510008
RESER
6 4840
7 2 242 4 220 6 194 8 160 10 120 12 75 4650/
RUN IA control C is given
,to exit input mode, and E
Iis given to exit.

Now type NETWK and when ask what the input

file is give
EDLIN EXAMP.DAT
* EXAMP.DAT
I puts EDLIN in input mode
1* When ask what file should receive the output give
$SPECIF PCHAR3=0,NODESP= 1,
COEFRO=.OO2 SEND EXAMP.OUT
NODES
1 .84600 You can subsequently display the solution on your
2 1 4530 monitor with the command TYPE EXAMP.OUT
3 .94520 or have it printed with the command PRINT
4 1.54550 EXAMP.OUT.

-vii-
 TUTOR FOR USU-NETWK

A tutor is available to assist you in getting distinguished. If your PC does not have a graphics
acquainted with the operation and input data card, or has a graphics card that is not compatible
requirements for USU-NETWK. This tutor is not with what TUTOR requires, you can still benefit
intended to replace this manual, but for an by use of this tutor.
individual who would rather sit at a PC terminal In making copies of this tutor diskette it is
than read a manual, the tutor will be helpful in necessary that the DOS DISKCOPY command be
getting started. If you have received the tutor in used rather than the DOS COpy command. The
the distribution of USU-NETWK and wish to use files on this diskette that have the extension .DAT
it, then you should read the rest of this page of the are input files for 8 small problems that are
manual. covered by TUTOR. You might want to print
To use the USU-NETWK tutor you should these input data files, and have the output that
do the following: USU-NETWK produces from them go to a file
1. Place the TUTOR diskette in a floppy disk with the same name but with the extension .OUT.
drive, i.e. drive A: If you print these output files, you can see if your
2. If this is not your default drive, make this program of USU-NETWK produces the same
drive the default drive by typing A: solutions as given by the TUTOR. The two
3. Type the command TUTOR to run the solutions should be identical.
program.
Most of the instructions given by TUTOR
are in text form. Sketch of small networks are
provided for which input data is subsequently
prepared. The PC that you are using must have an
IBM graphics card, and be fully IBM compatible
for these displays to work properly. If your PC has
a color monitor, numbers that denote pipes, and
nodes on these sketches will be in different colors
so that these values can be easily

(0.1500)
Pump Curve
Q (gpm) hp (ft)
1200 80
1800 73
3500 55

1--__::-1 470 '

(l}~ ______________________ ~~ __ ~

(1800.700)

450 gpm {7}

(3000,700)

all roughness coef. '" .002"

Figure 1. Small example network used in first 7 problems by TUTOR.

-viii-
 PREFACE

This manual describes the data that a HGL elevations through selected sequences of
versatile computer program USU-NETWK pipes, (b) display the pressure, pressure head, or
requires to complete a time dependent solution of HGL elevation as bars at the nodes of the
llowrates and pressures throughout a pipe network, or (c) draw a contour type map of the
neLwork. This program can be used for steady- pressure, pressure head, or HGL elevation.
state analysis and design, as well as simulation 4. A program OFPLOT that is contains the same
solutions through some period of time. In fact, calcomp graphics plotting capability in the full
unless specified otherwise, a steady-state solution program but is intended for off line use. 5. A
is assumed. program PRINT that is capable of efficiently
supplying requested specific information from a
A smaller version of the program is not direct access file written by the simulation version
capable of the time-dependent solution, but is of USU-NETWK, related to the solution at any
capable of doing most of the types of steady-state time. This last auxiliary program is described at
analyses described in this manual. If you have the in Appendix E of this manual. 6. A program
version of USU-NETWK for use on 16 bit IBM PLTTIM that will display variables from the
compatible PC's it may only do steady-state solution plotted against time, from time-dependent
analyses. soltuions.

You, as a user, may have acquired this Programs PIPEINPT and PLTNET are
smaller version of the program, because you described near the first of the manual after a
believe Lhis capability would meet your needs, your preview of the input to USU-NETWK is provided.
computer limitations, or for other reasons. This A TUTOR program is also available with PC's to
manual is still applicable for your program. All help users get acquainted with USU-NETWK and
basic information which describes the physical some of the flexibility it allows in input. Also a
features of the pipe network are identical for the documentation program for larger computers is
simulation and the steady-slate versions of USU- available that can be used in place of this manual
NETWK. The difference is that the simulation for help in using USU-NETWK. Help in using
version accepts additional data describing the time this TUTOR are given in a preliminary page of
dependent operation and characteristcs of the this manual.
network. In reading the manual you will be able
to identify those data items which are "additional The full version of USU-NETWK has the
time dependent data." Therefore, if your version ability to drive a calcomp plotter to produce a
handles only steady-state solutions this manual is graph of the network and its solution for inclusion
applicable to your program, if you ignore those in repons. If your installation subscribes to the
parts of the manual devoted to input data for DISSPLA graphiCS routines, then USU-NETWK
simulations. Also a few seldom used options are is also capable of producing a three-dimensional
not implemented for the steady-state version of perspective drawing of the network and its
the program. These are denoted by #. solution, as well as generating contour type maps
of pressure or HGL elevation lines. These can be
In addition to a description of data sent to a variety of graphics devices. Other
requirements for USU-NETWK, a pipe network auxiliary graphics programs are also available.
analysis and design program, this manual contains
descriptions of the use of auxiliary program that The complete program allows several
can be used in connection with USU-NETWK. forms for the input data. The standard form is
These programs consist of: 1. A preprocessor, discussed in the body of this user's manual with
PJPEINPT that assists in the preparation of other allowable forms described in Appendixes A
steady-state input data, 2. A post processor, and B. If you decide on a given form of input and
PLTNET that displays the network and selected wish to use it exclusively, the size of the program
solutions items to the monitor of a PC, and can be reduced. For example, if only the standard
3. A post processor program PROFILM that (a) form of input is desired, then subroutine AINPUT
will display profiles of pressure, pressure head, or and the references to it and its entry points may

-ix-
be removed. A reduced version of the simulation also perform time-dependent analysis is
program does not contain subroutine AINPUT dimensioned for 600 pipes, etc. A version for a
and other subroutines. If the command PIPE- PC that is dimensioned for 500 pipes, and 430
nodes is not to be utilized then the subroutine nodes will run slight faster than the version that
PNREAD may be removed. The size of the allows up to 1000 pipes. Since most all networks
program can be further reduced by descretion. can be skelonized to 500 pipes or less this smaller
For example three different solution algorithms program is often the one. distributed.
exist in the original program. The elimination of
the subroutine SOLSYM removes the solution In this user's manual there is no attempt
algorithms which establishes a symmetric positive to describe the methods used in USU-NETWK. to
definite Jacobian matrix for the Newton Method provide solutions to either analysis or design
of solution, which is ordinarily called upon if no problems. Rather this user's manual is devoted
pressure reduction valves, back pressure valves, or exclusive to describing the input data, and the use
other devices exist that destroy symmetry in this of USU-NETWK. in solving various types of
matrix. Solution subroutines SOLNYM and network problems.
BAND handle any problem. Also subroutines
SPARSE, SPARSA, SPARSB and EQUATS,
which utilize special sparse matrix solution
methods, handle any problem. Therefore, a
solution to any network could still be obtained
with SOLSYM removed; and/or with the sparse
matrix methods or SOLNYM and BAND
removed. The PC steady-state version of USU-
NETWK contains only the sparse matrix
subroutines, as does a smaller version of the
simulation program.

The program was initially developed using

pre-77 FORmAN that did not allow list directed
read statements. To allow input to be free of any
format requirements, USU-NETWK. has its own
free format input routines. List of options are
handled likewise by USU own equivalent
NAMELIST 1\0 capabilities, or they may be
replaced by NAMELIST 1\0 if this is allowed for
in the FORmAN being used.

The number of network components that

can be accomodated are established at the time
the program is compiled and linked. When using
a computer such as a VAX, MicroVAX, IBM
mainframe, or UNIX Engineering work stations
the maximum number of pipes, and nodes, etc that
may be accomodated is large such as several
thousand. A VAX version of USU-NETWK
dimensioned for up to 3,000 pipes, 2,580 nodes, 80
pumps, etc. is keep on file ready for distribution.
However, for use on 16 bit or 32 bit PC
computers under DOS the amount of memory that
can be addresses is 640 Kbytes and this limits the
size of networks that can be accomodated.
Currently for steady-state PC version of USU-
NETWK is limited to 1000 pipes, 860 nodes, 600
loops, 40 pumps, etc. The version of USU-
NETWRK designed for PC under DOS that will

-x-
 TABLE OF CONTENTS
USU-NETWK ON IBM-PC COMPUTERS iii
TUTOR viii
PREFACE x

TABLE OF CONTENTS xii

PREVIEW xiv
FLUID MECHANICS PROBLEMS xvii
UNIT AND CONVERSION FACTORS xx
SUMMARY MOST FREQUENTLY USED OPTIONS xxi
SUMMARY OF COMMANDS THAT ENTER DATA xxiii
INTRODUCTION 1

VERSATILITY OF COMPUTER PROGRAM 3

USU-NETWKS'S PROCESSOR: PIPEINPT 6

GENERAL SPECIFICATIONS (OPTIONS) 11

DETAILED STEADYSTATE DATA REQUIREMENTS 47
CHANGE DATA 67
COST INPUT DATA 71
DETAILED DYNAMIC SIMULATION DATA REQUIREMENTS
ILLUSTRATIVE EXAMPLES OF NETWORKS AND INPUT DATA USING THE
STANDARD FORM OF INPUT 111
APPENDIX A (1st alternate input to USU-NETWK) A-I
APPENDIX B (2nd alternate input to USU-NETWK) B-1
APPENDIX C - GRAPHICS ASSOCIATED WITH USU-NETWK C-l
APPENDIX D - IMPROVING DESIGN D-l
APPENDIX E - AUXILIARY PROGRAM PRINT E-l
APPENDIX F - PROGRAM PMPCUR for display of pump curves F-l

-xi-
-xii-
 PREVIEW
The program USU-NETWK is designed to PIPES
make the input as simple as possible, while Pipe no., Upstream node no., Downstream node
simultaneously allowing the user considerable no., Length, Dia., Coef.
flexibility in describing a network problem for a
computer solution. At the first rapid reading of
this manual, the new user may get the impression NODES
that a lot must be understood before preparing Node no., demand, elev.
the input data for USU-NETWK. To dispel this PUMPS
possible false impression this preview section is Pipe No., Qz' Hl , Q2' H 2, Q3' HH3' Sump elev.
given at the beginning of the manual. You should
read this section if you are a new user, and BOOST
prepare data for some simple networks such as Pipe No., Ql' H l , Q2' H 2, Q3' H3
those illustrated below, before reading the
remainder of the manual. That is you should get RESER
solutions to some problems before you concern Pipe No., Elev. of water surface in reservoir
yourself with all the possibilities that are allowed,
and are described in the rest of this manual. The network on the following page gives
From experiences in teaching undergraduate Fluid an example of how this is actually done. If the
Mechanics classes in the Col1ege of Engineering at data given below under "Input Data" were stored
USU, the writer has found that a 15-30 minute on a computer disk file, USU-NETWK executed,
lecture on the use of USU-NETWK, accompanied and this file name given for the input file when
by a two page handout similar to this preview prompted for it, then the output given under
equips students with sufficient understanding to "SOLUTION" would shows the output produced
use the program in solving assigned steady state by USU-NETWK for this network.
analysis and design problems. The input data
needed to define a network analysis problem is Input data
described below. and illustrated by a small example THIS IS A SIMPLE EXAMPLE
that follows. 1*
PIPES
MINIMUM INPUT DATA 1 0 1 500 8 .0102
REQUIREMENTS 2128006/
3421200/
Line No.1 - Provides the title for the 5 1 3/
network. This title is followed by a 1* in columns 4231000/
I and 2 in a new line after the title is complete. 6438008/
704500/
SUbsequent Lines The remaining lines of NODES
input contain command names followed by data of 1050
the type dictated by that name. Command names 2 1/
are words such as PIPES, NODES, PUMPS, 3 1.5/
BOOSTer, RESERvoirs, starting in column 1 and 40/
spelled to 5 upper case characters correctly. Each PUMPS
command is followed by lines of data which 7 1 55 1.2 54 1.4 5 50
provide information of the type denoted by the RESER
command. These data are free format, i.e. 1100
separated by one or more blanks, or a comma, and RUN
may be terminated with a / if the data given
previously for this item is to be used. The word
RUN or END denotes end of input data. The When USU-NETWK is executed it will
items contained in lines of input after the ask you what the name of the input data file is.
command names are as given below. You then give the name of the disk file that
contains the input data for the network. If you

-xiii-
-~. give TTY or REMOTE in response to this
prompt, then USU-NETWK expects you to enter
the data line after line directly from the keyboard.
Next you will be ask (prompted) to supply a name
for the output file. A file with the name you give
will be opened and the solution written to it for
later printing or examination by you with an
editor. If in response to this second prompt you
give TTY or REMOTE, then the solution output
will come directly to your terminal, or monitor if
you are working on a Pc. Example Simple Network

SOLUTION
100'
TITLE GIVEN TO NETWORK
THIS IS A SIMPLE EXAMPLE
ALL DEMAND FLOWS ARE MULTIPLIED BY 1. 0000 (1)
PIPES 7
NODES 4 8-500 [1]
SOURCE PUMPS 1
BOOSTER PUMPS 0
RESERVOIRS 1
MINOR LOSSES 0
PRVS 0
NOZZLES 0
CHECK VALVE 0
BACK PRES. v. 0 (7)
DIF. HEAD DEV 0
SPECIFIED PRES 0
PIPE 2ND ORDER COEF LINEAR COEF SHUT-OFF HEAD SUMP ELEV
7 -600.000 1315.000 -660.00 50.00 1.5
RES.(NOZZLE) PIPES & THEIR ELEV. ARE
1 100.0 50' -=:-1-::!!--t
JUNCTION EXT. FLOW PIPES AT JUNCTION
1 1 .000 -1 2 5
2 2 1.000 -2 -3 4
3 3 1.500 -5 -4 -6
4 4 .000 3 6 -7
FLOW FROM PUMPS AND RESERVOIRS EQUALS 2.500
ITERATION= 1 SUM= . 538E+00
ITERATION= 2 SUM= .547E-01
ITERATION= 3 SUM: .499E-02
PUMPS:
PIPE HEAD FLOW HORSEPOWER KILOWATTS KWATT-HRS/DAY
7 45.62 1.25 6.49 4.84 116.14
UNITS OF SOLUTION ARE
DIAMETERS - inch
LENGTH - feet
HEADS - feet
ELEVATIONS - feet
PRESSURES - (psi)
FLOWRATES - (cfs)
DARCY-WEISBACH FORMULA USED FOR COMPUTING HEAD LOSS
PIPE DATA
PIPE NODES HEAD HLOSS
NO. FROM TO LENGTH DIAM COEF FLOW RATE VELOCITY LOSS /1000
------------------------------------------------------------------------------
1 0 1 500. 8.0 .010200 1. 25 3.57 3.28 6.55
2 1 2 800. 6.0 .010200 .69 3.50 7.24 9.05
3 4 2 1200. 6.0 .010200 .34 1. 74 2.82 2.35
4 2 3 1000. 6.0 .010200 .03 .15 .03 .03
5 1 3 1200. 6.0 .010200 .56 2.85 7.27 6.05
6 4 3 800. 8.0 .010200 .91 2.61 2.85 3.56
7 0 4 500. 8.0 .010200 1.25 3.59 3.31 6.62
NODE DATA:
NODE DEMAND HGL
NO. (cfs) (gpm) ELEV HEAD PRESSURE ELEV
----------------------------------------------------------------
1 .000 . 00 50 . 46.72 20.25 96.72
2 1. 000 448.83 50. 39.48 17.11 89.48
3 1. 500 673.24 50. 39.46 17.10 89.46
4 .000 . 00 50 . 42.31 18.33 92.31

-xiv-
 ANOTHER POSSIBILITY 40502091 625
50500500
An alternative to providing pipe lengths is 6 0 50 2591 625
to give the x and y coordinates at the nodes of the PUMPS
network. These coordinates are necessary if a 6 1 55 1.2 54 1.4 5 50
plot of the network and its solution is requested. RESER
To inform USU-NE1WK that pipe lengthS will 5100
not be given, but rather these should be computed RUN
from x and y coordinates, the option LENGON =0 '
must be added to a list of options between In addition to the differences discussed
$SPECJF and $END i.e. a line such as, above, please note the following differences occur
$SPECIF LENGON=O $END in this input data and that used earlier because
is added after the /*. With the option NODESP=1 is given in the options list: 1. there
LENGON =0 included in the list of options, are 6 lines of data after the NODES command
lengthS are not given as the 4th item after the instead of 4; two extra lines for the two source
PIPES command, but rather x and y coordinates nodes, 2. the demands at these source nodes are
are given as items 4 and 5 after the NODES 0, e.g. the amounts of flow supplied by the sources
command. In order to give the locations, e.g. the are not demands, but will be flowrates determined
coordinates, of the sources of supply (the reservoir as part of the solution, 3. source PUMPS and
and the source pump) it will also be necessary to RESERvoirs are identified by their node numbers
number sources of supply as nodes. The now rather than the pipes that connects them to
reservoir that supplies the previous "simple the network, and 4. pipes 1 and 7 now list the
nctwork" will be numbered as node 5, and the source nodes 5 and 6 respectively, as their
source pump will be given the node number 6. If upstream nodes rather than O. In the previous
thc x and y coordinates are as given in the small "input data" O's for node numbers denoted "non-
table below (these existent" nodes, but are needed for if there were
omitted then the pipe lengths would be taken as
Node x-cord. y-cord. the downstream node numbers.
I 500 500 Why don't you write down the input data
2 1000 1125 that would be correct if the option LENGON =0
3 1591 o were not included in the $SPECIF list, but
4 2091 625 NODESP= 1 remains in the list.
5 o 500 lengthS for pipes, that will be used in the
6 2591 625 solution for computing head losses, etc. can be
given under the PIPES command and x and y
coordinates don't give the same length to all pipes coordinates, that will be used for plotting of the
as used previously), then an alternative "Input network, can also be given under the NODES
Data File" would contain the following: command by leaving LENGON =0 out of the
options list but adding NETPLT= 1 to this list
that indicates that a plot of the network is to be
THIS IS A SIMPLE EXAMPLE created.
1* Example Simple Network
$SPECIF NODESP= I,LENGON =0 $END
PIPES
I 5 1 8.0102 t--=-'=--i 100'
2 126/
342/
5 I 3/ (1)
423/
6438/ 8
764/
NODES
1050500 500
2 1 50 1000 1125
3 1.5 50 1591 0 (7)

1.5 (1591,0)

-xv-
-xvi-
 J

6. The pumps in Problem 5 have characteristics given in the table below.

FLUID MECHANICS PROBLEMS Find the flowrate in each pipe.
While USU-NETWK is designed to provide solutions to large looped systems JSump 2 J 4
of pipe it can also be used to solve small problems involving one or two pipes.
!lelow are brief statements and sketches for simple fluid mechanics problems
-------------------------------------------
Ql 0.3 0.3 0.1 0.15 cfs
similar to those that may be given to college students as homework aSSignments HI 60. 50. 40. 75. ft
to be worked using a pocket calculator, after applying the appropriate principles Q2 1.0 1.0 0.6 0.5 cfs
from flow in closed conduits. Host of these problems would require considerable H2 55. 45. 35. 60_ ft
effort 1f done by hand. You may find it interesting to sharpen you fluid Q3 1.8 1.8 1.0 1.0 cfs
mechanics skills by solving the problems by hand. Then use USU-NETYK to obtain H3 30. 25. 25. 40. ft
a solution to each problem. In solving these problems you w11l be using a number
of options, etc. that USU-NETYK allows. If you don't have a lot of time to do
computations you might set up the equations that need to be solved, and use the 7. What head and capacity should the pump produce in order to maintain a
solutions from USU-NETYK to verify that you have used the correct equations to
pressufo~,Of 40 psi at node 3?
describe the problems. The input data for each problem follows, but you should
see if you can make up the input data without looking first.

1. What head is required to produce the flowrate of 500 gpm in the

600 p.pm (3)"-:siY 0.6 c fs
below?
6" - 2500' -4~ ~O psi
8. What is the least cost combination of pump and pipe size to use for the
problem below? The following are basic economic data:
life of pump 20 yrs, capital cost of pumping station, etc.
Life of pipe = 80
$40,000,
500 g~ e ~ .005" of pipe is as program defaults to, costs of energy $0_l2/kilowatt-hour, in-
terest 10%. A single size pipe is to be used .
. - elev I: 40'

1"0-
2. What flowrate occurs between the reservoirs?

~
, 15 c"' _ 3000 m
e = .02 em
57
40 m
What diameters are~needed -14r-"- "Clt.: all nodes?
150 1 l~'L< I
<
..... 3. What flowrate in each pipe? (2)
cfs

,
.....
380'
(3)
, 0 4 cfs
e .004" 1900 ~1 (5) "\
all pipes elev 350' 151 1800 i '6l
10. Pressures were measureo'for the plpe flew below as shown on the sketch.
Determine both"the appropriate Darcy-Weisbach equivalent sand roughness coef-
4. Assume the head produced by the pumps are known and equal to I:1.p(l) =50 ficient, and the Hazen Williams coefficient for the two pipe.
and Hp(2)=40 ft_ What is the flowrate in each pipe? 1. 5 ds
100' 2S00' 28'
1. cfs
3" - 1000' 1--90' 2.
p ~ 36.4 psi
8" - 1500' 0'
t 522 psi
40 psi
p

l
5. The heads of the pumps are Hp(1)=50 ft, Hp(2):40 ft, Hp(3)=35 ft, and
Hp(4)~60 ft_ Find the flow in each pipe_

Q J cfS~ 4" _ 2800' --_ _-

-- \
Hr.L
Q
92.3'
3ds
FLUID MECHANICS PROBLEM 1 SIMPLE FLUID MECHANICS PROBLEM 7.
1* 1*
$SPECIF NFLOW=l $END $SPECIF COEFRO=.005,OUTPU1=2 $END
PIPES PIPE-
1 1 2 2500 6 .005 1 10. 1500. 1 1. 40.
NODES 2 8. 1200. 1 :2 .8 50.
1 -500 0 3 6. 900. :2 3 .6 50.
2 500 0 4 6. 1000. 2
RUN RESER
2 92.3 1 100
4 90
DREAD
SIMPLE FLUID MECHANICS PROBLEM 2 1 50. -3 4 40
1* RUN
$SPECIF NUNIT=3,NFLOW=3,OUTPU1=2 $END
PIPES
1 0 1 1500 15 .02 SIMPLE FLUID MECHANICS PROBLEM 8
2 1 0 1500 15 .02 1*
NODES $SPECIF ICOST=l,NFLOW=l,NPGPM=l $END
1 0 0 PIPE-
RESER 1 18. 1800. 1 -2050. O. :2
1 100 :2 18. 3000. 2 50. O.
2 40 RESER
RUN 2 132.31
RUN
INTEREST=.10
SIMPLE FLUID MECHANICS PROBLEM # 3 LIFE=20
/* PUMPS
$SPECIF OUTPU1=2 $END CAPI=40000.
PIPE- UNIT=.12
1 8. 1500 004 1 O. 360. LIFE=80
2 6. 1200 004 1 END
3 6. 1500 004 1
RESER
1 400 SIMPLE FLUID MECHANICS PROBLEM 9
2 380 1*
3 350 $SPECIF DESIGN~l,NOMSOL=l.IRGL=O $END
RUN PIPES
1 0 1 1000 0 .003
2 1 2 1800.1
SIMPLE FLUID MECHANICS PROBLEM # 4 3 1 5 19001
1* 4 2 3 16001
$SPECIF OUTPU1=2 $END 5 5 6 18001
PIPE- 6 3 4 19001
1 8. 1500 . 003 1 O. 50. 7 6 7 15001
2 8. 2500 003 1 2 O. 50. NODES
3 6. 2200 003 1 2 1 .2 45 40.
4 8. 1000 003 2 2 .5 38
RESER 3 .5 30
1 100 4 1. 20
4 90 5 .4 40
DREAD 6 .6 25
2 50 7 .9 14
3 40 RESER
RUN 1 150.
RUN
SIMPLE FLUID MECHANICS PROBLEM # 5
1* SIMPLE FLUID MECHANICS PROBLEM 10
$SPECIF COEFRO~150,OUTPU1=2 $END 1*
PIPE- $SPECIF DESIGN=3,IHGL=0 $END
1 6. 3500. 1 -3. O. 2 3. O. PIPES
2 6. 3000. 1 2 1 1 2 2500. 10 0
3 4. 2800. 1 2 2 2 3 3000. 8. 0
4 4. 4000. 1 2 NODES
DREAD 1 -2.5 0 52.2
1 50. 2 1.5 20 36.4
2 40. 3 1 0 40.
3 35. RUN
4 60. 3 92.31
RUN
2 92.3
-------------------------
SIMPLE FLUID MECHANICS PROBLEM # 6
1*
$SPECIF COEFRo-.005.0UTPUl-2 $END
PIPE-
1 6. 3500. 1 -3. O. 2 3. O.
2 6. 3000. 1 2
3 4. 2800. 1 2
4 4. 4000. 1 2
BOOSTER
1 .3 60 1 55 1.8 30/
2 .3 50 1 45 1.8 25/
3 .3 40 .6 35 1 25/
4 .15 75 .5 60 1 40/
RUN
2 92.3

-xviii-
 Problem from Fluid Mechanics Text
Below is a pipeline problem taken from a Fluid Mechanics Text Book. Solve
the problem using USU-NETWK.
"At a section of a pipe where it divides into two pipes the pressure is
300 kPa, Z = 40 m and V = 3 m/s. One of the parallel pipes, which is 0.2 m dia
and 400 m long, contains a globe valve, two closed return bends, and a turbine
which extracts 98 Nm/kg of energy from the flow. The other pipe, which is 400
m long, contains a pump which supplies 50 Nm/kg of energy to the flow. The pipes
both exist into a reservoir of surface elevation 50 m. The pipes are commercial
steel. The temperature of the water is 15 0 C. Find the diameter of the pipe
containing the pump if the Q entering the parallel pipes is 0.0877 m3/s."

The following could be done in defining the problem for USU-NETWK: (1)
This could be interpreted as. a 3 pipe network; pipe 1 joins into the 2 parallel
pipes and its other end is attached to a reservoir with the proper surface
elevation just slightly different from the specified head of, z+p/-y+V 2 /2g =
40+300/9.8+9/19.62 = 71.07m specified at the junction of the parallel pipes in
the problem. This pipe 1 will be small in diameter and long so it will carry
a negligible amount of flow, especially since the HGL at its two ends will be
about equal. (2) The two parallel pipes of the problem will feed reservoirs each
with a water surface elevation of 50 m, as stated in the problem. (3) The
turbine will be handled as a pump with a negative head. The information given
for the pump and turbine indicate units of Nm/kg of energy. If these units are
divided by g (the acceleration of gravity), then the more conventional meters
of head are obtained. Thus the extraction of 98 Nm/kg is equivalent to 9.9898
m of head, and the input by the pump of 50 Nm/kg is equivalent to 5.0968 m of
head.
One might use the type #1 DHEAD command for these heads, or alternatively specify
pumps, as suggested above. If pumps are used then either a pump curve must be
generated, or a guess about the flow rate mad~ the power computed, and if this
guess is not very close adjust the horsepower of the pumps. Notice for the
turbine the heads associated with the flow rates will be negative, and the power
will be negative if this option is selected. (4) The minor losses can be
combined into a single minor loss coefficient equal to about 11. With this
approach the for USU-NETWK to solve this problem might be:
Fluid Mechanics problem from text book
/ft
$SPECIF DESIGN=1,NUNIT=2,NFLOW=3,NPGPM=3 $END
PIPES
1 0 1 1000 .05 .00046
2 1 0 400 .2/
3 1 0 400 0./
NODES
1 .0877 40 71.07
RESER
1 71.08
2 50
3 50
MINOR
211
BOOSTER
3 .02 5.2 .04 5.1 .06 4.9 0./
2 .02 -9.5 .05 -10 .11 -14 0./
RUN

-xix-
 UNITS and CONVERSION FACTORS

BASIC CONVERSION FACTORS -- ES TO SI SYSTEM

j----------~---------------------------,.-------------------~
I
I

LENGTH 1 meter (m) :::;; 3.280 839 9 feet (ft) 1 ft = 0.3048 m

FORCE 1 Newton (N) :::;; 0.224 809 lb 1 lb = 4.448 23 N
MASS . 1 kilogram (kg) :::;; 0.068 521 78 slugs 1 slug 14.5939 kg
TEMPERATURE 1 Kelvin :::;; 9/5 Rankine OF = 32 + (9!5tC

LENGTH 1 ft = 0.3048 m 1 m = 3.2808 ft

1 inch = 30.48 m = 2.54 em 1 em :::;; 0.394 in
1 statute mile = 1.609 km 1 km :::;; 0.622 mile
1 nautical mile:::;; 1.852 km 1 km :::;; 0.540 N. mile
----------------------------------------------------------- ------------------------------------------- I
PRESSURE 1 psi = 6894.24 N/m2 1 N/m2 = 0.000145 psi !
2
1 Ib/ft == 47.877 N/m2 1 N/m2 :::;; 0.0209 Ib/ft2 j
1 in Hg = 0.491 psi = 3386 N/m2 :::;; 70.73 lb/ftl I
1 standard atmosphere:::;; 14.699 psi:::;; 2116 Ib/ft2

SPECIFIC 1 Ib/ft3 :::;; 157.07 N/m3 1 N/m3 :::;; 0.006 37 Ib/fl3

WEIGHT 1 Ib/ft3 :::;; 0.0160 g/em3 3
1 gm/em = 62.43 Ib/fl I
------------------------------------------ I
KINEMATIC 1 ftl/s :::;; 0.0929 ml/s 1 m21s :::;; 10.764 ft 2/s
VISCO CITY 1 stoke:::;; 1 em2/s :::;; 0.001 08 ft2/s

FLOW 1 efs :::;; 0.283 m3/s (ems) 1 m3/s = 35.313 cfs

RATES 1 slug/s = 14.594 kg/s 1 kg/s :::;; 0.068 slugsls
1 lb/s = 4.448 N/s 1 N/s :::;; 0.2248 IbIs

POWER 1 ft.lb/s :::;; 1.356 N.m/s 1 N.m/S(watt)=0.737ft.lb/s

1 hp = 33,000 ft.lb/min 550 ft.lb/s = 746 N.m/s = 0.746 kwatts
1 kwatt = 1.341 hp :::;; 738 ft.lb/s :::;; 0.947 Btu/s

KINEMATIC VISCOSITY of WATER

OF 2
ft21s c m /s

32 1.931 x 10-5 0 1.785 x 10"6

II II
40 1.664 5 1.519
50 1.410 " 10 1.306 "
60 1.217 " 15 1.139 II

70 1.059 " 20 1.003 "

80 0.930 " 25 0.893
II II
90 0.826 30 0.800

-xx-
 SUMMARY OF MOST FREQUENTLY USED OPTIONS
(These are included in the $SPECIF list. A * denotes default values)

COEFRO default roughness coef. (.oI02") NOMD1A enter new list of nominal diameters
0* use default nomial diameters
DESIGN analysis or special design sol. 1 read in list of nominal diameters
O' analysis solution
1 special design sol. NOMSOL follow special design sol. with analysis based on
3 special solution to solve for roughness nominal dia.'s (has meaning only if DESIGN""I)
coefficient O stop with special design solution
1 flow design sol. with analysis based
GAMMA specific weight of fluid 62.4 (for water) on nearest nominal diameters

HeiGL type of pressure specification at nodes (No NPLENG length of pipes

meaning if DES1GN=O) 0* in feet ( or meters)
0" psi (or kPa for SI units) 1 in 1000 ft (or 1000 meters)
1 head at node in ft (m for S1 units)
2 HGL at node in ft (m for S1 units) NPRINT level of intermediate output -3 to 10 allowed, 0 is
(in addition 3 & negative values indicate special default level
input will be given)
NUMPIP numbering of pipes
1NI'UTA form of input used 0* pipe numbers given in input data
O' standard in body of manual 1 numbers of pipes not included,
1 Appendix A input (numbered according to order of input
2 Appendix B input #1 data)
3 Appendix B input #2
NUNIT units of pipe data (sol. based on ES or SI units)
ISIML steady-state or time dep. sol. O dia. & e in inches, length in feet
O steady state analyses 1 all units in feet
1 time varying simulation 2 all units in meters (S1 units)(NFLOW>2)
3 dia. & e in cm, length in meters
LENGON 4 dia. & e in mm, length in meters
o length of pipe not given but other x-y
coordinate NSYMMT method of solution to be used
I" length of pipe given o nonsymmetric Jacobian, tighter banding
symmetric Jacobian
NETCHK single or dual definition of network 5* sparse matrix methods
O' single input of network data 2 or 7 smoothing to laminar flow for D.W. eq.
1,2,-1 or -2, duplicate input of network layout
OUTPUI form of solution tables
NETPLT o single table, 130 cols.
o no plot 1* two tables, 80 cols.
1 calcomp plot; write PLTNET plot file 11 pipes & nodes identified by 8 char. strings
2 or 3 dim. perspective plot; write special plot file 12 only pipes identified by 8 char. strings
13 or 14 formatted input from P1PEINPT, etc 13 only nodes identified by 8 char. strings
1000+ ,etc control of format in solution tables

NFI-DW units of demand PCHAR3 no. of pits. to define pump curves

O Q in cfs 5 Ibslsec o more than 3
I Q in gpm 6 Newtons/s 1* 3 points
2 Q in mjd 7 slug/sec
3 Q in m Is 8 kg/s PEAKF (FLOWFC) multiplier of given demands (1.0 default)
4 Q in Iitersls
PRESMA max. pressure that should not be flagged
NI'GI'M units of flows to define pump curves (defauIt--does not flag)
o Q in cfs 5 Ibs/sec
I Q in gpm 6 Newtons/s PRESM1 min. pressure that should not be flagged
2 Q in mJd 7 slug/sec (default --does not flag)
3 Q in m Is 8 kg/s
4 Q in Iitersls VELMAX max. velocity that should not be flagged
(default--does not flag)
NODESP designation of sources of supply
O by pipe that connects them to the VELMIN min. velocity that should not be flagged
network (sources are not given a (default--does not flag)
node no.)
by node assigned to source of supply VISC kinematic viscosity of fluid (1.41 x 10-5 ft 2/sec--ES units,

-xxi-
-xxii-
 SUMMARY OF COMMANDS THAT ENTER DATA
and default order of data after command

PIPES ITERA
No., upstream node, downstream node, length, dia., No. of iterations allowed for solution
coef.
UNITS
NODES designator of type of units
No., demand, elevation (pressure, x & y coord. and
check data may also follow) ERROR
convergence criteria
RESERVOIRS
designation(pipe--NODESP=O,node--NODESP=l), VISCOSITY
water surface elevation kinematic viscosity of fluid

PUMPS WEIGHT
designation, 0I' HI' 02' H2, 03' H3, elev. of water specific weight of fluid
surface
or PEAKF #
designation, power, On' elev. of water surface/ peaking factor, list (range) of nodes

BOOSTER (or BPUMPS) PRINT

designation, 01' HI' 02' H2, 03' H3 , 0 designator of level of intermediate output
or
designation, power, On' 01 DESIGN
or specifies that a special design solution is wanted
designation, coefficient of 2nd degree polyn. (same as DESIGN in $SPECIF list) SETPR &
SETHG pressure or HGL elev. known at some node
ROTATional speed but demand is unknown elsewhere, etc.
designation, ratio of speed to that of data
SETPR & SETHG (pressure or HGL elevation known at
SERiEs some node but demand is unknown)
designation, no. of pumps in series node of known pressure (HGL), node of
unknown,source pipe no., known pressure (HGL)
PARAL designation, no. of pumps in parallel
LPIPE, NLPIPE, LNODE, NLNODE (The lst L or NL refer
DHEAD (differential head) to list, or not list, respectively, and the PIPE and NODE
pipe, AH, node, source of supply, HGL part of the name indicate lines in the PIPE DATA or
NODE DATA table, respectively.)
VALVE (pressure reduction valves, PRV's) List (range) of pipe or node numbers that should, or
pipe, dist. to valve, HGL setting should not be listed in output solution tables

VALVC (check valves) FORMAT (This command allows USU-NETWK to be told

list of pipe numbersl that formatted data follows, and gives the FORmAN
FORMAT of this data.)
BPVAL (back pressure valve, BPV's) FORMAT specification of input data that follows
pipe, dist. from valve to end of pipe, HGL setting including the beginning and ending parentheses

NOZZLES (see manual if NODESP=l) CHECK (no data follows)

pipe, dia., discharge coef., elev., est. of HGL check data for errors but does not obtain a solution

J>IPE-(items below may be omitted, pipe and node number are END (no data follows)
integers; all other values must contain a decimal point) end of detailed input data

pipe, dia., length, coef., upstream node, demand, RUN (no data follows)
elev. (pressure, x-cord, y-cord), downstream node, same meaning as END, if command CHECK is
demand elev. (pressure, x-cord, y-cord) included in input END must be used

MINOR
pipe, loss coef, # implement only on simulation version of USU-NE'lWK

QGIVE #
pipe, flowrate

DFRAC
mult. of following demands

-:xx:iii-
-xxiv-
 INTRODUCTION
Analysis of flow in pipe networks is 2. node oriented input. With pipe oriented input
required to determine the ability of water the network's layout is defined by the end nodes
distribution systems to supply water to residential, associated with each pipe. These end nodes define
commercial, and industrial users at acceptable the network's connectivityand topology. In this
pressures and to meet emergency demands, such as form of input it is convenient to also supply
those needed to fight fires. The digital computer pipe diameters, lengths, and wall roughness
has made analyses of larger pipe networks feasible. coefficients on the same line as the pipe number
However, a good computer program should use and its upstream and downstream nodes.
computer resources wisely, accurately simulate Elevations of nodes and the demands at the nodes
hydraulics, and minimize the effort required by a are additional data, that are needed for an analysis
user to obtain solutions. Program USU-NETWK but which are not needed to define the network's
was developed to meet these requirements. layout. In the standard form of input these
additional node data are given on separate lines.
This manual provides the basic A command PIPE-nodes, however, allows this
information needed in preparing data needed to additional data to be given after any node number,
obtain analyses and design solutions, including as does the form of input described in Appendix
simulations through time, with USU-NETWK. A In preparing the data under the pipe oriented
The manual does not describe the methods used input, category 1 above, the user goes from one
by the program to obtain these solutions. pipe to the next of the network, until he includes
Methods for obtaining such solutions can be found every pipe in the network; therefore it is pipe
in the book "Analysis of Flow in Pipe Networksu1 oriented.
which is devoted to methods for solving large flow
distribution systems, as well as the other technical In the node oriented form of input,
literature on the subject. The material herein is category 2 above, Ute layout, or lopology, of the
intended to acquaint the user with the input data network is defined by giving the pipe numbers
requirements of USu. In brief, however, the which join at each node of the network. In this
program solves the corrective flowrate system of form of input the user moves from node to node
equations using the Newton method. of the network and jots down the pipe numbers
that join at each node. In this form of input the
USU-NETWK has been developed to demand at each node, as well as the elevation of
allow the user much versatility in obtaining the node are logically given on the same line as
computer solutions to any steady-state flow that node. However, such necessary auxiliary data
problems dealing with flow in a piping system. as pipe diameters, lengths and wall roughness
The program has been designed to make it as easy coefficients are given on separate lines. This
as possible to use. The amount of data need to latter node oriented form of input is described in
describe a network is minimal, and free of any Appendix B.
format requirements, e.g. items do not need to be
in any specified columns. The more general There are a couple of reasons for
program allows alternate forms of input. The discussing the different type of input here. One,
form which is considered most basic is described to let you know why the Appendixes are included
in the main portion of this manual. In this in the manual, but more importantly to alert you
standard form of input, names such as PIPES, to the fact that an effective way of discovering
NODES, PUMPS etc. are used to enter data of a errors that you make in input is to use the other
given type. Appendices A and B describe type of data in checking your network. USU-
alternative forms of input. While there are several NETWK prints out the node oriented form of
variations of input data allowed within each data that describes the network whether you use
category, it is useful to consider the input data as the pipe or node oriented input under the default
being in: 1. the category of pipe oriented input, or level of extra output. That is the pipes numbers
are listed that join at each node. Likewise, with
1Jeppson, Roland W., 1976, Butterworth the next level of extra output, a special table is
Publishers, 80 Montvale Avenue, Stoneham, MA printed that gives the node numbers at the ends of
02180-2422. each pipe. When you prepared data on the basis

-1-
of "pipe oriented" input a most effective way of option might also be used if a large number or
checking the network that you defined is to separate analyses are anticipated that do not
examine the tables giving pipes at nodes while change the network. Then the process of defining
looking at the map of the network, and vice the loops can be eiiminated but taking the loops
versus. defined by USU-NETWK from the output file and
editing them into the input file.
By giving USU-NETWK both the node as
well as the pipe oriented input you can have the Extensive internal checking of the data for
program check that the two definitions are consistency and correctness in describing a valid
identical. Any discrepancies will be identified, and network occurs, and meaningful error messages
processing will be terminated. Often giving such are provided to assist the user in locating errors
a dual definition of a network is preferred over should a problem be detected by USU-NETWK.
checking by hand, but it does require more input USU-NETWK has an engineering
data. (See the option NETCHK for a description economic analysis package attached to the
of how to define the layout of a network twice.) hydrauliC analysis packages. Therefore, if desired,
you can request as complete a cost analysis be
The input form described in the main done as you are willing to provide cost data for.
portion of this manual uses a description name, or Having an
command name, to enter data of a given type. engineering economic analysis done as part of the
These names are associated with the type of data
they enter. Some of these names are: PIPES, computer "run" provides you information related
NODES, PUMPS, RESERvoirs, VALVES. There to costs of alternative designs or configurations.
is a list of data items under each name, but the
order of items in the list can be changed with an The general features and capabilities of
option so that often input files that were USU-NETWK, are described in the next section.
developed for other network solvers, can be read This description should assist you in understanding
with minor changes by USU-NETWK. In the the types of networks that can be analyzed, and
alternative forms of input described in Appendix the relatively large number of options that are
A and B, no such command is used. Rather the available. This description will help you decide
different data must be ordered in accordance with what components of your distribution system to
the prescribed sequence. include in the analysis, but you may wish to read
it later, after you have had some experience in
Some users wish to identify pipes and/or using USU-NETWK. You may skip to the
nodes by Characters, and not just numbers. If following section, "Description of Input Data."
desired strings of any 8 characters, including
numbers and blanks, may be used to identify pipes
and/or nodes. When using this option, however,
the free format input is lost for this portion of the
data because field widths of 8 characters must be
used for these identifies. (See the option
OUTPUl for a description of how this is done.)

