DRAINS Manual February 2004
DRAINS Manual February 2004
DRAINS Manual February 2004
DRAINS
User Manual
February 2004
Watercom Pty Ltd
DRAINS
User Manual
A manual on the DRAINS program
for urban stormwater drainage system
design and analysis
by
Sydney
February 2004
CONTENTS
WELCOME
1. INTRODUCTION TO DRAINS
1.1 Outline 1.1
1.1.1 Basic Description 1.1
1.1.2 Modelling Aspects 1.1
1.1.3 Computer Aspects 1.4
1.1.4 Support 1.4
1.2 Installation 1.5
1.3 Examples of DRAINS in Operation 1.5
1.3.1 Running an ILSAX Model of a Simple Pipe System 1.5
Step (a) Defining Hydrological Models, Rainfall Data and Other Options 1.7
Step (b) Defining the Drainage System 1.12
Step (c) Running the Program 1.21
Step (d) Reviewing Results 1.23
Step (e) Recording Data 1.23
1.3.2 Running the Advanced Design Method 1.24
1.3.3 Running the Rational Method 1.26
1.3.4 Running Storage Routing Models 1.28
1.4 Comments 1.32
6. TECHNICAL REFERENCE
6.1 Introduction 6.1
6.2 The Development of DRAINS 6.1
6.2.1 General 6.1
6.2.2 The UK TRRL Method 6.1
6.2.3 ILLUDAS and ILLUDAS-SA 6.2
6.2.4 ILSAX 6.2
6.2.5 DRAINS 6.3
6.3 ILSAX Hydrology 6.4
6.3.1 General 6.4
6.3.2 The ILSAX Hydrological Model 6.7
6.3.3 Time-Area Routing 6.8
6.3.4 Catchment Surface Types 6.10
6.3.5 Overland Flows and Times of Flow 6.11
6.3.6 Loss Models 6.14
6.3.7 Combination of Hydrographs 6.18
6.4 Rational Method Procedures 6.21
6.5 Storage Routing Model Procedures 6.22
6.6 Pit Inlet Capacities 6.25
6.7 Pipe System Hydraulics 6.34
6.7.1 General 6.34
6.7.2 Pipe Friction Equations 6.35
6.7.3 Pit Pressure Changes 6.36
6.7.4 Tailwater Levels 6.38
6.8 Open Channel System Hydraulics 6.39
6.9 Detention Basin Hydraulics 6.40
6.10 Culvert and Bridge Hydraulics 6.45
6.10.1 General 6.45
6.10.2 Culverts 6.45
6.10.3 Bridges 6.47
6.11 File Formats 6.47
6.11.1 General 6.47
6.11.2 Drawing File Formats 6.47
7. EXAMPLES
7.1 Introduction 7.1
7.2 Examples provided with Demonstration Version of DRAINS
7.2.1 General 7.1
7.2.2 Examples 7.1
7.3 Example from Chapter 9 of the ILSAX Manual 7.10
7.4 Example from Chapter 10 of the ILSAX Manual 7.10
REFERENCES
INDEX
This manual, available in both printed and electronic forms, provides the information you
need to run the DRAINS program for design and analysis of stormwater drainage systems.
Together with the Help system that opens up from within DRAINS, it will lead you to an
understanding of what DRAINS can do, and how to use it to model many kinds of situation.
The data files included in the Manual Example Files folder on the data CD supplied with
DRAINS, which are also obtainable from the www.watercom.com.au website, can be used to
explore the operation of DRAINS.
DRAINS is a technical program with many features, using hydrological and hydraulic
methods developed by generations of engineers. If you have formal training in water
engineering and experience in using models and encountering practical problems, you
should find the program easy to use and to interpret.
If you are a beginner in the fields of hydrology, hydraulics or stormwater system design, or
are out of practice, this manual and the Help system will assist you towards understanding
the program’s operations and outputs.
This manual can be used as a learning guide for DRAINS or as a reference manual that you
can dip into. There is an index at the end, and the electronic pdf version has search and
indexing functions.
Chapter 1 describes what DRAINS is and does, and how it can be installed.
To get you started, it then provides an example that takes you through the steps of entering
data, running the program, and inspecting some of its outputs.
Chapter 2 presents the menus used to control operations in DRAINS, the tools used to
describe the components of drainage systems, and the databases used to store standard
data. To illustrate these, it provides an example of a system involving both pipes and open
channels.
Chapter 4 describes how DRAINS can be used for Design and Analysis tasks. It indicates
how runs can be made and results interpreted effectively.
Chapter 5 describes how DRAINS operates, covering computing and computational aspects.
It discusses the testing that has been used to develop DRAINS.
Chapter 6 provides the technical background to DRAINS and the methods used by it. There
are explanations and references relating to material on rainfall data, overland flows, pit inlet
capacities and pressure change coefficients, detention basins, culverts and bridges.
Since DRAINS is still developing and new features are being added, there will be frequent
revisions of this manual, available electronically and on paper.
Some sections of this manual refer to features in DRAINS that have been deleted or are
obsolete. They are included here to provide guidance when models created by earlier
versions of DRAINS are revisited. The descriptions of such features are set on a grey
background.
DRAINS can model drainage systems of all scales, from very small to very large (up to
10 km2 using ILSAX hydrology and greater using storage routing model hydrology).
It simulates the conversion of rainfall patterns to stormwater runoff hydrographs and routes
these through networks of pipes, channels and streams, integrating:
• design and analysis tasks,
• hydrology and hydraulics,
• closed conduit and open channel systems,
• culverts and bridges,
• stormwater detention systems, and
• large-scale urban and rural catchments.
Within a single package, DRAINS can carry out hydrological modelling using ILSAX, Rational
Method and storage routing models, together with hydraulic modelling of pipes and open
channels and automatic design procedures for piped drainage systems.
It is continuously being improved and expanded with additional functions being added. While
users need to adapt to new features and modes of operation in DRAINS, this steady
development process provides benefits from significant improvements in DRAINS’ modelling
techniques and breadth of coverage.
DRAINS is available in versions for 20, 50 and unlimited numbers of pipes or channels.
The ILSAX hydrology is standard in all versions. Additional Rational Method, storage routing
models and GIS capabilities are available at an extra cost. Details of prices are available
from Bob Stack of Watercom Pty Ltd on (02) 9587 5384 or [email protected].
The ILSAX Type Model, illustrated in Figure 1.1, is the main hydrological model used to
simulate the operation of urban stormwater drainage systems in DRAINS. It comes from the
ILSAX program (O'Loughlin, 1993), which in turn was based on ILLUDAS and the UK TRRL
Methods, as described in Section 6.2. This model uses time-area calculations and Horton
infiltration procedures to calculate flow hydrographs from sub-catchments. The various sub-
catchment flows are combined and routed through a pipe and channel system. Calculations
are performed at specified times after the start of each storm. The time intervals are small,
one minute or less. At each time step, a hydraulic grade line analysis is performed
throughout a drainage network, and flowrates and water levels are determined.
Design of a piped drainage system can be performed automatically by two methods and
results can be checked, viewed and exported as CAD (computer aided drafting) files, GIS
(geographical information system) files and spreadsheet tables.
As well as using ILSAX hydrology, peak flowrate calculations can optionally be performed by
the Rational Method, which has been the traditional method for calculating flowrates for
urban drainage design. Using the formula Q=C.I.A, it converts a statistical rainfall intensity I
to a flowrate Q using a runoff coefficient C and a catchment area A (see Section 6.4). The
Rational Method’s main drawback that it does not calculate flow hydrographs, and it is being
superseded by hydrograph-producing methods. In DRAINS, this method is implemented
using the same procedures as the ILSAX model - only three types of input are different, and
the outputs are very similar.
DRAINS goes beyond normal Rational Method calculations and includes a search procedure
that determines the time duration that gives the greatest value of Q = C.I.A, thus resolving
"partial area" problems.
DRAINS incorporates storage routing models of the type used in the RORB, RAFTS and
WBNM programs that have been used in Australia since the 1970s, and are applicable to
broad-scale rural and urban catchments of virtually any size. As shown in Figure 1.2, they
involve the division of a catchment into sub-catchments based on streams and internal ridge
lines.
Storage routing models treat sub-catchments and stream reaches as storages (similar to
reservoirs or detention basins) that can be modelled by the non-linear equation, S = k.Qm,
where S is the storage in an element, Q is the flow or discharge out of the element, and k
and m are model parameters. These models work downwards through a catchment, adding
runoff from the various sub-catchments and performing routing catchments that reshape the
hydrographs.
The runoff routing modelling facilities in DRAINS can be configured to emulate the RORB,
RAFTS and WBNM modelling structures. They can also be combined with ILSAX sub-
catchments and open channel hydraulic calculations, so that quite diverse flooding and urban
drainage systems can be described.
DRAINS follows Microsoft Windows conventions and PC users will find the standard main
window, menu bar, toolbar and Help system easy to navigate.
There is also a large choice of outputs – screen print-outs, CAD, GIS and spreadsheet files.
In most cases, revised data can be transferred back to the originating programs using these
files.
DRAINS is based mainly on two programs - the ILSAX model developed by Dr. Geoffrey
O’Loughlin and the PIPES programs created by Bob Stack. ILSAX is a MS-DOS design and
analysis program for urban stormwater drainage systems developed from the pioneering
TRRL and ILLUDAS programs. This was released in 1986 and attracted a large user-base in
Australia and New Zealand. PIPES and PIPES++ are water supply system analysis
programs that are also marketed by Watercom Pty Ltd. The graphical user interface from
these programs has been used in DRAINS.