For any of the forms of input, loops of

the network may be given if desired. Generally
users prefer to let USU-NETWK find the loops in
the network because giving the loops requires
additional data, and puts more burden on the
user, with more possibility for error. Therefore,
giving data for loops in the network is not a
recommended practice. When the user opts to
give loop data, then rather than determining the
loops internally, USU-NETWK will use the loops
that are given. This option may be used if the
program is unable to generate the loops properly
for a given network, which is highly unlikely. This

-2-
 VERSATILITY OF COMPUTER PROGRAM
Features of the program are: amount of additional computer time for a network
analysis its use is recommended.
1. A solution will be provided for any
system of connected pipes with (or without) 4. Loops in the network are determined
supply or source pumps, booster pumps, supply internally. If PRY's are present and a path of
reservoirs or tanks, minor loss devices, pressure connected pipes does not exist between
regulators (or pressure reducing valves, PRV), consecutively listed reservoirs and/or source
check valves and orifices at ends of pipes. The pumps, this is noted, and all alternative reservoirs,
network generally will contain loops, but may and source pumps are examined to find a path. If
consist only of pipes supplied by reservoirs and/or no path is available, it is noted that the PRY's
source pump but without any natural loops, or isolated the network into two or more pressure
may consist only of pipes (with or without loops) zones as the solution continues. If desired, the
with all inflows and outflows known. Branching end points for such pseudo loops between
networks without real or pseudo loops will also be reservoirs and/or source pumps can be specified.
accommodated. Special input allows p consist of This option permits the user to assist the program
of pipes in some branched system to be described in minimizing the amount of computations by
by just a few kilines and the solution for such will assisting the program in producing a narrower
size all pipes to give the specified sof the banded Jacobian matrix. An option can be turned
hydraulic grade line. on the informs USU-NETWK that the pairs of
sources of supply selected between which to form
2. Either English (ES) or International pseudo loops should be optimized to result in the
(SI) units may be used. If ES units are used, pipe least number of terms in these energy equations.
diameters may be given in inches or feet, pipe
lengths in feet of 1,000 feet, and demands (or 5. Reservoirs (or tanks) and source
consumptions) may be in cubic feet per second pumps may supply the network, and booster
(cfs), fallons per minute (gpm) or million gallons pumps may exist in any pipeline. Operating
per day (mgd). If SI units are used, pipe characteristics of pumps may be defined by (a)
diameters may be given in meters or centimeters, supplying three pair (or up to 10 pairs if selected
and lengths in meters or 1,000 meters and by an option) of flow rate versus head produced.
demands (or consumptions) in cubic meters per These values are taken from a pump's
second, or liters per second. In addition to the characteristic curve, (b) by giving the power
above volumetric flowrates, mass or weight (horsepower if ES units or kilowatts if SI units)
flowrates may be used. When using ES units that the pump supplies the flow and its normal
weight flowrates are in Ib/sec and mass flowrates capacity (flow rates may be cfs, gpm or mgd if ES
in slugs/second. When using SI units weight units are used, and m3 /s or liters per second if SI
flowrates are in newtons/second and mass units are used), or (c) by providing the coefficients
flowrates in kg/second. for a 2nd degree polynomial that becomes the
mathematical description of how much head the
3. Frictional head losses may be based pump produces as a function of the flow passing
on the Darcy-Weisbach formula, the Hazen- through it. Any number of pumps may operate in
Williams formula or the Mannings formula. The series or parallel at a given station. If not
program will determine internally whether the specified, one is assumed. The pump
Darcy-Weisbach or Hazen-Williams formula is to characteristics can be given for different rotational
be used by the magnitude of the first roughness speeds than the pumps are operating at, and the
coefficient. When the Darcy-Weisbach formula is rotational speeds of pumps can be changed during
used, no approximations are used. Rather, the time-dependent solutions, or series of solutions for
proper equation is determined and the friction the same basis network with components changed.
factor computed exactly dependent upon both the
Reynolds number and the relative roughness e/O. 6. Input data are free format, i.e., can be
Since the Darcy-Weisbach equation is more punched in any columns with a blank or a comma
fundamentally sound from a hydraulic viewpoint as the delimiter between individual data items.
and its use results in zero (or insignificant) The exception to this rule is that commas separate

-3-
option items in the $SPECIF and $1DATA lists. 10. The printed output is self-explanatory,
Any time the last items are identical to those and the amount of intermediate output is easily
given previously these need not be included in the controlled. 1\vo alternatives are available for the
line of data. The line is terminated by a /. final solution results (not available when doing
Termination of some lists, which may otherwise be time dependent simulations). One such
very long is with a /. When using the PIPE- alternative produces two tables: one with an entry
NODES command items are arbitrarily omitted for each pipe and one with an entry for each
within the record but not terminated with a /. node. (These tables are limited to 80 columns for
Generally pipe lengths are given. However, in teletype output. The other alternate gives the
place of giving lengths, the x- and y-coordinates at node data as part of the entry for each pipe.
nodes may be used to define the physical layout of (This output uses 131 columns.) If checking of
the network. problem specifications is desired before
completing a solution, this can be accomplished by
7. Pressure reducing valves and/or back examining the printed output if the "check" option
pressure valves may be inserted in any pipe of the is specified giving pipes joining at junctions,
network provided they don't exist in consecutive internally generated loops and other such data.
pipes with no other pipes at their joint junction or
the pipe supplied by reservoirs or source pumps, 11. Extensive checking of input data
Le., PR V's don't exist in pipes which are supply occurs, which will identity the majority of the
lines for then they become the effective reservoir. commonly made mistakes. To further verity the
The PRV's act as check valves if flow reverses in correctness of the network the user can select an
pipes containing them during the solution. If the option in the program that allows him to define
upstream head (or pressure) is less than the the layout of the network twice, once using the
valve's setting, the program will note this and pipe oriented form of input and once from the
replace the PRV with a minor loss device. Thus, node oriented form of input. USU-NETWK will
the 3 ways that PR V's operate are simulated, i.e., identity any descrepencies between these two
(a) They may operate in maintaining the definitions of the network.
downstream pressure and head constant. (b)
They may shut off preventing reverse flow but 12. Check valves are allowed in any pipe.
allowing the downstream pressure to drop below Should check valves shut-off, the system of
the valve's pressure setting. (c) They act as check equations being solved is altered to reflect this
values shutting of the flow in that pipe should the rather than setting the resistance of this pipe high
flow attempt to be in the direction opposite to the as is commonly done. Therefore, pipes containing
direction of the PRV. shut-off check valve, or PR V's have exactly zero
The same modes of operation exist of flow through them.
back pressure valves, BPV's.
13. Reservoirs and source pumps may be
8. A string of separate analyses of the given a node number or left unnumbered, i.e.,
same basic network is allowed in which selected given a zero when providing the input data.
components of the network are changed. Any of Generally, they are left unnumbered; thus,
the following changes are allowed: (a) reducing the amount of data required. If such
multiplication of all previous demands by a source nodes are numbered, then reservoirs and
specified factor, (b) the diameter, length or source pumps are identified in the input data by
roughness of selected pipes, (c) the demands at this node number rather than the usualy pipe
selected modes, (d) the water surface elevation of number which connects this source to the network.
selected reservoirs of water surface from which Furthermore, and option can be turned on that
source pumps obtain their supply, (e) the number allows pipes, nodes, or both pipes and nodes to be
of pumps operating in parallel or series may be identified by 8 character strings. Also remarks
changed, (t) the rotational speed of the pump can can be placed at the end of input lines, and these
be changes, (g) the elevation of any nodes, (h) all remarks can be printed in the solution tables.
pipe roughness to a new value, and (i) coefficients Also lines in the input may be comment out" of
for minor loss devices. active use by placing a % in front of them.

9. 1 Minor loss devices can be inserted in 14. The pipes and nodes may be
any pipe of the network. numbered arbitrarily (integers less than 5000), and

-4-
do not need to be entered in the sequence of properties such as the specific weight and
ascending numbers. Furthermore, pipes do not kinematic viSCOSity to that of any desired
need to be numbered. If not numbered, then they incompressible fluid.
are identified in the output by the sequence
number of the input data as well as the nodes at 19. If in using the two table form of
their two ends. output, remarks are to be printed to the right of
lines for certain pipes or node in these tables,
15. Nozzles (or orifices) are allowed at these remarks can be part of the input.
the ends of any pipe. The diameters of these
nozzles can be specified differently than the pipe 20. Three separate forms of input are
diameters and discharge coefficients may be given. allowed: 1) That described in the body of this
Otherwise, default discharge coefficients ar taken. users manual, which will be referred to as the
standard form of input, 2) that described in
16. The program is designed for batch or Appendix A, and 3) that described in Appendix B.
time-shared use. (Use of the program on a PC is there are variations allowed within these different
essential identical to time-shared use in that the forms as well. The standard form of input enteres
user responds to prompts for the input and output data of a given type through a command name.
files, etc.) If use occurs from a teletype under The other forms of input have a prescribed order
time-shared use, the program goes into an and do not use names. The first two forms are
interactive mode allowing input and output files to pipe oriented input data, i.e., the layout or
be designated dynamically by the user. topology of the network is defined by the pipe and
its end nodes. The Appendix B input is node
17. The program can be instructed to go oriented input that defines the layout or topoloyg
into simulation mode. If this vastly expanded of the network from the pipe numbers that join at
capability of the program is utilized, the each node.
performance of the network over any designated In addition to the different forms of input,
period of time is analyzed. options permit changes in the order of items lists
If operated in simulation mode, additional on the more important input records (or cards).
data are needed. These data give: (1) dimension- Consequently, the program can read data directly.
less demand functions for various uses and with few if any changes, that have been prepared
designates which nodes are assocated with each for another program.
such demand function, (2) elevation-storage
capacity curves for reservoirs and tanks and (3) 21. By specifying both pressure (or HGL)
operating rules or schedule for pumping stations, at nodes in addition to the demand, network
and many other commands related to controlling components can be sized. For each such dual
and specifying what changes with time or as a specification a component such as a pump pipe
function of conditions within the system. diameter, or pressure reduction valve is sized. In
In simulation mode considerable flexibility the limit it is possible to size as many components
of output is allowed. (1) Seperate tables such as as there are junctions in the network. In this
those for a single steady stage analysis may be limit an efficient algorithm can be selected that
printed for each time step, or a specified multiple does not require an iterative solution. With wise
of time steps. (2) Special tables giving pressures use of this design capability all pipes can be sized
are designated nodes as a function of time can be by first reducing the network t a branched system
selected. (4) A direct (random) access file that contains as many junctions as pipes, and after
containing the simulation solution can be created. these pip sizes are determined add the remaining
If number (4) is selected, any item of information pipes and have the program size them.
can be printed subsequently as the user, in time-
shared mode or batch mode, requests; that is by 22. A versatile engineering economic
means of simple instructions any desired analysis of the network can be performed. One
information at any time step can be obtained at time capital costs and/or unit coists can be given
near zero computer costs. for various system components. The results of
these cost analyses give the present worth, and
18. USU-NETWK assumes water at equivalent annual reoccurring costs for each such
standard temperature is the fluid. Other fluids component identifed.
can be accommodated by altering the fluid

-5-
 USU-NETWK'S PREPROCESSOR
In this section of the manual the use of prompts the user so he is constantly aware of what
an auxiliary program PIPEINPT is described that is needed next. The output file written by
assists in preparing input data files for use by PIPEINPT consists of input data that can be used
USU-NETWK. Because this program may be for the steady-state version of USU-NETWK that
used prior to using USU-NETWK for obtaining a runs on PC's directly, but if desired this file can
solution of a network problem it is referred to as also be used for the larger version of the program
USU-NETWK's preprocessor PIPEINPT. The use that also will performance a solution through
of this preprocessor program, is not required. time. If the later solution through time is desired
Some people will prefer to get acquainted with the the output file from PIPEINPT can be used for
input data allowed by USU-NETWK from reading the majority of the input, and the additional data
this manual, and then using an editor, or even that describes the time varying characteristics of
word processor that they are acquainted with in the problem can be added thereafter using a
preparing the input data. Others will find the use editor, or word processor.
of PIPEINPT helpful in preparing input data files.
However, PIPEINPT was never designed to do all The program PIPEINPT prompts its user
the editing, etc. that is generally done in relative to what it wants next, and therefore the
moditying a given data file for the various best method for getting acquainted with its use is
solutions that are typically obtained in a complete to run the program as you read this three page
study of a water system. There is an full screen description~
editor with a help file that describes how to use it
that can be acquired as part of PIPEINPT, but Since program PIPEINPT writes a formatted
this editor is generally not included in the file that can be read more rapidly by
PIPEINPT program since most individual will not USU-NETWK than the format free file that it will
want to learn another editor, but use the editor also read, a small amount of computer time saving
(or word processor) they are familiar with to edit will also result from use of PIPEINPT. The
USU-NETWK files. PIPEINPT was designed option NETPLT=13 should be set to take
specifically for users that have the PC version of advantage of this fixed format input. If a
USU-NETWK. It only runs on IBM compatible graphical display is desired then NETPLT must be
PC's, but can be used to prepare input data files changed to 14 (and also LENGON=O, unless the
for USU-NETWK that is running on a larger X- and Y- coordinates are to be added to the file
system. If this later mode of use occurs a that USU-NETWK will write for input to the plot
convenient method of operation is to let program PLTNET) in the $SPECIF list when this
PIPEINPT help prepare input data files on the PC file is to be used on an IBM compatible Pc.
and then transfer these files to the disk of the
larger computer system with a software The basis input to USU-NETWK might
communications package such as KERMIT, or be divided into the following three categories: 1.
CROSSTALK. The title of the network for which the data
applies. (This title is terminated by an /*.) 2. The
The description of the use of PIPEINPT list of options that are enclosed between $SPECIF
given below assumes that you have some and $END, and 3. The detailed data that
familiarity with the input data requirements of described the layout of the network, its pipe sizes,
USU-NETWK. As a minimum familiarity with the demands for which the solution applies, and
this input you should read the PREVIEW section the operation of pumps and reservoir to meet
in this manual and obtain solutions to the small these demands. The Preprocessor program
networks given there by executing USU-NETWK PIPEINPT utilizes these three categories of input
in solving them before reading this section, and in the order in which they are listed above.
using PIPEINPT.
The first question that PIPEINPT asks its
The preprocessor program PIPEINPT is user is to be supplies a file name that will contain
designed to help prepare input data for the its output file, or the input data for USU-
network analysis and design program NETWK. This name must conform to the
USU-NETWK. This preprocessor program MS-DOS file naming requirements. It can consist

6
of a drive designation, a path and a file name with the command name, or by typing the command
an extension following the file name. For example name. If you select to type the command name
C:/YY/FIL1.DAT. Generally, however, the default make sure that the "caps lock" key is on. Upon
drive and directory will be used and, therefore, the selecting the command it will appear at the top of
file name with an extension will be all that is the screen waiting for you to enter the data called
given, i.e. FILl.DAT. Next the user is asked for for under this command. The prompt at the
a title that he wishes to give the network. The bottom of the screen will list the items that must
title typed in by the user will be duplicated in the be entered under this command in the order in
file written by PIPEINPT. This title can consist of which they are to be entered.
up to 10 lines, and must be terminated on a new
line with a/*. Immediately upon pressing the In entering the data under any command
return key after typing the /* the user is prompted it is important to note that the following three
with a screen full of the available options that keys have special meanings: 1. the space bar
exist in the PC version of USU-NETWK, and (causes the value just entered to be placed into
what their default values are set to. To change the file and the input line at the bottom of the
any of these options type in the number that is screen now applies for the next item). 2. the
shown associated with an option. For those return key (causes this line of input to be
options that permit only two settings (I.e. a 0 or completed. If not all entries called for have been
a 1) the screen will show this option immediately entered the line will be completed with values
changed. For options for which 3. or more previously entered.), and 3. the divide key or I
possibilities are allowed, a prompt appears at the (causes the input under this command to be
bottom of the screen that shows the valid values, completed, and the bottom prompt lines to show
or indicates what the option means. For example the available commands again). Thus if you
if you type in 1, which is the number associated should press the return key at the beginning of
with the option NFLOW, you are shown that valid the line this line will be a duplicate of the
values are from 0 through 8. If you type in the previous line, except pipe and node numbers
number associated with VISC you are prompted which will be incremented by 1. Therefore, you
by the lower line on the screen telling you this is want to be careful to not press the space bar, the
the kinematic viscosity. For these latter type return key or the I key except when appropriate.
options you must next type in the appropriate However, by pressing the space bar after a value
value. Upon pressing the return key this value has been completely entered, or the return key if
will appear after the = following the option name. the remaining items are the same as on the
You must also press the return key after entering previous line you can enter the data rapidly with
the number associated with the option. To let a minimum of key strokes. The return key should
program PIPEINPT know that you are through be pressed when you complete a line rather than
changing options you type a number such as 43, the space bar. If you type a number in incorrectly
which is larger than the last number associated it can be corrected using the backspace key prior
with the last option, and press the return key. to pressing the space bar or return key. Simply
Only those options that are changed will appear in press the backspace key as many times as needed.
the options list of the output file. Should you Each time you press backspace the cursor will
want an option to appear in this output file with move backward over one more character. Now
the default value, you can either select it twice if type in the correct value. The value that appears
it is a two valued option or type in the same value in the input line at the bottom of the screen will
as its default is if it is a multivalued option. be entered in the proper column in the table
shown above on the screen.
Next you are prompted for the detailed
data that describes the physical characteristics of Up to twenty lines of input will appear on
the network. This data will be entered with a the screen under any given command. After this
command name such as PIPES, NODES, etc. To the top line will disappear from the screen as the
remind you what these commands consist of you last line is entered. A very limited editing
will see them listed on the bottom two lines, with capability of values in these 20 lines that appear
a prompt line above this that asks you to select on the screen can be accomplished as described
the command name for the type of input you wish below.
to give next. This command name can be selected
by either typing the number that is associated with The last command that should be selected

-7-
is END (or its item 12), or STOP (or its item 13). entered PIPEINPT permits you to correct values
Upon selecting the END command PIPEINPT in the detailed data that follows the command
assumes that you are through with the input for such as PIPES, NODES etc. as long as these
the network and supplies the RUN command at values remain on the screen, i.e. are within the 20
the end. It writes the rest of the data on its lines under any command that are not scrolled off
output file, i.e. the file name you supplied the top of the screen. Other changes, or editing
previously, and terminates. If the STOP is of the file must be done with an editor or word
selected instead, the RUN is not added to the end processor.
of the file in anticipation that you wish to supply
additional data later. If at a later time you give To utilize this limited editing capability
this same file name, you will be informed that the simply press one of the arrow keys on the
file already exists, and asks if you wish to add to extended pad portion of the keyboard. Pressing of
it. Upon responding with a Y or YES, any data these arrow key moves the cursor to an entry on
provided will be appended at the end of the the screen in the direction of that arrow key. The
existing file. If you terminate with an END during item on which the cursor is directed is shown
this session, the RUN command will be the last blinking. You simply move to the entry that
line written to the file. needs to be corrected by pressing the appropriate
arrow keys. When you are at this value press the
The data file written by PIPEINPT will space bar to indicate that this is the number you
conform with the requirements described in the wish to correct. At the bottom of the screen you
user's manual except in the case of source pumps, will see a prompt to enter the correct value. Type
or booster pumps if you select this option from in the corrected value and press the return key.
PIPEINPT. The data requested from you by If there are additional values to correct press one
PIPEINPT for pumps is identical to that described of the arrow keys again and repeat the above
in the user's manual, namely a pipe (or node if process. If you wish to delete an entire line this
NODESP=I) and three pair of flowrates and can be done by pressing the delete key in place of
corresponding heads produced by the pump giving the corrected value. The entire line that
followed by the water surface elevation of the the blinking value is on will disappear. When you
supply for a source pump. However, the output are through making such corrections, press the E
file will contain the coefficients of a second degree key to return to the position you were at in
polynomial that passes through the three points entering data prior to pressing an arrow key.
that you give, the water surface elevation and the
normal capacity of the pump, i.e. the middle, or
second flowrate value you give. If you wish to use ILLUSTRATIVE EXAMPLE IN USING
the option of giving the power supplied to the PIPEINPT
fluid by the pump and its normal capacity just As a simple example to get acquainted
enter a zero (0) for the next item called for. The with PIPEINPT you should execute this program
flowrates given in connection with pumps must be to prepare the input data file to solve the network
in the units designated by the option NPGPM. problem illustrated below. Read the following
USU-NETWK expects to read data for pumps as description as you execute PIPEINPT. The first
written by PIPEINPT when the option prompt you receive from PIPEINPT is the name
NETPLT=13 or 14. If NETPLT = 14, then when of the data file that the problem description
this data is submitted to USU-NETWK for a should be written to. You might give
solution it will write a file that contains TESTPI.DAT for this file name. You will next be
information needed by the post processor program ask to give a title for the network problem and to
PLTNET to graphically display the network on the terminate giving lines of title with */. Type in
screen. However, in order to use this graphical the following:
capability it is necessary that x- and y- coordinates
by given in describing the physical layout of the Illustration to use PIPEINPT
1*
network.
Next you will see the screen filled with the default
Limited Editing Capabilities options. Since all of these apply to this problem
type 43. The bottom of the screen now displays
In addition to allowing you to correct a the command names. Type 1 to select PIPES and
value with the backspace key before it has been enter: 1 0 1 3000 15 .005 Enter (press the space

-8-
bar between values, and the Enter key at the end TESTPI.DAT when ask for the input data file.
of the line). For the second line of PIPES data You might type TTY for the output file to have
type: the results from the solution displayed directly to
2 1 2 4000 12 Enter the screen.
Note that the coefficient for the 1st pipe has been
added upon pressing the enter key. For pipe 3
just press the Enter key, and note that the data
Q hp
from pipe 2 is taken to fill in this line. For pipe (cfs) (ft)
4 type
2.0 30
4 3 4 4000 8 Enter 3.5 24
5.0 15 1 cfs
and press only the Enter key for pipe 5.
Next press the divide (or forward slash) cfs
key I to get the command names to appear at the
bottom of the screen again. Now press 2 to select
10-6000
the NODES command and type in the following: " ,,\:)\:)
\..' I :;'V cfs
1 1 500 Enter ,<::l/
Enter All e ::: 0.005"
~~t:~ 700' All Elev. ::: 500'
Enter

Note the 4 press of the Enter key for nodes 2, 3,

4 and 5 fill in the proper dat for these nodes.
Now I again and type 4 to select BOOST. You
will see the prompt:
Should output contain: (1) 3 pts for P. Char; or
(2) Coef. for polyn.
Give 1 or 2
Type 2 and then enter the booster pumps
characteristics by typing:
2 2 30 3.5 24 5 15 Enter

Press the I key again, and this time type 5 to

select RESER and type
1 700 Enter

and press the I and terminate by selecting 12 for

END (selection of STOP assumes you want to
add to this file at a later time.) The data file that
PIPEINPT has written for you should look like
the following:
Illustration in using PIPEINPT
/*
SSPECIF NETPLT=13, SEND
PIPES
1 0 1 3000.0 15.00 0.00500
2 1 2 4000.0 12.00 0.00500
3 2 3 4000.0 12.00 0.00500
4 3 4 4000.0 8.00 0.00500
5 4 5 4000.0 8.00 0.00500
6 2 4 6000.0 10.00 0.00500
NODES
1 1.000 500.00
2 1.000 500.00
3 1.000 500.00
4 1.000 500.00
5 1.000 500.00
BOOST
-2 -0.66667 -0.3333 33.3333 0.0000
3.5000
RESER
1 700.00
RUN

Now execute USU-NETWK and type in the name

Table. 1. Order and nature of input data.
=======================================================
Category Pages in
of Input Description of Input manual

General Specifications
(a) Title of network up to 10 records long and terminated with a 1* (or *1) in columns
1 and 2 of final record. (For IBM computers the */ must be uSed). (b) The list of
1 options beginning with $SPECIF (or &SPECIF) in column 2 and ending with $END (or &END).
The list between these two consist of the parameter name followed by a comma. Only
those parameters whose values are to be set different from the default need to be
included.

Detailed Data of Network

This data consists of a command name such as PIPES, NODED, PUMPS, etc. followed on
the nest record by data of the type identified by the command name. The last such
2* command must be END or RUN. If special opetions are given some additional data may
be required after the END or RUN in addition to the data given in categories 3,4,
and/or 5.

Data for Alternate Steady-State Analyses

Givine Chanees to Basis Network
Each group of change data begins with the word CHANGE in column 1 and terminates with
the word END in column 1. Commands such as DIAMETER, LENGTH, DEMAND, etc. enter
3 changes to the existing network. These commands are followed by appropriate
change data, and occur between a change and end command. Any number of groups
change of data can follow each other. Such changes are accumulative.

Data for Time Dependent Solution of Network

This data is read only if the option ISIM=1 is set in the $SPECIF list and includes:
(a) A first record which is an options list entered with $TDATA (or $TDATA)
4 (b) The data giving the time dependent characteristics of the network which
is entered with command name suchn as DEMAND FUNCTION, PUMP RULES, etc. followed by
the appropriate data. The last command is END SIML or END. (Generally categories
3 or 4 are not used for the same problem. If both are given then category #3 data
follows category #4 data.)

Data for Cost Analyses

This data is read only if ICOST is set greater than 0 in the $SPECIF list of options.
It consists of costs associated with the various components of the network. It is
entered with commands such as INTEREST=i, LIFE=n, or ELECTRICITY, PIPES, etc. followed
by CAPI, or UNIT and the dollar amounts, and the designation of compoinents for which
the costs apply. The cost data is terminated by the command END. These cost data
5 precede each set of change data of category #3 and apply for the preceding analysis.
Cost data follow category #4 data, if called for. If ISIML= 1 and ICOST>O a cost
analysiS will follow each time step solution; if ICOST=2 every other time step; etc.
For each such new time step cost analysis called for at least an END command must
be given.

"'This second category of input is different if the forms of input described in Appendix A or B are selected.
The selection of these alternate forms of input is controlled by the option INPUTA in the $SPECIF list.

-10-
 GENERAL SPECIFICATIONS
(OPTIONS)

INPUT LINE # 1 - TITLE

The first line, or lines, of input data some older FORTRANS do). The terminating
consist of a title for the network. This title will $END may be on the same line, or the next line,
be written on the output file at its beginning for but must be given to terminate the list of options.
identification purposes. The title can be up to 10 If $END is not given the entire data file will be
lines long. Each line can be up to 80 characters examined looking for it, and when not found,
long. The line immediately after the title, which execution will be terminated. The $ sign can
terminates reading of the title must contain a /* generally be replaced by & sign if desired, at least
(or for larger IBM computers a *f) in columns 1 this is possible if the program does its own
and 2. In the examples in this manual the 1* will options input.
be used. It is desirable practice to put in the title The length of any line listing options is
enough to distinguish any given "run" from other limited to 80 characters. If the list is too long for
"run" that may be made so that it is not necessary one line continue onto other lines, but terminate
to examine detail in the solution to determine a line with a parameter and its value. In other
what a particular output is for. For example the words the parameter, the equal sign (=) and its
title may have extra lines added to the name for value should not be split between two lines. The
the water distribution system such as: parameters (which are listed and described in the
following pages) may appear in any order. Only
FIRE DEMANDS OF 1200 gpm AND 1000 gpm those parameters for which values different than
AT NODES 30 AND 54, RESPECTIVELY. those that are used as "default" values need to be
included in the list, but may be included if desired
The title can contain lower and uppercase even if they are given the same values that USU-
characters and numbers. Options, and command NETWK uses as default if not given. At the time
names, which are described in the following pages, the program is compiled it is easy to change the
must be upper case characters. default values. Generally, however, the default
values will be those described in this manual.
Any of the following variations are valid
INPUT LINE #2 - PROBLEM specification, or option's lists:
SPECIFICATIONS OR OPTIONS
$SPECIF NPRINT=2,NFLOW=1,NPGPM=1,
This line, or lines, starts with $SPECIF, VISC=.9E- 5 $END
contains any or all of the list of names,
e.g. parameters, followed by an equal sign and a $SPECIFNPRINT=2,NFLOW=1,VISC=.OOOO09
value that are listed in the following pages. NPGPM=1 $END
Should The version of the program you are using
utilize FORTRAN's NAMELIST capabilities, then $SPECIF NODESP= 1 $END
the $SPECIF must beginning in column 2. If the
version of the program you are using does its own (Note that upper case, or capital letters, are
input, then the $SPECIF may begin in column 1 required).
or any other column. The list of parameters can
be separated by either a blank or comma (unless
the FORTRAN NAMELIST requires commas,
as

-11-
Alphabetical list of parameter names that may occur in $SPECIF list.
=========================================================
COEFRO
ER8
12 CORNOD 12
16
'* CORPIN
FLOWFC
'* 12 CORPU I 13
12 GAM1A 16
CORPUM f 13
HLETTE I 16
DESIGN
ICOST
13
17
E~UDIA
I GL
16
INITAL
!RESID
'* 222416 ERROR
INPUT
ISIML
''**
22
24
INPUTA A/I
ITERA
23 IOEX
24 LENGON
34
24
roUT I
LENORG
23
25
roUTI #
LOOPRD
23
25
IOUTO /I
LOOPSE
17
23
26
MANNI 28 MAX 28 NCONTI 28 NELEV 28 N~UAT 29 NEQUDI I 29 NETCHK 29
NETPLT 30 NEWSTR I 31 NFLOW 31 NHAZEN 32 N HK # 32 NODESO 32 NODESP 32
NOMOIA 32
34
NOMSOL
Nl'RINT
32
34
NOPARP
Nl'Rl'RV
'* 33 NOSTOP'*
34 Nl'RPUM
33
35
NPERCT 33
Nl'RRES I 35
NPGPM
Nl'RTRM
34
35
NPLENG
NPSERI
34
35
36 NSYMMT AI 36 NTRAND I 38 38 NUNIT 38 OUTPU1 39 PCHAR3 43
44 PLOTH A$ 44 PRESMA 44 44 l'RVLOS 44 REYNUM 44 TVSUM 45
SUM1 # 45 TVSUM2 45 VELMAX 45 45 VISC 45

#These options are not included in the allowable $SPECIF list of the version of the program that
only does steady-state analysis. The Calcomp graphics plotting capability may be obtained with the steady-
state version of the program in which even NETPLT, HLETIE and PLOTH will be included.
'" These options are not included in the allowable $SPECIF list of the PC-DOS version of USU-
NETWK that does have the time-dependent solution capabilities.

The list of parameters, or options, that CORNOD#

may be include in this $SPECIF list are given in
the following pages in alphabetical order. Their Allows the standard order of data after the
effects on the program's behavior are described in NODES command to be altered. If CORNOD=l.
the paragraphs under these individual names. In then the altered order is specified. The four items
a few cases either of two names may be used with whose order can be altered are: 1. the demand,
identically the same effect. These alternative 2. the node elevation, 3. the pressure or HGL if
parameter names are given enclosed in DESIGN is larger than 0, 4. the x coordinate of
parenthesis. The above table lists summarized the the node, and 5. the y coordinate of the node.
valid parameter names. Note the node number must always be the first
You will find this table useful to scan for item. Also if NETCHK= 1 in the $SPECIF list
the spelling of the parameter, and then you can then it is also necessary that the node oriented
quickly turn to the indicated page to read about input that lists the pipes joining at the each node
the effect of the parameter, and values it may take must be given last. If CORNOD=1 appears in
on. the $SPECIF list, then immediately after this
Those names followed by # are valid only optional list, a list of integer numbers, giving the
for the version of the program with time new order of the items, is given. For example if
dependent solution capabilities, e.g. the simulation the elevation and demand are interchanged after
version of USU-NETWK. A PC version under the NODES command this extfa line of input
DOS of the simulation version implements most would consist of: 2,1,3,4 (or just 2,1/ if coordinates
of the options available in the full program are not given). The 2 at the first of this list
designed for larger computers, but those that have indicates that the standard second item after the
been deleted from this PC version are denoted by node number, i.e. the elevation, is first, and the
a "'. standard first item after the node number is in
second position, etc.
COEFRO Should CORPIN=1 also be in the
$SPECIF list, then these integer values will be the
The default roughness coefficient. This roughness second line after the $END.
coefficient may be the equivalent sand roughness,
e, for use in the Darcy-Weisbach equation, the CORPIN #
Hazen-William coefficient, C, or Manning's n. It
will be used as the wall roughness for pipes until Allows the order of items after the command
a values is given under the PIPES command, or it PIPE-nodes, or the input described in Appendix A
will be used whenever a roughness values is not to be altered. Of the four commands (CORPIP,
given on any line after the PIPE- command. CORNOD, CORPUM & CORPIN) that allow
(default, e = 0.0102 inches for use in the changing the order of items after commands, this
Darcy-Weisbach Eq.) one is the most powerful. Through its proper use
USU-NETWK can be made to read data that has

-12-
 been entered in almost any way. Depending upon pipe and node numbers, as the first three items,
= whether CORP IN = I, 2 or 3, the following is must be right justified in fields of 5 and the
allowed: remaining 3 values in fields of 10 as specified by
If CORPIN - 1. The order of data after the format. The difference is that whenever a
PIPE- is altered. If so, the first record after the roughness coefficient is not given by leaving
command PIPE- must contain the new positions columns 36-45 blank the value given by
for the following 10 items: (1) the pipe no., (2) COEFRO=.OO5 will be used rather than that
the diameter, (3) the length, (4) the roughness given for a previous pipe in the list.
coefficient, (5) the upstream node, (6) the demand
at the upstream node, (7) the elevation at the
upstream node, (8) the downstream node, (9) the ILLUSTRATION OF USE OF CORP IN =2
demand at the downstream node, and (10) the /*
elevation at the downstream node. This data must $SPECIF CORPIN=2,COEFRO=.OO5 SEND
be entered so that it can be read under the format NODES
(F5.0,3FI0.0,F5.0,FI0.0,F5.0,2FI0.0). 1 .8 100
If CORPIN - 2. The effect is the same 21.2/
if CORPIN=2, as that described above for 3 1./
CORPIN =I, except that in addition to the order PIPE
of the data being changed the format of the input (3F5.0,3FI0.0,4F5.0)
data must be specified using F,E or G fields. This 1,5,4,6,2,7,8,3,9,10
format specification is given before the integers 1 0 1 1200. 8. .015
that define the new order. For example, if 2 1 2 1400. 6.
CORP IN =2, the first two lines after the PIPE- 3 1 2 1000. 6. .025
command (or immediately before the pipe data if 4 3 2 1500. 6.
Appendix A input is used) could be, RESER
(3F5.0,3FI0.5,5X,4F5.0) 1200
1,5,4,6,2,7,8,3,9,10 RUN
These two lines will cause input to be in the order
as that after the PIPES command except that it As another example assume roughness
must be in the specified columns. (Formatted coefficients are given after the downstream node
input under the PIPES (and other commands) also and before the pipe length. Then if the $SPECIF
occurs if option NETPLT=13.) To understand list contains CORPIN=2, the first two lines after
how these last 10 values determine the new the command PIPE- would consist of:
positions, list the 10 input items in their usual (3F5.0,3FlO.5,4F5.0)
order as described under the command PIPE-. 1,6,5,4,2,7,8,3,9,10
The 10 given values indicate the new positions for With CORPIN=2 the format of the input
each of these 10 items, e.g. pipe no. is in position data is specified and, if desired, the usual order
1, diameter is in position 5, length is in position can be retained by giving 1,2,3,4,5,6,7,8,9,10 as the
4, etc. Items 6 (demand at the upstream node), 7 new order. Input efficiency can thus be gained by
(elevation at the upstream node), 9 (demand at allowing fixed format input over the free format
the downstream node) and 10 (elevation at input. (See NETPLT=13, IOUTI & IOUTO,
downstream mode) are given positions 7,8,9 and also.)
10, respectively, beyond the portion of the Jine If CORPIN - 3. The effect is to read the
that will actually contain any data. Thus these last input according to the standard format
4 items will be read in as zeros. Since the rules (F5.0,3FI0.0,F5.0,2FI0.0,F5.0,2FI0.0)
under PIPE- allow these items to be omitted the but not rearrange the order of the input. Thus if
zeros have no effect. The values for these items CORPIN =3 no additional lines are required after
can be filled in with the NODES command. the PIPE command. However, the data must
As a simple example consider the 4 pipe conform to the columns specified in the above
network supplied by one reservoir as defined by format. This restriction applies whether the data
the data below. In this data the option is entered via the command PIPE, or under the
CORPIN=2 is used. The data after the command input described in Appendix A resulting from
PIPE- is that under the PIPES command because including INPUTA=1 in the $SPECIF list.
of the new order specified by the second line of
integers after the PIPE command, except that the Since many items of input can be omitted

-13-
using the PIPE- command and it can be used in curve are provided (PCHAR3=O in the $SPECIF
conjunction with other commands, considerable list), they must occur after the reordered data. If
flexibility is available. If data files exist for other CORPUM=I, then the first line after the
network analyses programs these files may be $SPECIF list (or after the lines required if
readable by USU-NETWK. almost directly. By CORPIP=I, and CORNOD=I) must contain 7
having USU-NETWK. write data file, or have the integers that either designate (a) the new positions
post-processor, PLTNET write such a file, this for (1) QI' (2) HI' (3) Q2' (4) H 2, (5) Q3' (6) H3,
data can be converted to the standard input used ,and (7) elevation of water surface, or (b) the new
by USU-NETWK. for subsequent analyses after it positions for (1) the power the pump supplies to
has once been read, if desired. the fluid, (2) its normal capacity, and (3) the
elevation of the water surface. For case (b) only
CORPIP # the first three values have any meaning but 7
integer values are required and can be given as
This option, when set to 1 (Le. 4,5,6 & 7. Caution must be exercised in mixing
CORPIP= I), allows a new order of items after the methods of specifying pump characteristics
PIPES command to be specified. This new order between using 3 points on the curve and power
is specified by giving the new positions for the plus normal capacity, since altering the order
length, diameter and roughness coefficient alters the order for all pump input data regardless
immediately after the $SPECIF list. Note that the of the method used to designate pump
first 3 items, the pipe number, the upstream node performance. (See explanation under CORPIP for
and the downstream node cannot have their order understanding how the new order is defined.)
changed. The new order is specified by giving an For example, if the water surface elevation
integer value that represents the new position of is to occur immediately after the pump
items in the list whose order can be changed. For designation (pipe no., or node no. depending upon
example if data after the PIPES command consists NODESP) then the following integers would be
of: 1. the pipe no., 2. the upstream node, 3. the given: 2,3,4,5,6,7,1.
downstream node, 4. the diameter, 5. the length,
and 6. the roughness coefficient (Le. the length DESIGN
and diameter are interchanged from the standard
order), then this line would consist of: 2,1,3. This option is how you tell USU-NETWK
To understand how to designate these new that you want special design solutions rather then
position numbers, list the variables that can be an analysis solution. Values that DESIGN may
changed in their standard order, Le. I-diameter, have are: 0 (default - analysis solution), 1
2-length, and 3-coefficient. The first integer above (determine diameters for NJ pipes from pressures
designates the new position for the diameter, i.e. at all nodes), 2 (special "least cost" design
2. The second integer 1, since it gives position for described in Appendix D), 3 (determine wall
the length, indicates the length is first, and the roughness coefficients for NJ pipes), and 4 (and 5)
final integer 3 indicates that the roughness (specified velocities in branched pipes)
coefficient retains is 3rd position of the variables
whose positions can be changed. If DESIGN - 1. When DESIGN = 1, then
USU-NETWK is told that it is to determine the
CORPUM# diameters of as many pipes as there are junctions
(nodes), NJ, in the network, by means of a
This option, when set to 1 (i.e. computational efficient method. (NJ = number of
CORPUM=I), allows a new order of items after junctions, or nodes, of the network excluding
the PUMPS, BPUMP and BOOST commands. sources if these are numbered as nodes with the
The pump's designation must be first, but the option NODESP=1.) An alternative, but much
order of the 3 pairs of values that define the poorer choice than setting DESIGN = 1, is to
pump curve, and the pump's sump, or wet well introduce as many differential head devices as
water surface, elevation (or inlet pressure) can be there are network nodes and specify the
altered, or if the pump's performance is defined by pressures, or HGL elevations at all nodes by using
its power and normal capacity, the order of these the DHEAD command. Doing the latter adds NJ
and the water surface elevation can be altered. If additional equations to the system of corrective
more than 3 points on the pump characteristic flowrate equations being solved. With
DESIGN=I the solution is accomplished with NJ

-14-
 linear equations. prepared. A means of proceeding is to make
When DESIGN=l the pressure, the adjustments after the program has indicated that
pressure head, or elevation of the HGL, your specifications do not satisfy this second
respectively according to whether IHGL=O, 1 or restriction, and through this process acquire a
2, must be given at all nodes of the network. better understanding of the network's
When using the NODES command these pressure performance. A better alternative is to start with
values are given after the nodal elevation, as a 4th a network with as many junctions as there are
item in the list of input. When using the PIPE- pipe, because you can always then determine the
command these nodal pressure values are given as direction that a flow must be in and specify an
an additional data item containing a decimal point appropriate slope for the HGL. Thereafter add
after either the upstream nodal elevation, or the minimum pipe diameters to those pipes that form
downstream nodal elevation. A demand, and the loops of the system. The option IHGL allow
elevation must be given before the pressure when special input to make this easier for you.
using the PIPE- command. If DESIGN = 2. When DESIGN =2 in
Since NJ pipe diameters will be the $SPECIF list, then a special interactive
determined as part of the solution, it is necessary portion of USU-NETWK (the full program, but
to designate NJ pipes as having unknown not on PC's under DOS) is activated that assists
diameters. The remaining, NL=No. of pipes - NJ, in determining the "least cost" design. This design
diameters must be known. The pipes whose may suggest alternate pipe layout configurations.
diameters are unknown must be given a diameter Its use is described in Appendix D of this manual.
equal to zero in the data following the PIPES
command or the PIPE- command. An alternative If DESIGN = 3. When DESIGN=3 in
for branched networks with as many nodes as the $SPECIF list, then a special solution is
pipes is to use the special input described under requested to determine as many pipe wall
IHOL. The following additional options: IHGL, roughness coefficients as there are junctions in the
NOMSOL and NOMDIA provide control over the network. The same rules apply when DESIGN =3
special design solution activated by DESIGN = 1, as when DESIGN=l with the exception that NJ
or DESIGN =3. coefficients must be given zero values. When
In preparing data for this special design DESIGN =3 all pipe diameters must be given and
solution the following two restrictions must not be the roughness coefficients of the pipes are solved
violated. 1. At least one pipe at each junction so that the given pressures, heads, or HGL
must have its diameter unknown, i.e. not all pipes elevations (depending on the option IHGL) are
at a junction can have their diameters specified, satisfied. The roughness coefficients that are
and 2. The HGL's produced from the specified solved for can be: (1) the equivalent sand
nodal pressures must be consistent with the roughness of the pipe for use in the Darcy-
direction that the flow in the pipe must have to Weisbach equation, (2) the Hazen-Williams C, or
satisfy the demand here. A simple case of (3) the Manning's n. Since coefficients are given
violating this second restriction would be a dead as zero, it is necessary to provide NEQUAT or
end pipe with a positive demand at its end, but an NHAZEN in the $SPECIF list with the
HGL elevation at its downstream end above the appropriate value if either Corn are to be solved
HGL elevation at its upstream end. A slightly for.
more complex situation would exist at a node of
degree 2 with one pipe given a diameter and the If DESIGN = 4. When DESIGN=4 in
other not, and for which the flow rate in the pipe the SSPECIF list, then you must give the
of known diameter (which is determined by its end velocities that are to occur in pipes that are not
HGL elevations, length, diameter and roughness) included in any loops of the network rather than
is less than the demand at the node, yet the HGL their diameters, and the diameters for those pipes
elevations for the other pipe requires that its flow that are in loops of the network. These velocities
be away from the junction. Nodes of degree 3, or are given whether the diameters are called for
higher, may violate the second restriction, also if under the PIPES or PIPE- command.
at least one pipe does not have its flow in the DESIGN=4 is a special design in which pipes that
direction required by continuity. It is generally are branches, without loop, may be sized to give
difficult to not violate this second restriction, desired velocities.
unless considerable understanding of the network's If the network is a branched system all
performance exists when the data is being pipes will have their diameters determined to give

-IS-
 the specified velocities. However, for looped If DESIGN =4 is given, and the network
networks you must understand fully the layout of is not given any sources of supply (i.e. no RESER
the network so that as the data is prepared you or PUMPS commands occur in the input data),
will know whether you are giving a pipe diameter, then it is necessary to give a node number, and its
or a velocity in a pipe. USU-NETWK will assist HGL elevation after the RUN (or END)
you by providing you a list of pipes whose command since this is then a special network with
diameters must be given and whose velocities must all external flows specified.
be given. You request this assistance by setting
DESIGN =5, and prepare the input data without EQUDIA
being concerned whether a pipe is in a loop or See NEQUDI
not. This input could give a value for the pipe
diameter in the first line after the PIPES
command, and terminate all other lines under this ERROR (ER8)
command with a I. Mter this run with
DESIGN =5, from which you are given two lists of The error parameter that determines the
pipes by USU-NETWK; (1) those pipes that are accuracy of the solution. The iterative solution
part of loops and (2) those pipes that are not part process wilI terminate when the absolute sum of
of any loop, you edit this data file, giving either changes in Q's between consecutive iterations is
pipe diameters, or velocities as required. Should less than ERROR. The error parameter can also
you wish to specify other pipe velocities (or be set by assigning NDIGIT, an integer, equal to
diameters) than those that USU-NETWK defaults n in the equation ERROR = lO-n. The default
to by its built-in loop finding process, you can value is 0.01, for ES units and 0.001 for SI units.
utilize the option LOOPRD, and provide the loop For example the options list could contain:
data. However, when using LOOPRD valid and ERROR=.OOOl.
complete loop data must always be given. Should
you wish to specify the diameters of some of the
pipes that do not have loops passing through
them, this can be done by giving a diameter
FLOWFC
See PEAKF.
proceeded by a minus sign.
DESIGN =4 can be used in connection
with the options NOMSOL=1 (10 or 21) and
NOMDIA=1. If NOMSOL is not given a value GAMMA
(Le. the default of NOMSOL=O is left), then the
solution will consist of diameters computed exactly The specific weight of the fluid in Ib/ft3 or
to give the specified velocities. This solution will N/m3 depending upon whether ES or SI units are
be given in the usual output tables, rather than being used. If not given 62.4 Ib/ft3 or 9800 N/m3
the special output tables that are given when for water are assumed, respectively depending
DESIGN=1 or =3, in other words the table will upon whether ES or SI units are used. The
not contain the computed diameter plus a specific weight can also be entered by a command
standard diameter. With NOMSOL=I, however, WEIGHT. For example the $SPECIF list of
this solution will be followed by one in which the options could contain GAMMA=55 for a fluid
closest standard diameter, or nominal diameter, with a specific wight less than that of water.
will be used. If NOMSOL =10 thru 20 then the
next larger standard diameters will be used in this HLETTE#
follow on solution, and if NOMSOL=21 thru 30,
then the next smaller standard diameters will be Determines the vertical height of the
used. There is not provision that allows this pipe letters that will be used on a Calcomp plot.
to be divided into two segments one diameter Default =.13 inches. A menu will be displayed if
smaller than and one larger than the computed a Calcomp plot is ask for, and one of the items
diameter, as with a regular analysis solution. Use listed here allows the letter height to be changed.
of NOMDIA allows you to give a list of standard Therefore HLETTE is seldon used, but can be
diameters if the default values are not what you used so that the menu item need not be changed.
want. In using SI units, or diameter in other than For example HLETTE=.09 could be included in
inch sizes, a list of standard diameters must be the $SPECIF list of options.
given because the default values are in inches.

-16-
 leaST network that can be added to include the loop
forming pipes. These three functions are
If ICOST = I, then a cost, or engineering described separately below.
economic, analysis of the network is done. If
ICOST= I, then additional cost data must be Function # 1:
supplied after the RUN or END command, and If the value of IHGL is positive and less
after other data if options call for it. These than 3, then it designates whether: (1) nodal
additional cost data are described in a subsequent pressures, (2) nodal heads, or (3) HGL elevation
section in this manual under the title "Cost Input are given so that NJ pipe diameters, or roughness
Data." As a minimum this additional data must coefficients can be computed.
consist of END that indicates that the default cost
data are to be used. The only two costs that are If IHGL=O then nodal pressures are given
computed if other cost data are not provided is in psi, if NUNIT=O, or NUNIT=l, and if NUN IT
the default costs of pipes, and the default costs of =2 or NUNIT=3 then the units of pressure must
energy consummed by pumping. If ISIML= 1 also be in kPa (kilo Pascals, i.e. kilo Newtons/m 2). If
exists in the $SPECIF list, then the value given IHGL=I, then pressure heads, or the head above
ICOST determines the time step interval for which the node's elevation in feet or meters, respectively,
cost analyses will be performed. If ICOST=l, is given depending upon whether NUNIT=O or 1,
then a cost analysis will take place after each new or NUNIT >2 for SI units.
time step solution. If ICOST=2, then every other If IHGL=2, then elevations of HGL, or
such solution will have a cost analysis associated the absolute elevation position of the piezometric
with it, etc. For each such requested cost analysis surface is given, respectively, in feet or meters
new cost data can be supplied. As a minimum depending upon whether ES or SI units are used.
this cost data must consist of an END. Generally This latter HGL elevation at any node is the sum
cost analyses will be obtained with steady state of the elevation of the node, and the pressure
solutions, and therefore one would only supply head.
one set of cost data. However, if costs associated (default IHGL=2 for elevations of HGL)
with flow from tank, or other items that are
dependent upon flow rates are given then a series Function # 2:
of costs analysis may be obtained at specified time If the value of IHGL is negative or equal
intervals. USU-NETWK does not average, or to 3, then USU-NETWK is informed that special
provide a composite for all these time varying input follows. There are two different types of
costs. The user is responsible for this composite special input allowed, which are distinguished on
interpretation. the basic of whether IHGL is negative or
IHGL=3.
IHGL If IHGL = -2 or IHGL = - S, then USU-
NETWK is informed that special additional data
The option IHGL has meaning only if is given immediately after the $SPECIF list that
DESIGN > 0, and then it has three separate defines a branched network in which all pipe
functions. The first function designates the units diameters are sized so that a specified slope of the
that are associated with the nodal pressures that hydraulic grade line, HGL is satisfied. The
must be given with DESIGN=l or DESIGN=3, absolute value given to IHGL determines the
the second function is to inform USU-NETWK logical FORTRAN unit from which this additional
that special input follows that defines a branched data is read. If IHGL=-S and the program is used
network, or a branched portion of a network (for from a time shared terminal, or running on a
which additional loop forming pipes may be added stand alone PC, you supply this special data from
through the regular input), and the third function your keyboard in response to prompts from the
is to generate the HGL elevation needed with computer. These prompts correspond to the input
DESIGN=l or = 3 based upon HGL elevation data as described below, and thus act as reminders
values given at intersection points of a of what you should know from reading this section
rectangular grid system superimposed over the of the manual. If IHGL=-2, then this special
network. In addition the command IHGL can be input follows immediately after the $SPECIF list
used to have USU-NETWK write out a data file of options, Le. on the line after the $END before
that contains the branched portion of pipe the first command in the same file that contains
the other input data. The special data is

.17
 embedded in the regular data because when (2) The ending node number of the tree branch,
reading input from a disk file FORTRAN or trunk extension. These nodes must form a
LOGICAL unit 2 is used. At least one regular continuous sequence of numbers from the trunk
command, and END or RUN must be given in node number 1.
addition to this special data which terminates with (3) The slope of the HGL of this branch of the
and END. Therefore, two ENDs may exist. If tree. If the option NPLENG= 1 (length of pipes
IHGL is another negative number, such as -7 (-3 are given in 1000 ft or m), then this slope is in ft
is not allowed because it is the FORTRAN output (or m) per 1000 ft (or 1000 m) of pipe length.
unit number, and -4 is not allowed if remarks are (4) The pipe lengths within this branch of the
used at the end of lines after the PIPES or tree. Items (1) and (2) or items (1) through (3)
NODES commands, and -5 or -6 are not allowed can be terminated by a slash, /. If all lengths are
because they are generally the standard terminal the same, only one length needs to be given; that
assignments) then this special input can be placed is this list can be terminated anywhere after item
in a separate file that will be read as a default file (2) with the remaining items taken from the last
for that logical unit, i.e. on a VAX operating such value given for that item. Often the list of
under VMS the file name would be FOROO7.DAT, lengths are terminated with /. The list of lengthS
or under Microsoft FORTRAN as used for the PC is limited to 99 values. These succeeding lines of
version of USU-NETWK. the user will be input may spill over into several lines if the
prompted for the file name. Often this special number of pipes between (1) and (2) is large and
data is the only data given, but additional most of the lengthS are different. A new line is
components including, pipes and nodes can be given for each additional branch of the network,
added to define the network using the usual but any actual branch may be divided into two or
commands. This special input allows for large more lines if one desires to specify this branch as
branched network to be defined with just a few two or more branches.
lines of input. (See Example problem 18) If the option LENGON =0, then rather
This special additional input data consists than pipe lengths this item (4) consists of a list of
of the following: pairs of x & y coordinates for each succeeding
node along this branch of the network. These
First line contains the following 5 values: coordinates occur in pairs, and therefore, the list
can contain up to 49 pairs of values. When this
(a) The HGL elevation of the first node, which list of pairs is terminated by a / (as it must be),
must be the final trunk of the tree that forms the then all remaining pipes in the branch whose
branched system. coordinates are not given will be pipes of the
(b) The discharge from (or into) this tree trunk same length and direction as those of the last pipe
node. If positive the flow from the other nodes whose end coordinates are given. In other words
move toward this node; if negative this tree trunk the differences in the x and y coordinates of the
node supplies the network. If this given discharge remaining pipes will be the same as these
does not equal the sum of the external flow from differences for the last pipe with both end
the rest of the nodes, properly accounting for sign, coordinates given.
then a message is printed, but the discharge is For any branch of the tree in which items
corrected to equal the sum of all other external (c), (d) or (e) above change from the value given
flows. previously, a command ELEV, DEMAND or
(c) The elevation of this tree trunk node. COEF can be given followed by the altered value.
(d) The discharge from (or into, if negative) the These new values for the branch's end elevation,
next group of nodes until its magnitude is changes and its nodal demands and/or pipe wall roughness
with the command DEMAND. This value may be coefficients must precede the data for this branch
negative or positive, but is generally of opposite (1) through (4) above. Elevations are assumed to
sign to that in item (a) above. vary linearly from the beginning to the end of a
(e) The pipe roughness coefficient for pipes until branch.
its value is changed with the command COEF. Study the following example of a 43 node
network to better understand how this special
Succeeding lines consist of: input can be used to define a branched network,
or the branched portion of a network that may
(1) The beginning node number of the tree contain loops in other portions of it. Examples
branch, or trunk extension. of networks in a later portion of this manual

-18-
provide additional examples. branched network:
In addition to the above three commands
(ELEV, DEMA & COEF), the command TREE NE1WORK
/.
DHEAD (which stands for differential head, and
$SPECIF IHGL=-2,DESIGN=I,NFLOW=1,
can be truncated to the four characters DHEA) NPLENG=1 $END
can be given to include a pump or any other 152. 12900. 85. -300..005
differential head device with a constant head (i.e. 1 10 .0005 2. 04/
a # 1 DHEAD device described under this 220/
command name). After the command DHEA two 2 28 .0004 3. A/
COEF
values are required: 1. the pipe number that .012
contains the differential head, and 2. the 2137/
magnitude of this differential head (positive or 21 40 .0003 3. A/
negative). The command DHEA must precede 38 42/
38 43 .0003 1.5/
the branch data that contains the pipe number END
given in the line after DHEA. RUN

Generally when utilizing this special input

there is no reason why the node at the trunk of
the tree can not be given the number 1, and the
pipe connected to this node is given the number
1. However, if this is not acceptable it is possible
to number the nodes and pipes either starting with
a larger value, or have their values in descending
order. The starting node is established by the
number given the trunk, or beginning node
number, or the trunk of the tree, in the first line
identified as (1) above. If the ending node for the
first branch is less than this value then the node
numbers will be in descending order. In order to
have pipe numbering start with a value different
than 1 it is necessary to communicate this to
USU-NETWK. This communication occurs by
adding 1000 to the loss coefficient given in the
first line of input, item (e) above (and possibly
TREE NE1WORK
preceding with value with a minus sign). When
r 1000 is added to the loss coefficient, then USU-
$SPECIF IHGL=-2,DESIGN=I, NETWk will read an extra line, with one integer
NFLOW=I,NPLENG=1 $END value on it, that must contain the beginning pipe
152. 12900.85. -300..005 number. If the value given for the beginning pipe
1 2 .0005 2./
2 10 .0005 A/ number is positive, then numbering of the pipes
220/ will begin with the given value, and be
2 21 .0004 3./ incremented positively by one for each succeeding
21 28 .0004 A/ pipe in the branched network. If you desire that
COEF
.012
the pipe numbers should be in descending order
21 37/ from the given value, then place a minus sign in
21 38 .0003 3./ front this pipe number. Pipe numbers may be in
38 40 .0003 .4/ ascending order and node numbers in descending
3842/ order, or vice versus. (See the example below.)
38 43 .0003 1.5/
END
With this simplified form of input it might
RUN be desirable to be able to have the roughness of
the pipe wall dependent upon the diameter, e.g.
pipes 12 inches and larger may be made of
The following fewer lines of input could concrete, whereas smaller diameter pipes are made
be used to define identically the same small of steel. Since pipe sizes are part of the solution
it is not possible to enter these different

-19-
roughness coefficients under the COEF command prompt with:
described above. The case of two different Give no. corresponding to what you want on file:
roughnesses is accomodated, however. If the 1 - Length under PIPES data but no
roughness coefficient given in the first line of this coordinates under NODES,
special input is preceded by a negative sign, i.e. 2 Length under PIPES data &
item (e) above is given a negative value, then an coordinates under NODES after HGL,
additional line immediately following this line is 3 Length under PIPES data &
read that contains the following two values: (la) coordinates under NODES after elev.
the second roughness coefficient, and (lb) the 4 No lengths under PIPE & coordinates
diameter for which this second roughness after Elev.
coefficient should apply. The second roughness
coefficient will apply for all larger diameter pipes
as well.
If this value for the roughness coefficient
is less than minus 1000, then two extra lines are
requires, the first consists of the beginning pipe
number and the second consists of two values, a
second roughness coefficient and the diameter that
separates pipes with the two roughness
coefficients. Pipes larger than the given diameter
will have the second coefficient applied to them,
and smaller pipes will have the coefficient given in
the first line of special input apply to them.
Generally different roughness coefficients
will not be entered for other laterals with the
COEF command if this special extra input is
given. However, should other roughness values be
specified under the COEF command then they will
apply only to pipes with smaller diameters than
given by the input (lb) above, and pipes of the
diameter given by (lb) and larger will be given
roughness coefficients equal to the value specified
under COEF plus the difference between the value
given by item (la) above and the value given as
item (e) on the first line of input above.

Since this special form of input allow for

Exam~le USin~ifferent numbering
large branched networks to be generated with a EXAM LE OF ~CHED NETWORK wITH INFLOW
few lines of input, USU-NETWK allows you to AT ITS TRUNK. Elevations change.
/*
write the data that defines this network in the SSPECIF NFLOW=4,NPGPM=4~NUNITE2,NOMDIA=l,
NOMSOL=l,ICOST=l,IHGL=-4,NPLENG=l,
form required under the PIPES and NODES DESIGN=l,NPRINT=-2 SEND
150 -3520 100 220 1000.0005
command for future use, and modification. The 101
ELEV
usual form of data for the network can be written 60
117 110 .6 15. 5./
for a design solution with diameters equal ~ 0, or DREAD
contain the diameters solved for from the solution 108 20
116 106/
produced by this special input. To have this data DREAD
112 20
written to file gives values to IHGL that have - 116 101/
END
100, -200, or -300 added to its valve for special RUN
14 .28 .36 .565 .72 .87 1.075 1.17 1.32 1.475 1.6
input. If IHGL is between -100 and -200, then 1.8 2 2.2 2.4/
117 150
the file will be written with 0 diameters; if IHGL INTEREST=.08
LIFE=60
is between -200 and -300, then data in the file will PIPES
contain diameters computed from the design UNIT=14
.28 128 .36 162 .565 245 .72 306 .87 368 1.025 430
solution; and if IHGL is less than -300, then both 1.17 656 1.32 740
1.475 822 1.6 900 1.S 950 2 1000 2.2 1045 2.4 1090
of the above types of files can be written. For any PUMPS
UNIT-.12
of the three above cases USU-NETWK will CAPI=lS0000.
END

20-
- Example with ground elevations also changing
EXAMPLE OF BRANCHED NETWORK WITH INFLOW
AT ITS TRUNK. Elevations change.
there will be NOY separate lines under (d) and
each lines will contain 1+NOX values, the first of
/*
$SPECIF NFLOWc4 NPGPM=4 ANUNIT=2,NOMDIA=1,
which is the y that defines this y=constant grid
NOMSOL=l,ICOST= i ,IHGL=-~.NPLENG=l. lines, and the remaining NOX values are the HGL
DESIGN=1,NPRINT=-2 $END
150 -3520 100 220 -1000.0005 elevations at the intersections with the x=constant
.005 1.17
101 grid lines. In illustrating this input by a small
ELEV
60 example below data items (a), (b) and (c) are on
117 110 .6 15. 5./ the same line, i.e. the first line afler the $SPECIF
DHEAD
108 20 list of options, and item (d) is on NOY separate
116 106/
116 101/ lines.
END
RUN This special additional input is illustrated
14 .28 .36 .565 .72 .87 1.075 1.17 1.32 1.475 1.6
1.8 2 2.2 2.4/ by the small example network shown below. In
117 150
INTEREST"'. 08 this example the two x=constant grid lines are
LIFE=60
PIPES show by the two dashed vertical lines on the
UNIT=14 Sketch, and the 3 y=constant grid lines are shown
.28 128 .36 162 .565 245 .72 306 .87 368 1.025 430
1.17 656 1.32 740 as dashed horiwntal lines. However, the 2nd or
1.475 822 1.6 900 1.8 950 2 1000 2.2 1045 2.4 1090
PUMPS
UNIT=.12
middle y=constant lines coincides with the pipes
CAPI=180000. (1),(3) & (5) and therefore is not visible.
END

Function # 3:
If IHGL=3, then special additional input
data is provided that allows USU-NETWK to
generate the elevations of the HGL over the
network. These additional data define HGL
elevations at intersection points of a rectangular
grid system. When this third function is utilized
the following occurs: (1) The elevation of the
HGL at nodes are generated by interpolation of
surfaces over the network that defined the HGL
elevation. The HGL surface is defined by
additional data that are described below, (2) The
x- and y- coordinates must be given for each node
in the regular data entered with the NODES or
the PIPE- command, rather than the pressure,
head or HGL elevations. The position for the
nodes defined by these coordinates are used to
computer the HGL elevations from the HGL
surfaces of (1).
The special additional data must be given
immediately after the $SPECIF list of options and
consists of:
(a) The number of x = constant grid lines, NOX,
(b) The number of y = constant grid lines, NOY,
(c) The x values of each succeeding x equal EXAMPLE OF USING IHGL=3 TO
constant grid lines (NOX of these values must be GIVE HGL PLANES OVER NETWORK.
1*
given), $SPECIF IHGL=3,NFLOW=1,DESIGN=1
$END
(d) The y values of each successive y equal 2 3 O. 3700.
O. 800. 750.
constant grid line followed by the HGL elevations 1000. 780. 730.
at each of the intersections this y = constant grid 1900. 800. 750.
PIPES
line has with the NOX x = constant grid lines. 1000
1 1 2 1500 O. . 005
While this input is free format and may be on a 2 3 2 10001
3 2 5/
single line if it will fits within 80 columns, it is 4 4 5 1400/
5 5 6 1200/
useful to the separate (d) above into separate 6 7 6 1000/
7 8 6 1200/
lines, one for each y=constant grid line. Then 1 -800 720 0 1000
2 0 720 1500 1000
3 -300 720 1200 1900

-21-
4 -200 720 1490 0 values; an integer that is the pipe number and a
5 0 720 2500 1000
6 1900 720 3700 1000 real that consists of the flow rate. The units of
7 -300 720 3500 1900
8
RUN
-300 720 3400 0 this flow rate are determined by the option
NFLOW and this input comes after the RUN or
END command.
INITAL INITAL=-3 then the input is as with
INITAL = -1 except it must be given in another
In the version of USU-NETWK that also file with the name INITALDAT, and is not
does time-dependent solutions there are two placed after the RUN or END command. If this
subroutines to provide initial flowrates which file does not exist USU-NETWK will prompt for
satisfy all junction continuity equations. The the file name that contains the initializing flow
second such subroutine requires a previous valid rates.
solution. It is normally called upon to initialize INITAL=-4 then the input is as with
flows after the first solution for time zero has INITAL = -2 except it must be given in another
been obtained, or the first of a series of file with the name INITAL.DAT, and is not
alternative solutions has been obtained when using placed after the RUN or the END command.
the CHANGE command to defined a series of
separate solutions.. If INITAL=O is added to the If the initializing flow rates are given in
$SPECIF list, then all initializations will be basic units, e.g. cfs when the problem is in ES
produced by the subroutine which does not utilize units, or m3/s when the problem is in SI units,
information from a previous solution. (Default, regardless of the units used for the demand input
INITAL=l). as determined by the option NFLOW then 10
In addition to the usage of the option should be subtracted from the above values. In
INITAL with the version of USU-NETWK that other words INITAL -11 is the same as
also has the capabilities to do time dependent INITAL = -1 except that regardless of the value
solution in which this option determines whether of the option NFLOW, the initializing flow rates
the initialization of pipe flow rates is to be based will be in basic units, cfs or m3/s.
on the past solution, or generated by the same The initializing flow rates must satisfy
subroutine that produced the first initialization, it each and every junction continuity equation within
can be used to read in the initial flow rates. This the network. USU-NETWK will check that this
alternate use of the option INITAL occurs if its is the case, and if the values do not satisfy the
value is negative, and is also implemented in some condition that the sum of all flow rates into a
of the version of USU-NETWK that alJow only junction are the same as flow rates out of a
for steady state solutions. The different allowable junction (including the demand) then a message to
negative values that may be assigned to INITAL this effect is written, and the solution is
have the following meanings: termina ted.
INITAL=l then the initializing flow
rates are given after the RUN or END command INPUT #
e.g. are in the same file as the other input data.
The sequential order in which these flow rates Allows the FORTRAN logical input unit
must be given is the same as the order that the to be specified. If TTY or REMOTE is given in
pipes are entered under the PIPES or the PIPE- response to the prompt from USU-NETWK for
command. The direction of the flow is established the input file name then INPUT=5, the standard
by the first and the second nodes given after the terminal or keyboard input. If a file name is given
PIPES or the PIPE- command, and a negative to this response, then INPUT=2, and the input is
value indicates that the starting flow rate is in the read from the given file. Should your computer
opposite direction. The units of this flow rate are installation have different logical unit number
the same as that used for the demands, and is assigned, or if unit 2 cannot be used for input for
determined by tile option NFLOW. some reason, then by adding INPUT to the
INITAL=2 then each initializing flow $SPECIF list the flexibility is allowed to use these
rates is preceded by its pipe number. With units without recompiling USU-NETWK. Similar
IN ITAL=-2 the order of the initializing flow rates flexibility of the output unit is allowed with the
does not need to agree with the order of the other option lOUT.
pipe data input, but the input consists of pairs of

-22-
 INPUTA # used to write this file. For example if IOUTO=20
were indued in the $SPECIF list of options than
Allows three different forms of input data data would be written to a file using Fortran
to describe a network. If logical unit 20. Since data in this file is in
INPUTA -0, then the standard form of input specified columns, USU-NETWK can read the
described in this main body of the manual is used. data in this file faster than it can a file prepared
In this form a command name enters data of the by an editor, but unformatted. However, this file
specified type. (Default INPUTA=O) can be edited to change data it contains so that
INPUTA -1, then the form of input data described the problem read with IOUTI is different than the
in Appendix A is used. one for which the data was written. Thus the use
INPUTA=2, then the form of input data described of IOUTO, and IOUTI above, is an alternative to
in Appendix B is used. using the CHANGE capability of USU-NETWK
in obtaining a series of solutions for a network
lOUT # problem that is being studied. An important
differences is that the user can examine the
Allows the user to specify the logical solution just obtained and use this as a guide in
FORTRAN output unit on which the solution deciding what changes should be made for the
results will be written. If TrY or REMOTE is next solution. An alternative means of writing an
given in response to the prompt from USU- output data file is available with option NETPLT.
NETWK for a file name on which the output In making changes to this file, (1) the
should be written, then IOUT=6, the usual order of data as described below must be
terminal, or monitor output device. If a file name maintained, (2) the FORMATS for the data that
is given to this response then IOUT=3. If lOUT are described below must be maintained, and (3)
is included in the $SPECIF list of options then the pump characteristics must be defined by giving
the value assigned to lOUT will be the output the coefficients for the second degree polynomial
logical units on which the solution results are used to define the head versus flowrate curve for
written. Thus, if needed, assigning lOUT a value the pump. These coefficient are given as part of
allow flexibility at different computer installation the solution output, unless suppressed with a
in ~ending the output to various output devices, small value of NPRINT.
which may include tapes, for example, on which
the output is to be stored. The order of data in this file is as follows:
(1) A line that gives the number of the
following in the network using the FORMAT
IOUTI # (1615): (a) pipes, (b) nodes, (c) source pumps, (d)
booster pumps, (e) reservoirs, (f) minor losses, (g)
Reads the formatted data written by a
pressure reduction valves, (h) nozzles, (i) check
previous execution of USU-NETWK with the
valves, (j) back pressure valves, (k) sum of
option IOUTO equal to the same logical unit
reservoirs, nozzles, PRVs and BPVs, and if
number or prepared according to the FORMATS
NOCHK=1 then (I) total loops, (m) pseudo loops
indicated under the description below under
, and (n) real loops.
IOUTO. IOUTI can only be used if this data file
(2) The pipe data with the FORMAT
has been previously written. For example if
(315,2F15.3,F15.7)
during a previous execution of USU-NETWK
(3) The node data with the FORMAT
IOUTO=20, then ifIOUTI=20 during this run the
(I5,2F15.5,(915. If NOCHK is greater than 0
input for the network will be read from this file,
besides the node number, demand, and elevation
and this file must have the input in the FORMAT
the lines of the node data contain the pipe
described under the option IOUTO.
numbers that connect to each node.
(4) The reservoir data. This data depends
upon whether NODESP=O or NODESP=1. If
IOUTO # NODESP=O the FORMAT is (I5,F15.5), and the
items are (a) the internal pipe number i.e. pipe
The option IOUTO instructs USU- sequence number under the PIPES or PIPE-
NETWK to write an input data file under the command, and (b) the water surface elevation. If
built-in FORMAT. The value given to IOUTO NODESP=1 the FORMAT is (215,2FlO.5), and
is the FORTRAN logical unit number that will be the follOwing 4 items are needed on each line: (a)

-23-
 the internal pipe number that connect the flowrate and the energy loops. Only the corrective
reservoir to the network, (b) the node number flowrate loops are written, however. When using
given the reservoir, (c) the water surface elevation lOUT!, this second set of loop data will be
in the reservoir, and (d) the ground elevation at generated.
the reservoir node.
(5) The pump data. This data is also IRESID
different depending upon NODESP. If
NODESP=O, then the FORMAT is It may be possible to save a small amount
(I5,5EI5.8,I,EI5.8) and the following items are of computer time by not redefining the Jacobian
given: (a) the internal pipe number, i.e. pipe Matrix and completely solving the new linear
sequence number under the PIPES or PIPE algebra problem for the final few iterations of the
command, (b) the coefficient for the squared term Newton method. This option allows for use of the
for Q, (c) the coefficient for the linear term, (d) last solution of the Jacobian problem to be
the constant coefficient all for a second degree fit utilized if the sum of changes is less than IRESID
of the pump characteristics modified by the water (a real value). Experience has indicated little if
surface elevation and .5*QBNBP, (e) the any computer time is saved and use of IRESID is
transformation term QBA, (1) the normal capacity, discouraged. It can only be used in conjunction
and (g) the water surface elevation. If with the sparse matrix solution method (default
NODESP=I, the FORMAT is IRESID=O).
(2J5,4EI5.8J,3EI5.8) and the following items are
given: (a) the internal pipe number, i.e. pipe ISIML#
sequence number under the PIPES or PIPE-
command, (b) the node number (external value) Setting ISIML=l indicates that the
for source pumps and 0 for booster pumps, (c) the solution to the network is to be time-dependent,
coefficient for the squared term for Q, (d) the and therefore additional data as described under
coefficient for the linear term, (e) the constant the major section "Additional Data for a
coefficient all for a second degree fit of the pump Simulation Solution" is required (default,
characteristics modified by the water surface ISIML=O)
elevation and .5*QBNBP, (1) the transformation
term QBA, (g) the normal capacity;, (h) the water
surface elevation, and (i) the ground elevation at
ITERA (MAX)
the pump node.
The maximum number of iterations that
(6) The minor loss data according to the
are allows in order to obtain a solution. The
FORMAT (15,FlO.5) that consist of the pipe
number, and this loss coefficient. default number is set to 15 (but may be different).
For example if ITERA=6 were included in the
(7) The pressures reduction valve data
$SPECIF list then the solution process would
according to the FORMAT (15,FI5.9,FlO.5)
terminate after the 6th Newton iteration even if
consisting of the pipe number, the distance from
it upstream junction of the PRY, and the HGL the error condition ERROR were not meet.
elevation of the PRY setting.
(8) The nozzle data according to the
FORMAT (15,FI5.8) and consisting of the pipe LENGON
number and the nozzle coefficient, CNOZZ.
(9) The internal pipe numbers containing This option permits lengths of pipes to be
check valves under the FORMAT (1615), and determined by providing x and y coordinates for
(10) The back pressure valve, BPV, data the nodes rather then giving the pipe lengths.
according to the FORMAT (15,2FI0.5), and The default is that pipe lengths are provided, and
consisting of the pipe number, the distance from this default corresponds with LENGON = 1
the PRY to the downstream node, and the HGL (lengths of pipes on). If x and y coordinates are
setting. provided, i.e. LENGON =0, then NOD ESP must
If NOCHK is greater than 0, then in also be set to 1, because with NODESP=O there
addition to the above 10 types of items, the pipe is no means for providing the coordinates of
numbers defining the loops of the network are sources of supply and consequently no means exist
also given. If PRVs or BPVs are present two sets for USU-NETWK to compute the length of pipes
of loops are used in the solution; the corrective that connect sources of supply to the network. If