Installation and updating of DRAINS is quick and easy using a self-extracting file that can be
supplied on CD-ROM, or downloaded from the website www.watercom.com.au. The file
installs the latest release version of DRAINS, which also operates as a demonstration
program with a limit of five pipes or channels, and restrictions on changes to detention basins
and culverts.
When installed on a PC using Microsoft Windows, the program resides in the folder
C:\Program Files\Drains\Program. At present, the Drains.exe file is about 2.3 Mb
in size, and is accompanied by a HTML Help file of 1.9 Mb. To run with capabilities beyond
those of the demonstration version, a hardware lock or dongle must be inserted in the 25 pin
printer port of the PC on which it is being used, or in a USB port. DRAINS can be installed
and run on any PC or server system to which a hardware lock has been attached. Watercom
currently sells DRAINS in versions for 20, 50 and unlimited conduits.
1.1.4 Support
For DRAINS support, Watercom Pty Ltd can be reached at phone/fax (02) 9587 5384 or
[email protected]. Training workshops are conducted by Dr. Geoffrey O’Loughlin,
who can be contacted on phone (02) 9570 6119 or 0438 280 477, fax (02) 9570 6111 and
[email protected].
Training workshops on DRAINS are held frequently. These are notified to purchasers by a
DRAINS e-mail newsletter.
If you are setting up the demonstration version, or updating a DRAINS program that is
already installed, you only need the file drains_sef.exe, which can be downloaded from
the website www.watercom.com.au or obtained on CD.
From Windows, double-click your mouse on the icon for drains_sef.exe and wait for the
instructions to appear. When prompted, enter the password provided by Watercom Pty Ltd,
or enter “DEMO” to install the demonstration version. Then follow the instructions. You must
acknowledge the Conditions of Use.
For the first installation of DRAINS on a PC that is to be used with a hardware lock, you must
install from the CD-ROM provided when DRAINS is purchased. This installs a driver for the
hardware lock as well as DRAINS. Some drivers can also be downloaded from
www.watercom.com.au.
Installations can be made on any number of PCs. The hardware lock controls where a full
version of DRAINS can be used. Locks are programmed to model certain maximum sizes of
drainage network, 20, 50 and unlimited pipes, with ILSAX hydrology alone or with
combinations of ILSAX, Rational Method and storage routing model hydrology, and with the
ability to import and export data from GIS files. The lock also controls whether the advanced
design method can be used. Locks can be upgraded by obtaining a code number from
Watercom Pty Ltd.
DRAINS can be uninstalled using the Add/Remove Programs application in the Windows
Control Panel.
The version number of a DRAINS program in use can be found by opening the Help menu
and clicking About DRAINS…. The Upgrade function is also in this menu.
This relatively simple example illustrates how a pipe system, assumed to be located at
Orange, New South Wales, can be set up in DRAINS and design and analysis runs can be
made. You can follow the information below to set up a model using the demonstration or
full versions of DRAINS, or view this in the files Orange1.drn and Orange2.drn in the set
of examples accompanying this manual.
DRAINS is a menu-operated program, and you can perform work in many different
sequences. However, there is a basic pattern that you will follow in most situations, namely:
• to establish a datafile with a .drn suffix, containing a description of a drainage system,
the rainfall data from which the flows through the system are created and various design
parameters and specifications.
• then to run the program and review the results that are displayed, mainly in the form of
flowrates and water levels,
• and to export or copy results in forms that can be used to document outcomes and to
prepare design plans.
Some server systems will not add DRAINS to the Programs list in the Start menu.
In this case, a shortcut should be created on the desktop if DRAINS is to be used frequently.
When opened, the Main Window of DRAINS appears as shown in Figure 1.3.
Figure 1.3 The Main Drains Window with the Project Menu Selected
You can define a drainage system graphically on this blank screen by drawing components
such as pits and pipes, using the facilities in the menus and toolbar located at the top.
To operate DRAINS, you can enter data directly using the keyboard and mouse, or open or
import an existing file from the File menu. If you are entering an entirely new system, you
must follow the steps shown in the box below, which are explained in detail afterwards.
Step (a) Defining Hydrological Models, Rainfall Data and Other Options
Although you can start by drawing a system using the Toolbar tools, the hydrological model,
rainfall patterns and component data bases should be established first. This information can
be changed later, but there can be difficulties, and it is best practice to define it first.
Go to the Project menu at the top of the screen and select Hydrological Models…. The
window shown in Figure 1.4 appears, containing a dialog box that enables a process or
choice to be implemented.
Click the Add ILSAX Model button to open the window shown in Figure 1.5. This is a
property sheet in which the characteristics of some data base item or component can be
entered.
Enter the numbers shown. These will be explained later, but if you wish to see information
on them immediately, press the Help button, which will open the Help screen shown in
Figure 1.6.
Figure 1.6 Help Window opened from the ILSAX Hydrological Model Property Sheet
You should then click OK in the property sheet and the Hydrological Model Specification
dialog box, ensuring that "Orange Soils" is defined as the default model in the drop-down list
box in Figure 1.4.
You must now define the rainfall patterns to be used, using the Rainfall Data… option in the
Project menu. This opens the window shown in Figure 1.7, in which you can set up a data
base of rainfall patterns or hyetographs.
In this case, two patterns will be set up, both of 25 minutes duration, for average recurrence
intervals (ARIs) of 2 years and 100 years. These correspond to average intensities of 40.2
and 101 mm/h.
For most design tasks in Australia, the rainfall data required will come from the Institution of
Engineers Australia publication Australian Rainfall and Runoff (1987). These can be defined
easily in DRAINS by clicking the Add an ARR97 Storm button, which opens the dialog box
shown in Figure 1.8.
Figure 1.9 2 Year ARI, 25 Minute Rainfall Pattern for Orange, NSW
Next, change the antecedent moisture condition value in the Rainfall property sheet to 2.5.
Set up a second pattern, this time for an ARI of 100 years and an intensity of 101 mm/h.
Then click the OK button.
The next step is to select storms from the data base to be used for design and for analysis.
This is done using the Select Storms ► option in the Project menu. This opens a small
pop-up menu in which you can choose between Major Storms and Minor Storms. Choose
the last of these and the dialog box shown in Figure 1.10 will appear. Click the Selected
storms button in the top left corner and then click the downwards arrow on the first drop
down list box to show the names of the rainfall patterns in the data base. Click on the 2 year
ARI, 25 minute pattern. This action selects this storm as the one to be used to design the
pipe system.
Figure 1.10 Select Minor Storms for Design Runs Dialog Box
Close this dialog box by clicking OK, and then follow the same procedure to select the 100
year ARI, 25 minute storm for major storm analysis runs.
With this model you will be using New South Wales pits. To select these, choose the
Default Data Base option in the Project menu. The dialog box shown in Figure 1.12
appears, allowing you to select the data base that is to be used. (Note that this must be
done before any pits are entered into the Main Window. The selected data base stays in the
.drn file for each model. New pit, pipe and overflow route types can be added, but it is not
possible to remove pipe and pit specifications.)
Figure 1.12 Dialog Box for selecting a Data Base containing Pipe, Pit and Overflow Route
Specifications
At this point, you should save the file using the Save option in the File menu, or the diskette
symbol on the Toolbar, naming the file Orange1.drn.
The above processes create a template or shell, containing the base information that will be
used to run the model, and the pipe, pit and overflow route types to be referred to when
setting up drainage systems. The DRAINS .drn file can be saved at this stage, to be used
later as a starting point for models that use this base information.
This system can be drawn in the Main Window using five of the tools from the Toolbar:
If you click one of the Toolbar icons, the cursor will change to a pencil, which can be used to
place that component in the Main Window. You can use the pit and node tools, and ,
to draw five drainage pits and an outlet node in the Main Window, as shown in Figure 1.14.
As shown in Figure 1.15, pipes can be drawn between these by selecting the pipe tool
and clicking on the beginning and end points to locate each pipe. Overflow routes made up
of poly-lines can be added using the corresponding tool . Sub-catchments can be added
using the sub-catchment tool , making sure that they touch the pits.
The names of components (mostly given as ???? to start with) can be dragged to more
convenient locations. The components themselves can be moved round the screen. You
can select a component by clicking it to make “handles” appear, holding the left mouse
button down on it so that horizontal and vertical arrows appear, and then dragging the
component to the required location. A pipe or channel can be moved as a single unit by
dragging near its centre. Alternatively, their ends can be moved by dragging the handles.
As well as entering data onto a blank screen, as shown above, you can insert a background
from a CAD drawing file, together with a layout of pits and pipes. A drawing file for the
current example, Orange Base.dwg, is shown in Figure 1.16.
This can be entered into a DRAINS Main Window, after the hydrological, rainfall and options
settings have been established, using the Import DXF or DWG File… option in the File
menu, shown in Figure 1.17. This takes you through a set of dialog boxes in which you must
nominate layers containing data on the background, pipes (as lines) and pits (as circles), and
other information. The first of these is also shown in Figure 1.17.
Figure 1.17 Menu and Dialog Box for Nominating CAD Layers
The pipe system appears as shown in Figure 1.18. This version should be saved in file
Orange2.drn.
This view can be enlarged (using the magnifying tool or mouse wheel) and all pipe lengths
can be scaled by setting the length of one pipe. The other components of the pipe system,
such as sub-catchments, overland flow paths and outlets, can be added in the manner
described earlier in this section.
Data can be entered and edited using property sheets that appear when you right-click a
component, and select Edit Data from the pop-up menu, as shown in Figure 1.19. The
property sheet for Pit 1 is shown in Figure 1.20.