24
 LENGON =0 is included in the $SPECIF list of with LENGON =0 with the PC steady-state
options, then the following changes must be made version of USU-NETWK.
to the input data described in the section
Detailed Steady-State Data Requirements under
the PIPES, NODES and PIPE- commands. Under
the PIPES command the fourth item, or the pipe LENORG
length, is omitted so that only the following 5
items exist: 1. pipe no., 2. upstream node no., 3. In the final solution table the lengths of
downstream node no., 4. pipe diameter, and 5. pipes printed will be their actual lengths plus any
pipe waH roughness coefficient. Under the equivalent lengths for minor losses within them
command NODES the x and y coordinates of the under the default options. If their actual lengths
node are provided after the elevation of the node are to be printed then set LENORG=1.
so that instead of 3 items of input the input
consists of the following 5 items: 1. node no., 2.
demand at node, 3. elevation of node, 4. x LOOPRD
coordinates of node, and 5. y coordinate of node.
Should DESIGN = 1, then the elevation of the Is an option that allows for the loop data
HGL (or head or pressure) at the node is to be read in as input data rather than being
provided as the 4th item followed the x and y generated internally by USU-NETWK. If
coordinates. Should NETPLT be given a value LOOPRD=l then the loop data must be supplied
greater than 0 but less than 13, so that the x and as input in addition to the other required data.
y coordinates are provided as input already on the If the standard form of input is used (i.e. that
lines after the NODES command, then the only described in the body of this manual rather then
difference in input with LENGON=O is that pipe in Appendixes A or B), then this loop data
lengths are omitted from the data after the PIPES follows the RUN or END command immediately.
command. The first record of additional data consists of two
If the PIPE- command is used in place of integers; the number of pseudo and the number of
PIPES and NODES to enter data, then the pipe real loops (pseudo loops connect sources of
lengths are omitted from their usual position after supply, PRY, and/or BPV). This line is followed
the diameter on lines after PIPE-. However, the x by loop data, each line of which lists the pipe
and y coordinates are provided after the elevation numbers around each separate loop. One and
given for either the upstream or downstream only one line is used to define each loop. Pseudo
node. In other words there are now 4 real values loops must be given before real loops. The pipe
that follow the node numbers on lines after the number is given a positive value if the direction
PIPE- command. Therefore, the input data under transversed around the loop is in the direction of
the PIPE- command is similar to that if 0 < assumed flow, and pipe numbers must be preceded
NETPLT < 13, with the exception that the pipe by a minus sign if the direction traversed around
length is not given. Since sources of supply must the loop is opposite to the assumed direction of
have been identified by nodes when LENGON =0, flow. To see how these loops are defined you
it is not permissible to have only one node per should set NPRINT= 1 or larger, and study the
lines of data under PIPE- as is permitted with loop data that USU-NETWK places in the output
NODESP=O if the coordinates are to be used only file.
for plotting. Also whenever coordinates (or HGL If the network contains pressure reduction
elevations) are given it is not permitted to omit valves, or back pressure values, then it is necessary
the demand at a node. Therefore, with to repeat the loop data twice (including the line of
LENGON=O there must be 4 real values with a data that gives the number of pseudo and real
decimal point that fol1ow the integer node values. loops). The first set of loop data is for the
If DESIGN= 1 is also included, then the HGL corrective flow rate loops and the second for the
elevation (or head or pressure) at the node must energy loops. See modification # 4 of the input
follow the elevations giving 5 such values that for Example Problem # 1 for an example.
follow the node numbers. With LENGON left IfINPUTA=l (that is if the form of input
equal to the default value of 1 and DESIGN=l, described in Appendix A is used) and
then 3 real values must follow all integer node LOOPRD= 1, then the first record in front of the
values for data entered through the PIPE- corrective flowrate loops are not given since this
command. The use of PIPE- is not permissible

-25-
 data is given in the options under $ASPECI. then USU-NETWK assume that you are giving the
= INPUTA=2 requires loop data. The option pairs of sources between which pseudo loops are
INPUTA=3 is used when the form of input is as to be formed by the order in which these sources
described in Appendix B, with the exception that are given in the input data. The first pseudo loop
loop data are not provided but rather are will be formed between the first and the second
generated internally. In other words INPUTA=2 source pump; the next psuedo loop will connect
or INPUTA=3 has the same effect as LOOPRD the second to the third source pump, etc. The
:::: 1 or 0 if the data form described in Appendix last source pump will have one pseudo loop
B is used. between it and the first reservoir, the next loop
LOOPSE will be between the first and the second reservoir;
and finally there will be a loop from the last
This option allows control over the reservoir to the first PRY, etc. the last PRY to
pseudo loops that will be used in setting up the first BPV, etc. If only one source pump exists
problems for solutions. The amount of then the first pseudo loop will be between it and
computations required to obtain a solution the first reservoir etc. The above assumes that a
depends upon the length of loops (Le. the number connected path of pipes exists between the two
of pipes in a loop). Should some loops be very sources, but if pressure regulating valves exist then
long then the amount of arithmetic require for such connected paths may not exist between all
each Newton iteration in solving the linear system sources for the energy loops. USU-NETWK will
of equations become large especially if one of the determine this and find alternative sources to
banded methods of solution is used as described connect when it is necessary, while still
under the option NSYMMT. Pseudo loops maintaining the independence of the equations
between sources of supply (reservoir, source defined by these loops.
pumps, PRY and BPV) have the potential of In setting up the system of hydraulic
being long. A long loop will occur for example if equations that needs to be solved two separate
it is formed between two sources of supply that sets of loops are formed by the above procedure
are at opposite ends of a large network. In using if pressure regulating valves are present or other
USU-NETWK the user has four means by which special device exists in the network; namely the
he can control loops. corrective flowrate loops, and the energy loops.
(1) The sources of supply can be entered The corrective flowrate loops do not involve
in the input data in an appropriate manner, as PRY's or BPV's and the energy loops, about
described below, to keep psuedo loops short. which the head losses are equated to the
(LOOPSE::::O, the default). differences in HGL elevations at the sources, do
(2) The complete loops, both the include the PRY's and BPV's as if they were
corrective flow rate and energy, can be given by sources of supply. (See the book "Analysis of
using the option LOOPRD. Flow in Pipe Networks" by Jeppson for an
(3) The end pipes for psuedo loops explanation of these two different loops.) Thus in
betwcen sources of supply can be specified. entering the data one should enter the second
(LOOPSE::::1) pump after the PUMPS command as the pump
(4) USU-NETWK can be requested to closest to the first pump, and the third pump so
optimize the set of pseudo loops. (LOOPSE;:;:2) it is close to the second, etc. The last line after
The last 3 of these means can be selected among the PUMPS command should represent a source
through the use of the option LOOPSE. The that is close to the first reservoir entered under
default is LOOPSE=O which results in (1) above. the RESER command, and the second line under
Reasons why (1) above is the default are given RESER should be a reservoir close to the first
below. reservoir etc. Pseudo loops for energy equations
are form between the downstream side of PRY's
LOOPSE=O (default) and other sources, and between the upstream side
By selectively pairing reservoirs, source of BPV's and other sources. Therefore, the order
pumps and/or pressure control valves so that in which lines occur after the command VALVE
minimum sequences of pipes are needed to should be such that the downstream side of the
connect them the length of loops can be keep first PRY is close to the upstream side of the first
small allowing the banding subroutine to find a BPV, etc.
smaller band width for the Jacobian and result in For example if there are 5 source pumps
reduced computer execution time. If LOOPSE=O with pipes 8, 1, 30, 200 and 121 connecting them

-26-
 La the network, 4 reservoirs connnected to the LOOPSE=1
network by pipes 2, 10, 50 and 231, and 2 PRY's The method referred to as (2) above, of
in pipes 38 and 56, then as the default USU- contrOlling the formation of pseudo loops by
NETWK would attempt to form the corrective providing the pipe pairs between which both the
flow rate loops between the following pairs of corrective flow rate loops and the energy loops are
pipes: (8,1), (1,30), (30,200), (ZOO,121), (121,2), to be formed, is selected by setting LOOPSE=1 in
(2,10), (10,50), and (50,231) and the energy loops the $SPECIF list of options. This option allows
between the following pairs of pipes: (8,1), (1,30), much more flexibility than ordering sources as
(30,200), (200,121), (121,2), (2,10), (10,50), they are entered. Since several other sources
(50,231), (231,38), (38,56). Since the corrective (including PRVs and BPVs) might be paired with
flowrate loops might contain the pipes with PRVs, the same source, this option does require extra
but the energy loops cannot, the first of these thought and input data on the part of the user.
loops often are different even though they contain These pairs of pipes are given immediately after
the same end pipes. the RUN or END command.
Should there be no path of connected Each new energy pseudo loop thus
pipes between the end pipes selected in this specified must represent a new independent
manner, then a message to this effect is printed, equation, e.g. that is not redundant with those
provided NPRINT is large enough, and an already given. USU-NETWK assumes you know
alternate pair of end pipes is chosen. Should all what you are doing when you give pairs of pipes
such pairs be exhausted and still no pipe to a with LOOPSE=2, and mistakes in giving these
particular end pipe found then additional messages pair can cause unpredictable results. Giving pairs
arc wriLLen and the solution either terminated or between which no connected series of pipes exist
perhaps continued if it appears a valid single will result in termination of the solution, but for
neLwork may exist. If no path exists from a other mistakes on your part no checking is done.
particular source of supply to any other source of If PRVs or BPVs are presenL then two
supply it means that the network is actually two such sets of integer values must be given that
separate hydraulic systems. USU-NETWK provide the end pipes for (1) the corrective flow
assumes you have made a mistake, should this rate loops, and (2) the energy loops. The number
occur, and terminates the solution. You must of pair of integers required with LOOPSE=1
process this dual system as two separate analyses. equals the number of sources of supply (source
For many networks the user can easily pumps plus reservoirs minus one for the corrective
examine the map of his water system, and follow flow rate pseudo loops, and the number of sources
the above guideline in entering data. There are of supply including the PRVs and BPVs as
factors that do complicate the above, however. It artificial reservoirs (source pumps + reservoirs
is possible that PRY's and/or BPV's separate the +PRVs+BPVs) minus one for the energy loops.
network into 2 or more different pressure zones, When using the DHEAD command additional
in which event there will be no paths of connected pseudo loops will be formed between the node
pipes beLween all sources of supply and pressure whose HGL is specified and the source given, but
control valves. In some systems it is possible to these are not included in the above pairs.
form shorter pseudo loops by connecting two or Likewise specifying unknown demands, but known
more sources to a single source. It can occur that HGVs with the command SETPR, etc. will add LO
the ordering of sources as described above to keep the number of pseudo loops, but again you do not
the corrective flow rate loops short may result in include these as pairs. Modification 5 of the input
long energy loops or vice versus. (Other things data for Example #1 illustrates the use of
being equal it is best to favor keeping the energy LOOPSE=1.
loops short over keeping the corrective flow rate As a final means of forming pseudo loops
loops short.) In some systems it is possible to you can request that USU-NETWK optimize the
form shorter pseudo loops by connecting two or set of both corrective flow rate loops and energy
more sources to a single source. Therefore, even loops to minimum the lengths of all of these
if one gave full consideration to the above loops. One might assume this is the best solution,
ordering it may not represent a very good solution and should be the default. The problem is that a
LO the problem of getting the best set of pseudo considerable amount of computer effort is
loops defined for a given network. Setting necessary to perform this optimization. Often it
LOOPSE= 1 or 2 provides other possiblities. is much easier for an individual to do quite a
satisfactory job by simply looking at a map of the

27
 piping system. The computer does not view a specify that Manning's Equation is to be used in
map of the network. Rather it must form all computing the relationships between the
possible loops and then make a number of head losses and fiowratcs in the pipes of the
decisions about which of all of these possible network. From the magnitude of the wall
loops are the best sets to use. The table below roughness coefficient USU-NETWK cannot
shows the number of independent loops that are distinguish whether the Darcy-Weisbach Equation
potentially available (without pressure zones or the Manning's Equation is to be used, and the
occurring) as a function of the number of sources former equation will be used. Therefore, it is
of supply. mandatory to specify MANNI=l, or NEQUAT=2
With LOOPSE=3 all of these loops will if Manning's Equation is to be used.
be formed, or attempted to be formed, and
decisions made about which of these to use so
that they are all as short as possible, while still MAX
representing a set of independent equations. See lTERA
There are three methods that could be used to
handle the information related to these loops
within the program: (1) the information could all NCONTI
be retained in memory, (2) it would be written to
a direct access file, and retrieved as necessary NCONTl=l causes USU-NETWK to act
later, or (3) the loops could be repeatedly similar to when NOSTOP=l, however, does not
redetermined whenever they are needed by the terminate execution when error conditions
optimization scheme. Increased computer times associated with appurtenances of the network are
are definite disadvantages of (2) and (3). For a detected.
large number of sources, method (1) will require
a large amount of memory since each loop may
require a large number of word or information.
NELEV
The decision was to use (1) above, but discard, Le.
replace information as soon as it is apparent that
This option has a dual function; one
a loop will not become part of the final set of
associated with the standard input, and the second
pseudo loops. Furthermore, since it is not
when the form of input described in Appendix B
possible to determine in advance how much
(Le. INPUTA=2 or 3 is given in the $SPECIF
memory may be required, and it would be costly
list).
to check each time new loop information is
In connection with the standard input
obtained whether memory is still available, no
NELEV can be given the value -1 to specify that
checking is done within the
an extra table should be written that distinguishes
program to terminate execution if adequate
pressure heads and velocity heads for the different
memory is not available.
pipes that join at each node of the network.
If a number of solutions for a network are
Ordinarily NETWK does not make any distinction
anticipated one might first use this option, and
between the elevation of the energy line, (the sum
then later use LOOPSE= 1, or better yet to extract
of the elevation head, the pressure head, and the
these loops from the output obtained with
velocity head), and the elevation of the hydraulic
NPRINT= 1 or larger, and then add them to the
grade line (the sum of the elevation head, and the
input data with the option LOOPRD=l, or use
pressure head). However, if it is important that
the options IOUTO and IOUTI in obtaining these
some difference be made then this can be
additional solutions.
accomplished by setting NELEV=-l. With
NELEV=-l a special extra table will gives the
No. sources 2 5 10 15 20 30 50 75 100 160 pressure head and pressure at each node that joins
Pseudo loops 1 10 45 105 190 135 1225 2775 4950 12720 every junction of the network by substracting the
different velocity heads in the pipes from the
elevation of the energy line. Since the pipes
joining at a junction will generally have different
MANNI velocities in them these pressure heads will be
different. The reported values in the NODES
Manni set equal to 1, is an alternate to DATA table will be the elevation of the energy
setting NEQUAT=2 in the $SPECIF list to

-28-
line, and the head reported in this extra table will NEQUDI (EQUDIA)
be the sum of the pressure head and the velocity
head. In addition if NELEV=-l then a loss will This option only has meaning if
be assume equal to one-half the velocity head, for differential head devices of type # 2 are entered
flow in any pipe leaving a reservoir, and a full with the command DHEAD. If the differential
velocity head will be assume to be lost for any head is of a magnitude so that the combined head
flow in a pipe entering a reservoir. caused by the frictional head loss plus the
USU-NETWK is not designed to correctly differential head can be achieved by another size
solve pipe flow problems for which velocity heads pipe, then USU-NETWK will compute the
are dominate over pressure head, as is the case for diameter of this equivalent pipe. These equivalent
a small diameter venturi, or aspirator. For such diameters are the pipe sizes that the command
problems concern about what controls the DHEAD can be used to determine. If this special
elevation of the energy line in pipes upstream and table of equivalent pipe diameters is not to be
downstream of nodes must exist. USU-NETWK created NEQUDI=O should be given. (Default
assume the energy lines are the same for all pipes NEQUDI=l)
that connect at a node.
If NELEV is given a value in conjunction NETCHK
with INPUTA=2 or 3, then it can have the values
o or 1. The default is NELEV=1. If NELEV=O, This option allow the network's layout, or
then the elevation of the node is read in topology to be defined by both pipe and node
immediately after the demand on the line of group oriented input data, and therefore is useful in
No. 27, described in Appendix B, before the pipes verifying that the input data is correct that is
which join at that node. If NELEV=l, then the being used by USU-NETWK in providing
elevation is read on a separate line as described in solutions. The default is NETCHK = 0 which
Appendix B. indicates that the node oriented data is not given.
The following four values can be given to
NEQUAT NETCHK to allow different methods for
providing this additional data that usually defines
Determines what equation will be used to the network's layout: -1, 1, -2 or 2. If NETCHK
define the relationships between the headlosses is positive, i.e. 1 or 2 then the node oriented data
and flowrates in pipes. If NEQUAT=O then the is provided after the elevation of the node after
Darcy-Weisbach Equation will be used; if the NODES command. If NETCHK is negative,
NEQUAT=l then the Hazen-Williams Equation i.e. -lor -2, then this additional data follows the
will be used; and if NEQUAT=2 then the RUN or END command. The difference between
Manning's Equation will be used. absolute value of 1 and 2 is whether the assumed
USU-NETWK will determine whether to direction of flow should be checked in every pipe
use the Darcy-Weisbach Equation or the Hazen- between this data and that given in the pipe
Williams Equation based on the first pipe wall oriented data. If NETCHK= III directions of flow
roughness coefficient given under the PIPES or are not checked whereas if NETCHK= 121
the PIPE- command, and therefore, NEQUAT is directions of flow must be the same in both
generally not included in the $SPECIF list. If no definitions of the network.
pipe roughness values are given so that the default The extra data requires if NETCHK is not
value of COEFRO is used, then this value will equal to zero consist of the pipe number that join
determine the equation that will be used, and at each node of the network. If the value is 2 or -
since the default value of COEFRO=.Ol02, if it is 2, then these pipe number must be preceeded by
not incJuded in the $SPECIF list, the Darcy- a minus sign (Le. be the negative of the pipe
Weisbach equation will be used. However, if the number) if the assumed flow direction is into the
Manning's Equation is to be used either junction.
NEQUAT must be given a value of 2 or MANNI To summarize the following is available
a value of 1 because the magnitude of Manning's with this option:
n and the equivalent sand roughness, e cannot be 1. If NETCHK=-l, then after the RUN or END
distinguished. Selling NHAZEN = 1 is equivalent command extra data must be given that lists the
to NEQUAT=l, as is MANNI=l equivalent to pipes that joint at the nodes of the network.
NEQUAT=2. These data are on separate lines, and the data on

29
these lines is described as Card No. 27 using the or possibly none of the graphics routines, and
alternate node oriented form of data described in therefore only the formatted input feature
Appendix B. Each line of this data consists of (a) described above may be controlled by NETPLT.
the demand at this node, (b) the elevation of this If the absolute value of NETPLT=1 (i.e.
node if NELEV =0, but if NELEV = 1 then the lor -1) then a computer generated Calcomp Plot
elevation is not given, i.e. (b) is skipped, (c) the of the network will be generated. Depending
pipe numbers that join at this node. These lines upon the installation and the graphics drivers, this
must be in the same order as the node data given Calcomp Plot may be directed to other graphic
after the NODES command, and if the PIPE- devices. This plot consists of the layout of the
command is used the order of these data must be network, with pipes and nodes labeled if desired,
the same as the order in which node numbers first as well as pipe diameters and length printed.
appear. Demands at nodes may be printed. Flowrates in
2. If NETCHK=l, then pipe numbers that join at pipes may be printed, and pressure at nodes may
the node are listed after the elevation on lines be printed. A menu will appear before the plot
after the NODES command. It is not possible to file is written that allows selection of the items
set NETCHK= 1 and use the PIPE- command to you desire to have on the plot, as well as its size,
enter the basis data for the network. Specified and the height of the lettering on the plot. The
directions of flow are not checked i.e. it is not version of USU-NETWK designed to run on PC's
necessary to give minus values if the assumed under DOS will write data to a file PLTNET.DAT
direction of flow is into the junction. if NETPLT=l, that can be used by the post
3. If NETCHK=-2, then the input is identical to processor PLTNET to display the network, and
that described with NETWK=-l, except that if the selected item from its solution. See Appendix C
assumed direction of flow for a pipe is into that and the explantion in using program PLTNET.
node its number is preceded by a minus sign.
4. If NETCHK=2, then the input is identical to Summary of option NETPLT
that described with NETWK= 1, except that if the
assumed direction of flow for a pipe is into that This option controls:
node its number is preceded by a minus sign. 1. Output files for graphics
For an examples that uses this option see 2. Writing data files that can be used as
Modification # 1, # 2 and # 3 for Example No. input to USU-NETWK, or other graphics
1. programs.
3. Reading formatted input files created
NETPLT with PIPEINPT.

The option NETPLT is used to create

several types of plots, or written data to files for THE FOLLOWING OCCURS IF NETPLT
later plotting, depending upon its value. It also is = 0 No graphics output, or special data
used to indicate that the input data file is in the files written, free format input.
format written by the preprocessor PIPEINPT, and = 1 on a VAX: Calcomp routines are
therefore can be read faster as a formatted file. If called to:
NETPLT=13 or NETPLT=14 then the input data (a) Write Calcomp plot file for network &
file must has the data in the columns that its solution. A menu will allow you to
PJPEINPT writes in. If NETPLT equals zero, the determine plot size, what to include, height of
default value, then no plot is made. Appendix C lettering, etc.
describes graphics that can be produced by USU- (b) Write a file that contains basis
NETWK. This graphics is controlled by NETPLT. information to do graphics with another graphiCS
You should read Appendix C in connection with software package.
this description to understand how some of the (c) Both (a) and (b)
graphic output can be requested, and what you = 1 On PC: Writes a file for program
must do for this to be possible. The PC version PLTNET to use to display network & selected
of USU-NETWK does not have some of the items on monitor.
graphics capability described. Rather post = -1 on a VAX: Same as 1 except you
processors may be used to do some of this same can give message to calcomp operator(dependent
graphics, as described in Appendix C. Other upon installation) Message may be: USE INDIA
versions of USU-NETWK may contain only some, INK, PEN SIZE # 2

-30-
 = 2 Same as 1 except original input next solution.
data does not have x & y coordinates for nodes of
network. These are, therefore, provided in NFLOW
another file.
= - 2 Same as 2 except you can give This option tell USU-NETWK what units
message to operator. are associated with the demands that are given
= 3 More general graphics output file under the NODES, or the PIPE- commands. In
will be written that contains information for addition to several volumetric units for flowrate,
auxiliary programs to use such as: USU-NETWK also allows weight and mass
(a) PLTNET3 (Requires that DISSPLA flowrates to be used.
software be present) - makes: Depending upon the value given to
(1) 3-D perspective display of NFLOW the following units apply to demands:
network & HGL. NFLOW = 0, in cfs if ES units are used,
(2) Contour map of pressure, in m3/s if SI units are used.
etc. NFLOW = 1, in gpm (allowed only with
(3) 3-D perspective over contour ES units).
map. NFLOW = 2, in mgd (allowed only with
(b) PROFILM - makes ES units).
(1) HGL-elev. & elev. profiles NFLOW = 3, in m3/s (allowed only with
thru selected sequences of pipes. SI units).
(2) Network layout with pressure NFLOW = 4, in liters/s (allowed only
bars at nodes. with SI units).
(3) Contour map type map that NFLOW = 5, in pounds/s (allowed only
shows lines of constant pressure, with ES units).
pressure head, or HGL-elevations. NFLOW = 6, in Newtons/s (allowed only
with SI units).
(c) OFPLOT - draws the network and NFLOW = 7, in slugs/s (allowed only
selected solution items similar to the plot with ES units).
routines included as part of the full NFLOW = 8, in kHograms/s (allowed only
simulation version of USU-NETWK with SI units).
OFPLOT is a separate program for PC
(or larger computer use) that consists of In solving a network problem, USU-
the same plotting subroutines that are NETWK assumes the fluid is incompressible, Le.
built into the full simulation version of the internal energy or enthalphy of the fluid is not
USU-NETWK. included in the energy equation. For some
problems a reasonable assumption is that the
= 13 & 14 Informs USU-NETWK that internal energy does not change, but since the
input data is according to format from PIPEINPT fluids density, or specific weight, might change
(& PLTNET). If = 14, then a data file written significantly because of large temperature or
for PLTNET to use will be written from the pressure changes, an overall volumetric flow rate
solution. balance does not exist. For such problems USU-
NETWK can produce a valid enough solution for
many purposes by giving demands in units of
NEWSTR# weight or mass flowrates. Then continuity is
When a series of solutions for a basic satisfied, since all computations will be based on
network are being obtained using the CHANGE these weight or mass flowrate units. If cfs (or
command, or by running a time dependent m3/s) are already in an input data file, weight or
solution, the usual procedure is to leave PRY's mass flowrates can still be specified for demands
etc., that have shut off in that condition for the by giving PEAKF the value of 62.4 or 1.94 for ES
next solution or time step solution. By setting units, or for SI units setting PEAKF=9800 or =
NEWSTR = 1 this procedure will be altered so 1000, respectively. Likewise PEAKF may modify
that all devices will be opened, or put in the gpm or mgd, etc. to weight or mass flowrates.
condition denoted by the original data, prior to An example of a problem in which weight
beginning the iterative process of obtaining the or mass flowrates should be used is the feed water
train system of a steam turbine power plant where

31
the water leaving the condensor may have a NODESP
temperature of UXfF or less and be at a pressure
of 14.0 psia so that here the fluid's specific This option must be set equal to 1
weight is 62.0 Ibs/ft3 . As the fluid passes through (NODESP=I) if source pumps and reservoirs are
several stages of heaters and pumps and arrives at given node numbers. If these sources of supply
the boiler end, its temperature may be 400F or are given node numbers, these node numbers must
more, and be at a pressure of 1000 psia or more, be given in the node data after the NODES
so that its specific weight is now 53.3 Ib/ft3 . The command. If they are not numbered (NODESP
large change in the specific weight makes =0, the default) then O's are given to denote
volumetric flowrates meaningless. nonexistent nodes for pipes that connect a source
Another example is the flow of natural of supply to the network when using the PIPES
gas within a portion of the distribution network command. If nodes are numbered with
where internal energy is near enough constant to NODESP=I, then nozzles just like reservoirs (and
be assumed constant. source PUMPS) are identified by their node
number, and must also be numbered as nodes.
NHAZEN
NOMDIA
If NHAZEN = 1 USUNETWK is told to
use the Hazen William equation for the hydraulic This option has meaning only if DESIGN
computations. If NHAZEN > 1, then Manning's = 1, and then indicates to USUNETWK that a
equation is used. The same is achieved by setting list of standard pipe diameters will be given from
NEQUAT=1. which to select standard pipe sizes from. If
NOMDIA=O, the default, then the default
NOCHK# diameters given below wiU be used for this
purpose. If NOMDIA= 1, then immediately after
This option has meaning only when used the RUN command, this list of standard diameters
in conjunction with IOUTO and IOUTI. It must be provided. This data is provided by giving
determines whether the input data will be checked the number of new standard diameters in the list,
for errors, etc., or not. If NOCHK is greater than followed by the values of these diameters all on
o then the data read in from file IOUTI will not the same line. Because of the dimensions in the
be checked for errors or whether a network is program, the number in this list must be 20 or
properly defined. Thus with NOCHK > 0 less. When using SI units this list must be given
computer time can be saved. Not only is the data since the default standard diameters are in inches.
not checked for errors, but also the corrective The list of standard diameters, if the default
flowrate loops are not regenerated. Thus if the NOMDIA=1 is: 0,4,6,8,10,12,15,18,20,24,30,36,
data file written by setting IOUTO >0 is changed 42,48,54,60,72,84,96,108 inches.
so that loops must be different, then it is
necessary that this loop data which is stored in the NOMSOL
file also be changed, or NOCHK=O so that these
loops will be regenerated. It is not possible to set Has meaning only if DESIGN=l, and
NOCHK=O with IOUTI > 0 if NOCHK was set then determines whether USU-NETWK will stop
equal to 1 when the file was written with IOUTO when the special design solution is completcd, or
> O. The reverse is possible, however, that is whether it will foHow this design solution with an
NOCHK can be equal to 1 when using IOUTI > analysis solution in which the standard diameters
o even if NOCHK=O when the file was created will be used for the pipes whose diameters were
with IOUTO>O. (Default NOCHK=I) not specified. If NOMSOL=O, the default, then,
USU-NETWK stops when the solution for the
NODESO unknown diameters is complete. If NOMSOL= 1
then an analysis solution follows using the closest
Determines whether or not the output in standard diameter to those computed.
the NODES DATA table will be listed in the
order of ascending nodes or in the order Used in NOMSOL can also be used for special
the input data file. NODESO=llist in ascending computations associated with standard diameters
order (Default NODESO=I) and the deSign solution of pipe diameters.
Generally no distinction is made between the
energy line and the hydraulic grade line. If
velocity heads are significant, and these head will
cause significant head and pressure differences in in the network as the equivalent pipe, thus
pipes that join at a junction, then USU-NETWK parallel pipes will be easily identified in the
can be instructed not to ignore the velocity heads output Table of PIPE data. The default for
in computing nodal pressures and heads by setting NOPARP=O, since many networks do not contain
NOMSOL negative. If NOMSOL=-l, then this parallel pipes and considerable internal checking
distinction is made and an analysis based on the is necessary to determine which pipes are parallel,
standard diameters does not occur. If and thus this checking is eliminated. If parallel
NOMSOL=-2, then an analysis of the network pipes do exist in the network, especially if many
using the standard diameters is performed. parallel pipes do exist, NOPARP should be set
The standard diameter is generally taken equal to 1. It is obvious that any parallel pipes
closest to the computed diameter. If the diameter that are removed from the network cannot contain
just larger than the computed diameter is to be special devices such as PRVs, BPV, check valves,
selccted, then set NOMSOL equal to a value form booster pumps, differential head devices, and
10 thru 20. If NOMSOL= 15, then the analysis minor losses. If such devices should exist in only
based on the standard diameters will not be one pipe of a group of parallel pipes the option
performed. If the diameter just smaller than the NOPARP=1 can be used provided this pipe
computed diameter is to be selected, then set occurs in the input data before the other pipes
NOMSOL=21 through 30. If NOMSOL=25 then that are parallel with it, since the position of the
the analysis solution based on the next smallest first pipe is used for the equivalent pipe. Since
diamter will not be performed. If NOMSOL=99, the pOSition for the other parallel pipes are
then instead of assuming a single diameter exist removed, i.e., these pipes are not assumed to be in
for the links between the nodes, the links will be the network, diagnostic messages will indicate the
composed to two pipe sizes with the standard devices have been specified in a nonexistent pipe
diameters just smaller and just larger then the and the solution terminated. In using either the
computed diameter. The lengthS of these two Hazen-Williams or the Manning's equations minor
pipes win be computed. No analysis will follow if losses will be handled properly with NOPARP=1
NOMSOL=99 since the specified HGL elevation provided only the first pipe of the group contains
are satisfied exactly. such a device. Since the equivalent pipe must be
redetermined depending upon the total flow
through a group of parallel pipes if the Darcy-
NOPARP # Weisbach equation is used, minor losses will not
be handled completely correct even if the first
If a network contains a number of groups pipe contains this minor loss. It is also not
of parallel pipes it is possible to save possible to change any of the characteristics of
computational effort in the solution by forming an removed pipes through the CHANGE command.
equivalent pipe for these groups of parallel pipes
and thus reduce the number of pipes and number
of loops used in the solution. For example, if ten NOSTOP#
pipes are eliminated in this process the number of
simultaneous equations that must be solved will be If the correct number of energy loops is
reduced by ten also, since each such removed not determined to equal the number of unknowns,
parallel pipe removes a 2 pipe loop from the basic then USU-NETWK stops execution. This
loops of the network. NOPARP allows control condition occurs when there is an error in the
over whether an equivalent pipe will be created to specifications of the network:. By setting
duplicate the hydraulic characteristics of each NOSTOP= 1 this error condition will be ignored.
group of parallel pipes or not. If NOPARP=O Generally, however, execution will terminate later
equivalent pipes will not be formed. If NOPARP because of a singular matrix condition or some
= 1 the network will be examined for all parallel other condition.
pipes and any such group of pipes with the same
upstream and downstream nodes will be replaced
by an "equivalent hydraulic pipe" in carrying out NPERCf
the solution. The removed pipes will be restored,
however, at the time the solution tables are USU-NETWK may be assisted in
printed. All pipes in the group will be entered in obtaining a better initialization by providing
the table in the position of the pipe that remains

-33-
estimates for the amount of inflow (or outflow as NPLENG
a negative inflow) from each of the sources of
supply. To provide such estimates NPERCf must If set to a value greater than zero, i.e.,
be given a value of 1 or 5. If NPERCT= 1, then NPLENG=l, then the length of the pipes are to
estimates are in fractions of the total demand. If be given in 1000 feet or 1000 meters.
NPERCT=5, then these estimates are in the units
used in specifying the demands. The order in NPNRES #
which these estimates are given is the same as the
order of reservoirs in the input data, followed by Applicable only if ISIML= 1, and then
the order of source pumps in the input data, i.e. denotes whether pump rules and changes in pump
according to the internal reservoir and source rotational speeds are to use pressure at designated
pumps numbers. In the absence of such assistance nodes, or elevations of the water surfaces in the
in initializing each reservoir is assumed to supply reservoirs to control the number of pumps which
an equal inflow to the network if the pipes that will be operating, i.e. whether the rule depends on
connect reservoirs to the network have their the pressure at a node of a reservoir water surface
downstream nodes on the network. The elevation.
unassisted initialization assumes source pumps If NPNRES=O, then the pressure at the
supply an amount equal to their normal capacities, designated node control the number of pumps
generally. which will operate at a station. If NPNRES=l,
If NPERCf is greater than 0, the then the water surface elevation in a designated
estimates of the inflow data come after the RUN reservoir controls the number of pumps which will
or END command, i.e, are the last data line operate at a station. This same option can be
required in the input data besides the CHANGE placed in the $TDATA list.
data or SIMULATION data. This line comes
after the data described if LOOPSE= 1.

NPRINT (IOEX)
NPGPM
Determines the amount of extra outpul
This option indicates the units of flow that will be printed. A NPRINT=3 suppresses all
associated with describing the pump such extra output and NPRINT=lO gives the
characteristics. These flows may be in different maximum amount. The default value for
units than the units of demands. Depending upon NPRINT=O.
the value of NPGPM the following are assumed
regarding lhe units associated with any flowrates
associated with defining pump curves:
NPGPM=O, flowrates in cfs if ES units
NPRPRV
are used; in m3 /s if SI units are used.
Allows for the setting of pressure
NPGPM=l, flowrates in gpm (allowed
only with ES units). reduction valves and back pressure valves to be
NPGPM=2, flowrates in mgd (allowed given in terms of the elevation of its downstream
only with ES units). HGL elevation, or its pressure setting in psi, when
using ES units, or Pascals when using SI units. If
NPGPM=3, flowrates in m3/s (allowed
only with SI units). NPRPRV =0, then this setting is given in ft (or
meters) of head. This HGL elevation equals ph
NPGPM=4, flowrates in liters/s (allowed
only with SI units). +elevation of the PRY or BPV. If NPRI)RV=l
then this setting is given in psi, when using ES
NPGPM=5, flowrates in pounds/s
(allowed only with ES units). units, or Pascals, when using SI units. When this
NPGPM=6, flowrates in Newtons/s option is used, the elevation of the, PRY or B~V
(allowed only with SI units). is assumed at the point determmed from Its
NPGPM=7, flowrates in slugs/s (allowed location in the pipe as though the pipe were a
only with ES units). straight line between its nodes. (Default
NPGPM=8, flowrates in kilograms/s NPRPRV=O)
(allowed only with SI units).

34
 NPRPUM less than the value specified for the device, then
the device is removed from operation. If this
Allows for giving pressure at the inlet of process is not finding the proper status for devices
source pumps instead of the water sump elevation. effectively, i.e. devices are repeatedly shut-off and
If NPRPUM=l, pressure in psi is given, when then opened again, or removed from operation,
using ES units, and pressure in Pascals is given and then placed back in operation, then NPRTRM
when using SI units. These pressures will be the may be given a value different from the default
last item given after the command PUMPS. This value of 0 to allow the user to effect some control
option is useful when analyzing gas flows in of this process. The values of NPRTRM have the
networks by USU-NETWK, assuming that following effects:
compressibility (i.e. thermodynamic) effects can be
neglected. In such situations, the pressure of inlet NPRTRM=O Do the usual thing in shutting off
sources of supply to pumps are known, and the the flow if conditions seem to call for this, or
sump elevation is an artificial quantity obtained by removing PRY's and SPY's from operation.
dividing the pressure by the specific weight of the
gas plus the elevation of the pump location. NPRTRM=l Print the results from the solution
just obtained which indicates that the PRY should
be removed from operation before redoing the
solution with the PRV(s) removed. Examining
NPRRES this solution can give the user insight into the
operation of the PRV(s), and whether its setting
Allows reservoir sources of supply to have should be changed. (Simulation version only)
pressures specified rather than water surface
elevations. If NPRRES=l, then pressure is given NPRTRM=2 Shut-off only one PRY at a time
in psi when using ES units, and Pascals when before performing additional iterations toward the
using SI units, in place of the water surface solution. The PRY that will be shut-off, if more
elevation described after the command RESER. than 1 PRY has a negative flow rate through its
When using this option with NODESP=O, the pipe is the one in the pipe with the largest
elevation needed to convert the given pressure negative flow rate.
into an HGL elevation is taken as the elevation of
the node downstream from the reservoir since no NPRTRM=3 Display information on the monitor
ground elevation is given at the reservoir. This showing the flow rates, etc through pipes that
option is useful for pipe networks containing gas contain PR V's, and ask the user to response
flows (when thermodynamic effects are whether any PRY with a negative flow rate
insignificant) and heads are artificial quantities through its pipe should be shut-off or not. Thus
computed by dividing pressure by the specific the user can prevent a PRY from shutting off the
weight. If NPRRES=l, then pressures for all flow by indicating that its flow should not be shut-
reservoirs in the network must be given as the off. This prompting will occur during each
second item on all lines after the command subsequent iteration for which the residual is less
RESER. then TVSUM and a negative flow rate occurs in
the pipe during this iteration. Also the user will
be able to specify what PR V's should be removed
from operation, or restored to operation.
NPRTRM NPRTRM=4 The same occurs as with NPRTRM
=3 with the addition that the value of the current
Allows the user control over the manner
TVSUM (that determines when to check if PRY's
in which pressure reduction valves and back
should shut-Off) is displayed and the user is
pressure values are shut-off, or removed from
requested to give a new value that will be used
operation, etc. The usual mode of operation of
thereafter.
PRY's and SPY's is to maintain the pressure, or
HGL elevation setting that is given in the input
data. However, should the direction of flow try to
reverse, then the device will shut-off the flow NPSERI
when the residual becomes less than TVSUM, or
if the HGL elevation supplied by the system is This option allows pumping stations to be
references for purposes of specifying number of

-35-
pumps in series or parallel by: (a) their number is to introduce a transformation for pumps (see
(like the order of their occurrences in the input Jeppson's book ftAnalysis of Flow in Pipe
date), (b) by the pipe connecting them to the Networks"). This transformation generally
network, or (c) by their node number, if given a accelerates convergence, but destroys the symmetry
node number with NODESP=1. See data under of the Jacobian matrix used in the Newton
commands SERIES and PARALLEL. If Method. If NSYMMT=O, then method # 1 is
NPSERI=O, pumping stations are used in solving the problem.
identified by the pump number, i.e. by the Another method, called method # 2,
sequence number of the pump as determined by maintains symmetry in the Jacobian matrix, and
the order of the pump data after the PUMPS consequently does not introduce the pump
command. transformation. Since the Jacobian cannot be
NPSERI=l, pumping stations are made symmetric when PRY's, BPV's or
identified by the pipe number that connects the differential head devices introduced by the
source pump to the network or which contains the DHEAD command, and when pressure at nodes
booster pump. (Default, NPSERI=I) are specified with the SETPR command method #
NPSERI=2, source pumping stations are 2 can not be used if any such device exists in the
identified by their node number. NPSERI=2 can network. Method # 2 is allowed only for
be used only if these sources of supply are problems for which a symmetric positive definite
numbered as nodes by giving NODESP= 1 in the Jacobian Matrix is possible from the equations
$SPECIF list also. Booster pumps will still be that solve for the corrective flowrates in the loops
identified by the pipe they are in. of the network. USU-NETWK will determine this
if you select method # 2, and inform you.
Method # 2 is selected by giving NSYMMT= l.
NSORTP A third solution method, called method #
3, utilizes special sparse matrix methods in solving
This option control the order of pipes in the linear system of equations for each Newton
the output table. If NSORTP=l the pipes in the iteration. Method # 3 allows a symmetric, or a
output tables will be arranged in ascending order nonsymmetric matrix, and is the only method built
of their numbers, i.e. the smallest pipe number into the PC version of USU-NETWK.
will be the first entry in this table, and the largest To summarize the use of NSYMMT in
pipe number will be the last pipe. If NSORTP=O, selecting the method of solution the following
then the order in the PIPE DATA table will be occurs:
that of the pipe data contained in the input data. IfNSYMMT=-l, USU-NETWK will select
(Default NSORTP=I) whether to use method # 1, or method # 2 based
on whether symmetry of the Jacobian matrix can
be maintained or not.
NSYMMT# If NSYMMT=O, method # 1 will be used.
If NSYMMT= 1, method # 2 will be used
This option allows selection between three if pOSSible, otherwise method # 1 will be used if
separate solution methods that the full version of devices in the network do not allow symmetry, and
USU-NETWK has built into it. The PC version a message will inform you that method # 2 cannot
has but one solution method, as may be the case be utilized for this network.
with other reduced version of USU-NETWK, and If NSYMMT=5, method # 3, the sparse
therefore this use of NSYMMT will have no matrix routines will be utilized. (Default
effect. In addition NSYMMT can be used to NSYMMT=5, generally, but may be NSYMMT=-
control how USU-NETWK handles computing 1)
friction factors for use in the Darcy-Weisbach Note: If convergence is not achieved in
equation for turbulent flows that are close to using method # 2 or method # 3, and PR V do
laminar, where a discontinuity exists in the Moody not exist, you should set NSYMMT=O and see if
diagram. See description below for this use of convergence does occur. The use of method # 1
NSYMMT. Valid values for NSYMMT are: -1 may require more computer time than method #
(the default), 0,1,2,5 & 7. 2, but generally it also is capable of converging to
One method of solution that USU- a solution when pump characteristics adversely
NETWK may use, which will be called method 1, influence convergence. It is possible, however, to
create a network containing pumps for which no

-36-
 solution exists. For example, if two pumps with between being computed by the Colebrook-White
flat curves supply a network in close proximity, equation and the laminar flow equation. Should
but produce quite different heads, may result in a this occur even for one pipe in the network, then
saddle point in the mathematics of the pump with convergence to the solution will not take place.
the lower head (i.e. an iminary solution) because By specifying that NSYMMT=2, or
it can not provide the needed head. For such NSYMMT=7, a continuous relationship between
situations convergen to a solution can not occur the friction factor and Reynold's number, i.e. the
or the flowrate will be negative through one of the flowrate in each pipe is established. Use of this
pumps. Of course such situations should not be relationship should prevent the lack of
allowed in real piping systems. convergence that would occur if the regular
In obtaining a time dependent solution relationships are used. Before you can assume
USU-NETWK. will get in trouble should an that the lack of convergence is due to the above
altitude valve shut off flow from (or to) a switching between laminar and turbulent flow you
reservoir and it is attempting to obtain the should verify that the flowrates in some pipes may
solutions based on a symmetric matrix, i.e. method be laminar. This determination can be done by
# 2. The program does not know in advance if an computing the Reynold's numbers associated with
altitude valve will shut of and in the interests of the flowrates given by the nonconverged solution.
efficiency selects method # 2 if NSYMMT=-1. Another easier means for determining whether the
Should this happen you will need to rerun the flow is laminar is to set NPRINT larger, so that
problem with NYSMMT=O, or NSYMMT=5. the coefficients and exponents from the
In addition to the above use of NYSMMT exponential relationship are output, or requesting
it has the following use. By setting NSYMMT=2, the wider single table of output by setting
or NSYMMT=7 you can request that USU- OUTPUI = 0 or 2. (The PC version of
NETWK. use an alternative method for evaluating USU-NETWK. does not print the exponents when
the frictional head losses. This alternative method NPRINT is set large.) If an exponent is 2 (or one
is only applicable for use with the Darcy-Weisbach for the exponent minus 1 that is printed with
equation, and will give result that differ from the NPRINT set larger) it indicates that laminar flow
regular solution only if the flow in some pipes is exist in that pipe.
laminar, or very close tb laminar. However, the To provide a continuous relationship
use of this alternative method may be necessary to between the friction factor and the Reynold's
achieve convergence because of the discontinuity number (or flowrate), the laminar relationship
that exist in the friction factor f associated with f=64/Re is modified as follows:
Reynold's number around 2100. When the 1. The friction factor is computed from
Darcy-Weisbach equation is used, then the Colebrook-White equation for a Reynold's
USU-NETWK check to determine whether the number of 2100 and the value for the relative
Reynold's number associated with the flow in any roughness e!D for that pipe.
pipe is less than or greater than 2100 after each 2. The values of a and b in the equation
iteration. Should this Reyn<:>ld's number be less f = a/Q**b are computed so this equation
than 2100 (Le. the flow is' laminar) then the produces the f computed by step 1 for Re equal to
friction factor is computed equal to 64 divided by 2100, and f = .128 for Re = 500 (This is the f
Reynold's number, otherwise the implicit produced by f=64/Re), i.e. a straight line
Colebrook-White equation is solved (or the wholly relationship is assumed on a log-log plot between
rough equation is solved should this be the laminar relationship for a Reynold's number
applicable). The friction factor solved from the of 500 and that produced by the Colebrook-White
Colebrook-White equation has quite a different equation for a Reynold's number of 2100.
magnitude than the friction factor computed under 3. Since the laminar flow relationship will
the assumption that the flow is laminar. For produce an infinite value for the friction factor as
example at a Reynold's number of 2100 the the Reynold's number approach zero, a linear
Colebrook-White equation gives f = .050 for a relationship is assumed to apply between Re=500
6-inch cast iron pipe (e=.0102 inches) whereas the & f = .128 and an f = 1.0 for a Reynold's number
laminar relationship f = 64/Re gives f = .030. of O. The switching between the relationship from
Therefore, a discontinuity in the friction factors step (2) and step (3) however, may not occur until
occurs for flows with small Reynold's numbers the Reynold's Number is less than 200 or possibly
about 2100. In the iterative solution process the 150 depending upon what is set at the time the
computation of the friction factor may alternate program is compiled.

-37-
 In summary if the solution fails to of 100 is added to the above values of NTRAND
converge, and it can be determined that the flow then USU-NETWK will complete N time step
may be laminar in some pipes of the network, solutions between each solution whose results will
then the problem of lack of convergence may be be written in the usual manner of output. For
corrected by setting NSYMMT = 2 or NSYMMT example if NTRAND=200, only every second
= 7. With NSYMMT=2, solution method used solution will have the usual table output given. If
will be based on banding the Jacobian matrix just NTRAND=202 then every other time step will be
as if NSYMMT=O, and if NSYMMT=7, the added to the file TABLE.DAT, but the results for
solution will be based on the sparse matrix each time step will be written to the direct access
methods just as when NSYMMT=5. It is file.
important to note that the above applies only
when the Darcy-Weisbach equation is used. This If NTRAND is minus 100, or minus 200
special solution may be used for other problems a print frequency for both the direct access file,
for which no laminar flows occur, but with a slight and regular output tables can be established.
increase in computation times, generally. Giving NTRAND values such as: -101, -102, -103,
-201,-202, etc. tell USU-NETWK that the absolute
multiple of 100 is the frequency that should be
NTRAND# used to add to the direct access file, and the last
digit determines the frequency of writing regular
This option applies only if ISIML=I, and tables.
allows control over whether the usual tables of
output will be written, or a special direct access NUMPIP
data file will be created, or both. In addition
NTRAND can be used to determine the This option must be set equal to 1 if the
frequency, i.e. time increments at which, these files pipes are not given a number in the input data
will be added to. The direct access file mentioned under the command PIPE or PIPE-. If
above can be used by the auxiliary program NUMPIP= 1, then the sequence order in which the
PRINT, described in Appendix E, to provide pipe data occurs in the input file determines the
interactively most any desired portion of the pipe number, i.e. pipes are actually numbered
solution, or complete tables at any time. sequentially starting with 1 for the first pipe.
NTRAND determines the type of output for a When NUMPIP=I, pipes are designated only by
time-dependent solution as follows: their two nodes in the input, but in the output are
NTRAND=O, The solution tables for given their sequence numbers. Also for other data
each time step will consist of the regular tables that requires a pipe number to be given, such as
obtained with OUTPUl=1. A direct access file a pipe that contains a booster pump, the sequence
will not be created. (Default NTRAND = 0) number must be given. (Default NUMPIP=O, for
NTRAND=I, only a direct access file will giving pipe numbers in the input data.)
be written that contains information for auxiliary
program PRINT to provide the solution at any
time requested. NUNIT
NTRAND=2, both a direct access file and
the creation of another file will occur which This parameter is used to distinguish
contains solution tables for each time step in the problem using the International System of units,
usual format of the two table form of output. SI, from those based on the English units e.g. the
Only the tables portion of the solution will be ES system of units. It also allow for variations in
written to this extra file. The additional units for diameters, and equivalent sand roughness
information usually printed to the output file coefficient with these two systems of units. The
besides the solution tables will still be directed to value of NUNIT has the following effect:
the output file given in response to the prompt NUNIT=O, then the diameter and wall
requesting the name of the output file. The name roughness, e (for the Darcy-Weisbach Eq.) are in
of this extra file that will be created is inches, and the length in feet (or 1000 feet if
TABLE.DAT. NPLENG= 1). This is the default.
In addition to the above uses NTRAND NUNIT=I, then the diameter, wall
controls whether the results will be written to the roughness, e, and the length are in feet (or the
files after each time step, or not. If a multiple N

-38-
length in 1000 feet if NPLENG=l). and the coefficient and exponent in the
NUNIT=2, then the diameter, wall exponential equation for defining the flowrate
roughness, e, and the length are in meters (or the head loss relationship. For output control the
length in 1000 meters if NPLENG=l). option OUTPU1 indicates the following:
NUNIT=3, then the diameter and wall OUTPUl=O, then the single wider table is
roughness, e are in centimeters, and the length in used for the output of the solution.
meters (or 1000 meters). OUTPUl=l, (the default) then the two
NUNIT=4, then the diameter and wall separate tables, one contain PIPE DATA and on
roughness, e are in millimeters, and the length in containing NODE DATA are used for the output
meters (or 1000 meters). of the solution.
If the Hazen-Williams, or the Manning's OUTPUl=2, then the solution is
equation is used, then the above units designated duplicated by writing the solution to both the
for the wall roughness, e, are not applicable. This single, as well as the two tables. For time-
same control may alternatively be exercised by dependent solutions the two table form is required
using the command UNITS. if special output or the direct access file is to be
written. Also selection of string for pipe and
node identification automatically selects the two
OUTPUI table form of output as described below.

This option has three roles. The first is USE TO IDENTIFY PIPES AND NODES BY
to specify the output table(s) that should be used ALPHA-NUMERIC STRINGS
for the solution. The second is to allow pipes
and/or nodes to be identified by 8 character strings PIPES and NODES may be identified by
rather than numbers. The third is to control the alpha-numeric strings of up to 8 characters in
precision, or formatting, Le. the number of digits length, if desired, rather than using numbers
to the right of the decimal point of values printed (integers). To communicate this usage to USU-
in the output solution tables. The valid values for NETWK the option OUTPU1 in the $SPECIF list
OUTPU1 are: 0, 1, 2, 11, 12, 13 (-11,-12,-13 in the should be given a value of 11, 12 or 13. The
case of the simulation version) and these values following applies:
added to 1000, 2000,. . 9000 to control the 1. If OUTPUl=ll then both PIPES and
formatting in output tables. The first three are NODES are identified by 8 character strings.
for the first role, the values 11, 12 & 13 denote 2. If OUTPUl=l2 then only PIPES are
whether "pipes & nodes", just "pipes", or just identified by 8 character strings.
"nodes" will be identified by characters, 3. If OUTPUl=l3 then only NODES are
respectively, and adding the thousands control identified by 8 character strings.
formatting. When using the version of USU-NETWK
that also allows time dependent solutions, then
USE TO CONTROL OUTPUT SOLUTION the above values can be preceded by a minus sign
TABLES with the effect that in the solution tables the
USU-NETWK can use either or both of internal pipe and node numbers will be given in
two different forms to report the usual solution addition to the alpha-numeric identifiers. Thus, if
results. One form, the default forms, consists of OUTPUl=-ll then in the PIPE DATA table
two separate tables for the solution. The first (given as part of the solution) the first three
table provides PIPE DATA, i.e. there is an entry columns will contain: 1. the internal pipe number,
for each pipe of the network in this table, and the 2. the upstream internal node number, 3. the
second table provides NODE DATA, i.e. there is downstream internal node number, 4. the pipe
an entry for each node of the network in this string identifier, 5. the upstream node string
table. Unless the option of using 8 characters to identifier, 6. the downstream node string identifier,
identify pipes, and nodes is used, these tables are and the other usual items of data. Each line in
restricted to 80 column in width. The second the NODE DATA table will contain the follOwing
form of output consist of a single table that can columns: 1. the internal node number, 2. the node
used up to 132 columns across a line. This single identifiers, and the other usual items of data. If
table does, however, contain information not in OUTPUl=-l2, then the internal pipe numbers will
the two table form of output, such as the friction be given in the first column before the pipe
factor, if the Darcy-Weisbach equation is utilized, identifiers. If OUTPUl=-l3, then after the pipe

-39-
 numbers in the PIPE DATA table two extra new command, including the RUN and END must
columns will be included for the internal upstream be preceded by the word STOP. The exceptions
and downstream node numbers, and the NODE are the first command, NODES, and if only pipes
DATA table will contain the internal node are identified by character strings, then the first
numbers in addition to the node string identifiers. command PIPES is not preceded by STOP.
Presently the PC version of USU-NETWK accepts S. It is not possible to use the
only positive values for the option OUTPUl. "CHANGE" command to define alternate analyses
Furthermore, if PIPES and/or NODES are by using PIPE or NODE character identifiers.
identified by alpha-numerics strings then the input If CHANGE data does change pipe diameters,
data file to USU-NETWK must adhere to the lengths, or demands at nodes, then the internal
following special requirements: number for the pipe or node must be given.
1. NODES must be the first command Internal numbers are described later.
that is given after the $SPECIF list of options. 6. When giving lists of pipe identifiers
The exception is if only PIPES are given as under such commands as VALVC (check valves),
character strings (OUTPU1=l2); then the PIPES or MINOR (minor losses) , it is necessary that
command may be first. each succeeding identifier be within the next field
2. Since up to 8 characters, including of 8 character width. This list of pipe identifier
blanks, can be used for the character identifiers it must be terminated by a / in the first character
is necessary that this data be given in 8-character position of the next 8 character field. If pipes are
field widths. This requirement means that after numbered in the usual manner (OUTPU1 = 13),
the PIPES command that the first 8 positions of then this list may be the usual free format for the
each input line are for the pipe identifier, and if integers it contains.
NODES are also identified as character strings 7. The options NSORTP and NODESO
then positions 9 through 16 and 17 through 24, for having entries in the PIPE DATA and NODE
respectively, are reserved for the two nodes at the DATA output tables listed in ascending order,
ends of this pipe. In identifying devices such as now have no meaning. Both of these tables will
source PUMPS, RESERvoirs, and NOZZLes the retain the order of the input data under the
character string identifier for the pipe or for the PIPES and NODES commands, respectively.
node must be in the first 8 columns of these lines Furthermore, it is not possible to use the single
of input data depending upon whether the option table output (e.g. that given if OUTPU1=0, or
NODESP=O, or NODESP=l, respectively. If =2).
NODESP=O, and pipes are numbered (not 8. If the option NODESP=O (the default
identified by a character strings), then the usual unless set to 1 in the $SPECIF list), then
free format applies to the later integer data, and upstream nodes for pipes that connect sources of
if NODESP=l, and nodes are numbered, then the supply to the network do not have an upstream
usual free format applies to these lines, and the node. In using numbers for nodes these non-
node number (integer) is the first value on the existent nodes are identified by a zero in the data
line. For devices such as VALYES (PR V's), and after the pipes command. When using alpha
VALVC (check valves), that are always in pipes, numeric identifier for nodes (IOUTPUll =11 or
the 8 character fields must be used for the pipe 13) the same zero can be used provided it occurs
identifies for these input lines if OUTPU1=11 or in the first column of the 8 width field for this
= 12, but these lines of input retain their usual node identifier. The other alternative is to leave
free format input if pipes are numbered, i.e. the full 8 character field for this non existent node
OUTPU1=13. A / at the beginning of the next identifier blank. If NODESP=O, and pipes are
field terminates the list. denotes with a alpha numeric identifier
3. The PIPE-nodes command cannot be (I OUTPUll =11 or 12), then the identifier for the
used to enter any data if either PIPES or NODES pipe that connects a reservoir or a source pump to
are identified by character strings. Rather the the network must be used in identifying pipes
commands PIPES and NODES must be used. under the RESER and PUMPS commands.
4. Since any character string is accepted 9. The PIPE DATA output table will
in the first column for the identifier, except a % exceed 80 columns and, therefore, if the output is
which denotes that this line does not constitute a sent to a monitor with only 80 columns across a
line of input, it is necessary to terminate data of line directly by giving TTY for the output file
a given type. This termination is done by the name then each line will be broken with the last
word STOP in columns 1 through 4. Thus each portion of the line displayed below the first

-40-
 portion of the line. Thus the table will be the option NODESP=O of not having sources of
difficult to read. The same is true if output files supply numbered as nodes are given below.
are displayed with a TYPE, or COPY DOS
Both PIPES and NODES are identified by
command. (A terminal or PC with the capability character strings.
of displaying 132 characters across a line can be {*
$SPECIF NODESP=l,NFLOW=l~NPGPM=l.
set to this mode and show the tables properly.) OUTPUT1=11,PEAKF=2.5 $ENu
NODES
When printing output files written by USU- NODELN1 a 301
NODELN2 2001
NETWK with abs(OUTPUl) > 10 (but without NODELN3 2001
NODELN4 1501
thousands added) either wider paper than 8 1/2 NODELN5 200/
NODELN6 1501
inch paper must be used, or a font should be used NODELN7 0/
that prints 16.5 characters per inch. This applies NODELN8 0/
STOP
only if PIPES are designated by character strings. PIPES
PIPELN1 NODELN8 NODELN1 500 8 130
See the description of the auxiliary program PIPELN2 NODELN1 NODELN2 5001
PIPELN3 NODELN2 NODELN3 500/
PRNT to accomplish this. PIPELN4 NODELN1 NODELN4 1500 61
PIPELN5 NODELN3 NODELN4 2000 61
10. Much of the extra output that can be PIPELN6 NODELN4 NODELN5 25001
PIPELN7 NODELN6 NODELN5 2500 81
requested by setting the option NPRINT larger PIPELN8 NODELN6 NODELN3 15001
PIPELN9 NODELN7 NODELN6 5001
than 0 will show pipes and nodes as internal STOP
integer numbers rather than their character RESER
NODELN8 100
strings. These internal numbers are established by NODELN7 120
STOP
the order in which the input data occurs after the VALVC
PIPELN3 PIPELN7 PIPELN6 /
PIPES and the NODES commands. These STOP
RUN
internal numbers can be displayed associated with
Only PIPES are identified by character strings.
the alpha-numeric identifiers by setting the option
NPRINT=5 or larger, and as noted above can be ~;PECIF NODESP=l ANFLQW=l,NPGPM=l,
OUTPU1=12,PEAKF=~.5 $END
shown in the solution tables by placing a minus in NODES
1 0 301
front of 11, 12 or 13 when using USU-NETWK 2 2001
3 2001
on larger computers. 4 150/
5 200/
11. When specifying number of pumps in 6 1501
7 01
series or parallel at stations with the SERIES or 8 0/
STOP
PARAL commands, the option NPSERI will PIPES
PIPELN1 8 1 500 8 130
automatically be set to 0, and it is necessary that PIPELN2 1 2 I
the data after these commands use the pump PIPELN2 3 2 /
PIPELN4 1 4 1500 6/
station numbers, e.g. the order of the data after PIPELN5 3 4 2000 6/
PIPELN6 4 5 2500/
the PUMPS and BOOSTer commands. Also it is PIPELN7 6 5 2500 8/
PIPELN8 6 3 1500/
necessary that the command BOOSTer be used PIPELN9 7 6 5001
STOP
rather than BPUMP for booster pumps. RESER
8 100
Use of character strings does not change 7 120
STOP
the manner in which USU-NETWK internally VALVC
keeps track of pipes, nodes, loops, etc. for the PIPELN3 PIPELN7 PIPELN6
STOP
network. This internal book keeping is done with RUN
the internal numbers. These internal numbers are Only NODES are identified by character strings.
1*
established by the order in which PIPES and $SPECIF NODESP=l ANFLOW=l ANPGPM=l,
ICOST=1.0UTPU1=1~.PEAKF=~.5 SEND
NODES are first entered in the input data. The NODES
NODELN1 0 301
use of character strings is a convenience to the NODELN2 2001
user who wishes to have a more easily understood NODELN3 2001
NODELN4 1501
mean for separating pipes and nodes in different NODELN5 2001
NODELN6 150/
parts of the distribution network. An alternative NODELN7 0/
NODELN8 0/
to the use of character strings is to use numbers STOP
PIPES
starting with different hundreds, or possibly 1 NODELN8 NODELN1 500 8 130
2 NODELN1 NODELN2 500/
thousands, but this limits the number of zones, or 3 NODELN2 NODELN3 500/
4 NODELN1 NODELN4 1500 6/
pipes within a designated zone. 5 NODELN3 NODELN4 2000 6/
6 NODELN4 NODELN5 1500/
Example of input data files for a small 7 NODELN6 NODELN5 2500 8/
network that identify: (1) both PIPES and 8 NODELN6 NODELN3 1500/
9 NODELN7 NODELN6 500/
NODES, (2) only PIPES, (3) only NODES with STOP
RESER
character strings and (4) the same as (1) except NODELN8 100
NODELN7 120
STOP
VALVC RESER
376 PIPELN1 100
STOP PIPELN9 120
RUN STOP
INTEREST=.12 VALVC
LIFE=80 PIPELN3 PIPELN7 PIPELN6
RESERVOIRS STOP
CAPI=3000 RUN
WATER RESER
UNIT=3000 RESEER,1 PIPELN1 100
UNIT=4500 6,RESER,2 PIPELN9 120
UNIT=12000 ,NODES , 1-5 STOP
OPERATING VALVC
CAPI"'2000 PIPELN3 PIPELN7 PIPELN6 /
UNIT=2000, PIPES, 1-4 STOP
ADDITIONAL RUNINTEREST=.12
MAINT. LIFE=80
UNIT=4000,RESER RESERVOIRS
UNIT=3500,FLOWS CAPI=30000
UNIT~2500,PIPES,1-4 WATER
END UNIT=30000,RESER,1
UNIT=45000,RESER,2
Pipes and Nodes iden~ified by charac~er UNIT=12000,NODES,l-5
s~rings OPERATING
/* CAPI=20000
$SPECIF NFLOW=l,NPGPM=~~ICOST=l UNIT=200Q. PIPES , 1-4
OUTPU1=12,PEAKF=2.5 $ENU ADDITIONru.
NODES MAINT.
NODELNl 0 30/ UNIT=4000,RESER
NODELN2 200/ UNIT"'3500,FLOWS
NODELN3 200/ UNIT=2500 , PIPES , 1-4
NODELN4 1501 END
NODELN5 2001
NODELN6 1501
STOP
PIPES
PIPELN1 0 NODELN1 500 8 130
PIPELN2 NODELN1 NODELN2 500/ USE TO CONTROL FORMATTING IN TWO
PIPELN3 NODELN2 NODELN3 500/
PIPELN4 NODELN1 NODELN4 1500 6/ TABLES FORM OR OUTPUT
PIPELN5 NODELN3 NODELN4 2000 61
PIPELN6 NODELN4 NODELN5 25001
PIPELN7 NODELN6 NODELN5 2500 8/
PIPELN8 NODELN6 NODELN3 1500/ It is possible to change the number of
PIPELN9 0 NODELN6 500/
STOP digits printed beyond the decimal for the following
quantities when using the two table form of

Table showing quantities whose printed accuracy is controlled by OUTPUl.

PIPE DATA: NODE DATA;

units digit (3) controls length format units digit (3) controls pressure format
10th digit (2) controls diameter format 10th digit (2) controls elevation format
lOOth digit (1) controls flowrate format 100th digit (1) controls demand format

Table showing which parameters can have their printed formats altered using values of the option OUTPUI
equal to 1000 or greater.

Thousands Controls FORMAT (digits beyond decimal point) in position*

Digit PIPE DATA NODE DATA
1000 + length-3,diameter-2,flowrate-l
2000+ demand-3,elevation-2,pressure-l
3000+ flowrate-l demand-3,elevation -2
4000+ diameter-2,flowrate-l demand-3
5000+ diameter-2 demand-3,pressure-l
6000+ length-3 elevation-2,pressure-l
7000+ length-3,flowrate-l elevation-2
8000+ length-3,diameter-2 pressure-l
9000+ diameter-3 elevation-2,pressure-l
* OUTPUI =
x,xxx (Value from 1000 to less than 9999)
tttt
123

-42-
Examples
OUTPUl-3101 - digits beyond decimal: l-flowrate in PIPE DATA; l-demand & O-elev. in NODE DATA
OUTPU1=6210 - digits beyond decimal: O-length in PIPE DATA; l-elev. & 2-pressure in NODE DATA

output in (a) the PIPE DATA: length, diameter, corresponding head, H, as a pair), or more then
and pipe flowrate; and (b) the NODE DATA: first three points are to be used to define pump
column of demand, elevation, and pressure. The characteristics. The operation of pumps is defined
control over these formats is by means of the by passing a second degree polynomial through 3
option OUTPUI as given in the tables below. pairs of values of Q and H. This polynomial
Incontrol over these formats is by means of the understanding the information in these tables the
option OUTPUI as given in the tables above. In option OUTPUI is given a value between 100 and
understanding the information in these tables the less than 9999. Starting from the left side of this
option OUTPUI is given a value between 1000 value the digits after the thousands are numbered
and less than 9999. Starting from the left side of from 1 to 3. The value give to these 3 individual
this value the digits after the thousands are digits specifies the number of digits that will be
numbered from 1 to 3. The value give to these 3 printed after the decimal point. This becomes the
individual digits specifies the number of digits that mathematical description that replaces the "pump
will be printed after the decimal point. curve" and defines its operating characteristics.
Thethousands denotes what will be controlled, i.e. Three points are generally adequate to properly
1000 indicates only PIPE DATA formats will be define the operation of a pump if for the solution
controled; 2000 indicates only NODE DATA requested it is operating near its normal capacity
formats will be controled, etc. as shown in the (where the pump achieves it maximum efficiency).
second table below. The values given to the digits However, if a pump is poorly suited to do the job
numbered 1 to 3 control different parameters. needed in a network, it may be forced to operate
The first (or No.1) digit, or the hundreds digit, with a flowrate, and head quite different from its
controls either the format of the tlowrate in the maximum efficiency point. Under such conditions,
PIPE DATA, or the format of the pressure in the or under vastly varying conditions at different
NODE DATA The second (or No.2) digit, or times during a time-depend en solution, 3 pairs of
the tens digit, controls either the format of the points on a pump curve may not be adequate. By
diameter in the PIPE DATA, or the format of the adding the option PCHAR3 = 0 to the $SPECIF
elevation in the NODE DATA output solution list of options up to 10 pairs of values of Q versus
table. Finally the No.3, or units digit, controls H can be given to define pump characteristics.
either the format of the length in the PIPE When more than 3 pairs of values are provided,
DATA, or the demand in the first column of the USU-NETWK still fits a 2nd degree polynomial
NODE DATA between 3 pairs of consecutive points, but the 3
Anytime the value given to OUTPUI is points will be altered as the solution progresses so
1000 or greater, two things are communicated to that the flow in the pipe containing the pump
USU-NETWK. First it should use the default two during the iterative solution processes will be
table form or output, and second that the number bracketed by the 1st and last of these three points,
of digits printed after the decimal point of three unless the flowrate is less than or larger then the
parameters is specified by the values in the three smallest Q and largest Q in the list. When giving
numbered digits. Thus if it is desired that the more than three points they must be arranged in
length have two digits printed beyond the decimal ascending order of magnitude of Q.
point, and the diameter have zero decimal points, If PCHAR3 = 1 (the default), only three
then OUTPU1=1202 should be included in the pairs of values for flowrate Q and pump head H
$SPECIF list of options. The first 2 after the 1 can be given. The order of these points if then
for the thousand simply keeps the standard 2 unimportant.
digits of precision for the pipe flowrate. The 1000 If a solution shows negative heads
indicates the format of the parameters in the first produced by some pumps, or fails to converge
column of the table below are being specified, i.e. properly the problem might be eliminated by
the parameters in the PIPE DATA defining the pump characteristics with more than
three points. When doing this it is even
permissible to provide some heads that are
PCHAR3 associated with negative flowrates, as an extension
of the pump characteristic curve to the left of the
This option is used to communicate if origin. This extension may be made artificially
three pairs of points (flowrate, Q, and a steep, but should be smooth enough to ensure
that a 2nd degree polynomial fits the data wiLhout

-43-
 wild behavior between the points. An alternative inches.) See Appendix C for more information.
= to using more than three points is to spread the
points out over a wider range of flowrates. USU-
NETWK assumes that the middle of the 3 points PRESMA
represents the normal operating conditions for the
pump, and if the solution shows considerable If a value is given to PRESMA other than
deviation from this operation a message is written zero, then the pressure at each node will be
informing the user of this. examined, and any pressure greater than this value
A utility program PUMCUR, which is will be flagged. All such flagged values will be
described in Appendix F is designed to help you written to the output file before the summary
easily visualize how the pump data you give is table(s) which give the solution results. The
interpreted by USU-NETWK in defining how pressure given by PRESMA is in psi, when using
much head a pump produces as a function of the ES units, and Pascals when using SI units.
flowrate passing through the pump. Should
convergence to a solution not occur, or negative
flows occur in pipes containing pumps, then you PRESMI
might well be specitying pumps that are
incompabible in their combined operation. Such If a value is given to PRESMI other than
situations should be avoided in practice since a zero, then the pressure at each node will be
pump under these conditions will only consume examined, and any pressure less that this value will
energy but not contribute to the flow through the be flagged. All such flagged values will be written
network. to the output file before the summary table(s)
which gives the solution results. The pressure
given by PRESMI is in pSi, when using ES units,
PEAKF (FLOWFC) and Pascals when using SI units.
All external demands are multiplied by
this value. Giving this option different values for PRVLOS
consecutive solutions allows the total demand level
to be changed without preparing new input data. Under special conditions the upstream
Only one value needs to be Changed. This option pressure may drop below the pressure setting of
might also be used to convert from any units PRY's. Should this condition occur the PRY
which the demand may be given in, into cfs, etc. becomes inoperative, but acts like a minor loss
For example, if demands are in ac-ft per day device. PRVLOS is the minor loss coefficient, K,
PEAKF should equal 1.9835 to convert to cfs with in the equation HL = K(Vi2g) for such minor
NFLOW=O. Likewise giving PEAKF=l.55 is losses. Default is PRVLOS = 10)
equivalent to setting NFLOW=2. Also if the
convention of the data is minus for demands
outfrom the network (opposite to the convention
of USU-NETWK) this can be allowed by setting
REYNUM
PEAKF=-l. (See the commands PEAKF,
If the Darcy-Weisbach equation is used to
andDFRAC as well as command under
determine head losses as a function of flowrate
CHANGES for other possible method for
and if minor losses are present, the equivalent
changing demands from those in the input data
length of pipe to duplicate the minor loss device
file on a more selective basis.)
is computed on the basis of a specified ReynOlds
Number. This equivalent length is not adjusted
during the solution process if LENORG=l.
REYNUM is the Reynolds for which the
PLOTH # equivalent length is first computed. If
LENORG=O, the value given to REYNUM will
Determines the height (in the y direction) have no effect on the final solution results.
of the plot on the Calcomp Plotter in inches. The (Default REYNUM=500,OOO.
maximum and minimum y-coordinates will be
found, and the x and y coordinates scaled so that
this specified height is achieved. A menu will
appear just before the plot is made that allows
you to change this value. Default is PLOTH=20

-44-
 TVSUM VELMAX
If a value other than zero is given to
This option contains the value of the VELMAX, then the velocity in each pipe will be
absolute sum of changes in flowrates between examined. Any velocity whose magnitude exceeds
consecutive iterations at which the flow in pipes the value of VELMAX in feet per second, when
containing PRY's is checked to see if the flow using ES units, or in m/s when using SI units, will
direction is reserved. If a negative flow occurs in be flagged. All such flagged values will be written,
any such pipe the PRY, etc. is shut-off. No and identified in the output solution file between
check is made to determine if a PRY, etc. should the summary tables which give the PIPE DATA
shut-off if this sum of changes is greater than results and the NODE DATA results. These
TVSUM. The default value is 2 + (No. pipes)/10, velocities will be examined only if the two table
but not more than 40. form of output (OUTPU1=1 or 2) is selected.

VELMIN
TVSUMI #
If a value other than zero is given to
In order to prevent an occurrence in VELMIN, then the velocity in each pipe will be
which a PRY, or check valve, continues to cycle examined. Any velocity whose magnitude is less
open and then close between consecutive than the value of VELMIN in feet per second,
iterations, the value of TVSUM is reduced each when using ES units, or in m/s when using SI
time a device changes its condition of operation. units, will be flagged. All such flagged values will
TVSUMI is the factor by which TVSUM is be written, and identified in the output solution
multiplied if a device shuts-off. (Default file between the summary tables which give the
TVSUMI =0.85) PIPE DATA results and the NODE DATA
results. These velocities will be examined only if
the two table form of output (OUTPU1=1 or 2)
TVSUM2 # is selected.

This option serves the same purpose as VIse

TVSUM1, except it is the multiplying factor when
the device opens again after being closed. The kinematic viscosity of the fluid in
(Default TVSUM2=0.85) ft2/sec or m2/s depending respectively whether ES
or SI units are used. The default value is
0.00001217 ft 2/sec (water at 60F). if ES units are
used or 0.000001131 m2/sec (water at 15.6q in SI
units are used.