From Figure 1.13, you can see that any overflows from Pit 1 will flow to Pit 2, so that this first
pit should be selected as an on-grade pit. Note how pit types and sizes are selected from a
data base of pit types using two drop-down list boxes. (When entering data, you can move
from box to box using the Tab key on your PC’s keyboard, or you can use the mouse.)
The pit pressure change coefficient value of 4.5 is suitable for a pit at the top of a drainage
line. You can obtain further information about these factors, which influence the water levels
in the pipe system, from the Help system or from Section 6.7.3.
After closing the Pit property sheet, you will find that the pit name has changed and the
question marks have disappeared. As data is entered, the allocated names appear in the
Main Window. DRAINS will not run until all the required data is entered. Even if the data for
all components are entered, question marks will remain if the connections between
components are incomplete.
Table 1.1 defines the data for the pits in this system. If you are following this example, enter
the appropriate values for the five pits. The names of pits and other components can be up
to 10 characters long. Assume all pits to be NSW RTA (Roads and Traffic Authority) SA2
pits, at the slopes shown in the table.
Pit 2 is a sag pit, in a hollow in which stormwater can form a pond over the pit. For this type
of pit, two additional pieces of information have to be provided - a maximum ponded volume,
and an allowable ponding depth, here taken to be 5 m3 and 0.2 m.
Next, you can enter the data for sub-catchments using their property sheets opened from the
pop-up menu in the same way as for pits. The form of the property sheet is shown in Figure
1.21. In this example, the simplest form of data is used. The user must divide the sub-
catchment draining to each pit into three types of land use:
• paved area (impervious areas directly connected to the drainage system),
• supplementary areas (impervious areas not directly connected to the drainage system),
and
• grassed areas (pervious areas).
A time of entry is assigned to each land-use. This is the time that it takes for stormwater to
flow from the furthest boundary of each type to the point nearest to the pit. The
supplementary area drains onto the pervious area. If the grassed area does not extend to
the pit, a lag time is specified to account for the time taken for this grassed area runoff to
pass over the section of paved area near the pit. Times are calculated from the technical
relationships presented in Section 6.3.5.
The parameters for all sub-catchments are shown in Table 1.2. After entering values for
Catchment Cat 1, you can enter parameters for the four other sub-catchments. The names
of all should change from “?????” to the names you provide - if not, check that the icon for
each sub-catchment touches that for the related pit.
Now you can enter data for pipes, opening the property sheet for each by right-clicking on
the pipe. (Be sure to click on the object and not on its name, which will bring up a dialog box
in Figure 3.24 for customising or changing the information presented in the Main Window.)
The sheet for the first pipe is shown in Figure 1.22.
Since the pipes here are to be designed, it is not necessary to specify invert levels or slopes;
DRAINS can calculate these during the design. You must, however, specify the pipe name,
length and type, selecting the type from a list box. (If you have inserted a background, pipe
lengths can be automatically scaled after the first length is entered.) When you close the
property sheet, you will find that the pipe name is prefixed with “??", indicating that the data
are still incomplete.
Pit Total Paved Supp. Grass. Paved Supp. Grass. Lag Time
Name or Area Area Area Area Time Time Time or
Node (ha) % % % (mins) (mins) (mins) Factor
Cat 4 Pit 4 0.35 75 5 20 9 1 15 0
Cat 5 Pit 5 0.02 90 0 10 2 0 3 0
Cat 1 Pit 1 0.125 28 5 67 8 1 13 0
Cat 2 Pit 2 0.231 33 5 62 9 1 15 0
Cat 3 Pit 3 0.025 90 0 10 3 0 4 0
You can now complete the entry of the remaining pipe data, shown in Table 1.3.
The next components to be defined are the overflow routes. You must define a name and an
estimated time of flow, as shown in Figure 1.23, and you must also define overflow route
cross-sections from the overflow route data base, with slopes and percentages of
downstream areas, as shown in Figure 1.24. The information for this example is shown in
Table 1.4.
Figure 1.23 The Overflow Route Data Property Sheet - Page 1 (Top Portion)
The percentages of downstream areas are used to define flow characteristics along the
overflow route, as explained in Section 2.3.6.
Finally, the outlet names and levels must be defined. Here we will assume that the main
outlet operates as a free outfall. The tailwater level for the pipe system will be the higher of
the normal and critical depths for the outlet pipe, unless this is running full. Outlet surface
levels have been given in Table 1.1.
You can now run the program in Design mode using options in the Run menu. In this
example, the standard design method is employed by clicking on the Standard Design
(pipes only) option. (Analysis runs are not possible until the pipe invert levels have been
specified, so that the Analysis options in the menu are greyed out. If the Advanced design
method is enabled, this can be run using the Advanced Design (pits and pipes) option.
When a Standard Design run starts, you are given a warning that the file will be changed.
The run then proceeds and a report is displayed, as shown in Figure 1.26. After you close
this window you will see that the names of components have changed to coloured numbers
as follows:
• the black numbers are the maximum flowrates from the sub-catchments, in m3/s,
• the blue numbers are the greatest flowrates in each pipe, in m3/s,
• the red numbers are the greatest overflows from pits, in m3/s,
• the green numbers are the highest levels reached by the hydraulic grade lines (HGLs)
throughout the pipe system, in m elevation, defining the highest water levels during the 2
year ARI, 25 minute storm event considered.
DRAINS calculates hydrographs or time series of runoff flowrates from the rainfall
hyetographs, so it is possible to view what happens at all times during the storm event, as
shown in Figure 1.27.
You can now inspect the results and check the pipe inverts and sizes determined by
DRAINS. There are a number of ways of doing this, the most useful probably being the
transfer of information to a spreadsheet, using options within the Edit menu. The data
spreadsheet for the Orange Example is shown in Figure 1.28. There is also one or more
corresponding results spreadsheets.
The major/minor system is usually employed in Australian drainage design. Pipes are sized
to carry flows of a minor ARI, from 2 to 10 years, and a check is made to ensure the safe
working of the system during a major storm event, with an ARI or about 100 years. So far,
DRAINS has performed the design, determining pipe sizes and invert levels. You can now
also perform an analysis using the 100 year ARI, 25 minute rainfall pattern. Simply run
Analyse major storms from the Run menu to produce the results shown in Figure 1.29.
The flowrates are now larger and the overflows are more significant. Many of the
hydrographs have lost their peaked shape and have a flat or hollow top. The HGL plots
indicate that water levels in the pits may have reached the surface and overflows occurred.
This last step involves the storage of results. The input data is all stored in the DRAINS
datafile Orange2.drn. There is plenty of opportunity to make comments,
in the spaces provided in the property sheets for individual components, and in a
Description … option in the Project menu.
The spreadsheet results can also be stored, and, as will be shown in Chapter 3, it is also
possible to transfer the results via a DXF file to drawing programs that can print plans and
longitudinal cross-sections of pipe systems.
While the advanced method is more complex in its calculations, the procedure is simple in
DRAINS, involving the following steps, all of which can be performed with the Orange2.drn
example if the Advanced Design Method is enabled with your hardware lock. The first step is
to choose the Advanced Design (pits and pipes) option in the Run menu. This runs and a
message appears advising that the process is complete and that you should analyse with
minor or major storms to assess the results. You can do this my choosing the Analyse
minor storms and Analyse major storms options.
Results for the Design run are shown in Figures 1.30 and 1.30. The final diameters and pit
inlet sizes will probably differ from those specified in the inlet data. This is derived from an
Analysis run with major storms, performed after the Design run.
Although this process may appear to be similar to the Standard Design procedure, it involves
a greater amount of calculation to establish pipe sizes and to determine inlet types in the
Design run. This is done on the basis of the allowable flow along the overflow route. The
method determines this flowrate, taking into account the effects of the sub-catchment
immediately downstream of each pit. It then works backwards to define a set of pit inlets and
pipe sizes that will limit overland flows to safe levels in both minor and major storms. Safety
requirements are defined in terms of flow depths and velocity-depth profiles in the Overflow
Route Data Base.
To illustrate the Rational Method procedure, the same Orange system can be modelled using
a Rational Method hydrological model in the file named Orange2Rat.drn. You can run this
if your hardware lock is enabled to run the Rational Method. Only three of the property
sheets for data entry differ from those for the ILSAX hydrological model. In Step (a), the
setting up of project options, the Hydrological Model for a Rational Method called from the
dialog box in Figure 1.4 takes the form shown in Figure 1.32. There are three choices on the
form of the Rational Method procedure to be used. The version from Australian Rainfall and
Runoff, 1987 is selected, as shown in Figure 1.33.
Figure 1.32 Hydrological Model Property Sheet for the Rational Method
The runoff coefficient for the impervious area is set at 0.26 using a formula in Section 14.5.5
of Australian Rainfall and Runoff, 1987, based on a 10 year ARI, 1 hour rainfall intensity of
37.2 mm/h.
Into this you can enter the minor and major ARIs you require and the corresponding design
rainfall intensities for durations of 6 minutes, 1 hour, 12 hours and 72 hours. This information
will be available from Australian Rainfall and Runoff, 1987 or from a council’s drainage
design manuals or codes. The ARR Wizard button allows this information to be obtained
from nine factors which can be found from Volume 2 of Australian Rainfall and Runoff, 1987.
These are entered in the property sheet shown in Figure 1.35.
Figure 1.35 Factors from Australian Rainfall and Runoff, 1987, Volume 2
The intensities are calculated from this information, and are used to establish the full
intensity-frequency-drainage relationship used by DRAINS for Rational Method calculations.