NOTES RELATED TO OPTIONS

-45-
-46-
DETAILED STEADY-STATE DATA REQUIREMENTS
Commands that enter data The latter method is most convenient,
especially if subsequen t analyses may wish to
Numeric data which describe the network utilize this data again. NETWK will ignore any
in detail are introduced by a command, or line in the input file that contains a % in column
descriptive name, that precedes the numeric data # 1. USU-NETWK also will ignore any
of a specific type. The first 5 characters of these information that occurs after a / anywhere in a
commands must be identical to those described in line unless this I is followed immediately by an R
this portion of the manual. Furthermore, these in which case the text thereafter wil1 be taken as
commands must be given in upper case letters, so a remark as explained in the next paragraph.
when making up the input data file, etc. make
sure that the "caps Lock" is on. The individual Use of "remarks" at the end of input lines
commands are (listed in alphabetical order): (Not implemented in Steady-State PC program)
BOOSTer, BPVALve, CHECK input, DESIGN,
DFRAC, DHEAD, END ,ERROR, ITERAtion, Remarks can occur on any data line after
LPIPE, NLPIPe, LNODE, NLNODe, MINOR, a /. If these remarks occur on lines after either
NODES, NOZZLes, PARALlel, PEAKF, PIPES, the PIPES, NODES, or PIPE- commands they will
PIPE-nodes, PRINT, PUMPS, QGIVE, be printed in the solution tables at the end of
RESERvoir, ROTATional, RUN, SERIES, these same pipe or node lines, provided that an R
SETPRessure, UNITS, VALVCheck, VALVE follows the I. In other words to tell USU
(prv), VISCOsity, WEIGHT. Characters beyond NETWK that you wish to have these remarks
the fifth character are not examined by the retain and printed in the solution tables an R
computer and, therefore, command names may be must follow the I without any blanks between.
added to if additional characters help clarify the Thus if a line after the PIPES command is:
name. There must be at least one line containing 12 4 7 2400. 20./R THIS PIPE NEEDS TO BE
data after the command if this name implies data REPLACED the remark "THIS PIPE NEEDS TO
follows. The three commands that do not call for BE REPLACED" will appear at the end of the
data are: CHECK, RUN and END. Generally all line for pipe 12 in the solution table for PIPE
the data of this type will follow a single command. DATA. Likewise, any remark on a NODES line
However, as many of the same commands as preceded by an /R after the required data will be
desired may be included in the input data file. written in the output file after the line for this
The exceptions to this rule are that the data given node in the NODE DATA solution table.
on the last of the following lines will be used: Similarly remarks after a /R on a line after the
lTERA, UNITS, ERROR, VISCO and WEIGHT, PIPE- command will appear on the appropriate
hecause these commands introduce a single value. line in the table for PIPE DATA. Remarks are
These same values can be introduced by options restricted to 32 (or 36 depending on the computer
in the $SPECIF list. R UN and END are used) characters after the /R, and if they occur on
exceptions also since they indicate that the data lines that are not truncated by a I, then the /R
for describing the network has all been given, and must follow the last number without any blanks,
no more commands of those listed above are read or a comma, between. These remarks can be used
after an END or RUN only in connection with the two table form of
output, that is the option OUTPU1=1, or
Use of % in column 1 OUTPU1=2 must be in effect. The ability to
retain, and write remarks in the solution tables
In doing several different analyses of a does not exist in the PC steady-state version of
given network in which some components of its USU-NETWK
description are changed it may be desirable to
keep the original data available. Two methods of General requirements. free format. and
keeping data from previous analyses in the file truncation of Jines
are: (1) Use an editor to move these lines of the
file to after the last END of the data file, and Every data file must contain a PIPES and
(2) placing a % in the first column in front of NODES command, or the command PIPE- with
the lines. appropriate data following them, since every

-47-
 network has pipes and nodes. The other item. Many pipe will have the same diameter, and
= the above rule indicates that they need not be
commands are optional, depending upon the
configuration of the network, and the type of given provided the diameter last given is still
analysis being requested. The data after all correct. The first item in the list, the pipe
commands is free of any format requirements, number, is different for each pipe, and therefore
unless special options are included in the it can never be truncated.
$SPECIF which alter this. If OUTPU1> 11 an
exception occurs so that data fields 8 characters in List of commands and what they enter
width are reserved for the pipe and/or nodes that
are identified by a string of characters rather than The type, and order of data, which follow
by integers. Another common exception is if each of the commands are described in the
NETPLT=13 or =14, indicating that the data is following paragraphs. The items described by the
formatted according to that produced by the numbers 1, 2 ... under each of the heading, which
preprocessor PIPEINPT. The free format data are the commands, established the default order of
can separate individual items in the list with a data under this command. In the case of the
blank or more between them, or a comma, or a simulation version of USU-NETWK this default
comma followed by one or more blanks. order can be changed by the options whose names
It is not necessary to provide all of the begin with COR. The descriptions of these
items call for under the commands. The rule that commands are given below in alphabetical order
USU-NETWK follows is that if a line of input is to make it easier to find a given command, rather
terminated with a / before all of the items called than according to their frequency of usage. See
for under the command are given, then the the first paragraph in this major section for a list
missing items should equal the value last given to of commands that are available.
that particular item. In other words a / in a line
of input causes two things to occur: 1. the BOOSTer pump or BPUMP
reading of this line is stopped, and 2. items called
for are given values equal to those last given in The input under the command BOOST or
the input file. Generally it is not necessary to BPUMP is identical to that described under
terminate a shorten list with a /, since for most of PUMPS with the exception that nodes cannot be
the input USU-NETWK expect all called for items used to identify booster pumps even if
10 be on a single line of input. After reading 20 NODESP= 1, and no elevation of the source of
consecutive blanks termination in reading a line supply exists. The direction of flow through the
occurs as if a / existed. However, for some lists, pump must be specified by the "upstream" and
such as the list of pipes containing check valves, "downstream" nodes in the pipe data for this pipe
several lines may be read, and these lines must be as entered under the command PIPES or PIPE-.
tcrminated with a /. The input consists of use (a), (b), or (c) below.
The exception to the rule that data at the
end of a list can be omitted by giving a /, is the (a) using three (or more points, see PCHAR3)
data after the command PIPE-. Under this points to define the pump's characteristics,
command decimal are required to be given with all 1. The pipe number (which is the pipe
real values, and decimal points are not given with identifier if OUTPU1 = 11, or OUTPU1 = 12) that
integer values such as pipe and node numbers. contains the booster pump,
Here data in the list can be left out from within 2. The flowrate Ql'
the line. With the other data after a command, 3. The head, H 1 produced by the pump
the default order of item under the command are corresponding to Ql'
arrange so that generally the most frequently 4. The f1owrate, Q2' corresponding to
repeated items occur at the end. For example most efficient operation of pump
after PIPES the last item in the list is the 5. The head, H 2, produCed by the pump
roughness coefficient. If, therefore, all pipe in the corresponding to Q2'
network have the same roughness coefficient, it 6. The flowrate Q3,
need only be given on the first line after PIPES. 7. The head, H3, produced by the pump
Then if a subsequent analysis is to use a different corresponding to Q3.
roughness coefficient, only one value need be This 7th item is terminated with a / or a 0 as an
changed in the input data file to define this new extra value for the water surface elevation.
problem. Diameter is the second from the last

-48-
 (b) giving power the pump supplies to the fluid meters depending upon whether ES or SI units
- and its normal capacity, are used, respectively.
1. The pipe number(or identifier) that 3. The pressure setting of the BPV,
contains the booster pump, expressed in elevation of the hydraulic grade line,
2. The power of the pump times its HGL, immediately upstream from the valve. The
efficiency in units of horsepower when using ES HGL elevation is in feet or meters depending
units, and in units of KiloWatts when using SI upon whether ES or SI units are used respectively.
units, If the option NPRPRV=1 then pressure is given
3. The normal capacity of the pump, i.e. instead of the elevation of the HGL, but this
its flowrate at peak efficiency. HGL is computed on the basis of the pipe being
This third item should be terminated with a I a straight line between its end nodes.
For additional explanation about the input
data after the BOOST or BPUMP command see Example of back pressure valve data
the description under PUMPS. BPVAL
5500 340
(c) giving coefficients to 2nd degree polynomial. 15300 360
(See explanation under command PUMPS for this
input.) CHECK
Example of booster pump data This command does not have any lines of
BOOST data following it. If the command CHECK is in
20 1.2 400. 1.0 420..8 430./ the input data file, it informs USU-NETWK that
it should not attempt to obtain a solution, but
BPVAL (back pressure valve) rather check the data to find any error. This
command may be used when you first try
A back pressure valve, BPV, is used to "running" a new network problem, and you know
maintain a constant pressure at the upstream side from your past experience that you will
of it, thus maintaining a pressure in the upstream undoubtedly have made some mistakes. The
portion of the network that would become too command END must be the last of the detailed
small if the BPV were not installed. A BPV data requirements when a CHECK is given. RUN
accomplishes this by discharging only sufficierit and CHECK imply opposites.
flowrate through the pipe in which it is installed
to maintain the desired upstream pressure. DESIGN
The maintenance of a constant upstream
pressure may be considered the normal mode of The command DESIGN is an alternative
operation of a back pressure valve. Obviously, to setting DESIGN=1 in the $SPECIF list of
should the pressure on the upstream side of the options. The command DESIGN must appear
BPV drop below its pressure setting, the value can before the NODES, or PIPES command or the
only shut-off the flow, but cannot maintain the PIPE-nodes command since USU-NETWK must
pressure setting. A shut-off BPV is a secondary be informed that pressure are provided following
mode of operation. Finally, as a third mode of the nodal elevations and it is not to terminate
operation, some back pressure valves permit when a zero is given for a pipe diameter.
reversed flow from the downstream to the DESIGN indicates to USU-NETWK that a special
upstream sides of the valve. In this mode the design solution is requested in which it is to solve
BPV becomes effectively inoperative, but since it for NJ pipe diameters or roughness coefficients,
is an obstruction in the line it becomes a minor etc. as defined under the option DESIGN.
loss. USU-NETWK determines the mode of (NJ=the number of junctions in the network).
operation and does what is necessary to Values after the command DESIGN can be: 0, 1,
accommodate how a BPV actually works in 2, 3, 4 or 5. The input data accompanying this
practice. special design solution requested with DESIGN
Data for a back pressure valve consists of: equal to 1 or 3 must (a) give NJ pipe diameters
1. The pipe number (integer) that equal to zero, and (b) provide the pressures (or
contains the back pressure valve. HGL elevations), as well as, the demands at all
2. The distance between the back pressure nodes of the network.
valve and its pipe's downstream node in feet or

-49-
 DFRAC is introduced in a pipe somewhere in the network,
which is the differential head device.
The command DFRAC has one value that The type # 2 differential head device is
follows it. This value is a new peaking factor, or very useful in designing a network in that it allows
multiplier of demands. Any nodes entered in the both the pressure (or HGL elevation), as well as
input list after DFRAC will have their demands the demand, to be specified at a given node. In
multiplied by the given value. Thus the command order to specify both of these, something in the
DFRAC allows the value given to PEAKF (or network must increase or decrease the head, such
FLOWFC) in the $SPECIF list to be altered as a pump, or a pressure reduction valve.
repeatedly through the input data. An additional However, the amount of this increase or decrease
NODES command (or the PIPE-nodes command) in head is unknown. The magnitude of the
would follow each occurrence of the command differential head will be provided as part of the
DFRAC, unless DFRAC is used as a replacement solution. From this magnitude an appropriate
of PEAKF, and occurs before the NODES or device can be selected to provide the incremental
PIPE- command. head indicated. If the magnitude is positive, a
booster pump is called for, or if the frictional loss
DHEAD is greater in magnitude than the positive
differential head, a larger size pipe can be used.
This command is used to introduce a If the magnitude is negative a pressure reduction
differential head device which allows a solution to valve can be inserted in the pipe or a smaller size
situations in which both demands and pressures pipe used to give the same frictional head loss as
are specified at some nodes in the network. This the given size pipe, plus the differential head loss.
command is useful to size network components, If the option NEQUDI=l, then USU-NETWK
such as heads and capacities of pumps, setting for provides a table of such equivalent pipe sizes,
pressure reduction valve, or determine pipe whenever an equivalent pipe can be used. Thus,
diameters, i.e. design network components. in addition to providing design information for
A differential head device may create a pumps, or head reduction devices, the type # 2
positive head in a pipe line such as a booster differential head device can be used to size pipes.
pump, or a negative head, such as a pressure The user must use sound judgement in
reduction valve, in addition to the fluid frictional specifying differential head devices of type # 2,
head loss. There are two types of differential because a problem can easily be specified for
head devices, as far as USU-NETWK is which no solution exists. For example, it is
concerned. The first type, or type # 1, is one for obvious that an HGL elevation, or pressure, at a
which the amount of the differential head is node may be specified too small for even an
specified. The second, or type # 2 is referred to infinite resistance to flow in a pipe, to meet,
as a design differential head device for which the especially if this pipe is far removed from the
amount of the differential head is unknown. The node where the pressure is speCified. The flow
magnitude of this differential head is to be simply comes to the node through other paths,
determined such that the pressure, or HGL and arrives with too large a pressure.
elevation, at a specified node equals some given There are many other possible situations
amount. The demand is also specified at this for which solutions are not possible. For instance,
node. The command DHEAD enters both types the pressure at the node downstream from a single
# 1 and type #2 differential head devices. The supply source for a network cannot be specified
device is type # 1 if input items 3 and 4 below are unless the differential head is inserted in that
given zero values. For a type # 2 device an pipe, since this is already uniquely determined by
additional energy equation will be written between the reservoirs water surface elevation, the total
the node of the specified HGL elevation (or demand which must come through this pipe, and
pressure) and one of the sources of supply. You the pipe headloss properties. Likewise, it is not
must give both the source of supply as well as the possible to specify pressures at two nodes from
HGL (or pressure) at the node where this is to be this node, if no other pipes join at this node,
specified. That is, a differential head device of without these pipes containing a differential head
lype # 2 allows both the head and demand at a device. USU-NETWK does not attempt to
designated node to be specified. In order to determine all the situations that you might specify
satisfy this dual condition, an additional unknown for which solutions are not possible. This burden
is left to you. However, if convergence to a

-50-
= solution does not occur it is likely that you have meters, depending upon whether ES or SI units
an impossible situation specified. The user is, are used, respectively, at the node in item # 3. If
therefore, wise to Hmit the number of iterations item # 3 is negative this value will be taken as the
allowed by using ITER or MAX as options. pressure (in psi or Pascals) and USU-NETWK
Before using a large number of differential head converts the given value to an HGL elevation by
devices in a single solution the user should have dividing the pressure (appropriately converted to
some knowledge about the network's performance. psf) by the specific weight of the fluid to get
Input data consist of: pressure head and adds this amount to the
1. The pipe number containing the elevation of the node. If type # 1 device is
differential head device, intended, this item is given a zero, or truncated by
2. The amount of this differential head. a I.
In the case of type # 1 devices, this head (in feet
or meters depending upon whether ES or SI units Example of input under the DHEAD command
are used, respectively) will be added to the DHEAD
frictional head loss in the pipe. If type # 2 5 -10.5 01
devices are being introduced this is only an 10 5. 4 11 165.
estimate of the differential head loss. A head
loss is negative and an increase in head, or a END
positive differential head, increases the HGL
elevation from the upstream node of the pipe The command END denotes the end of
toward the downstream node of the pipe. the detailed data, and upon encountering END,
Obviously, upstream and downstream nodes for USU-NETWK stops looking for additional
pipes with DHEAD devices in them are commands. Depending upon options that may be
important, as are specified flow directions for set there can be data after the END.
pipes containing booster pumps, PRY's, check If NPERCf= 1 in the $SPECIF list, then
valves, and back pressure valves. the fraction of the total demand coming from each
3. The node number where the HGL source of supply is required as further data after
elevation (or pressure) is specified for a type # 2 END.
device. If this value is preceded by a minus sign Also if end pipes for pseudo loops are
(I.e. is negative), then USU-NETWK assumes that specified as denoted by setting LOOPSE= 1 in the
pressure is being given in psi for ES units, and in $SPECIF list, then this data also follows the END.
Pascals (N/m2) for SI units, rather than as a HGL If all external flow are specified to a network, i.e.
elevation. A zero for this third item indicates to no reservoirs, or source pumps exit then it is also
USU-NETWK that a type # 1 device is intended necessary to supply a node number and an HGL
and, therefore, it does not set up the additional elevation as a starting value for the HGL
equation needed to solve for the differential head elevations and pressures of nodes to be computed.
for a type # 2 device. These two values also follow the end for such
4. The designation of a source of supply special networks.
for type # 2 devices (i.e. a source pump, or
reservoir) that USU-NETWK should use to form ERROR
an energy loop to, from the node given in item #
3. If the option NODESP= 1 this designation is This command introduces a single value
a node number assigned to the source. Otherwise after it, which is the error parameter that
it is the pipe number that connects the source to determines when the Newton iterative solution
the network. In order to save computations this process will be terminated. When the absolute
source should be in the proximity of the node of sum of the changes in flowrates is less than this
item # 3, since it is used to obtain an additional amount, the iteration stops. Use of the command
energy equation to allow for a differential head in ERROR is an alternative method to setting
the pipe of item # 1 to be solved for. It should ERROR= to this value in the $SPECIF list of
also not be a source that has little influence on options.
the flow in the pipe of item # 1 or the pressure
at the node of item # 3. If this is to be a type # Example of the use of ERROR
1 differential head device, the value should be ERROR
zero. .0005
5. The elevation of the HGL in feet, or

-51-
 FORMAt GAMMA
If the option NETPLT=13, or = 14 then Has same action as WEIGT. (see
USU-NETWK is told that the input data is WEIGT)
formatted. The default formatting for this input
data, that is expected ordinarily, is that written by
the preprocessor PIPEINPT. For input under the
command PIPES this expected format is as
follows:
(3I4,F8.2,FlO.2) if LENGON =0 ITERAtion
(3I4,F8.1,F8.2,FlO.2) if LENGON=1
and if NUMPIP= 1 then the PIPES format is: This command is an alternative to using
(2I4,F8.2,FlO.2) if LENGON =0 ITERA or MAX in the $SPECIF list of options,
(2I4,F8.1,F8.2,FlO.2) if LENGON=1 and as such it sets the maximum number of
The default formatting for input under other Newton iterations that will be allowed in obtaining
commands such as NODES, RESER, PUMPS, the solution. If not given in either place, the
etc. that have one integer followed by real default value of 8 or 15 will be assumed.
variables is:
(I5,7FlO.2) Example of the use of lTERA
The command FORMAT allows any other ITERA
format to be given prior to reading data that is in lO
columns specified by this given format. The given
format information must begin with a left LPIPE,NLPIPE,LNODE,NLNODE
parentheSis, (, and end with a right parenthesis, ),
and contain the information as it would in a These four commands are listed together
FORTRAN FORMAT statement. The length of because they are all designed to permit the user to
this format information is restricted to 48 not have all pipes or nodes of the network appear
characters or less. For example if the data is in the two tables form of output. Generally it
formatted that contains the items expected under will be desired to have all pipes and nodes listed
the PIPES command and is in the following in the final solution tables. However, for large
columns: (1) diameter 1-10, (2) length 11-20, (3) networks especially if the analyses are later ones
roughness coefficient 21-30, (4) pipe number 31- of a series, or steps in time dependent solutions it
35, (5) upstream node number 3640 , and (6) may be desirable to not have all pipes, or nodes
downstream node number 41-45, the following two appear in the solution tables. The commands
lines of input must precede the PIPES command LPIPE (standing for list pipe), and NLPIPE
for USU-NETWK to properly read this input: (standing for not list pipe) allow the user to either
give a list of pipe numbers that should be listed in
FORMA the final pipe data table from the two table form
(T31,3I5,Tl,3FlO.2) of output, or give a list of pipe numbers that
should not be contained in this final pipe data
The FORMAT command followed by the table, respectively. The pipe numbers can be
proper format information can precede any other listed as individual values separated by a blank or
command, and the given format information will comma, or a range of pipes can be given by
apply until it is changed by another FORMAT including a - between the beginning and ending
command. For the given format to be utilized it pipe numbers of the consecutive series, or any
is necessary that the option NETPLT also be set combination of these. Pipe numbers given in this
to 13 or 14, informing USU-NETWK that the list that do not actually exist in the network will
input is in given columns, and FORTRAN be ignored. Likewise the commands LNODE
FORMATTED input is to be utilized rather than (standing for list node) and NLNODE (standing
free format information. Otherwise the format for not list node) allow for specifying node
information will be ignored and the input will be numbers to include, or exclude from the node data
read as if it is free format. When using formatted table from the two table form of output,
input it is not possible to truncate items from !he respectively. If NODESP=I, such that the
ends of input lines, as can be done when usmg sources of supply are numbered as nodes, then
free format input, however.

-52-
lhese extra nodes will always appear in the node of the pipes here so that pressure and pressure
data table and cannot be deleted with the heads, etc. can be computed. These two items of
command NLNODE. Furthermore, when using information are not contained in the items after
the NODESP= 1 option it is necessary that these the PWES command, and therefore constitute
nodes occur in the input list of nodes after all additional information not logically entered under
real nodes. If it is desired that sources of supply the pipe oriented form of input. In addition if x
not be listed in the output table, then the option and y coordinates are to be given for plotting of
NODESP=O should be used. The program the network, or used to define the layout of the
anticipates that you will either use LPIPE or network from which pipe lengthS are computed
NLPIPE, but not both. Attempts to use both they are also items that are supplied to USU-
commands will result in unpredictable results. NETWK with the NODES command.
Likewise either LNODE or NLNODE should be Furthermore, with the option DESIGN=1 or
used but not both. DESIGN=3 HGL elevations are given, and these
become an item entered with the NODES
Example of using LPIPE & LNODE command. Also "data to check the network"
LPIPE might be given with the option NETCHK=2, and
1 3-7 30-50 140/ these come after the NODES command.
LNODE Therefore, the list of items under the NODES
4-7 24 40 50-70/ command will be expanded upon depending upon
the options in the $SPECIF list. The first three
MINOR items listed below are always entered with the
NODES command, and the others dependent
The command MINOR provides the loss upon whether an option call of it (them) or not.
coefficient for devices, such as possibly globe
valves, that produce extra losses in a pipe in 1. The node number (an integer unless
addition to fluid frictional losses. Devices that OUTPUl=l1 or 13, in which case an 8 character
cause small losses such as large radius bends, gate identifier replaces the number) for which this data
values, etc. are handled as part of the frictional applies. The node numbering does not need to
loss in a pipe, generally. In fact, the common contain all numbers from 1 to the lotal number of
practice is to use "extra" losses such as those nodes in the network, or be7. Re in consecutive
introduced by MINOR sparingly in defining a order. Depending upon the dimensions of the
network. Often the magnitude of loss coefficients program these integer node identifiers must be
are "guessed at", and consequently an equally less than some limit such as 5000, 10000 or 15000.
satisfactory procedure is to accommodate extra 2. The external flow, or demand, at this
"minor losses" as part of the pipe wall roughness node. If inflow occurs at this node, the demand is
coefficient. The following two items occur on a negative value, and outflows are positive. The
each line after the command MINOR: units of this demand must be those designated
1. The pipe number which contains the with the option NFLOW. If the option
minor loss device, and NODESP= 1 in the $SPECIF list, such that source
2. The coefficient K in the minor loss pumps and reservoirs are identified as nodes in
equation, the network the demands given for these
2 "artificial" nodes must be zero. In olher words
hL = K (V /2g).
you do not provide an estimate of the flowrate
Example of the command MINOR from supply pumps, or reservoir, or into reservoirs
MINOR as a demand. Flowrates from sources of supply
22.5 are part of the solution. Therefore, a demand is
18.5 considered an external flow from (to) the network.
(If you want to supply estimates of flowrates from
NODES sources see the option LOOPSE.)
3. The elevation of the node with units of
The command NODES supplies feet or meters, depending upon whether ES or SI
information about the demands (consumptions or units are used, respectively. If NODESP= 1, then
if negative inflows) that are assumed to take place the elevation given for sources should be the
at the junctions of the network, and the elevations ground elevation. The water surface elevations for
reservoirs, and the sump elevations for source

-53-
= pumps are given under the commands RESER Example of data after NODES with the option
and PUMPS. The default of NODESP=O does NETCHK=2, but DESIGN=O
not provide an entry in the NODES DATA table NODES
when giving the solution for sources of supply 1 1.2 500. -6 2 4 7/
since they are not numbered as nodes, and no 2.8500 -28 10/
ground elevation is needed then. 5 1.1 52036/
8 O. 520 7 6 -31
Example of node data without options set
that require more data after the NODES Example of data after NODES with the option
command LENGON =0 (The last two items are the x & y
NODES coordinates)
1 1.2 500. NODES
2.8/ 1 1.2 500. O. 1000.
5 1.1 520. 2 .8 500 1200. 500.
80./ 5 1.1 520 1000. -400.
8 O. 520 1800. -300.

4. The next possible item after the Examples of data after NODES with the option
NODES command is the elevation of the HGL LENGON=O & DESIGN=l
elevation ( or pressure head, or pressure NODES
depending upon IHGL), if and only if the option 1 1.2 500 620 0 1000
DESIGN=l or DESIGN=3 in the $SPECIF list. 2 8 500 618 1200 500
If DESIGN=O, but other options call for more
information under the NODES command, the
items listed below move into this fourth position.
NOZZLes
5. If either the option LENGON =0, or 0
< NETPLT < 13 , which indicate that pairs of x The command NOZZL introduces data
and y coordinates are going to be given, then the describing a sprinkler nozzle, or orifice at the end
x coordinates in feet, or meter, respectively of a pipe. Besides the more obvious uses of this
depending upon whether ES or SI units are used, command, it might be used to determine how
is given as extra data after the NODES command. much flow the network might be able to provide
6. The y coordinate of the node at a fire hydrant if it were fully opened. One
corresponding with the x coordinate given as item would need to know the hydrants opening size,
# 5. and its discharge coefficient. Data after the
7. If the option NETCHK is given in the NOZZL command consists of:
$SPECIF list then the pipe numbers that join at 1. The pipe number (integer) at whose
this node are includes as this item of data. end the nozzle exits. Nozzles exit at the ends of
pipes just like sources of supply do. One might
8. Remarks may occur at the end of any view a nozzle as a negative source, but with
line of input if /R precedes this remark. The R different characteristics. Therefore, if
after the / tells USU-NETWK that you want the NODESP=O the ends of pipes where nozzles exist
remark recorded. Otherwise any comments you are not given a node number and after the PIPES
place on lines of the input file after a / will be command this end would be denoted by a O. If
ignored. Remarks are restricted to 32 (or possibly NODESP=l, then the nozzle ends of pipes must
36) characters. Similar remarks are permitted be given a node just as the sources of supply are
after the PIPES and the PIPE- commands. This given a node. Under this option this item is the
input is listed as item 8. However, regardless of node number given to the nozzle end of the pipe.
the options it is always the last item if given. For 2. The diameter of the nozzle, or orifice
example if a line under NODES were as follows: in the same units that are used for pipe diameters.
8240. 220./R NODE AT INTERSECTION #10 This diameter is assumed to be short in length, i.e.
then in the NODE DATA table node 8 will have no frictional Joss is computed for a pipe of some
NODE AT INTERSECTION #10 at the end of length with this diameter.
its line. 3. The discharge coefficient of the nozzle
or the orifice. For an orifice this will be

-54-
approximately 0.6. line is the peaking factor and the second is the list
4. The elevation of the nozzle or orifice. of node numbers for which this peaking factor
(Given only if NODESP=O) applies. The command PEAKF thus accomplishes
5. If desired, an estimate of the HGL the same as the option PEAKF (FLOWFC) only
upstream from the nozzle may be given. If not if all nodes are listed such as 1-320, but allows
given the starting HGL elevation is generated greater flexibility in that different groups of nodes
internally in USU-NETWK. If this item is not can have the given demands multiplied by
given, then # 4 is terminated by a I. Likewise, if individual factors. Any number of groups of two
other entries are as previously given they may be lines can occur after the command. The first of
omitted by ending with a /. each pair consists of a single real value, the
multiplier or peaking factor. The second of the
Example of nozzle data with NODESP=O pair, containing the node numbers, may consist of
NOZZLES single integer values separated by blank(s) or a
82..58 100./ comma, or by a - (minor sign) between two
12 1.5 .62 80. 120. integers to denote the range from the first through
the last node. If the - is used no blank spaces can
If NODESP=1 (Le. Nozzles numbered as nodes) occur between the two integers. The second line
If node numbers are used to identify must be terminated with a I.
nozzles, then the input item no. 4 above will be The command PEAKF must be entered
given on the NODES line for this node, and after the NODES command (or PIPE-command)
consequently item # 5 above replaces item #4, used to enter the nodes for which it gives peaking
but might be deleted by ending item #3 with a I. factors. The command PEAKF can be used in
connection with PEAKF in the specification list
Example of nozzle data with NODESP= 1 $SPECIF if desired. When PEAKF appears in
NOZZLES both places all given demands are multiplied by
152./ the value given to PEAKF in the $SPECIF list
182.3 .56/ first. Thereafter, the new resulting demand is
again multiplied by the value entered for the
PARALlel designated nodes through the command PEAKF.
The same node number may appear under more
It is possible to specify that any number than 1 group of two lines under the PEAKF with
of pumps of the same curve characteristics as the effect of having the multiplication
given by the PUMPS of BOOST command are accumulative.
operating in parallel (or in series) at a station.
Two items of data are needed to specify pumps in Example of using the command PEAKF
parallel, and are: PEAKF
1. The designation of the pumping 1.5
station according to whether NPSERI=O, 1 or 2 1 3 10-16 20 31-451
by pump number, pipe number, or node number .8
(if a source pump station and NODESP=l). 2221
2. The number of pumps operating in
parallel at this station. PIPES
Example of the PARALLEL command The command PIPES enters data for each
PARAL pipe in the network as well as its nodes, length,
32 diameter and roughness coefficient and this data
41 describes the systems connectivity and physical
characteristics. The items under the pipes
PEAKF command consist of:
1. The number (integer) associated with
This command does not replace the this pipe. If the option NUMPIP=l is included in
option PEAKF, but could be used to do so. After the $SPECIF list, then this first item is omitted,
the command PEAKF there are two lines of data and the pipe numbers associated with pipes in the
that constitute the needed information. The first output data will be according to the sequence
number of the pipe in the input data. Pipes do

-55-
 not need to be numbered sequentially, nor do all 6. The wall roughness coefficient for this
numbers from 1 to the total number of pipes in pipe. Depending upon whether you wish the
the network need to exist in the input data file. flowrate, headloss relationship to be bases on (a)
Duplicate pipe numbers will be identified by a the Darcy-Weisbach, (b) the Hazen-William, or (c)
message to this effect, but processing of the the Manning's equation, this coefficient will be:
solution will continue. The maximum number (a) the equivalent sand roughness in the units
that can be assigned to a pipe will vary depending designated by NUNIT, (b) the Hazen-William's C,
upon the dimensions allowed by the version of or (c) Manning's n for the given pipe. If you
USU-NETWK that you are using, but will be don't use NEQUAT in the $SPECIF list, then
5000, or larger. USU-NETWK will decide whether the coefficient
If the option OUTPUI > 10, then this is for use in the Darcy-Weisbach equation of the
first item under the PIPES command must be an Hazen-Williams equation according to the
8 character string (which may include blanks). See magnitude of the first coefficient encountered in
the description under OUTPUI in the section of the input.
this manual that describes options for more detail 7. Remarks. Generally remarks are not
about this string of 8 characters as a pipe given associated with a pipe, but if you wish to
identifier. have a remark in the solution table for the PIPE
2. The number (integer) of the node at DATA you must follow the last given item with a
the upstream end of this pipe. If OUTPUl=l1 or /R. The R after the / tells USU-NETWK that you
13, then this will be an 8 character string. For want the remark recorded. Otherwise any
most pipes the user may not know the direction of comments you place on lines of the input file after
flow, and for these pipes upstream and a / will be ignored. Remarks are restricted to 32
downstream have no significance. USU-NETWK (or possibly 36) characters, and similar remarks
will determine which is the correct upstream and are permitted after the NODES and the PIPE-
downstream nodes as the solution is obtained. commands. For example if a line under pipes
However, for pipes that contain booster pumps, were as follows:
valves, etc. upstream and downstream nodes must 1247 2400. 20./R THIS PIPE NEEDS TO BE
be given correctly. REPLACED
In the two table form of output an then in the PIPE DATA table pipe 12 will have
asterisk (*) precedes the numeric data for the pipe THIS PIPE NEEDS TO BE REPLACED at the
if you select the incorrect direction for flow, as a end of its line. (Only the simulation version of
help should you want to know if you have the USU-NETWK allows for remarks.)
direction of flow reversed in some pipes.
3. The number (integer) of the node at Example of pipe data with LENGON=1 (the
the downstream end of this pipe. The same default)
applies as stated above for item # 2. PIPES
4. The length of this pipe in feet, or 3 8 2 1500. 6..008
meters depending upon whether ES or SI units 7 1 42000. 8/
are used, respectively. Two options have an effect 924 1000/
on this length. If NPLENG= 1, then these lengths 10 3 8/
must be given in 1,000 feet or 1,000 meters. If the 12 4 7 2500 10/
option LENGON =0, then this item is omitted
from this list, and the pipe length will be Example of pipe data with LENGON =0, and
computed from the x and y coordinates of the NUMPIP=1
nodes which are required as other input under the PIPES
NODES command. 826..008
5. The diameter of this pipe in the units 148/
designated by the value given to the option 2 4/
NUNIT. (If the option DESIGN=4 (or =5) is 38/
used then for any pipe that is not part of a loop 4710/
this item will be the velocity specified for this
pipe. The pipe diamete for this non-looped pipe PIPE-nodes
will be solved for to provide teh specified velocity.
See the option DESIGN =4 for further The command PIPE- introduces both the
explanation.) pipe and node data on the same line, one line for

-56-
 each pipe. Therefore, it can replace both the for PIPE DATA at the end of the line for the
PIPES and the NODES command in defining the given pipe. A line cannot exceed 80 characters in
physical characteristics of a network. The length. By examining items with decimal points
command PIPE- may be used exclusively for and those without decimal points, USU-NE1WK
supplying all the information for pipes and nodes permits many of the items listed above to be
or it may be used to provide only a portion of this omitted for any line. The items which may be
data, in which case the omitted, or must be omitted are as follows:
commands PIPES and NODES would be needed Item # 1, the pipe number, NOP(I), must
to supply the remaining data. be omitted if and only if NUMPIP= 1 in the
The general form of data entered by the $SPECIF list of options.
PIPE- command is given by the numbered items 1 Item # 4, the roughness coefficient, E(I),
through 10 below. However, depending upon may be omitted on any line. Whenever E(I) is
options in the $SPECIF list, this list of items is not given on any line the standard value given by
added to or deleted from. Many of the items the option COEFRO will be assigned as the
listed may be omitted as descibed later. With all roughness coefficient for this pipe. The default
default options the list of item entered by the value for COEFRO is 0.0102 inches for a cast
PIPE- command are (FORTRAN variable names iron pipe in ES units.
are given for each item that are referred to in a Item # 5, the upstream node number,
table that follows): Lll, if a source of supply (or sink) such as a
1. The pipe number, NOP(I) reservoir, souree pump, or nozzle exists at an end
2. The pipe diamter, D(I) (If the option of this pipe, and NODESP=O (the default) in the
DESIGN =4, then instead of pipe $SPECIF list. If desired a 0 (zero) may be
diameter this will be the specified entered for this now existent node.
velocity for the non-looped pipe as Item # 6, the demand, QJ(Lll) may be
described under the option DESIGN.) omitted on any line. If QJ(Lll) is zero this
3. The pipe length, L(I) demand does not need to be given on any line. If
4. The wall roughness coefficient, E(I). QJ(Lll) is not zero, it must be given at least
This value will be the equivalent sand roughness, once, but may be duplicated on other lines if
c, for the Darcy-Weisbach equation, C for the desired. If different values of QJ(Lll) are given
Hazcn-Wiliams equation, and n for Manning's on different lines, the last value given will be
equation. taken as this demand.
5. The upstream node number of the Item # 7, the nodal elevation,
pipe, Lll ELEV(Lll), may be omitted on any line if ilS
6. The demand, or external flow, at the value is given on another line. IF NODESP=O
upstream node, QJ(Lll) ELEV(Lll) must be omitted as must QJ(Lll) if
7. The elevation of the upstream node, this upstream node number Lli is omitted.
ELEV(LlI) Item # 9, the demand at the downstream
8. The downstream node number of the node, QJ(L22), may be omitted on any line. If it
pipe, L22 is zero this demand does not need to be given on
9. The demand, or external flow, at the any line. If QJ(L22) is not zero, it must be given
downstream node, QJ (L22) on at least one line, but may be duplicated on
10. The elevation of the downstream other lines if desired. If different values are given
node, ELEV(L22). for QJ(L22), the last value will be taken.
In providing these data decimal points Item #10, the elevation at the
must be included with real variables, and cannot downstream node, ELEV (L22), may be omitted on
be included with integers. Items 1, 5 and 8 are any line. The same rules apply as for ELEV(Lll).
inlegers. The rest are reals. Therefore, decimal One important rule that must be kept in mind for
points must be given with items 2, 3, 4, 6, 7, 9 and QJ(Lll), ELEV(Lll), QJ(L22), and ELEV(L22)
10. is that if only one decimal point number follows
As is the case with other data, this data is a node number (LlI or L22) , then this will be
free format; that is, the data items are separated assumed to be the elevation and not the demand.
by one or more blanks, or a comma which may be The reason for this rule is that many demands
followed by blanks. If a line is followed by a !R may be zero, and therefore need not be included
any remarks following this up to 32 characters will as zeros, but may be included as a zero if desired.
be keep track of, and written in the solution table A summary of the items which may occur

-57-
~

Table of possible entries and the order of these items (with CORPIN =0) after the PIPE- command. (Note
that options LENGON=O dictates that pipe lengths L(I) be omitted; NUMPIP=1 dictates that NOP(I) be
omitted; DESIGN=1 or =3 dictates that an HGL elevation, pressure head, or pressure be given, and this
will value will follow the elevation ELEV(Lll) or ELEV(L22); and if 0<NETPLT<9 or LENGON =1 then
x and y coordinates are needed, and these follow the HGL elevation of elevation of the node. Furthermore,
the roughness coefficient E(I) may be omitted from any line.

Q
Item of input
Case no. #1 #2 #3 #4 #5 #6 #7 #8 #9 #10

1 NOP(I) 0(1) L(I) E(I) Lll QJ(Lll) ELEV(Lll) L22 QJ(L22) ELEV(L22)
2 NOP(I) 0(1) L(I) E(I)Lll ELEV(Lll) L22 ELEV(L22)
3 NOP(I) D(I) L(I) E(I) LII L22 QJ(L22) ELEV(L22)
4 NOP(I) 0(1) L(I) E(I) Lll QJ(Lll) ELEV(Lll) L22 ELEV(L22)
5 NOP(I) 0(1) L(I) E(I) Lll ELEV(Lll) L22 QJ(L22) ELEV(L22)
6 NOP(I) 0(1) L(I) E(I) Lll L22 QJ(L22) ELEV(L22)
7 NOP(I) 0(1) L(I) E(I) Lll L22 ELEV(L22)
8 NOP(I) 0(1) L(I) E(I) Lll ELEV(Lll) L22
9 NOP(I) 0(1) L(I) E(I) Lll L22
10 NOP(I) 0(1) L(I) E(I) L22
11 NOP(I) 0(1) L(I) E(I) L22 QJ(L22) ELEV(L22)
12 NOP(I) 0(1) L(I) E(I) L22 ~

a NOP(I) _ pipe number; 0(1) - pipe diameter; L(I) - pipe length; E(I) - roughness coefficient; Lll -
upstream node no.; L22 - downstream node no; QJ(Lll) - demand at upstream node; QJ(L22) - demand at
downstream node; ELEV(Lll) - elevation at upstream node; ELEV(L22) - elevation at downstream
node.

on any line after the PIPE- command is given in given a value greater than zero, but LENGON=1,
lhe table above. In this table items are lined up then L(I) remains but the x and y coordinates
in columns, but in the input data need be follow either ELEV(Lll),or ELEV(L22), or both.
separated only by a single blank, or comma. When giving x and y coordinates the restriction is
Since E(I) may be left out of any line this that both the demand and elevation must be given.
table actually represents 24 different cases that are If DESIGN=1 or DESIGN=3, then the elevation
allowed under the PIPE- command. Furthermore, of the HGL, the pressure head, of the pressure
with the default of NODESP=O (source of supply (depending upon the option IHGL) must follow
nol numbered as nodes) it is possible to enter a the nodal elevation ELEV(Lll), ELEV(L22) or
o for LI1 to denote this non existent node if this both. If NETPLT > 0 or LENGON =0 along
is preferable or leaving it out. It is important to with DESIGN> 0, then after the HGL elevation
nOle that the 12 cases given in this table apply the x and y coordinates are given, so that 5 values
only under the default options, and that several containing decimal points may follow either the
options do change items in the list. With upstream node, L11 or the downstream node, L22.
LENGON=O item # 3, L(I), must be omitted, The restriction for this latter situation is that x
and x and y coordinates may be given after the and y coordinates cannot be given unless the
elevation for the upstream, or for the downstream, nodal pressure is given first, and both the nodal
or both nodes (not permitted with the PC steady- demand and elevation must also be given.
state version of USU-NETWK). These elevations
must contain a decimal point. If NETPLT is
Example of use of the PIPE- command under under it by lines under the RESER and BOOST
default options commands, since a source pump consists of a
1 12. 5000..01 3 1.2 148. 4 3.0 150. pump in a pipe line, like a booster pump, plus a
2 8. 1000. 1 5 1.5 165. reservoir with a known water surface elevation.
5 14.6000. 7 1.6 180. The data after the PUMPS command
mathematically describes the operation of a pump,
Example of use of the PIPE- command with i.e. provides information so that the head
NUMPIP=l produced by the pump can be defined as a
12. 5000..01 3 1.2 148. 4 3.0 150. function of the flowrate in the pipe that contains
8. 1000. 1 5 1.5 165. the pump. There are two means for doing this.
14. 6000. 7 1.6 180. The best is to provide pairs of flowrate Q and
corresponding head H produced by the pump to
Example of use of the PIPE command with define this relationship. The second method is to
DESIGN=l provide the power the pump supplies the fluid (its
1 12. 5000..01 3 1.2 148. 220.43.0 150. 248. power rating multiplied by its efficiency) and its
2 8. 1000. 1 5 1.5 165. 255. normal capacity. These are described below
5 14. 6000. 7 1.6 180. 275. starting with item 2.
Data after the PUMPS command consists
Example of use of the PIPE command with of:
LENGON=O 1. The pump station identifier. Under
8 24. 140. 1 2. 1000. O. O. 2 1.5 1040. 1500. O. the default options this identifier is the pipe
9 18. 135. 2 3 1.3 1010. 1500. 2000. number that connect the source pump to the
10 12. 130. 4 1.2 1040. O. 2000. 3 network, i.e. the pipe that contains the pump. If
NODESP= 1 then this identifier is the node
number assigned to the source pump, instead of
PRINT the pipe number. If NUMPIP=O and
NODESP=O then the pipe number given to
The command PRINT allows an identify this pumping station is the sequencia I
alternative means for controlling the amount of number of this pipe after the PIPES or PIPE-
output just as the option NPRINT (or IOEX) in command.
the options list does. The command PRINT is If the option OUTPU1= 11 or 12 is in the
followed by a single integer value from -3 to 10. $SPECIF list, such that pipes (and/or nodes) are
The command PRINT changes the print option identified by 8 character strings, then the identifier
after it is entered, and may appear as many times given for this item is that identifier i.e., the 8
as wanted. Thus if one desires that only the character string for this pipe (or node). Not only
NODES information that is given in the input are the characters used important but also their
data file be written to the output file PRINT position in the field of 8 character width, i.e. a
followed by 4 may preceded the NODES data, and blank in the pipe identifier before PIPEN02 is
PRINT followed by 0 may follow the NODES not the same as PIPEN02 with a blank after it.
data. If one desired all of the input data file to be See the description under the option OUTPU1 for
written to the output file the option NPRINT= 10 more explanation about use of character strings
could be in the $SPECIF list, and then just before for identifiers.
the RUN command the command PRINT could
be followed by -3.