The remaining property sheet that is different is for sub-catchments, shown in Figure 1.36.
The Rational Method does not distinguish between directly-connected and non-directly-
connected impervious areas. The paved and supplementary area percentages are added to
produce a percentage impervious, and the ILSAX model grassed area becomes the pervious
area. The data that needs to be entered for sub-catchments is presented in Table 1.5.
In Step (b), the other data, for pits, pipes overflow routes and nodes, is entered in the same
way as for runs using the ILSAX hydrological model.
In Step (c), when a Design run is made, the results appear similar to those obtained with the
ILSAX hydrology, as shown in Figure 1.37.
A difference is that only peak flows are generated, rather than a full hydrograph, as shown in
Figures 1.27 and 1.29, so that detention storages cannot be modelled using the Rational
Method.
Steps (d) and (e) proceed in the same manner as with the ILSAX hydrology. An Analysis run
can be made to determine major storm effects.
Storage routing models can be implemented using many of the same features and processes
used with the ILSAX and Rational Method programs. To illustrate this, consider the RAFTS
Model shown in Figure 1.38, modelling a hypothetical creek at Shepparton, Victoria.
This rural catchment has been divided into four sub-areas, and a RAFTS model has been
superimposed on this. The four sub-catchments shown by the symbol are sites where
conversions from rainfall to runoff and routing occur. Routing can also occur, if required, in
the dashed stream routing reaches.
Property sheets for the sub-catchments and the stream routing reaches are shown in Figures
1.40 and 1.41. The stream reach sheet offers a choice of translation of the hydrograph
(movement of flows without changing the hydrograph shape) or an approximate routing
procedure based on kinematic wave hydraulic principles.
A name must be entered for nodes, but surface levels are not required, as the routing is not
tied to particular elevations or datum levels.
The results from a major storm run involving four storms is shown in Figure 1.42.
The black numbers at the sub-catchments represent the routed sub-catchment flows, while
the pairs of red numbers represent the peak flowrates at the top and bottom ends of a reach.
Hydrographs can be examined easily, as shown in Figure 1.44, and data cab be transferred
to a spreadsheet.
1.4 Comments
The examples in the previous section have taken you through the basic workings of DRAINS,
showing how it can be used for design and analysis of new pipe systems. Many matters still
need to be explained - what other features are available, how the program works, and the
details of the factors used in the program. Most of these are explained in the on-line Help
system, and in the following chapters of this manual.
2.2 Menus
DRAINS employs seven drop-down menus that appear when you click the items in the menu
bar:
Where an item in a menu list is followed by “…” or “ ►” it opens another menu, or a dialog
box or property sheet.
2.3.1 General
The remaining thirteen buttons can be used to draw components in drainage systems in the
Main Window. The first group of six are all nodes or junctions; the next group of six are links,
and the remaining sub-catchment button provides a source of water. If you hold your mouse
arrow over each button, a ScreenTip will appear to indicate its function.
Clicking on these buttons changes the cursor from an arrow to a pencil, which is used to
place components in the Main Window. Holding down the Shift key while entering a
component retains the pencil cursor after you have entered a component, allowing you to
add another component of the same type. If you become “stuck”, with the cursor still in
pencil form when you no longer want to enter a component, simply enter it and then delete it.
It is important that the components should connect properly. Links should be placed near the
centre of nodes, and sub-catchments should clearly connect to pits and nodes. The layout
should be tidy, to enable components to be viewed and accessed easily. Names and
positions of components can be shifted to clarify the layout.
A background layer imported as a DXF or DWG file from a CAD program, as described in
Section 3.2.2, or as part of the importation of GIS files (Section 3.2.4) provides a guide for
locating pits and other components
Lying behind each of the components provided in DRAINS is a computer algorithm (logic +
equations + data) that is employed within calculation frameworks. There are often alternative
ways to describe the operations of components such as pits or detention basins. When
performing analysis work with DRAINS, you should assess the methods and equations used
in the program (detailed in Chapter 6) and examine results, to confirm that the program
operates as you expect it will.
You may encounter situations that are not fully described by the standard components, such
as a complex system of detention basins. It will then be up to your judgement and modelling
skills to use the available tools to fit the situation. This may involve “tweaking” of the model.
An example of this type of manipulation is the use of detention basins to simulate stormwater
infiltration systems or pumps.
The following sections describe the features of each component, starting with those that you
are likely to use most frequently.
2.3.2 Pits
Pits, like other forms of node, act as entry points for water into the pipe system.
They can represent a street gully pit, a manhole, a junction, a flow diversion or other
components of drainage system. They also define the framework of a pipe system, and most
users will draw pits and nodes in first when creating a system diagram.
In April 2001, DRAINS changed the method of specifying pits, to accommodate the
Advanced Design Method. Since it is still possible to use files created with the previous
method, both methods are covered here, with the obsolete older method in a section marked
in grey.
The Drainage Pit property sheet can take three forms, depending on the type of pit selected.
If an ILLUDAS type pit is specified, the box appears as shown in Figure 2.2.
The pressure change allowance for part-full flows may be adjusted if required, or the default
choices may be accepted. Pit pressure changes are explained in Section 6.7.3. Typical
values are presented in Table 2.1.
The ILLUDAS pit type is derived from ILLUDAS, one of the programs on which DRAINS is
based (see Section 6.2). It accepts all inflows that come to it, so there is no provision for
overflows. If the pipe system downstream becomes pressurised and the hydraulic grade line
rises to the surface, water is stored at the pit, and then released back into the pipe system
when flows diminish.
Type of Pit ku
Pit at the top of a line 5.0
Pit with a straight through flow, 0.2
no sidelines
Pit with a right angle direction change, no 1.0
sidelines
Pit with a straight through flow, 1.5
one or more sidelines
Pit with a right angle direction change, 2.0
one or more sidelines
This type of pit should be used with care, since it may not realistically model a situation.
Most stormwater pits can overflow, and they should be modelled as on-grade or sag pits.
On-grade pits are located on slopes, while sag pits are in hollows or depressions, as shown
in Figure 2.3.
When stormwater runoff reaches an on-grade pit, at smaller flowrates all flows will be
accepted. As the approach flowrates increase, a point is reached where some bypass flow
will occur. This will flow away from the pit, perhaps to another pit downstream, with
additional flows joining it along the way. To model on-grade pits, a relationship between the
approach flow and the flow captured by the pit must be specified. These cannot be
established by theory and are usually determined from modelling studies or tests on installed
pits. This is discussed further in Section 6.6.
In the new DRAINS pit formulation, this inlet capacity information must be set up in advance
in the Pit Data Base, as described in Section 2.4.6. This is then referenced in the Pit
property sheet shown in Figure 2.4. Even if a pit type is a “one-off”, a pit type must be
established for it in the Pit Data Base. The pit can be excluded from the advanced design
process using the buttons at the bottom left of the property sheet.
On this property sheet there is also a check box with the label “Pit has bolt-down lid” that
allows pits to be sealed, in which case the HGL may rise above the surface.
A sealed pit cannot accept flows at the surface, and cannot overflow. In the sheet there is
also provision for blocking factors, default values of which can be set in the Options property
sheet opened from the Project menu as shown in Figure 1.11. The inlet capacity calculated
from the relationship obtained from the Pit Data Base is multiplied by 1 minus the blocking
factor. Thus a factor of 0.2 will reduce the inlet capacity or capture rate by 20%. More
restrictive blocking factors are usually applied for sag pits than for on-grade pits. Values of
0.5 for sag pits and 0.2 for on-grade pits are typical.
For sag pits, the Drainage Pit property sheet appears as shown in Figure 2.5, with additional
ponded volume and depth data required for water that might form a pool over the pit. Again,
the pit type is defined by referring to the Data Base. It is also necessary to provide
information on the ponding area over the pit. The maximum ponded depth and the volume of
ponded water corresponding to this must be specified. Further information on technical
aspects of pit inlets is given in Section 6.6.
Versions of DRAINS released prior to April 2001 described pit inlet capacities using the
equations described in Section 6.6, while the new procedure define relationships as a series
of points in a table. DRAINS can still open older files and run with the superseded
relationships, but the old and new methods cannot be mixed.
Figure 2.6 shows the older property sheet for an ILLUDAS type pit:
Figure 2.6 The Old Drainage Pit Property Sheet with ILLUDAS Type Pit Selected
Figure 2.7 The Old Drainage Pit Property Sheet for an On-Grade Pit
If a pit type is chosen from the Data Base, the selected type will be shown in the list box and
the factors will be shown with a grey background.
For a sag pit, the inlet capacity relationship requires three factors named VCAP1, VCAP2
and VCAP3 plus a ponding volume. These factors define a general fitting equation for inlet
flow rates. They can be developed from theory or from laboratory or field measurements.
Figure 2.8 shows where they are entered into the Pit Property sheet with a sag pit selected.
The factors can be:
(a) entered directly into the boxes,
(b) selected by choosing a pit type from the list box in the Pit Property sheet, or
(c) defined using a wizard that determines factors from the geometry of the depression
(the slopes of two gutters leading to a depression and the cross-fall slope of the road)
and the perimeter of the pit.
For the wizard to operate, the pit type selected must be “You specify inlet parameters”.
Clicking the Inlet Capacity Wizard button opens the dialog box shown in Figure 2.9.
Further information on the equations defining pit entry capacities is provided in Section 6.6.
It is possible to update files that use this older system to the new procedure by attempting to
run the file using the Advanced Design method. A message appears asking whether you
wish to change to the procedure using pit families and sizes. If you agree to this, you are
asked to nominate a default .db1 pit data file from those stored in the c:\Program
Files\Drains\Program folder.