Example of use of the command PRINT

PRINT
5

PUMPS
The command PUMPS enters data for
source pumps. An alternative to the use of the
command PUMPS would be to replace each line

-59-
First method (pairs of 0 versus H) Second method (power & normal
capacity)
2. The flowrate, Ql' at the first point on
the pump's characteristic curve in the units This alternate is used if a pump curve is
designated by the option NPGPM. not available but the pumps power rating is
3. The head, HI' produced by the pump known, and its normal capacity is known (or
corresponding to Ql' assumed). When using this alternative USU-
4. The flowrate, Q2' at the second point NETWK generates a 2nd degree polynomial
of the pump's characteristic curve. This flowrate equation that defines the head produced by the
is assumed to be the normal capacity (i.e. point of pump as a function of its flowrate typical of most
greatest efficiency) for the pump, provided that common mixed flow pumps. However, each pump
on Iy 3 points are given to define the pumps has it unique characteristics of operation, and
operating curve. consequently the results may differ from your
5. The head, H2 , produced by the pump pump when you use this option, and the solution
corresponding to Q2' shows a flowrate through this line quite different
6. The flowrate, Q3' at the third point of from the normal capacity of the pump.
the pump's characteristic curve. 2. The power supplied the fluid by the
7. The head, H3 , produced by the pump pump in horsepower, or kilowatts, depending upon
corre.<;ponding to Q3' whether ES or SI units are used, respectively.
(All heads are in feet when using ES units and in 3. The normal capacity of the pump in
meters when using SI units.) the units designated by the option NPGPM.
4. The elevation of the source of supply
This list of pairs of Q versus H can be as described under 8 above. This fourth item
expanded to 10 pairs if the option PCHAR3=0 in must be terminated with a I so that USU-NETWK
the $SPECIF list of options. Under this option can determine that you are using this alternative
the next item, i.e. the water surface elevation, method of input.
must be terminated with a I, and the normal
capacity is taken at the flowrate nearest the Example of data after PUMPS (using the power
middle of the list of pairs. Furthermore under the and normal capacity)
option PCHAR3 it is necessary that the flowrates PUMPS
be listed in ascending (increasing) order of 1 12. 1.5 32001
magnitude. 15 15. 1. 3100./
8. The elevation of the well, or reservoir
water surface from which the pump receives its Third method (giving coefficient of 2nd degree
supply in feet or meter, depending upon whether polynomial
ES or SI units are used, respectively. This supply
surface is often call the sump water surface If the pump station identifier is negative,
elevation. If the option NPRPUM=I, this is the then USU-NETWK will read in the constants a, b
pressure at the pump's inlet instead of the and c in the second degree polynomial equation
elevation of the water surface. given below to define the pump curve.
See Appendix F, that describes a utility
program PMPCUR, to help you understand and Hp = aQ2 + bQ + c
visualize how a given set of imput data actually
defines a pump's operating characteristics, and in which H is the head produces by the pump
how you might provide data to more nearly and Q is thl flowrate. Both head and flowrate are
duplicate the opertion of a pump. in basis units. The preprocessor PIPEINPT will
Example of data after PUMPS (first method) obtaine, and write the pump curve giving these
PUMPS coefficients rather than 3 points. When this is
5 1 300 1.2 290 1.4 270 3000 done then the preprocessor does the work of
9 2 350 1.8 355 1.6 357 2800 filling the second degree polynomial to the 3 pairs
of data rather than having USU-NETWK do these
computations. Upon using the utility program

-60-
PMPCUR described in Appendex F of this manual this pipe specified. Therefore, by providing
you may wish to also provide these values. If so negative pipe numbers as input under QGlVE the
the input data consists of: user communicates to USU-NETWK that he is
1. The pump identifier (pipe no. if not certain whether the flowrate in a pipe is
NODESP=O, or node no. if NODESP=l actually specified. It will only be specified if it is
preceeded by a negative sign. possible to control the flowrate by a valve.
2. The coefficient a in the above Should the flowrate in this pipe be less than that
equation. desired, then the actual achievable flowrate is
3. The coefficient b in the above accepted.
equation. The solution for problems that use the
4. The coefficient c in the above QGIVE command to specify flowrates in interior
equation. pipes will report, as extra data, the heads (positive
5. The elevation of the fluid source for or negative) required in these pipes for the
the pump. selected flowrates to be possible. Data after the
USU-NETWK will print the coefficients a, band command QGIVE include:
c under the default level of NPRINT for each 1. The pipe number that contains the
pump. specified flowrate.
2. The specified flowrate. This flowrate
QGIVE will be assumed from the upstream node to the
downstream node. If a reversed direction is
This command permits flowrates in desired this flowrate can be negative.
selected pipes to be specified. It is sometime Important note: If remarks are provided
desirable to be able to specify the flowrate that after data following the PIPES or PIPE-
must exist in a pipe. While QGIVE permits such commands, then it is necessary that the pipes with
Oowrates to be specified, there are situations for specified flowrates must occur in the input data
which QGIVE is not applicable, and therefore the file after the pipe with the last remark associated
following explanation if provided. QGIVE cannot with it. The reason for this restriction is that
be used to specify the flow in a dead end pipe, USU-NETWK does not consider pipes with
because the demand at its end, that is not specified flowrates the same as regular pipes and
connected to another pipe, already specify what they are removed in its counting of pipes.
the flowrate in it must be. If a specified flowrate
is to come from a source of supply it should be Example of the QGIVE command
handled as a dead end pipe with a negative QGIVE
demand. If the pipe is "interior" so that loops of 1 1.1
the network pass through it, then its flowrate can 100 1.2
be specified through the command QGIVE. A
special application for which QGIVE might be
used is in connection with a valve in a pipe that
controls the flowrate by being partly closed. For RESERvoirs
such applications there is a maximum flowrate
that is possible when the valve is fully opened. In The command RESER is used to enter
other words no positive head (or head increase in data that define the water surface elevations in
the direction of flow) is permitted to achieved the reservoirs or tanks. As such a reservoir serves two
specified flowrate; rather only negative heads, or functions from the viewpoint of hydraulics of the
headlosses in addition to the fluid frictional losses network. First, it provides a source of supply
arc allowed. This application can be (which may be out from the network, as well, but
communicated to USU-NETWK by placing a whose magnitude is unknown), and second, it fixes
minus sign immediately before the pipe number the elevation or the HGL at this point. For
described below. If a negative pipe number is steady-state analyses, this elevation is constant.
given and the solution for this pipe indicates a For time-dependent solutions the initial reservoir
positive head is required, then USU-NETWK: (1) water surrace elevation is specified in the
will reject this as the desired solution to the definition of the network, and a stage-volume or
network, (2) remove this pipe as a pipe whose storage relationship (given under the time-
flowrate is specified, and (3) recompute the dependent data) controls how this elevation may
solution to this new network without the flow in change with time. Only one pipe may connect a

-61-
 reservoir to a network, and this pipe cannot and corresponding heads that you read from the
= contain devices such as a pressure reduction valve, pump curve before entering the data for USU-
for example. If it did then the PRY would NETWK to use under the PUMPS or the BOOST
effectively specify the head at this position. Data command. Alternatively you can give the data as
after the RESER command consist of the read from the pump curve, and then use the
following two items: ROTAT command and have USU-NETWK scale
1. The identifier for the reservoir (an up the flow rates and heads.
integer unless pipes or nodes are given as 8 Dimensional analysis can be used to
character strings). If the option NODESP=O (the develop the following two similarity relationships
default), then this identifier is the pipe number for the operation of homogenous pumps:
that connects this reservoir to the network. If
3
NODESP= 1 in the $SPECIF list, then this Flow rate - {Q/ND3 }1 = {Q/ND }2
identifier is the node number assigned to this
source of supply. Pump head - {H/N2D2} 1 = {H/N2D2} 2
2. The elevation of the water surface in
the reservoir, or tank in feet or meters, depending in which Q is the flow rate through the pump, N
upon whether ES or SI units are used, is its rotational speed, D is its impeller diameter,
respectively. If the option NPRRES = 1 in the and H is the head produced by the pump. The
$SPECIF list, then the pressure is given instead of subscripts 1 and 2 represent two different
the water surface elevation operating conditions, or two homogenous pumps.
The assumption USU-NETWK makes in The applications of the above two similarity laws
initializing the problem for the Newton iterative in USU-NETWK is_ limited to changing the
solution is that the flow is from reservoirs to the rotational speed N. The first similarity law above
network unless you indicate otherwise. Should indicates that the flow rate Q varies proportional
you assume the flow is from the network into the to the rotational speed N, and the head produced
reservoir, i.e. it is filling, there are two method by the pump H varies as the square of the
you can use to tell USU-NETWK this. The first rotational speed.
is to give a negative pipe number for # 1 above The input under the command ROTAT
(negative node numbers are not permitted). A consists of pairs of numbers. The first of each
second better method is to give this as the pair consists of the pump designation. and the
downstream node in the data entered with the second the ratio of the desired rotational speed of
PIPES or PIPE- command. Denoting flows into the pump to the rotational speed for which the
reservoirs, where this is most likely the case, will input data under the PUMPS or BOOST
result in USU-NETWK providing a better flow command apply. The designation of the pump is:
initialization, and generally result in fewer (a) according to pump number, i.e. its order in the
iterations for a solution, therefore. input data, (b) the pipe number that connect the
source pump to the network, or that contains the
booster pump, or (c) the node number for source
ROTAT pumps if NODESP= 1 in the $SPECIF list
according to whether NPSERI=O, = 1, or =2,
The command ROTAT stands for respectively. Thus pumps under the ROTAT
rotational speed of pumps. This command can be command are identified the same as they are
used in entering original data for a network. The under the SERIES or the PARALlel commands.
same command, under CHANGE data, and similar This list of pairs of values must be terminated
commands associated with time dependent data with a / and generally will only be one line long,
allow for different rotational speeds for pumps to but may extend unto other lines if desired. If you
be specified. Use of this command with the prefer each pair of values may be on a separate
original input data allows the information line. For the example above for which pump
associated with pumps under the PUMPS or the data is given for 1200 rpm, but the operation of
BOOST command to be for a different rotational the pump is at 1500 rpm, and if the pump
speed than the pumps are operating under for the designator is 1, the input line after the ROTAT
requested solution. For example the data you command would be 1 1.25/
have from the pump curve may be for 1200 rpm,
but the pump is actually operating at 1500 rpm. Example
One method would be to scale up the flow rates, ROTAT

-62-
2 1.4 5.9 7 1.3/ These data are described in the next section of
this manual under the heading "Change Data."

SERIES
RUN
The command SERIES is used to specify
The command RUN tell USU-NETWK that any number of pumps are in series at a given
to: 1. terminate reading data that defines the pumping station. Two items of data are need in
network, and 2. to obtain a solution. The END each line after this command. They are:
command caused the same. For any problem it is 1. The designation of the pumping
possible to have the following data appear after station according to whether NPSERI=O,l, or 2 by
the RUN or END commands, depending upon pump number, pipe number or node number (only
options set, and the nature of the network: if NODESP=l, also). If a booster pumping
1. Node oriented input used to station has more than one pump in serier (or
duplication the definition of the network's layout parallel), it cannot be designated by node number.
to assure correctness of the input data (see the If NPSERI=2, booster pumps are still designated
option NETCHK=-1 or =-2). by pipes.
2. Data needed for time-dependent 2. The number of pumps in series.
solutions. This data is expected only by the
simulation version of the program, and then only Example of the command SERIES
if the option ISIML= 1 in the $SPECIF list. This 2 3/
input data is described in the subsequently section 42/
entitled "Detailed Dynamic Simulation Data
Requirements." SETHGL and SETPRessure
3. The node number and the HGL
elevation at this node that gives the starting value In the analysis of a pipe networks
for hcad and pressure computations. This pressures may be known rather than demands.
additional line of data is required if, and only if, The command SETPR and SETHG allow for a
the network is the special simple network without nodal pressure, or HGL elevation to be given and
any sources of supply such as a source pump or have USU-NETWK compute the demand (which
reservoir. For such networks all exterior flows are is assumed unknown) at some other node, or the
given and the sum of positive demands equalS the same node, for each entry. The commands
sum of negative demands. Fluid mechanics text SETPRe (standing for setting pressure) and
books often give these simple network as SETHGI (standing for setting HGL elevation) can
illustrative problems. This node and HGL be used to either specify the pressure or the
elevation one needed if IHGL=-l, for example if elevation of the hydrauliC grade line at nodes,
no PUMPS or RESER are given and respectively. If the command SETPR is used the
NOMSOL=l is included in the $SPECIF list of pressure will be given in psi when using ES units
options. for pipe diameters and lengths (e.g. NUNIT less
4. Pairs of end pipes to be used to define than 2) and in kilo pascals, kPa, when using SI
the pseudo loops. These values are given if, and units for diameters and lengths (e.g. NUNIT= 2
only if, LOOPSE = 1 in the $SPECIF list. Should or =3). When pressures are given the pressure
PRY's exist in the network, it is necessary to give head is added to the elevation of the node to
both pairs of end pipes for the corrective flowrate establish the hydraulic grade line elevation. If the
loops, and the energy loops. data is entered with the command SETHGI, then
5. Estimates of the fraction of total the elevation of the hydraulic grade line is given
demand, or the flowrates coming from each in feet or meters dependent, respectively whether
reservoir and source pump of the network. These ES or SI units are used for the problem. In
values are given if, and only if, NPERCT =1 in order to satisfy the specified pressure at a node it
the $SPECIF list. is necessary that a demand is adjusted somewhere
6. The cost data called for if ICOST=1 in the network. The input data under these
or greater in the $SPECIF list. commands consist of:
7. CHANGE data for an alternative 1. The node number where the pressure
analysis to define a network with some elements (or elevation of the HGL) is specified.
altered but which is still the same basic network.

63-
 2. The node whose demand is to be = 1, then the diameter, wall roughness,
adjusted, e.g. solved for as part of the solution. and length are in feet.
This node can be the same node as the node in # = 2, then the diameter, wall roughness
1, but may be any other real node (e.g. not a and length are in meters,
source node if these are numbered with the use of = 3, then the diameter, wall roughness
the NODESP= 1 option). are in cm, and the length is in meters.
3. The pipe number that connect a source
of supply, e.g. a reservoir or source pump, to the VALVCheck
network. This item is required to save effort in
the program to find the source with the smallest The command VALVC enters data for a
paths from the nodes given in # 1 and # 2. This check valve, which prevents flow in a direction
source of supply should be as close as possible to opposite to that specified from the upstream to
the nodes given in # 1 and # 2 to minimize the downstream node of the pipe. The data are:
computations involved in the solution process. 1. The list of pipe numbers (integers)
The pipe number is required and no option allow that contain the check valves, terminated with a /.
for the source number, or its node number if Since only one item of information, the
NODESP=l. pipe number, is need for each check valve, this
4. The pressure, or elevation of the HGL, data is different in that it is not necessary to have
that is being specified at the node given by # 1, a different line for each additional check valve, but
respectively depending whether SETPR or they can, and generally are all on a single line. A
SETHG is used. check valve is different than a pressure reducing
The use of the commands SETPR and valve (PRV). The PRY holds the pressure
SETHG is different from the use of the DHEAD constant downstream from it regardless of the
command. With the latter command the solution upstream pressure, whereas a check valve only
determines a differential head needed in a pipe, prevents reverse flow. A PRY acts as a check
such as a pump if positive, or a loss device if valve should the flow in its pipe reverse it flow
negative, or a different pipe size that will satisfy direction.
the specified pressure. With the commands
SETPR and SETHG and external flow to or from Example of data after the VALVC command:
the network will be solved for in order to satisfy 4682/
the given pressure. Generally the demand at the
node of # 3 above will be given a value of zero VALVE (pressure reduction valve, PRY)
under the NODES or PIPE- command, since this
demand is designated as unknown. However, a Pressure reducing valves (PRV's) or
demand may be given if desired. If a demand is pressure regulators are used to reduce pressure to
given the value will be utilized in the initialization a specified amount regardless of the flowrate
of the flowrates needed for the iterative solution through the pipe. If the flow reverses in a pipe
process, and a good guess of the amount may line that contains a PRY, then it shut off the flow,
reduce the number of iterations needed for a and acts like a check valve. Should the pressure
solution. upstream from the PRY be less than the pressure
setting, the valve becomes inoperative, i.e. opens
Example of using SETPR and SETHG as wide as it can, but it can not maintain the
SETPRE SETHGL pressure setting. Under wide open operation a
2 5 1 40 2 5 1 592.31 PRY has the effect of a minor loss, very similiar
7 7 3 45 7 7 3 603.85 to a globe valve. USU-NETWK will determine
which of the above 3 conditions should exist at a
UNITS PRY, and obtain the correct solution accordingly.
PRY's cannot exist in consecutive (connected)
The command UNITS enters a single pipes, or pipes from reservoirs or source pumps
integer value. It is an alternative to the use of the for obvious reasons. Data entered after the
option NUNIT in the $SPECIF list. If the value command VALVE consists of:
following the command UNITS equals: 1. The pipe number (integer) which
= 0, then the diameter and wall contains the pressure reduction valve.
roughness, e, are in inches, and the length in feet 2. The distance between the upstream
(or 1000 feet),

-64-
node of this pipe and the valve in feet or meters, will be the same for all pipes, i.e. it is not possible
depending upon whether ES or SI units are used. to enter VISCO intermittently between pipes and
3. The pressure setting of the PR V have its value changes.
expressed in elevation of the hydraulic grade line
(HGL-elevation), immediately downstream from Example of the VISCO command
the valve. This HGL-elevation is in feet or VISCO
meters, depending upon whether ES or SI units 1.275E-6
are used. If the option NPRPRV=1, the pressure
is given instead of the HGL elevation. To WEIGHt
compute the HGL-eIevation from this pressure
USU-NETWK assumes that the pipe is a straight The command WEIGHT allows the
line between its end nodes to find its elevation at specific weight of a fluid to be given as an
the location of the PRV. alternative to using the option GAMMA in the
$SPECIF list. One value follows the command
Example of PR V data after the command VALVE WEIGHT, e-j. the specific weight of the fluid in
VALVE Ib/ft3 or N/m , depending upon whether ES or 'SI
2 100 340 units are used, respectively. If not given, 62.4
3 160370 Ib/ft3 is assumed if ES units used, and 9800 N/m3
is assumed if SI units are used by setting
VISCOsity NUNIT=2 or greater.

The command VISCO enters the Example

kinematic viscosity of the fluid in units of ft 2/sec
or m2/sec, depending upon whether ES or SI units WEIGHT
are used. The kinematic viscosity can alternatively 47.5
be specified different from the default with the
option VISC in the $SPECIF list. The viscosity

NOTES RELATED TO COMMANDS

-65-
-66-
 CHANGE DATA
USU-NETWK allows changes to be made detailed input data is that in the body of this
tn the basic network, and obtains a solution for manual, or according to the rules described in
the new network thus defined without having to Appendix A or Appcndix B.
makc thosc changes in the original input data filc. Changes to the network are specified by
Solutions obtained in this manner are rcferred to providing pairs of lines for each type of change.
as alternate solutions. Generally less The first line of each pair consists of a command,
computations is required to obtain an alternate or descriptive name, which describes the change,
solution than the original solution. The computer and the second provides the "changc" data that are
effort in reading the input data file is always introduced with this command. The commands
saved. Depending upon the Changes specified, the which are allowed, and a brief explanation of what
alternate solution may be obtained with only a they change are given in Table 1 below. These
fraction of the amount of computations needed for namcs must be spelled correctly to 5 characters,
thc original solution. This saving is accomplished and if desired more characters can be added.
by branching back to an appropriate point in thc Thesc first 5 characters must be upper case, and
program. Thc point of this branching depends are shown in upper case in Table 1 with added
upon thc nature of the changcs. For example, if characters in lower case to identify what the
the change consists of multiplying all demands by command does better. A command can appear
a constant, then initialization is obtaincd by between the CHANGE and END any number of
multiplying all previous flowrates by this constant, time for a given alternate analYSis. The units of
and then branching to the equation solving part of the changes are the same as those used in the
USU-NETWK. If only the elevation of the original detailed input data for the network, e.g.
junctions arc changed, then the previous solution those indicated in the $SPECIF list of options.
is used to get the new heads and pressurcs, etc. If
individual demands, and/or pipe lengths or Table 1. Commands for making changes
diameters, arc changed, then a new initialization to the basic network.
and solution are obtained. The version of USU-
NETWK that can also do time-dependent analyses i Command What it Changes
contains two subroutines to provide initial 1
Ilowrates in all pipes that satisfy the junction
continuity equations. The second such subroutine IALLCOEeffic. All roughness coer.
is designed to use the previous solution for the ICOEFFicient Individual roughness cocff.
initialization process, and it will be call on to !DEMANdS Individual demands
generate the initializing Ilowrates for the alternate !DFRACtion Multiplier of all dcmands
solutions, as well as for subsequent time step
solutions. None of the changes will result in the I DIAMEters Individual diameters
basic loops being determined again, or a new iDHEAD Dirrerential heads; both
banding of the Jacobian matrix, unless a PRY, etc. i type # 1 & type # 2 handled
shuts-off or opens again in the alternate solution. iNLNODes Elevation of nodes
Any number of alternate analyses can be IELPUMps w.s. elev. of pump sump
obtained of the basis network. For each such new 1 ELREServoirs W.s. elev. of re..<;crvoirs
alternate analysis the word CHANGE introduces iLENGTh Individual pipe lengths
the changes, and the word END terminates the MINOR coer. for minor losses,(not PC)
Changes. In other words each group of changes I
, PARALlel Numbcr of pumps in parallel
arc included bctween the commands CHANGE IPEAKF Multiplier of demands for
and END and these must start in column 1 of I ranges of nodes
separatc Iincs. Thc changes arc accumulative; that IPRVHG HGL selling of PRY's
is whatevcr changes wcrc madc to obtain thc IQGIVE Specified tlowrates in pipes
previous alternative solution will be carried over
into the next altcrnate analysis. If you want to go
IROTATional Ratio of rotational speed of
Pumps to that of pump curve
back to the original problcm, it is nccessary to SERIEs Number of pumps in series
change back as part of the new change SETPR Specified pressure at nodes
instructions. The change option is exercised the SETHGI Specified HGL-elcv. at nodes
same regardless of whether the form for the

67
 A single value follows the two commands 1 .34 5 .55 14 1.33{
ALLCO and DFRAC, and these values are the SETPR (Pair are: 1. node of specif. pres., 2. new pres.)
30 50.360./
new coefficient, and the new peaking factor,
respectively. All roughness coefficients are
The commands PARALlel and SERIEs
changed to the value given after ALLCO,
change the number of pumps operating in para1lel
regardless of whether all pipe had the same
or series at a pumping stations. Three items are
coefficient in the previous solution or not. The
needed (grouped together) instead of two as with
value after DFRAC is a multiplier of all demands
the commands above. These three values consist
from the previous solution. Thus if a peaking
of: (1) the number of the pumping station, (2) the
factor of 1.5 were used in the original input data
old (or previous) number of pumps in series or
using the option PEAKF= 1.5 in the $SPECIF list,
parallel, and (3) the new number of pumps to be
or the command PEAKF, then a value of 0.5 after
placed in series or parallel operation. Note if
DFRAC will result in the peaking factor of 0.75
simultaneous changes in the water surface from
for the alternative analysis. It is not possible to
which a source pump obtains the fluid and the
change computations from being carried out from
number of pumps which operate in parallel or
the Darcy-Weisbach equation to the Hazen
series at this same station are to be made, then it
Williams equation or vice versa.
is necessary that the Change in the number of
pumps in parallel or series by given before the
Examples using ALLCO and DFRAC
ALLCO change in water surface elevation.
.O()5
DFRAC Example of the SERIEs command
.5 SERIES
112332210{

The commands DHEAD, DIAME,

Two lines of data are required after the
LENGT, COEFF, DEMAN, ELNOD, ELRES,
command PEAKF, instead of the single line like
ELPUM, PRVHG, MINOR, QGIVE, SETPR and
the other commands. The first line consists of lhe
SETHG are followed by pairs of values. The first
peaking factor, i.e. the multiplier of the demands
value of each pair is the pipe number, node
at the nodes given in the next line, and the second
number or device number, and the second value of
the list of nodes. This list of nodes may consisl of
each pair is the new value. The command names
groups of nodes identified by a - between the
denote what the new value is. The first of each
beginning and last value of the group, or
pair must be an integer without a decimal point,
individual node numbers. This command is not
and the second may contain a decimal point.
implemented in the PC version of USU-NETWK.
These pairs of values can extend across a line up
to and including column 80. The list of these
Example of use of the PEAKF command
pairs of values must be terminated with a /. In PEAKF
using DHEAD the pipe designated on the first of 1.5
the pair of values must contain a differential head 202850-70 10 80-85{
in the original data. If this is a type # 1 device,
then the second of the pairs of values is the new The command TITLE is used to introduce
specified differential head. If this is a type # 2 a new title for the alternate analysis. The exact
differential head device, then the second of the manner in which this command has been
pair of values is the new specified HGL-elevation implemented is different in USU-NETWK
(e.g. the changed value for the last, or 5th, item in currently than was done previously. Earlier
the original input data) at the node given as the versions of USU-NETWK had the new tiUe follow
3rd item in the original input data. If this second the command TITLE on the same line. If more
item of the pair is negative, then it will be taken than a Single line of new title was desired, then
as a pressure (psi for ES units, and Pascals for SI the command TITLE was repeated on several
units) rather than an HGL-elevation. See lines. However, to make this title change
alternate to Example Problem # 2 for an example. consistent with the title at the beginning of the
input data file USU-NETWK now has the lines of
Examples using the above commands new title follow the command TITLE. The new
DIAME title is terminated by a 1* or */ (the latter a
2 8. 6 12. 30 10./
must for large IBM computers) on the next line,
DEMAN
as is the original title terminated. the original problem.
CHANGE
Example of new title for alternate analysis TI1LE
First alternate analysis, all previous demands multiplied by 2.5
TI1LE
It
FIRST ALTERNATE ANALYSIS RELATED TO THE DFRAC
STIJDY OF THE KERNS SYSTEM 2.5
Fire demands at nodes 2, 7 and 8 END
1*
CHANGE
TI1LE
Second alternate analysis, diameters of pipes 1 8 and 50
The following gives an example of change changed
data that will cause two alternate analyses to and the demand at node 10 changed to 1.5 cfs
follow the original solution to a network problem. 1*
These lines of input occur after the RUN DIAMETERS
1 6 810 5012/
command, or the last line of input required for DEMANDS
10 1.5/
END

NOTES RELATED TO CHANGES

-69-
-70-
= COST INPUT DATA

Introduction default values can be overridden by providing

values to replace them. Thus, as the user, you
In addition to providing solutions to have control over how detailed the cost analysis is
hydraulic problems giving flowrates and pressures to be. At the simplest level, the default values
throughout pipe networks, USU-NETWK will also will be used for capital costs of pipe and unit
provide an engineering economic analysis of the costs of electricity at the default interest
system for which the hydraulic solution has just
been obtained. The costs given as part of this rate and life will be used. At a much greater level
analysis can be as applicable to any given system of complexity, you can supply different interest
as arc the data provided as Hcost input data. H Or rates, life expectancies, and/or different capital and
only default cost data built into USU-NETWK unit costs to various individual components or
may be utilized, without providing any data, in groups of components of the network
which case the results from the cost analysis would In writing this section of the manual the
provide only a rough comparison of relative costs assumption has been made that you have
associated with various alternatives. Stated completed a course in Engineering Economics, or
another way, the cost analysis can be as complete at least have had experience in evaluating
and realistic as you, as a user, are willing to alternatives by making comparisons of costs on
provide appropriate input data, or if no data are either the basis of the "present worth" or on the
provided, beyond the request to perform an basis of "reoccurring series costs." However, to
economic analysis, then only two items are refresh you memory about basics such as capital
included: the recover and present worth factors, the following
costs of pipes and the costs of energy, if pumps subsection is included in this manual.
exist, and these are based on default values for (a)
cost per unit length of different size pipes, (b) Synopsis of Engineering Economic
energy costs, (c) life expectancy, and (d) interest Analysis
rate.
The capability of USU-NETWK to The costs associated with a water
provide an engineering economies analysis makes distribution system roughly fall within the category
it easy to determine the cost effectiveness of of a capital investment, or a reoccurring cost. The
various alternatives. Therefore, as a user of USU- acquisition of the right of way, installation of pipe,
NETWK, at least as you approach the final design purchase of pumps and their installation, and
of a system you will probably find it very useful to construction of storage tanks or reservoirs arc
request that a cost analysis also be done. Cost examples of capital investments. With these there
data should be collected and prepared so that is a present cost to acquire and/or get the facility
USUNETWK's cost analysis is meaningful. on line. Often the facility will be paid for over
This section of the manual describes the many years. If so the total amount paid will
input data that are needed for such an engineering include the initial cost plus interest on the money
economic analysis. This input data has been borrowed. Paying for energy consumed by pumps,
designed to be as simple as possible, while buying water, maintaining the system, and paying
rermitling a maximum amount of flexibility in salaries of personnel associated with the system's
assigning both reoccurring costs, such as operation are examples of reoccurring costs.
maintenance, energy and costs for water and/or In order to have a convenient basis of
one lime capital investment costs to entire groups comparison of the costs of alternatives in the
of network components, or provide separate values design of a water system, either the capital
to each individual component. Basic default investment costs need to be converted to
values for capital costs for pipe and unit costs of reoccurring costs and then added to these costs,
electricity are built into USU-NETWK, allowing that do actually reoccur; or the reoccurring costs
a minimum cost analysis to be performed by the need to be converted to present worth values, and
single command END, which denotes the end of added to the capital investment costs. USing the
the cost data. Other costs can be added to these former as the basis of comparison, the alternative
by providing the appropriate data, and these with the smallest reoccurring sum of costs would

.71
be selected if all other consideration are equal, The amount of the payment, $16,274.54 shown in
and based on the latter, the most attractive column 4 of this table is obtained by multiplying
alternative would have the smallest composite the crf, which from Eq. 1 equalS 0.0162754 by
present worth. The choice of the reoccurring $100,000. Interest shown in column 3 has
composite cost, or the composite present worth accumulated over the year from the balance due
comparison is a matter of personal choice. after the last payment. The charges for the use of
A capital investment cost is converted to the money are listed in column 3, and are
a reoccurring cost by multiplying its amount by obtained by multiplying column 2 by 0.1. The last
the capital recovery factor, crf. The formula that column, which is the difference between the
gives the crf is, payment, column 4, and the interest, column 3, is
subtracted from the first column to obtain the
i(l+it next value in the first column for the next year.
crf = ---------- . (1 ) The 10 payments exactly equalS the amount of
(l+i)n-l interest plus the initial $100,000. Therefore, the
reoccurring amount of $16,274.54 is equivalent to
in which i is the interest rate as a decimal, and n the $100,000 capital costs of the pump. If the
equalS the number of reoccurrences of the pump needs to be replaced after these 10 years of
payment. A reoccurring cost (that is assumed to operation, then the $100,000 present worth of the
be constant in amount) is converted to a present pump is identical to a $16,274.54 annually
worth by multiplying it by the present worth reoccurring cost.
factor, pwf, which is the reciprocal of crf, and is The annual payment amount of
given by the formula, $16,274.54, however, is not the total cost for
(l+i)n - 1 having the new pump on line. It uses energy and
pwf = ------------- . (2) requires maintenance that, let's assume equalS
i(1 +i)n $20,000 per year. One method for evaluating the
yearly costs for the pump is to add these two
For example assume it costs $100,000 to reoccurring amounts to give $31,228.80/yr. An
get a new pump on line. However, since $100,000 alternative is to multiply the $20,000 by the pwf
is not available the money must be borrowed at an and add this amount to the $100,000 capital cost
interest rate of 10%, with an agreement to pay the of the pump. The pwf based on a 10 percent
loan back over a 10 year period with equal interest rate and number of payments equals
payment at the end of each year. The table below 9.090909 so that the total composite present worth
shows the cash flow associated with this capital of the pump equals $281,818.18.
investment. Any number of other capital investment,
or reoccurring costs can be added to get the total
Table showing cash flow for example acquisition cost. Before they can be added it is necessary to
of pump. (Life=lO yr, interest rate=O.lO, convert them to the same base, e.g. a reoccurring
capital investment amount=$100,OOO.) cost or a present worth amount. Since the values
of crf and pwf depend upon the interest rate, and
Yr. Amount Interest Payment Reduction the number of payments i and n must be known.
owed at Charges amount in amount If the amounts are to be interpreted as those
start of owed necessary to keep the water system in continual
year service, then the number of payments can be
(1) (2) (3) (4) (5) interpreted as the life expectancy of that
1 100000.00 10000.00 16274.54 6274.54 component of the system. If payments are made
2 93725.46 9372.55 16274.54 6901.99 twice a year, the crf can be computed using n:;:
3 86823.47 8682.35 16274.54 7592.19 2 x (life expectancy) and i = 1/2 x (annual interest
4 79231.27 7923.13 16274.54 8351.41 rate). Life expectancies of different components
5 70879.86 7087.99 16274.54 9186.55 of the system need not be the same, nor do
6 61693.31 6169.33 16274.54 10105.21 interest rates associated with them need to be the
7 51588.10 5158.81 16274.54 11115.73 same. The crf and/or pwf must however be
8 40472.37 4047.24 16274.54 12227.30 computed based on the proper nand i for that
9 28245.07 2824.51 16274.54 13450.03 network component.
10 14795.04 1479.50 16274.54 14795.04 There are numerous variations that might
Total 0 62745.40 162745.40 be used to the above. Salvage values may be

-72-
included, but if the salvage value is subtracted shown below.
from the cost of the replacement the same final INTEREST=i simple annual interest rate
result is obtained. Inflation may be factored into (without compounding)
the analysis by subtracting its rate from that of LIFE = life expentency in years
actual interest rate. For practically all water PAYMENT=time between payments in
distribution system analysis constant reoccurring years (or fractions thereOf)
series costs is realistic. USU-NETWK assumes To get the i used in equations 1 and 2 above
that all reoccurring costs are constant for all ~NTEREST is multiplied by PAYMENT, or
payment periods. I=INTEREST x PAYMENT, and the value of n
is computed by dividing LIFE by PAYMENT. In
other words if PAYMENT is 1 the values entered
by the commands INTEREST and LIFE are equal
Nature of Cost Data USU-NETWK to j and n in Equations 1 and 2; otherwise they
are not. For example, if the desire is to compute
Understands costs based on making monthly payments and an
annual interest rate of 10 percent exits, then i
You communicate to USU-NETWK that
.10/12 ::;; 0.00833333, and payment 1/12 =
~n en~ineering economic analysis is to be done by
.0833333. Every new entry of any of these three
IncludIng ICOST=1 in the $SPECIF list of
commands will result in new values for crf and
options. With ICOST set greater than zero, USU-
pwf being computed, and these new values will be
NETWK expects to read cost data after the RUN
used in computing costs and summing each new
command.
computed cost into the accumulated "present
The input data for an engineering
worth" and accumulated "annual cost" until any
economic analysis of a network utilizes the same
one of them is re-entered, at which time a new crf
general format as other input data. A name
and pwf will be computed. The default values will
(command) enters data of a given type. The cost
be used in any such computations until one of
dat~ arc entered with any, or all of 18 following
these commands is first entered.
available commands. The names are words you
The commands EFFI, DAYS, RESP and
can relate the type data they enter to. These
NPRI are followed by a single value that have the
commands consist of: INTEREST, LIFE,
following meanings:
PAYMENT, ELECTRICITY, PIPES,
EFFI - This command enters a combined
RESERVOIRS, APPURTENANCES, WATER,
motor pump efficiency that will be used to
OPERATING,POWER,PUMPS,ADDITIONAL,
compute the energy used by the pumps. In other
NPRINT, EFFICIENCY, DAYS, RESPOWER,
words, the energy that a pump adds to the fluid
DEMCOSTS, and END. The first 4 characters in
w.ill be divided by this value in computing the
each command are significant, and each may be
kilowatt-hours of energy that will by multiplied by
truncated to only 4 characters. For example
the unit cost values entered by UNIT= after the
ADDl, EFFI, etc are sufficient. In addition the
ELEC or PUMP commands. The default for
wo:ds CAPI= and UNIT= preceded values being
EFFI is 1. Therefore, if EFFI is not given as part
assigned to capital and unit costs, respectively.
of the cost data the values entered under pumps
Funhermore, the names EQ-C and EQ-A are used
by UNIT= should equal the electrical rate paid
to introduce coefficients for equations that will be
per kilowatt-hour divided by the combined
used t~ computer capital, and reoccurring costs,
efficiency of the pump and its motor. The value
rcspeclively. The uses of these 18 commands and
entered by EFFI will be used in all energy cost
the additional 4 names are summarized in Table 1.
computations until it is given a new value.
All commands and names must begin in column 1.
DAYS - This command enters the number
The first three commands listed above
namely INTEREST, LIFE, and LIFE are in ~
of days of operation for pumps throughout a year.
~ifferent category than the rest in 'that they The default is 365. Therefore, if a pump operatcs
50 % of the time the command DAYS followed
Introduce parameters that are used to compute a
on the next line by 182.5 should be given before
new crf and pwf, and they are followed
the ELEC ?r PUMP command for this pump.
immediately with an equal sign, which in turn is
An alternative would be to reduce the unit energy
followed immediately with a value. These
costs to one-half the actual amount.
parameters can occur individually, or together, and
RESP - The default meaning of
may be repeated any number of times in the form
UNIT=amount after the RESER command is the

73
cost per unit (feet or meters) above the ground pumps that are included in the input data under
elevation of the reservoirs. This ground elevation the commands PUMPS and BOOSTER can be
is provided under the NODES command if included, or negative demands at nodes might be
NODESP= 1. When NODESP=O this ground considered to be pumps. There is only one line
elevation is taken as the elevation at the node on (which may extended into several lines) that
the other end of the pipe that connects the fol1ows the command DEMC. The items on this
reservoir to the network. An alternative meaning line consist of:
for UNIT=amount is achieved by giving the name 1. KW1im - The limit of kilowatts that can be
RESP, that denotes the flow from reservoirs is to used by the pump without a demand charge. The
be considered water that has been pumped into amount of this demand charge is computed by the
them and, therefore, the outflow is to cost the equation below. (An exception to having zero
same as pump power or energy, a value different demand charges occurs if the kilowatts, KW, used
from zero. This alternative meaning is by the pump are less than KW1im, but the next
accomplished parameter Tde is not zero).
by including the following in the cost data: 2. Tde - A demand charge amount that will occur
if the quantity (KW-KWlim ), as defined below, is
RESP negative. The amount Tde is multiplied by the
1 days of operation to obtain the demand charge.
3. Cde - The constant in the quadratic equation
To change back to the default meaning, a 0 can below that computes the demand charge.
follow RESP subsequently in the list of cost data. 4. Bde - The linear coefficient in the quadratic
After the command RESP followed by a 1 the equation below that computes the demand Charge.
amount after UNIT is the cost of the energy 5. A de - The coefficient that multiplies the
associated with this water flowing out of squared term in the equation below that computes
reservoirs. Should the flow be into the reservoir, the demand charge.
no cost is associated with it, since this cost will be 6. Sde- A demand Charge that is a dollar amount
taken into account by the pumps that supply the for each pump included in the list, but docs nOl
now to the network. The energy is computed as depend upon the kilowatts of power consumed hy
for pumps, e.g. the flowrate is multiplied by the the pump, and is not multiplied by the number of
head (the difference between the water surface days of operation.
elevation and the ground elevation) and an 7. A list of pump numbers (i.e. the order of the
appropriate conversion factor to give pumps in the input data with booster pumps
kilowatt-hours. The number of hours per year can following source pumps), or the list of node
be reduced from 365 by providing an amount after numbers preceded by a minus sign if a negative
the command DAYS. Also if an amount is given demand represents pumped water. This list of
after the command EFFI, then this given efficiency pump numbers, or node numbers) must be
is divided into the above computed energy before terminated by a I. The 1 allows this list to be
multiplying by the unit costs for energy. extended unto several lines if needed.
NPRI - This command changes the The equation that computes the demand
amount of extra information that will be written charge rate is as follows:
10 the output file. In other words this command,
as part of the cost data, overrides the value set by
the option NPRINT in the $SPECIF list, or the
last value entered by the command PRINT. Its in which KW is the kilowatts of power required hy
usc as part of the cost data allows more, or less, the pump.
extra output be obtained. For example, if Should the quantity (KW-KWUm)' which
NPRINTis used to enter a 6 just before the PIPE is the kilowatts of power required by the pump
command, and again a 0 before the next minus the limited value given or parameter #1, be
command, it is possible to get the costs printed negative, then the amount Tdc given as item # 2
out for every pipe in the network on a separate above is taken instead of the amount of $/day
line, but not this amount of detailed cost data for computed by the above equation. To get the
other components of the network. reoccurring costs the amount computed by the
DEMC - This command allow for demand above equation is multiplied by the days of pump
charges associated with pumps to be included as operation per year, or 365 unless the command
part of the engineering economic analysis. Either DAYS appears before the command DEMC.

-74-
 In addition to the cost computed by the pumps. The word VALVE indicates flowrates
last multiplication the value of Sdc or item # 6 through those pipes containing pressure reduction
above (without multiplying it by the number of valves.
days of operation) is added to the above amount. Mter the commands UNIT=value or
This latter value Sdc is added whether the CAPI=amount a comma may be given followed by
kilowatts of power required by the pump exceed a range of values. In Table 1 this is shown as
the limit KWlim or not. Any of the coefficients in ,<range of . . > in which the . . are two integers
the above equation may be assigned zero values separated by a - that identify the range of pipes,
should this be appropriate in defining the methods reservoir, etc. The range is given by an integer
used to obtain the demand charges. If all negative value, a dash followed by a second integer without
demands represent pumped water, and all of these any intermediate spaces. This range is optional,
arc to have the demand charges associated with but if given then only those pipes, etc. within the
them then rather than listing all of these node range will have the given costs attached to them.
numbers, as item 7 above, this item can be given Should the range include pipes, etc. that do not
as -9999 following by I. exist, the non exist numbers will be ignored. If
END - The command END tells USU- the dash and second integer of the range are
NETWK that no more cost data are being omitted then only the one given value constitute
provided, and therefore the cost analysis should be the range. For example UNIT=.05,PUMPS,1
terminated. In obtaining time-dependent indicates the unit cost of .05 applies only for the
solutions an additional END command is needed flowrate through the line containing pump 1. A
for each succeeding time step solution for which 1-1 in place of 1 would have exact the same effect.
a cost analysis is requested. Giving the following after the command WATE:
The remaining seven commands must be UNIT=.04,NODES would indicate that a unit cost
followed on the next line or lines with the name of $.04 per cfs of flowrate should be charged for
UNIT=value (unit cost), or CAPITAL=amount all demands throughout the network, whereas
(or CAPI=amount). The values following the UNIT=.04,NODES,5-25 would indicates that a
cq ual signs are the amounts to be used as the unit unit cost of $.04 per cfs of flowrate should be
cost or amount of the capital investment. An charged for the demands at nodes 5 through 25
exception is the line after the command PIPE if it for those node numbers that exist within this
is UNIT=N. In this exception N indicates the range. Mer the command ELEC a line
number of pairs that follow giving the cost of pipe UNIT=.09 would indicate that the energy in
per foot of length (or meter of length if SI units kilowatt-hours used by all pumps in the network
are used) and the diameter in inches (or meters) should be multiplied by .09 to compute these
for which that cost applies. The other exception costs, but a line UNIT=.09,
is the command ADDI. Mter ADDI the next line 2-4 would indicate that this unit cost applies only
provides a name for that additional cost. This for pumps 2 through 4.
name is followed by UNIT=unit cost or It is important to understand that in doing
CAPI=amount in the usual manner. Commands an engineering cost analysis, USU-NETWKcarries
WATER, OPERATING and ADDITIONAL, out the computations in the same sequence as the
when followed by UNIT=unit, have a comma data are given, and that the costs are accumulated
followed immediately by one of the words: FLOW, as the computations are completed. If no PIPES
PIPES, NODES, RESER, PUMPS, or VALVE or ELEC commands are included, then USU-
(all 5 characters long). All of these are unit costs NETWK will carry out these two costs analyses
that multiple flowrates in basic units (cfs, m3ls, based on the default costs for these items. Since
Ib/scc, slug/sec, N/s or Kg/s). The word FLOW commands may be repeated as many times as
indicates the total demand on the network. The desired, with the cost for that item accumulated,
word PIPES indicates the flowrate in pipes. The you have considerable flexibility in defining the
word NODES indicates the positive demands from costs and what should be included. Both CAPI=
the nodes (flowrates into the network, i.e. negative and UNIT= may follow a command, and either or
demands, are ignored in computing costs under both of these may be repeated any number of
these commands). The word RESER indicates times after a given command. Furthermore,
flowrates fromlto reservoirs. This cost for ranges may be given. If ranges overlap between
reservoirs flowrate is accumulated regardless of consecutive UNIT='s the effect is to add the two
whether the flow is from or into the reservoir. separately given unit costs to the overlapping
The word PUMPS indicates flowrates through numbers. Thus for example if you desire you

-75-
could use the PIPES command to separate the commands (except pipes) the multiplication of the
installation of pipes into: (I) the actual cost of the unit cost by the item indicated will be considered
pipe, (2) the instaUation of the pipe in the a reoccurring cost, or constant series cost that
ground, and (3) the cost of acquiring the right-of- occurs each payment period in years. If
way. For item (1) there would be considerable PAYMENT=I (the default), then this will be a
different unit costs for different pipe sizes and the cost each year.
given data would reflect this. For item (2) there 3. CAPl= enters a single capital
would likely be smaller differences in unit costs investment cost. When CAPl=value is given after
for the small versus large diameter pipes, and for the commands ELEC, PUMP and RESER without
item (3) all unit costs would be the same giving a range, then this amount will be applied to
regardless of the pipe diameter. all such components in the network. For example
You should carefully study Table 1. After if there are 5 pumps in the network, and "value"
you have become familiar with cost input data to is 15000, then the total capital cost will be 5 x
USU-NETWK you will find that a glance at this 15000 = $75,000 for the 5 pumps. CAPl=value
table is about all you need to decide on what you after other commands such as WATER, etc. will
want to include in the cost data portion of the cause the amount of only one "value" to be added
input. The following explanation of the meaning to the costs. If you wish to have the amount of
of terms will help you in understanding Table 1 "value" included only once after a command such
hetter. as ELEC and several pumps exist, then simply add
a comma followed by 1 for the range.
Ex~mplc of cost data
4. In the case of the energy costs
PIPES
CAPI=25000 associated with pumping no efficiency of motor
UNIT=8,1-40 and pump is used unless you have previously given
4 2 6 4 8 6 10 8 12 10 15 12 18 14 20 20 the command EFFI and an effiCiency (fraction)
UNIT=8,41-60 following it. Therefore, if you don't use the EFFI
44 6 6 8 10 10 12 1220 15 25 1845 20 60
RESER
command the actual costs for electricity should be
CAl' I = 12000.1 divided by the combined efficiencies of the motors
UNIT=200 and the pumps.
ELECTRICITY 5. By changing the interest rate, and life
CAP 1=5000,2 repeatedly with the commands INTEREST=rate
UNIT=.08
WATER
and LIFE=life before giving other cost data it is
UNIT= 1.,I'LOWS possible to use different interest rates, etc. as the
UNIT=.5,PUMPS,4-5 basis for computing reoccurring costs from given
CAPI=I0000 capital costs, and present worth, from reoccurring
OPERATING cost so that the interest rates, and lives can
UNIT=.05,PIPES
ADDITIONAL duplicate what the actual situation is. The
REPLACEMENT default for LIFE is 50 years, and the default
UNIT=.02.FLOWS interest rate is 0.10.
END 6. The value given after the equal sign
following the command PAYMENT is the
fraction, or multiple of years when the reoccurring
1. ELECTRICITY represents electrical uniform payments occur. The default is I,
energy consumed by pumps, and must be used for meaning that uniform series costs occur once each
costs associated with any pump in the original year. If you want to have the computations based
data. If pumps exist but they have not been on monthly compounding of interest, the
entered in the original data with the command command PAYMENT=.08333 could be given for
PUMPS or BOOSTER but rather have been given example.
as a negative demands, then the command PUMPS 7. The default capital costs for pipes
must be used rather than ELEC. consist of the following:
2. UNIT= enters a unit cost, or a
mUlliplier of the item indicates by the command
to get the reoccurring costs. When given after the
command RESER this multiplication will produce
a one time cost that is added to the initial capital
investment cost. When used after any other

-76-
 POWER provides the capability to determine
Pipe Cost/unit revenue that can be obtained through generation
dia. length of power. The amount given after CAPI=
following the POWER command represents the
capital costs associated with the installation, etc.
4" $ 3.67 of the power generating facilities, and is a costs.
6 5.33 The amount provided after the UNIT=is the
8 7.67 amount that each kilwatt hour of power can be
10 10.67 sold for. This energy is computed based on the
12 16.67 flowrates, i.e. demands at the nodes and heads
15 24.00 available at the nodes as determined by the
18 43.33 solution. The range of nodes (which will generally
20 56.67 be a single value) must be given. No efficiency of
24 80.00 the hydraulic turbine-generator is used in the
30 100.00 program. Therefore, the unit cost data supplied
36 120.00 must reflect this combined effeciency plus any
42 145.00 transmission losses.
48 170.00 9. It is possible to provide costs
54 200.00 associated with ELECTRICITY, PIPES,
60 235.00 RESERVOIRS, POWER, and PUMPS by
72 270.00 providing parameters in equations that are
84 300.00 commonly used for this purpose. Table 2 below
% 350.00 gives these equations. The name EQ-C (which
108 390.00 can replace CAPI) following the above commands
enters the parameters K, a and b shown in the
equations, and the given equation is used to
The command PIPE is used to supply a compute a capital cost. The name EQ-A (which
different list of pipe diameters, and associated unit can replace UNIT) following the above commands
costs. This is done by following the command enters the parameters K, a and b shown in the
PIPES on the next line with UNIT=N, in which equations, and the given equation is used to
N is the number of different pipe diameters for compute a reoccurring cost. The parameters are
which data will be supplies. N cannot exceed 20 entered by leaving a space after the name EQ-C
with current program dimensions, and then on the or EQ-A and then giving:
next line (or lines) provide the list of N pairs of A=value,B=value,K=value; <range>. Note that
values with the diameters followed by the costs. capital A & B in the actual input correspond to a
This list can be spread across a line up to and & b in the equations in the table below. If any of
including column 80, or can be listed with only the parameters (letters) followed by an = is not
two values per line. If a pipe diameter in the given, then the current default value for that
network is not equal to one of the given pipe parameter will be used, and if the range is not
diameter, i.e. between two of the entries of the given all devices of that type are assumed. The
given pipe data, then the costs associated with that starting default values are shown in the table
pipe will be interpolated linearly between the costs below, and any value given for a device becomes
for the two pipe diameters that bracket its size. the default thereafter.
Costs associated with pipes with either smaller The following is an example of using
diameters, or larger diameters than in the list of equations to compute costs.
diameters will have their unit costs equal to the PIPES
end values of the cost data, i.e. the pipe costs data EQ-C K=25,A=1.5
ELEC
are not extrapolated. The diameters must be UNIT=.09
entered in ascending order of magnitude, i.e., the EQ-C A=.4,K=580,B=-.35;1-2
smallest diameter first and the largest diameter CAP I =32000,1-4
last. RESER
EQ-C K=15000,A=500,B=.5j2
8. The POWER command is the only one
EQ-A K=750,A=500,B=.5;2
that produces a revenue, or negative cost, END
associated with the network unless negative values
are entered following CAPI= or UNIT;;;:.

-77-
 Table 1. Summary of input to provide data for an engineering economic cost analysis of the network. (Commands and names
that enter data are in upper case letters.)

1. IN1EREST=rate (Enters the interest rate, default=O.l) CAPI=amount

II. L1FE=Iife (Enters the life of the network XI. OPERATING

components, default=50 years) UNIT=cost/Unit of water,FLOWS
UNIT=cost/unit of water,PIPES,<range of pipes>
III. PAYMENT=x (Enters the time in years, or fractions UNIT=cost/unit of water, NODES, < range of nodes>
thereof between payments of the uniform reoccurring UNIT=cost/unit of water,RESER,<range of
series amount, default=1.0) reservoirs>
UNIT=cost/unit of water,PUMPS,<range of
IV. ELECTRICI'Y (For pumps in detailed data of network) pumps>
UNIT cost/kwhr, < rangeofpum CAPI=amount
pnumbers>(default =.10)
CAPI=amount,<range of pump numbers> XII. ADDITIONAL (Up to 5 additional costs may be deCined)
EQ-A eq. parameters; < range of pump numbers> NAME OF ADDITIONAL COST (up to 12 char.
(see Table 2) long)
EQ-C eq. parameters; < range of pump numbers> UNIT=cost/unit of water,FLOWS
UNIT=cost/unit of water,PIPES,<range of pipes>
V. PUMPS (For negative demands at nodes, assumed UNIT=cost!unit ofwater,NODES,< range of nodes>
pumped) UNIT=cost!unit of water,RESER,<range of
UNIT=cost/kwhr, <range of node numbers> reservoirs >
CAPI=amount,<range of node numbers> UNIT=cost/unit of water,PUMPS, <range of
EQ-A eq. parameters; <range of node numbers> pumps>
EQ-C eq. parameters; < range of node numbers> CAPI=amount

VI. POWER (For income received from power generated) XIII. EFFICIENCY (Enters combined efficiency of pumps &
UNIT=revenue/kwhr,<range of node numbers> motors) decimal value (default=!.)
CAPI=amount,<range of node numbers>
(is a cost) XIV. DAY (Enters days of pump operation through ye.1r)
EQ-A eq. parameters;<range of node numbers> value (default=365)
EQ-C eq. parameters;<range of node numbers>
XV. RESPOWER (Water coming from reser. is assumed
VII. PIPES (Enters unit costs of different pipe diameters) pumped)
UNIT=N,<rangeofpipes> (N=numberofdifferent 1 or 0 (1 denotes power, 0 costs based on
pairs of diameters and unit costs in list flowrate)
that follows) -
dial cost/ft (or m) XVI. NPRINT (Changes amount of output written) integer-3
dial cost/ft (or m) thru 10 allowed, with -3 least)
(see table in text for
default unit costs for pipe) XVII.END (Terminates reading of cost data and writes
diaN cost!ft (or m) summary table giving present worth and annual costs)
CAPI=amount,<range of pipes>
EQ-A eq. parameters;<range of pipe numbers>
EQ-C eq. parameters; < range of pipe numbers>

i VIILRESERVOIRS
I UNIT=cost!ft of elev. above ground,<range of
reservoirs>
I
I CAPI=amount,<range of reservoirs>
EQ-A eq. parameters; < range of reservoir numbers>
I EQ-C eq. parameters; < range of reservoir numbers>

IX. APPUR1ENANCES (PRV,BPV,check valves & other I

controls)
CAPI=amount
I
I
Ix. WA1ER (Costs of water) I
iI
UNIT=cost/unit of water,FLOWS
UNlT=cost/unit of water,P IPES, < range of pipes> 'I I
J
UNIT=costlunit ofwater,NODES,<range of nodes >
UNIT=costlunit of water,RESER,<range of I
I __
reservoirs>
U_N_I_T_=_COS_p_t~_:_~_~_~_f_w_a_ter,PUMPS,< range of _____ 1______________ .. _______ . ____________________ _
 Table 2. Allowable Equations that can be used to define costs

Itcm Equation Initial Default Values Comments

ELEC $=KQI! or K=635,a=.642,b=.453 Q & H are f10wrate and head in basic units,
i.e., cfs, ems, ft or m.
i
$=KE" if b=O or E is kilowatt of power.
$=K~Qlbl if b<O K=635,a=.642,b=-.453
POWER (same as ELEC)
PUMP (same as ELEC)
PIPE $=K.D"EXP(bD) K=23.18,a= 1.6875, D is dia. in basic units
i.e., feet or meters
RESE $=Ki K= 12912,b=.556,a=500 a = vol. of tank in 1000 gal for example.

If the command IS EQ-C then the $'s are the present worth values; If the command IS EQ-A then $ stands for cost
of reoccurring payment.

NOTES RELATED TO COSTS DATA

~79~
-80-
 DETAILED TIME DEPENDENT DATA
REQUIREMENTS
(Applies only to the simulation version of USU-NETWORK)

Introduction ordinary differential equations, numericall~. In solving

differential equations a "corrector" IS generally
From its first implementation in 1976, USU- subsequently applied. The corrector technique is not
NETWK has allowed the performance of a plpmg
system to be simulated over any designated period of used by USU-NETWK to adjust conditions at the new
lime. The literature dealing with the subject of time step since this would result in an iteration of
modelling water distribution systems now often refers to iterative solutions by the Newton Method, and greatly
such time-dependent solutions as "extended time increase computer times required. Further~ore, the
simulations." Such lime dependent solutions perform needed logic would be considerably more complex and
a series of steady-state analyses at specified time if not extensive enough could result in a fliptnop
increments until the total time desired in covered. situation of repeatedly turning a pump on and then off
Reservoir elevations are adjusted for each succeeding again, for example. Therefore, should conditions vary
solution based on the elevation-storage characteristics very rapidly in time, the user must specify. an
of the reservoirs; demands are adjusted to agree with appropriately small time step for the results obtamed
time varying magnitudes as determined by dimensionless from USU-NETWK to agree well with what actually
time dependent demand functions; or rules based on occurs.
nodal pressures or tank water surface elevations; the
number of pumps that operate at pumping stations are When should time-dependent solutions be
adjusted according to time schedules; and/or rules based obtained?
on pressures at nodes or water surface elevations in
reservoirs; etc. so that the actual performance of the The answer to the question ''when should I
system over the designated period of time is simulated. obtained a time-dependent solution to my network
The accura<-1' of the simulation is dependent upon how problem" is very subjective. Considerably morc data is
well such "time dependent" input data reflect the actual required for such simulations than a steady-state
conditions under which the network is performing. analysis, and the implication is that you know what the
USU-NETWK does not do a transient analysis time varying demands, etc. actually are. Physical
of pressure surges that result from very rapid closure of information about a network such as pipe diameters
valves, or on/off operation of pumps. Such transient and lengths, locations of pumps, and tanks, etc. arc
analyscs must take into account the inertia of rapidly relatively easy to obtain versus information related to
decelerating-accelerating the fluid, as well as the time-dependent demands. When a steady-state solution
clastic properties of the fluid, and the pipes that is obtained it represents a hypothetical condition such
contain this fluid, and accommodate surge tanks, air as an average daily demand condition, which likely
chambers, etc. that are used to control the large "water seldom if ever occurs since the time when the average
hammer" pressures resulting therefrom. In fact USU- demand occurs at one location is likely differcnt than
NETWK docs not account for the pressure differences when it occurs elsewhere. When fire demands are
needed over the length of a pipe to accelerate, or extracted at selected nodes, the solution is used to
decelerate a "solid column" of water that does not answer a "what if' question, that helps evaluate the
compress.' Such analyses based on solid water column adequacy of the network to deliver under unique
theory require the solution of a system of first order circumstances, that again likely never will occur.
ordinary differential equations. USU-NETWK solves However, a simulation over a day's time period implics
the same system of nonlinear algebraic equations that that you have information that not only describes how
dennc a steady-state problem. The difference is that demands change over a typical day, but that you have
this system of equations is modified for each new time information related to tank water surface levels, pumps
step to reflect a new steady-state condition at this time. operating, etc. at the beginning of that d~y. True ~he
The determination of the conditions at each new time day may be hypothetical, but the time varyIng
arc based on projecting conditions, as defined by the conditions through the simulation period should
current solution, ahead one increment in time. This represent a typical real condition, or else the results
technique is analogous to a "predictor" in solving

-81-
~.= from the computer solution have little resemblance to in defining the problem. However, you can provide
what happens in the actual network. some of this information interactively as you examine
Therefore, generally time-dependent solutions the solutions from previous time steps, i.e. you can turn
are not obtained first in studying the performance of a pumps on or off depending upon what you see is
piping system. Rather steady-state solutions are occurring during the simulation, for example.
obtained first, and they are used to verify that the The time-dependent input data have two major
model thus created by the computer does described the categories similar to that of the initial input data;
actual system reasonably well. The knowledge acquired namely a list of options followed by the detailed data
from real data as well as studying individual steady-state that describes the time variations. The data in each of
solutions, might then be used to define what the time these major categories are described in separate
varying problem most like should be. sections below. The time dependent data follows the
While considerably more information is RUN or END command fro~ the detail data, provided
required for a time-dependent solution than individual options have not dictated that other data follow this
steady-state solutions; the amount of information command.
obtained about the networks performance is also much
greater. In fact the information included in a single
time-dependent solution supercedes many indivIdual TiIne Dependent Options
steady-state solutions, because conditions, such as water
surface elevations in tanks, and pumps on or off are Options, or speCifications, that provide general
automatically adjusted to reflect what should actually information related to the time dependent-solution are
occur. Thus some of the hypotheticalness of the included between $TDATA and $END, and this list
individual steady-state solutions is replaced by what must preceded the other detailed data for the timc-
might actually occur. Perhaps the best answer to the dependent solution. This $TDATA list of options is
question of "when should time-dependent solutions be similar to the $SPECIF list of options. Valid
obtained?" is only if you have considerable knowledge parameter names followed by an equal sign and the
and insight into actual operations and conditions value given to the parameter are given for those
imposed on the system. However, once this knowledge options for which the default values are not appropriate
exists time dependent solutions should be used rather for the problem being solved. The difference is that
than individual steady-state analyses because the results the list under $TDATA apply only for the time-
therefrom will provide more information about the dependent solution. The parameter entered in the
network's performance. $SPECIF list still apply. For example an abbreviated
The time-dependent input data to USU- list of options may be,
NETWK has been made as flexible as pOSSible, while $TDATA DTIME=l,INCHR=l,ALTV=l $END
still making it as simple as possible for you. The best The names that can be included in this options list are
way for you to decide what information, should be given in alphabetical order below.
assembled for a time-dependent analysis is to study the
input allowed, and gain some experience in obtaining Parameters that may be included in $TDATA
time- dependent solutions for simple example problems list of options:
that you make up. You will also need to put more
effort in understanding what the time dependent AL1V: This option permits water surface elevations in
solutions tell you about the performance of the reservoirs to be limited by altitude valve settings at the
network. USU-NETWK provides you tools that assist top of the tank, and the bottom of the tank or not be
in this regard as well. You can have special extra limited, i.e. assume that the tank has no top and
"Lime-dependent" tables created to examine pressures at bottom. If ALTV=O, then the storage elevation data
selected nodes that are most important, flow rates in given for reservoirs will be extrapolated as necessary
selected pipes, and/or water surface elevations in tanks, under the assumption that the reservoir cannot overtop
etc. Also the auxiliary program PRINT (described in or become empty. A time-dependent analysis with
Appendix E) allows you to interactively ask questions ALTV=O, for example, could indicate what elevations
about selected pressures, flow rates, head losses, etc. at the top and bottom of the storage tank should have to
any time. The auxiliary graphics program PLTTIM (see accommodate conditions described in this simulation so
Appendix C) permits you to display the variation of that it will not empty or overfilL If ALTV = 1, then the
pressure, flowrates, water surface elevation in tanks, etc. largest elevation given for a given reservoir's storage
as functions of time. Generally most information elevation curve (entered with the command STORAGe)
related to time variations are prepared before hand, and will be assumed to be the highest water surface
placed in the input data file that USU-NETWK reads elevation allowed. Should the water surface elevation

-82-
attempt to rise above this value, then the altitude valve 4. The next time for a changed increment.
will shut off the flow into the reservoir. Likewise, the 5. The corresponding new time incre-
smallest elevation given in the storage elevation data ment.
will limit the lowest water surface elevation that is 6. etc. with the number of pairs given
allowed. Flow out from this reservoir after the smallest equal to the integer given in # 1 above.
elevation exists will result in the flow from the tank
being shut-off. The value given to ALTV applies to all (C) If either INCHR or INCMIN is negative, and
reservoirs in the network, e.g. it is not possible to have input 1. from (B) above is given a zero value, i.e. no
some tank "topless" and "bottomless" and flows from pair of time and new time increment is given, then after
other tanks controlled by altitude valves. (Default each time-step solution the user will be prompted for
ALTV=I) the new time increment that should be used for the
next time step. These time increments must be given
DTIME: The total time of analysis in days that the in seconds.
time-dependent solution should cover. A value of
DTIME= 1 for example is equivalent to HTIME=24. INCHRP: The time increment, in hours, between
(Default DTIME=l) which the results will be written to a file (printed).
This increment cannot be smaller than INCHR, but
llTlME: The total time of the analysis in hours that could be twice as large as INCHR if the results from
Lhe time-dependent solution should cover. (Default only every other solution are to be written to the
HTIME=24) output file. See the option NPRTAB for a description
of what will, or will not, be written for every other such
INCIIlt: The time increment, in hours, between time increment.
consecutive time-dependent solutions. (Default
INCHR=l) INCMIN: The time increment, in minutes, between
consecutive time-dependent solutions. (Default
It is possible to change the time increment that INCMIN=60)
will be used during different portions of the time- See the description under INCHR for how
dependent solution by giving either this option INCHR variable time increments can be used.
or the option INCMIN a negative value. The effect is
as follows: INCMIP: The time increment, in minutes, between
(A) If INCHR & INCMIN are positive then the which the results will be written to a file (printed).
simulation will use a constant time increment for the This increment cannot be smaller than INCMIN, but
cntire time-dependent solution. could be twice, or three times as large as INCMIN if
(B) If either INCHR<O or INCHM<O, then the the result from every second, or third solution are to be
first time increment will be the absolute value of the printed. See option NPRTAB for a description of what
negative value(s) given these options. In addition at will, or will, not be written on this increment.
least one line of additional input is needed that defines ISUNIT: This options indicates what units are being
when the time increment(s) should be changed, and used to give reservoir storage volumes under command
what the new time increment(s) should be. This added STORAGE.
line must come immediately after the $TDATA list on If ISUNIT
the next line and this added input consists of: =0, storage volumes are in cubic feet.
1. The number of new time increments = 1, storage volumes are in acre-feet.
that will be supplied in the list that follows. =2, storage volumes are in million gallons.
This list is limited to 8 of less pairs. If this =3, storage volumes are in billion gallons.
number is zero, then the list below is not given
but rather after each new time step in the LINEAR: Permits interpolation of demand function
solution the user will be prompted for a new curves, storage function curves, and flow rule curves to
time increment in seconds. This invokes be linear between consecutive pairs of values, or based
option (C) below. on a second degree polynomial passing through 3
2. The time when the new time increment consecutive values. If LINEAR =1 interpolation is
should stan. The units of time used for this linear, if = 0 the interpolation is quadratic. (default=O)
time are those designated by the option
NTIMED. MTIME: The total time of the analYSis in minutes that
3. The new time increment in the units the time-dependent solution should cover. For example
denoted by the option NTIMED. if one wished to cover a time period of 12.5 hours, this

83-
could be done by setting HTIME= 12.5, or by setting solution results printed at all. Therefore, if NTRAND
MTIME=750, or by setting HTIME=12 and is used to control printing frequencies, then don't use
MTIME=30. INCMIP, INCHRP and NPRTAB.

NOPUMF: This option allows you to control whether NPNRES: This same option is permitted in the
the flow rate in a pipe that contains a pumps should $SPECIF list with the same effect as when it is included
shut-off, e.g. be exactly zero, if all of the pumps at the in the $TDATA list of options. The option determines
station in this pipe shut-off. If NOPUMF=I, then the whether a nodal pressure, or reservoir water surface
flow rate is not set to zero when the number of pumps elevation controls pump rules. If NPNRES=O, then
in series or parallel that are in this pipe becomes zero. nodal pressures at selected nodes control the number of
In other words even though the pumps are all shut off pumps on and off in PUMP rules and in FLOW rules.
in a given pipe there can be flow in this pipe. A by If NPNRES=I, then the water surface elevation in
pass line as often exits at a pumping station would be selected reservoirs control these given rules.
modelled by setting NOPUMF=1.
NTIMED: This option allows different units of time to
NPUNOD: Allows pumps and reservoirs to be be used in the demand functions and pump schedules
referenced by three options when giving time-dependent given under the commands DEMAND FUNCTION and
data: If NPUNOD PUMP SCHEDULE. If NTIMED
=0, pumps (including booster pumps) are referenced by =0, time is in minutes.
number, e.g. the order in which they occur in the input =1, time is in hours (default)
data. The numbering of booster pumps occurs after the time is in days.
last source pump. Thus if there are 3 source pumps,
the first booster pump after the command BOOST will PRINTT: This option allows for special (extra) tables
be number 4. to be written to files under the filename
= 1, source pumps and reservoirs are referenced by SPECIAL*.DAT (where * = 10, 11, 12 .. ). (When
node number (allowed only if NODESP=1 in the running USU-NETWK under DOS on a PC this name
$SPECIF list of options). Booster pumps are is shortened by leaving out the L to stay within the 12
referenced by their pipe number. (Default = 1 if characters limit allowed.) These special tables contain
NODESP=I) time as the first column, and flow rates or head losses
=2, pumps and reservoirs are referenced by the pipe in selected pipes, or pressures at selected nodes as
number that connects the source to the network. described below. These tables allow easy examination
Booster pumps are in pipes and therefore are of what happens over time at selected positions within
referenced by pipe number whether NPUNOD=1 or = the network. Graphics described in the graphics
2. Appendix C of this manual allows data from these
special tables to be displayed using the utility program
NPRTAR: This option has meaning in conjunction with PLTIIM. If PRINTI is different from 0, then special
options INCHRP and INCMIP by indicating what tables are requested as described below.
tables these latter options have control over writing at If PRINTI > 0, then it is necessary that the
the indicated frequencies. If NPRTAB =0, then the appropriate extra commands, PIPE TABLE and/or
specified frequencies apply for both the special tables NODE TABLE follow the $TDATA list immediately as
requested by PRINTI>O and the regular tables of out- dictated by the value given PRINTI. If PRINTI
put. If NPRTAB=I, then the specified frequencies = 1, tables of pressure at designated nodes are
apply only to the regular tables of output. For example generated. In this (ese) table(s), and all other special
if NPRTAB=I, INCMIN=15 and INCMIP=30, then tables, the first column contains time. The remaining
any requested data from PRINTI>O would be written columns are pressures at the nodes given under the
to the special tables every time step of 15 minutes, but NODE TABLE special command.
only every other time step, or every 30 minutes would =2, tables of flow rates in designated pipes are
the solution be written to the regular tables of output. generated.
If NPRTAB=2, then INCMIP has no control on the =3, tables of both pressure at designated nodes and
regular tables, but controls the frequency of writing flow rates in designated pipes are generated.
data to the special tables requested by PRINTI>O. =4, tables of head losses in designated pipes are
Note control of printing intervals can also be had with generated.
the option NTRAND in the $SPECIF list of options. =5, tables of pressure at designated nodes and head
Conflicting specifications between these two controls losses in designated pipes are generated.
can result in unpredictable printing of SOlutions or no If PRINTI>O then according to the tables

-84-
 requested the following special commands must occur: RESER. TABLE can be used with any combination of
special tables, i.e. PRINTT can be given any value from
1 through 5. If a special reservoir table(s) is to be
PIPE TABLE created without any other special tables, then PRINTT
should be given the value of 10.
The command PIPE TABLE is used to introduce Example (PRINTT=3)
pipe numbers that are to be included in a special table PIPE TABLE
of flow rates, or head losses, versus time. This com- 15 10 12-16/
mand or one of the following special commands must NODE TABLE
occur immediately after the STDATA options if 5-15/
PRINTT>O. A list of pipe numbers, or the word ALL, RESER. TABLE
follows PIPE TABLE. ALL indicates that all pipes in ALL
the network should be included in the special tables. END TABLES
In order to give PIPE TABLE, the option PRINTT Example (PRINT= 10)
must = 2, 3, 4 or 5. RESER. TABLE
Example 1-10/
PIPE TABLE BOTTOM
1 10-1521/ END TABLES
(see examples 10, 11, 12, 20 and 21)
NODE TABLE
END TABLES
The command NODE TABLE is used to introduce
nodes to include in the special table called for with This special command indicates the end of list of
PRINTT= 1, 3 or 5. A list of node number must follow numbers for special tables. It must be given if and only
the command or the word ALL. if PRINTT>O.
Example
NODE TABLE veOEMA: This options introduces the maximum
4 7 ]5-20/ minor loss coefficient that can exist for a valve or
minor loss devices introduced by commands such as
RESER. TABLE SMINOR, SVALVE, etc. before the flow in the pipe is
shut-off completely. The default value is 1000. This
The special command RESER. Table allows the user loss coefficient, is the multiplier of the velocity head
to designate which reservoirs this (ese) speCial table(s) that produces the head loss due to the minor loss. The
is (are) to be created for in the same manner as PIPE reason for this maximum loss coefficient is to prevent
TABLE or NODE TABLE cause special tables of data. numerical problems should a resistance for a pipe
Depending upon the value of NPUNOD the becomes far out of line with those of other pipes in the
designation of the reservoirs will be by: (0) number, network. If a few resistance coefficient do becomes
(I) node provided NODESP=I, or (2) by pipe number. extremely large, then the solution will be subject to
In the special output table(s) the heading for water truncation errors associated with the Jacobian matrix
surrace elevation or water depths in the reservoir will used in the Newton iterative method of solution. If the
be this reservoir designation. The list designating the loss coefficient exceeds the value of VeOEMA, then
reservoir, which is the second record after the RESER. the same routines are called on to shut-off the flow in
TABLE command can be the word ALL, if all a pipe if it try to reverse in a pipe containing a PR V or
reservoirs are to be included. If the reported values in check valve. During the simulation if the loss
this special table for reservoirs is to be the depth of coefficient becomes less than VeOEMA, and the flow
watcr rather than the water surface elevation, then the in a pipe was previously shut-off, then the flow will be
command BOTTOM (with no data following it) should restored.
be one of the commands used for special tables. If
BOTTOM is given, then the bottom of the reservoir
will be taken as the smallest value provided in the data
Commands that Enter Time
[or this reservoir's STORAGE FUNCTION. Should no Dependent Data
STORAGE FUNCfION be provided for this reservoir,
thcn the water surface elevation will be reported even
if the command BOTTOM is given. The command As with the original data that defines a
network, commands enter data that describe what is to

-85-
 happen with time for a time-dependent solution by 7. Special level and pressure control
= USU-NETWK In the pages that follow there is a algorithms: LCALGO and PCALGO
complete explanation of the nature of the data that 8. Specitying lists of pipe and/or nodes that
each command enters under the command as a heading. should, or should not, be included in
These headings are listed in alphabetical order to make the solution tables: LPIPE, LNODE,
it as easy as possible for you to turn to a given NLPIPE and NLNODE.
command. However, the alphabetical listing has the
disadvantage that commands that accomplish similar The description of the commands understood
functions are not together except for a few cases where by USU-NETWK for entering time dependent data, and
the commands are very similar. Therefore, to give you what this data consists of are described in the following
an overview of what the commands do a numbered list pages. Each command must be spelled correctly to 6
of information that can be provided to USU-NETWK characters.
to control a time dependent solution is given immed- Before describing individual commands let's
iately below. The commands are capitalized. examine a possible small time-dependent simulation.
1. Demands changing, Before describing individual commands let us examine
(a) according to time: DEMAND a possible small time dependent simulation problem
FUNCTIONS that involves the more commonly encountered
(b) according to pressures at commands. As you read this you should look ahead
nodes or water surface elevations in and read the descriptions under the commands that are
reservoirs: FLOW RULES used in providing time dependent data to USU-
2. Volumes stored in reservoirs, or tankS, as NETWK The network problem that will be dealt with
functions of water surface elevations: is Example 1 given in the subsequent section under
STORAGE FUNCflONS ILLUSTRATIVE EXAMPLES, and the original input
3. Changing the number of pumps that are that describes the network will be as given in this latter
operating in series or in parallel at section using the PIPE- command in describing the
pumping stations, basic network. This network consists of 11 pipes, 6
(a) as a function of time: PUMP nodes, two reservoirs, and a source pump. Assume that
SCHEDULES the demand at node 1 in this network remains constant
(b) as a function of the pressure at a at 1 cfs, and that the demands at nodes 2 5 and 6 vary
node, or water surface elevation in a in time such that the given values need to be multiplied
reservoir: PUMP RULES by the factors given in the first table below to define
(c) by changing the rotational speed of these time variations, and that the demands at nodes 3
a pump as a function of time: and 4 vary as given in the other table below. The
ROTSCH variation in demand between table entries is linear.
(d) by changing the rotational speed of
a pump as a rule: ROTRUL Multiplier of Multiplier of
4. Changing differential head devices: DHEAD Demands at Demands at
and HGLSET nodes 2,5 & 6 nodes 3 & 4
5. Changing the flow rate specified in interior
pipes,
(a) as a function of time: SQGIVE Time Peaking Time Peaking
(b) according to a 2nd degree (hrs) Factor (hrs) Factor
equation whose coefficients are given
based on pressures at nodes or water 4 1.2 3 1.3
surface levels in tanks: RQGIVE 8 1.0 6 1.5
6. Change valve setting and/or minor loss 12 0.8 10 1.0
coefficients, 16 0.5 14 0.6
(a) as functions of time: SVALVE, 20 0.6 18 0.5
SVALVC, and SMINOR 24 1.0 24 1.0
(b) as functions of pressure or water
surface elevations in tanks: RVALVE,
RVALVC, RVAL+P and RMINOR. Assume that four pumps exist in series at the pumping
The command DATVAL enters data station, and that for the steady state analysis given in
that provides the loss coefficient for a the example problem # 1 that the given pump curve
valve as a function of its opening. applies for three of these pumps operating. The
operation of the pumps is controlled by the water 1l/
surface elevation in tank (reservoir) attached to the 23920397 314160 402 628320/
12/
network through pipe 11, such that when its water END SIMULATION
surface elevation gets down to 348 ft the fourth pump
is put into operation. The complete operating rule for You should study each line of this input data
this pump station is:
in connection with the descriptions under the individual
Water Surface Elevation, Tank 11 348 352 355 358
No. of pumps operating in series 4 3 2 1 commands so you understand what each value
The tanks are circular with the one attached through communicates to USU-NETWK. Note the following:
pipe 11 with a diameter D = 70.7 ft (a cross-sectional The constant inflow to the larger tank (reservoir) is
area of 3927 ft 2) and the other tank has a diameter D handled by adding an extra node 7 (with a negative
= 282.4 ft (a cross-sectional area of 62832 ft2). The demand of 8 cfs), and pipe 12 with a 1 ft length to the
smaller tank (off pipe 11) has its bottom at elevation at input data description in Example Problem # L
345 ft, and it is 15 feet high. The other larger tank has Special Tables are requested with the option
its bottom at elevation 392 ft, and it'is 10 feet high. A PRINIT=3 in the $TDATA list. Since the PUMP
constant inflow of 8 cfs occurs to this larger tank. The
input for this problem can consist of: RULE is tied to the water surface elevation in tank 11,<