As well as receiving surface flows, a pit can receive a constant baseflow or a user-provided
inflow hydrograph, specified using the buttons at the top of the Drainage Pit property sheet.
These can be introduced at the surface or inside the pit. Flows introduced inside the pit are
not subject to the pit inlet capacity relationship.
When the Baseflow… button is clicked, the property sheet shown in Figure 2.10 appears.
Only a single flowrate needs to be entered.
When the Inflow Hydrograph… button is clicked, the property sheet shown in Figure 2.11 is
opened. A number of hydrograph ordinates, or flowrates at particular times are to be
entered. Entry is in the text boxes labelled “Time (mins)” and “Flow (cu.m/s)”. The graph
aids the entry of data, and specific ordinates can be located and altered using the arrows in
the spin box associated with the times.
The type of node called a simple node can be used for several purposes:
• to represent an outlet,
• to act as a node in an open channel drainage system,
• to provide a junction for stream reaches in a storage routing model, and
• to act as a closed junction in a pipe system.
DRAINS detects whether a node is at an outlet to a system, and if so, it presents the property
sheet shown in Figure 2.12.
As explained in Chapter 5, for part-full pipe flows DRAINS projects hydraulic grade lines
upward through a drainage system. If a free outfall is specified, the starting point for this
upwards projection at each time step is the higher of the pipe’s normal and critical depths for
the current flowrate. If a tailwater level higher than these depths is specified in the Outlet
Node property sheet, this becomes the starting level.
If the outlet node connects to a stream routing reach used in a storage routing model, only
the node name is required. Figure 2.12 will appear for all nodes. Intermediate nodes in pipe
or open channel systems appear as shown in Figure 2.13. For an open channel, the surface
level corresponds to a high bank level.
2.3.4 Pipes
The Pipe property sheet shown in Figure 2.14 requires, as a minimum, that you enter a
name, length and number of parallel pipes (default value 1), and specify a pipe type from the
drop-down list box. This information prepares the pipe for a Design run, in which DRAINS
will specify the pipe diameter and invert levels. The pipe type chosen must be defined
beforehand in the Pipe Data Base located under the Project menu options. Rectangular
pipes and dummy pipes can be used, though not for design, and minimum pipe diameters
can be set for design as described in Section 2.4.5.
The Survey Data… button at the bottom of the sheet opens the Survey Data property sheet
shown in Figure 2.15. Levels can be entered along the line of the pipe, so that the design
procedure can allow for minimum cover all along the pipe, and intermediate points can be
plotted in a long-section drawing, as described in Section 3.5.5. It also allows the positions
of other services to be defined so that DRAINS can avoid these (allowing for a vertical
clearance defined in the Options… property sheet in the Project menu.
Figure 2.15 Survey Data Property Sheet for Defining Intermediate Levels
along a Pipe Line and Positions of Services
In Design runs, DRAINS tries to locate pipes between services, going under them if no other
route is possible. If this is unacceptable, the designer can selectively remove services, or
make manual adjustments to the pipe cross-section and/or alignment.
The long section display in Figure 2.16 shows how the ground levels and service positions
appear after a Design run is carried out, using the Long Section option in the pop-up menu
for the pipe.
In some cases, a non-return device such as a flap gate may be installed in a pipe, preventing
flows from moving upstream. This can be modelled by ticking the Include Non Return
Valve box in the Pipe property sheet.
2.3.5 Sub-Catchments
The form of the property sheet for a sub-catchment depends on the hydrological model
defined in the Hydrological Models… option in the Project menu, as shown in Figures 1.3
and 1.4. If an ILSAX type model is chosen, the sub-catchment can be divided into the three
land-use types illustrated in Figure 2.17:
• paved area (impervious areas directly connected to the drainage system),
• supplementary areas (impervious areas not directly connected to the drainage system),
and
• grassed areas (pervious areas).
The supplementary area models impervious surfaces from that drain onto pervious or
grassed areas, where the runoff might be absorbed into the soil. These could be sheds,
swimming pools, parking lots and other impervious areas that do not drain through pipes and
on impervious surfaces to the sub-catchment outlet.
The full form of the Sub-Catchment property sheet for the ILSAX Model is shown in Figure
2.18, with the “comprehensive data” option chosen in the check boxes labelled Use. Figure
1.19 in the previous chapter displayed the “abbreviated data” option.
In both, you must enter the total area in ha, and the percentages of the three land-uses.
Figure 2.18 requires additional information to establish times of entry using different flow-
path components, applying the kinematic wave equation described in Section 6.3.5.
For the paved part of the sub-catchment, there are three sets of inputs. The first is a
constant time, the second a set of values for flow path length, slope and the third a set of
gutter length, slope and a factor named GUT.
For example, in Figure 2.19 a flow path consisting of three sections is shown. For paved
areas within this sub-catchment, the time of entry would be the sum of:
• a constant time, for the path from the roof of the furthest buildings to its property
boundary,
• an overland flow time, to be calculated from the overland flow path length, slope and
Manning’s roughness, using the kinematic wave equation, and
• a gutter flow time, calculated using the length, slope and the GUT factor that includes
information on the gutter cross-sections and roughness.
Other sub-catchments may involve only a constant time, so that a single section is all that is
required, and the “abbreviated data” form can be used. Some may require the overland flow
section alone, in which case the “more detailed data” option is selected, and the constant
time is set to zero. A time of entry for a short property drainage path draining directly to a
gutter could be modelled with the “comprehensive data option” and a constant time, an
overland path with zero length and a gutter section with appropriate data. In, all there are
seven ways of defining flow paths that provide considerable flexibility.
For the supplementary and grassed areas, the times of entry are obtained by adding two
components:
If only the first of these is required, the flow path length can be made zero; if only the second
is required, the constant time can be set as zero.
A lag time is specified if the “abbreviated data” and “more detailed data” options are
selected. If the “comprehensive data” option is selected a lag factor is employed. This
defines the lag as a proportion of the calculated paved area (overland + gutter flow time).
The property sheet has a very similar format to the ILSAX model Sub-Catchment property
sheet, the main difference being that sub-catchments are being divided into pervious and
impervious areas, instead of paved, supplementary and grassed. An example is shown in
Figure 2.21.
The sheet similar for the three available types of Rational Method model (General, Australian
Rainfall and Runoff, 1987, and Standards Australia AS 3500.3.2). The only difference is the
need to enter roofed percentages for the AS 3500.3.2 method.
The three types of storage routing model described in Section 6.5, RORB, RAFTS and
WBNM, require different inputs, due to their structures and parameters being different.
Figure 2.22 shows a RORB sub-catchment input. Only loss calculations occur in a RORB
sub-catchment, so that the only information required is the area of the sub-catchment and
the percentage of impervious area.
The sheet for a RAFTS sub-catchment shown in Figure 2.23 requires more information. In
addition to catchment area and percentage impervious, a sub-catchment slope and a
Manning's n for the pervious portion of the catchment are required. This information is used
to calculate hydrological losses and to define a routing parameter.
The property sheet for the Watershed Bounded Network Model (WBNM), shown in Figure
2.24, is the same as that for the RORB Model. However, routing does occur in WBNM sub-
catchments, using equations based on the sub-catchment area.
The Customise Storms button near the bottom of the ILSAX model Sub-Catchment
property sheet allows special features to be chosen, using the property sheet shown in
Figure 2.25. These features are useful in special studies involving gauged rainfall and flow
data, or where you wish to explore the effects of varying the rainfall intensity, pattern and
timing of storms over the catchment area.
You can select a particular storm to apply to this sub-catchment, in Design or Analysis
calculations. The storm is selected from the rainfall pattern data base using the list box
shown. A time lag can also be specified and the storm patterns can be multiplied by a
constant multiplier.
These options allow you to specify a different rainfall pattern and intensity at every sub-
catchment in a drainage network, if required.
These paths define the routes taken by stormwater flows that bypass on-grade pits and/or
overflow from pressurised pipe systems. DRAINS needs this information to calculate flow
characteristics. Clicking on an overflow route opens the two-page property sheet shown in
Figure 2.26. A name and an estimated time of travel must be entered on the first page.
Figure 2.26 The First Page of the Overflow Path Property Sheet (Top Portion)
DRAINS can define flow characteristics at a selected critical location, which may be at a pit
receiving overflows from this overflow route, combined with flows from its local sub-
catchment. This location could also be just downstream of the pit from which the overflow
occurs. The position is effectively defined by the percentage of the downstream catchment’s
flow that is carried by the cross-section, which must be entered into the property sheet in
Figure 2.27. By changing the value in the box labelled “Channel slope (%)”, the slope can be
varied along the entire flow path length to reflect a concave or convex longitudinal profile.
Figure 2.28 shows how a downstream sub-catchment may contribute to flows. Here the
critical point is the downstream pit, which is in a sag. An estimated 65% of Catchment 2
drains to this point, on the left side of Pit 2. An example of the output for such a situation is
given in Figure 3.31 in the next chapter.
As described in the next section, the overflow path from a detention basin acts as a high-
level outlet to the basin, and requires additional information to an overflow from a pit, this
being set out on a third page.
DRAINS can incorporate large or small detention and retention basins into drainage
networks. To fully define a basin or storage, two, or perhaps three components are required.
The first is the Detention Basin property sheet, an example of which is shown in Figure 2.29.
This includes a Basin Name, an initial storage, an elevation-storage (or height-storage)
relationship, and a low level outlet specification. The elevation-storage relationship must use
levels at the same datum as for the rest of the drainage network.
In January 2003, the form of the Detention Basin property sheet was changed, to add the
pit/sump outlet facility and to eliminate ambiguities that occurred when a detention basin was
specified with a circular or rectangular pipe outlet. Formerly, it was possible to specify the
pipe in the Detention Basin property sheet and also in the property sheet for a pipe passing
out of the basin.
Pipe is modelled as
the basin low-level
outlet with a
finite length
Headwater Flow goes into
Level open channel
Tailwater Level
This older method is still recognised by DRAINS so that models that have already been
created can continue to run. If the property sheet for an existing basin with a pipe outlet is
opened, it will appear as in Figure 2.31.
While DRAINS will cope with basins created in the old format, it is advisable to re-enter
basins and associated pipes in the new format if the file is to be used extensively.
For new detention basins, the requirements for a pipe will be simpler. As shown in Figure
2.29, it is only necessary to specify the entry and bend losses. The rest of the information is
included in the pipe property sheet.
Note, however, that in the case where a detention basin outlet pipe discharges to an open
channel, it is now necessary to specifically include the pipe, connecting it to the open
channel at a simple node. Where a basin pipe outlet connects directly to another pipe, the
old system only required the downstream pipe to be drawn. The information for the upper
pipe was entered in the Detention Basin property sheet. Now both pipes must be drawn,
connecting through a pit or node.
If an orifice outlet is selected, the property sheet takes the form shown in Figure 2.32. You
must supply a diameter (mm) for a circular orifice, and the elevation of its centre. The check
box labelled High Early Discharge allows the modelling of a high early discharge pit, a
special type of OSD system. The user must provide a crest level and length for an internal
weir that is a feature of this kind of storage. Further details of the hydraulics of these options
are given in Section 6.9.
If the pit/sump outlet type is selected, the outlet changes to that shown in Figure 3.33. A pit
family and size is to be selected using the same drop-down list box as in the Pit property
sheet.
If the circular or rectangular pipe options are chosen, you need only supply head loss factors
for the entrance and for any pipe bends, as shown in Figure 2.29.
The fifth and last type of outlet is a "None" option. If this is selected, water can only leave the
basin through a high-level outlet (to be described below), and the outflows will not be
affected by downstream hydraulic grade lines or backwater effects. If a height-outflow
relationship is specified for a high level outlet, the detention basin modelling will be carried
out in the relatively simple way used in ILSAX, rather than having HGLs projected upwards
through the basin.
The pipe leaving a basin is specified in the same way as a normal pipe. If this is rectangular,
it may be necessary to set up a special pipe type and size in the Pipe Data Base, as
The last component required to define a detention storage is the high level outlet, which is
described in the property sheet for the overflow route from a basin. When an overflow route
originates in a basin, the property sheet has three pages instead of the two shown in Figures
2.26 and 2.27. Two of these pages are the same as those in those figures.
On the third page, labelled "Weir Data" you have the choice of specifying a weir outlet, as
shown in Figure 2.34, or an elevation-discharge (or height-outflow) relationship, as shown in
Figure 2.35.
For a weir, you must provide a weir coefficient, a width (m) (at right angles to the direction of
flow) and a crest level (m). Further details are given in Section 6.9. A suitable coefficient for
the earth embankments used as high-level outlets for many detention basins is 1.7.
There can be any number of overflow routes from a detention basin, representing high level
outlets at different levels as well as pumped discharges and stormwater infiltration.
The Prismatic Open Channel property sheet, shown in Figure 2.36, enables easy entry of the
parameters needed to define channels of uniform cross-section and slope. If calculations
determine that the channel depth exceeds that specified in this property sheet, the sides of
the channel will be extrapolated upwards, and a warning message will be provided. DRAINS
does not allow for overflows from channels. In the majority of cases, where overflows will
follow the same route as the main stream channel, they can be accommodated by defining a
channel cross-section large enough to carry them. If necessary, the channel should be
defined as an irregular open channel, as explained in the following section, to include
overbank flow areas.
Where an overflow from a channel will cause a breakout that follows a different path to the
main stream, special ways of modelling the separation of flows are required (see Table 2.2 in
Section 2.3.15).
There is provision to place a roof over the channel, using the check box at the top.
If the water rises above the top level, the channel section will be assumed to run full, and can
be used in pressurised flow situations.
This component, with the property sheet in Figure 2.37, allows you to model:
(a) a stream or channel with varying cross-sections and slopes,
(b) closed and open conduits with cross-sections other than circular, rectangular or
trapezoidal.
It is necessary to define channel reaches over which flowrates are the same, and to define
for each reach at least two cross-sections, at the upstream and downstream ends of the
reach. At each cross-section, you must enter:
• the channel name, total length and chainages or lengths of reaches along a stream;
• a set of X-Y coordinates (m) that define the cross-section, with the X datum being at an
arbitrary point on the left bank of a channel, and the Y datum being Australian Height
Datum (AHD) or some other standard datum (as shown in Figure 2.38);
• distances from the upstream node (m) and Manning’s roughnesses for the left overbank,
main channel and right overbank areas;
• coordinate locations of the left and right banks (m), and expansion and contraction
coefficients (dimensionless).
Various facilities facilitate the entry of cross-sections. The top section of a reach must be the
same as the bottom section of the reach above it. If reaches are entered in a downwards
direction, DRAINS will automatically enter data from the previous reach. Sections can be
viewed and checked using the View Cross Sections and View maximum water level
profile options in the pop-up menu for an irregular channel component, as shown in Figure
2.39.
2.3.10 Multi-Channels
The prismatic and irregular channel types do not adequately cover the situation where two or
more channels with different characteristics connect the same two points. This is handled by
multi-channels that use the property sheet shown in Figure 2.41 to call up the boxes for
prismatic or irregular channels, or a box for circular channels.
The data required is similar to that for other open channels. Conduit lengths, roughnesses,
slopes and even starting and ending levels can vary. DRAINS distributes flows between the
different conduits. At present, DRAINS does not report on the separate flowrates.
This link type, shown as , is used to connect nodes and sub-catchments that form parts of
storage routing models described in Section 6.5. Its exact function differs with the type of
storage routing method selected. For a RORB or WBNM model it performs a non-linear
routing, while in a RAFTS type model it performs a translation or a hydraulic, kinematic wave
routing.
If a RORB storage routing model is selected, the property sheet for its stream routing
reaches appears as shown in Figure 2.42.
A reach name and length must be specified, and a Channel Condition selected. If a channel
condition of "excavated unlined" or "lined or piped" are selected, it is also necessary to
provide the reach slope.
Figure 2.43 First Form of RAFTS Stream Routing Reach Property Sheet
If the second option in the Flow Routing Method box is chosen, the property sheet changes
to the form shown in Figure 2.44. It is now necessary to provide a reach length and, using
the second page of the property sheet shown in Figure 2.45, a cross-section is to be selected
from the Overflow Route Data Base. This section is meant to be representative of the whole
stream reach and to be used in a kinematic wave routing procedure derived from Chapter 9
of Open Channel Hydraulics by F.M Henderson (Macmillan, New York, 1966).
Figure 2.44 Second Form of RAFTS Stream Routing Reach Property Sheet
The WBNM stream routing reach property sheet shown in Figure 2.46 requires only a name
and a stream lag factor. When this factor is not zero, routing occurs along the reach using
parameters based on the area of the sub-catchment at the node at the end of the reach.
Figure 2.46 WBNM Stream Routing Reach Property Sheet (Top Portion)
2.3.12 Headwalls
The headwall allows open channels to be connected directly into a pipe system.
The corresponding property sheet is shown in Figure 2.47. In older DRAINS models
this situation had to be handled by terminating an open channel at a detention basin
and starting a pipe from there.
2.3.13 Culverts
Culverts that are located where roads or railway lines cross streams or channels can be
modelled using the Culvert component with the property sheet shown in Figure 2.48.
The inputs allow for flow through the culvert, and under the road above, and also for
overflows that cross over the road. For the former, it is necessary to provide:
• conduit name, and number of parallel conduits;
• conduit diameter (mm), or if rectangular, width (m) and height (m);
• length (m) and slope (%);
If you are specifying river chainages for an open channel system, you should allow for the
length of the culvert.
A culvert can bank up water like a detention basin, so DRAINS presents the upstream and
downstream water levels on the screen display after a run. The formulae used in culvert
calculations are discussed in Section 6.10.
The term “culvert” can also be applied to a long, roofed channel. This is best modelled as a
pipe or as a roofed open channel.
2.3.14 Bridges
Because of the differences in shapes, abutment and pier arrangements and approach
conditions, bridges are more complex that culverts. In DRAINS, calculations are performed
using a method provided in the AUSTROADS (1994) manual, which is based on the US
Federal Highway Administration report by Bradley (1970). The property sheet for Bridges is
shown in Figure 2.49.
You will need to refer to the original references to fully understand the inputs required. It is
necessary to specify:
• the name of the bridge, and the levels of the deck (m) and the soffit (underside of deck)
(m);
• the weir coefficient for overflows over the bridge deck, typically 1.7;
• pier width, locations of piers (as noted in Figure 2.50), and pier type;
• the abutment type and the X-Y coordinates at the bridge section (m); left overbank, main
channel and right overbank Manning’s roughnesses, and the X locations of the left and
right banks that divide the zones of different roughness (m).
More complex bridge modelling procedures are available in HEC-RAS, MIKE-11 and other
programs. DRAINS results should be checked using these programs if the accurate
determination of levels is critical.
The components described above are represented by objects that can be placed in the Main
Window and connected together as links and nodes.
Sub-catchments can be connected to all types of nodes: simple nodes, pits, detention
basins, culverts and bridges.
Because some arrangements are not logical, because they create computational difficulties,
some linkages cannot be modelled. Table 2.2 describes the possible connections between
nodes and links. It notes those that cannot be made. The footnotes provide suggestions as
to how you can get around some of these limitations.
Notes:
1 - If a node has pipes both upstream and downstream, it acts as a closed junction, and can be
pressurised, with the HGL rising above the surface. Generally, however, it is better to connect
pipes through sealed or unsealed pits, where a head loss can be modelled.
2 - You need to be aware that nodes will accept all flows coming to them, and check whether this is
realistic. Where there are likely to be overflows, a pit should be substituted if the node is in a pipe
system, and a detention basin if the node is in an open channel system.
3 - Overflows are permitted from a node, but not if there is also a pipe or channel leaving the node.
For an open channel where overflows will run along its banks, you should raise the height of the
channel cross-section so that overbank areas are included. The open channel downstream will
need to be defined as an irregular open channel. Where channel overflows are to be directed
out of a channel, you should place a detention basin at the location of the low point where
overflows might occur. An elevation-storage relationship based on the storage within the channel
upstream of this point can be defined. High-level outlets with weir data or a height-discharge
table can be used to control the overflows.
4 - With the older, now obsolete way of defining detention basin outlets, you had to be careful not to
double-define pipe outlets. If the outlet pipe led to another pipe, you needed to define the
detention basin low level outlet as a pipe with zero or nominal length, and specifically draw the
basin outlet pipe as connecting to a pit or node. Now it is necessary to draw in a pipe outlet, but
outlet pipe details are not required in the Detention Basin property sheet. If the basin outlet pipe
connects to an open channel, the drawn pipe can be connected to the channel via a node.
5 - Overflow links from a detention basin require more information than a normal overflow link, to
define high level outlets. It is possible to have several high-level outlets from a basin.
6 - Culverts and bridges must span open channels. Where a road is located at a point where
stormwater emerges from a pipe, or goes from on open channel into a pipe, it is probably
inappropriate to model this situation as a bridge or culvert. If cross-sections change under road in
these circumstances, the transitions can be modelled by pipe or open channel sections.
7 - While water may pond behind a bridge or culvert, and even overflow over the top of the road,
DRAINS does not allow for any diversion of flows away from the downstream channel. This
might be modelled by locating a detention basin upstream of the device, or perhaps by modelling
a culvert as a detention basin.
8 - Bridges have more restrictions than culverts in DRAINS. You cannot have two upstream
channels meeting at a bridge, as they can at a culvert. It is necessary to insert a section of
combined channel upstream of the bridge. A multi-channel cannot be placed downstream of a
bridge - a short section of single channel can be interposed, however.
9 - You cannot have an open channel coming out of a pit. However, a short section of pipe and a
simple node may be used to link the pit and a channel.
Example file Sydney OSD.drn provides such a comparison for an on-site stormwater
detention system. As shown in Figure 2.51, the natural catchment is set out on the lower left,
and the developed drainage system around a house and backyard occupies most of the
window. These two systems are run together using the same project specifications, allowing
a direct comparison of results.
2.4.1 General
By storing standard data sets in five data bases that are easily accessible from drop-down
list boxes, DRAINS makes it easy to select and alter hydrological models, rainfall patterns,
The role of data bases is particularly important in the advanced design procedure. Pipes and
pits are both organised into types or families of different sizes from which DRAINS can
select.
When you start DRAINS it checks for the presence of the file StandardStorms.drn in the
C:\Program files\Drains\Program folder that contains the file Drains.exe. If
found, this file is loaded in place of the standard data base file. The intention is to let you use
this file to avoid the need to redefine storms for each new job. If you work in only one
geographical area, where the average intensity for each storm (ARI and recurrence interval)
is constant, this file need not change. In that case, when you first use DRAINS you should
set up the storms and hydrological model for your area, make any changes you require to the
pipe, pit and overflow route data bases, then save the file as StandardStorms.drn into
the Program folder. Note that this file contains no pipes or other components.
If you work in different locations, you will need one standard file for each geographical area.
When you first use DRAINS in a particular area you should set up the storms and
hydrological model(s) you require, make any adjustments to the data bases and save the file
(e.g. as SutherlandStandardStorms.drn). The next time you work in this area you
would open this file. You should then immediately save it as a different name to preserve the
main file. Alternatively, you could replace the file StandardStorms.drn from Windows
Explorer before you open DRAINS.
The basic set ups for ILSAX, Rational Method and storage routing hydrological models are
described Chapter 1.
It is possible to develop a number of different ILSAX models, say for different soils, and to
mix these in a model. (This is done in the example file Ilsax9.drn described in Section
7.3.) The storage routing models can be mixed with ILSAX models, although it is only
possible to have one type. You cannot mix RORB and WBNM models, for example.
However, it would be possible to create a DRAINS model that used three kinds of ILSAX
model and two kinds of RORB model
It is also possible to specify three different kinds of Rational Method model, as shown in
Figure 2.52. These different models can be inter-mixed.
A large number of diverse models can be stored in the Hydrological Model data base. The
hydrological model finally selected in the Hydrological Specifications dialog box acts as a
default model that applies to all sub-catchments.
The setting up of standard Australian patterns in the Rainfall Patterns dialog box has been
demonstrated in Section 1.3.1 (Figures 1.6 to 1.8). It is also possible to set up non-standard
patterns by clicking the Add a New Storm button in the Rainfall Data property sheet to open
the box shown in Figure 2.54.
The duration of the pattern and the time step can be set, and the rainfall intensities entered
directly in the text box labelled “Intensity (mm/h)”. Corrections can be made be locating a
value using the spin box for the intensities, and altering the contents of the text box.
Figure 2.55
Spreadsheet Columns
2.4.5 Pipes
The Pipe Data Base property sheet shown in Figure 2.57 is opened by clicking the Pipe Data
Base … button in the Project menu. This operates in two stages. The first is to define a
pipe type, and to specify its name, whether it is circular or rectangular, its roughness
(according to the pipe friction formula set in the Options property sheet called from the
Project menu), and its minimum cover (m).
The second stage is to provide data for specific pipe sizes in the property sheet shown to the
right in Figure 2.57. For circular pipes, the nominal diameter, internal diameter (I.D.) and wall
thickness must be supplied in mm. For rectangular pipes, the width (m), height (m) and wall
thickness (mm) must be supplied.
The check box labelled “Not available for selection in design runs” allows you to omit pipe
sizes that are considered too small or are unavailable.
If you wish to vary cover depths with pipe sizes, or to have different classes of pipes (with
different wall thicknesses), specific pipes classes should be entered as pipe types.
Care is needed in entering values, but once established, the data base is easy to apply.
Pipe types and sizes are readily accessed from the Pipe Data property sheet. Editing of the
data base, and deletion of pipe types and sizes, is not from design calculations possible if
drainage system components are present. It has to be done on a standard file or a file used
as a template.
When a DRAINS file is opened and closed, its pipe, pit and overflow profile data bases
remain in DRAINS and some may be inherited by the new system. Standard or template
files also supply suitable data bases.
2.4.6 Pits
There are different procedures for the advanced design method and the equation-based
procedures originally employed in DRAINS. Since both are still operable, both facilities are
set out below.
The Pit Data Base is accessed through the Pit Data Base… option in the Project menu. As
shown in Figure 2.58, pits are organised into types or families of different sizes, in a similar
way to pipes. The pit type is described in the Pit Type property sheet, also shown in Figure
2.58.
Data for each individual pit is entered in the triple property sheet shown in Figure 2.59. The
relationships are entered directly, as tables. This provides flexible relationships, particularly
at the “top end” of the curves. It is possible to set an upper limit on inlet capacities if
required. A Table Wizard is available for sag pits using the weir and orifice equations from
the U.S. Federal Highway Administration's Drainage of Highway Pavements, Hydraulic
Engineering Circular HEC-12, 1984. An interpolation procedure is used to estimate
relationships in the gap between the weir and orifice operations.
The relationships can be set out in a spreadsheet, plotted to check for errors, and then
transferred directly into DRAINS using the Paste Data buttons associated with the On-Grade
Data and Sag Data pages in Figure 2.59.
As part of the introduction of the Advanced Design method, sets of relationships for pits in
New South Wales, Queensland and Victoria have been established, and are included in the
.db1 data files contained in the C:/Program Files/Drains/Program folder, under
names such as Drains_NSW Feb04.db1. Relationships for the Australian Capital Territory
and South Australia will be added in 2004. These can be selected for a new project using
the Default Data Base option in the Project menu, in the dialog box shown in Figure 2.60.
If the Pit Data Base is opened in these files, and a nominal change is made to one of these
pits, the following message appears when this is closed by clicking on the OK button:
Clicking the Yes button sets up the file’s pit data base as the standard one.
(New South Wales, Queensland And Victorian pit types are available as .db1 files in the
C:/Program Files/Drains/Program folder. If the Drains.db1 data file is deleted and
Drains_NSW.db1 or Drains_Qld.db1 is renamed the Drains.db1, the NSW or
Queensland pit types will be installed.)
DRAINS still accepts old files that use the ILSAX equations for pit inlet capacities. However,
if you attempt to run the Advanced Design method, the following message will appear:
It would then be possible to import a new Pit Data Base and to alter each pit’s type to
conform to this.
This property sheet, shown in Figure 2.61, allows for entry of on-grade and sag pit factors
together, adding to the Pit Data Base. The equations and factors are described further in
Section 6.6.
As with the Pipe Data Base, with both the standard and old Pit Data Base procedures there
are restrictions on altering data or deleting pits when components are in the Main Window.
The property sheet opened from the Overflow Route Data Base… in the Project menu is
shown in Figure 2.62. Using X-Y coordinates, a cross section can be defined for a roadway,
footpath or other route that may operate as a path for overflows.
At present, the section may be divided into two zones with different Manning’s n
roughnesses, by specifying these, and the X value of the dividing line. As X-Y values are
entered, the picture shows the section being produced.
Figure 2.62 Overflow Route Cross Section Data Base Property Sheet
It does this by providing pits and pipes with the appropriate capacities to do this, following
procedures within the Queensland Urban Drainage Manual (Neville Jones & Associates et
al., 1992).
• Input Options,
• Display Options,
• Run Options,
• Output Options, and
• Help Options.
3.2.1 General
The example file in Chapter 1 was established using the screen tools provided on the
Toolbar, and their associated property sheets. Other options are available that allow a
substantial part of the information required to be inputted by other means. These are mainly
implemented through the File menu shown in Figure 3.1, and the two additional menus that
are opened using the Import ► and Export ► options.
Figure 3.1 The File Menu and Sub-Menus showing Import and Export Options
In design work, it is likely that considerable data will be available from CAD files created by
surveyors and used by designers to set out street layouts, cadastral (land boundary) data
and positions of services. Some of this data can be taken directly into DRAINS through the
facility for importing CAD files in DXF or DWG formats.
This includes a background showing streets, lot boundaries and other information. Other
data obtainable from CAD drawing files, such as sub-catchment areas, will have to be
entered into property sheets, or via a spreadsheet.
Where a new system has been drawn in a drawing package or digital terrain model, it can be
imported into DRAINS in DXF and DWG formats. The former is one of the oldest drawing
formats, which can be imported by almost all technical graphics packages. DWG is the
widely-used AutoCAD binary format.
Figure 3.2 shows a drawing created in AutoCAD LT, representing a drainage system
assumed to be at Brisbane.
Information that is in this file can be imported by opening DRAINS and selecting Import a
DXF or DWG File… from the File menu. You will be requested to nominate a file with a
.dxf or .dwg suffix. You will then see a dialog box that asks you to nominate the names of
the layers on which pipes, pits and background are located, as shown in Figure 3.3. This is
saved as a file Brisbane.dxf.
Using the drop-down list box, you can select the appropriate layers. Pits and pipes can be
placed on the same layer if you wish. Once layers are selected, a number of information
windows appear. The first one shown in Figure 3.4 allows pipe lengths to be automatically
scaled off the DXF drawing, according to the length allocated to the first pipe for which full
data is entered.
The DRAINS Main Window then appears with the drainage system and background shown
as in Figure 3.5. This can be enlarged if necessary using the Zoom tool, and the colour of
the background can be changed using the Background Colour… option in the View menu.
You must now enter information for pits and pipes, and draw in sub-catchments, overflow
routes and outfall nodes. Directions of pipes may have to be changed, and it is likely that the
connections will have to be re-arranged to assist viewing of results.
Information can also be imported from a spreadsheet file. Since this file will usually be
created by outputting information from a DRAINS file, spreadsheet processes are described
in Section 3.5.6 of this chapter, dealing with output options.
This process enables you to import data into DRAINS from one to six sets of ESRI or
ArcView files, plus an optional background from a DXF file. The procedure is the reverse of
the exporting process for ESRI files described in Section 3.5.7(a).
If you wish to model an existing drainage system in DRAINS, importing data from ArcView
records that are available, you must edit these into the format required by DRAINS, which is
described in Section 6.11.3(b). You can also see this format by exporting a small drainage
system and examining the resulting DBASE tables.
The six sets of files contain data for nodes (pits and outlets), pipes, overflow routes, sub-
catchments, survey levels along pipe routes and the positions of other services along a pipe
routes. Each set includes three files with SHP, SHX and DBF suffixes.
The transfer must include the files for nodes, the rest are optional. In an initial transfer, it is
unlikely that all the information required by DRAINS will be available in the GIS. Information
that is already in the GIS should be included In the files to be transferred. You can then
choose whether to add additional data in these files, or to use dummy values and enter the
required values later, in DRAINS.
The data for nodes (including pits and outlets) and pipes, each contained in a "theme", needs
to be entered in the data base tables shown in Figure 3.7, which can be created by editing in
ArcView, Microsoft Access or Excel. A DXF file containing a background for the DRAINS
model, shown in Figure 3.8, can be created from GIS layers.
To make the transfer, you must place all files to be transferred into the same Windows folder,
set up a DRAINS model with the ILSAX hydrological model and pit and pipe data bases that
you require, and then use the File -> Import -> ESRI Shapefiles… option, which will display
the following message: After entering "Yes", you must select one of the ESRI files to be
transferred, as shown in Figure 3.9:
With this setup it is possible to import a new background or to exchange the current
background with another. Using the File -> Import -> Import DXF background… option
brings up a dialog box from which a DXF file can be opened. When a file is selected, the
following window appear: When a choice is made, the background is replaced.
This process enables you to import data into DRAINS from one to six sets of MapInfo files,
plus an optional background from a DXF file. The procedure is the reverse of the exporting
process for MapInfo files described in Section 3.5.7(b), and is similar to the ESRI transfer
process described in the previous section of this chapter.
To transfer MapInfo data to DRAINS, you need to edit the available MapInfo data into pairs
of MID and MIF files in the format required by DRAINS, which is specified in Section
6.11.3(c). This is the same as the format generated in the export process that creates
MapInfo files from DRAINS data, which you can see by exporting a small system and
examining the resulting tables.
All the required information that is already in the GIS should be included in the files to be
transferred. It is then a matter of choice as to whether you add additional data in these files,
or enter dummy values, and enter the missing data later in DRAINS.
To illustrate the process, consider the following example, which parallels the ESRI file import
example. Figure 3.11 shows the Oldtown System in MapInfo, with the data for one node
being displayed. This can be set up in MapInfo or a text editor. The corresponding pipes
table is similar.
From the MapInfo layers, a DXF file containing a background for the DRAINS model can be
created. This will appear in the same form as Figure 3.8.
To make a transfer, you will need to place all the files to be transferred into the same
Windows folder, set up a DRAINS model with the ILSAX hydrological model and pit and pipe
data bases that you require, and then use the File -> Import -> Mapinfo MIF files… option,
which will display the following message:
As with the ESRI transfers, with this setup it is possible to import a new or additional
background. Using the File -> Import -> Import DXF background… option brings up a
dialog box from which a DXF file can be opened. When a file is selected, the following
window appears:
There is an option in the File menu to merge DRAINS files together. As this first involves an
output process, it is described in Section 3.5.8.
DRAINS is partly based on the ILSAX program, which has been used widely in Australia
since 1986. Many Government organisations have ILSAX files describing their drainage
systems. These can be converted to DRAINS files using the Import ILSAX Files … option
in the File menu.
Since ILSAX does not include any positions or levels of components, these have to be
provided. The positions in the DRAINS Main Window have to be established as X-Y
coordinates. This is done by setting up a DRAINS file in which pits are defined, as shown in
Figure 3.13 for a drainage system assumed to be in Melbourne.
Figure 3.13 Part of Melbourne Drainage System, prior to Importing ILSAX Files
The simplest way to establish this file is to draw in the pits using the Pit Drawing Tool and
then defining these as “simple sag pit (as per ILLUDAS)” pits in the Drainage Pit property
sheet shown in Figure 3.14. Only the pit name, pressure change coefficient and surface
elevation need to be entered.
The pit names must correspond to those used in ILSAX, which are in two parts - a 1 to 3
character Branch Name and a 1 to 3 character Reach Name. The name given to a DRAINS
pit must be the branch name, followed by a full stop, followed by the reach name. For
example, DRAINS Pit A.2 corresponds to ILSAX Branch A and Reach 2. When DRAINS
imports data from ILSAX files, it searches for this name and attaches the data to it.
The pit pressure change coefficient can be left at the default value of 1.5 prior to a data
transfer, and more suitable values can be entered later. Dummy values can be provided for
the pit surface levels, but it is preferable that the actual values are entered before the
transfer.
Once the pits have been defined as in Figure 3.14, the Import ILSAX Files… option in the
File menu can be employed. This opens the dialog box shown in Figure 3.15. ILSAX
employs two data files. The Run & Rainfall file contains the parameters for making a run,
together with the rainfall input data, while the Pipe file describes the drainage system. You
should nominate the first of these. In this example; this is the file Melrain1.dat.
Another dialog box of the same4 form as Figure 3.15 then appears. The pipe file,
Melpipe1.dat should than be nominated. Two rainfall boxes then appear. Click OK to
these, and the transformed DRAINS file shown in Figure 3.16 will appear.
Sub-catchments, pipes, overflow paths and outlet nodes are defined. If you open the
property sheets for these, you will find that ILSAX information has been transferred over.
The file then needs to be tidied up and additional information added. The hydrological model
and rainfall data will probably need to be added or modified, as well as the components and
their positions.
Not all ILSAX formats and information are transferable, but most pipe systems can be readily
transferred. Even if error messages and warnings appear, it is likely that the bulk of the
ILSAX data will be converted to DRAINS format.
Pits in the transferred data will be in the old format described in Section 2.4.6(b). To transfer
these to the new format, select the Advanced Design option in the Run menu. (This will be
available, even though the data input is incomplete.) This will display the message below.
You can then select a new data base from the c:/Program Files/Drains/Program
folder in the window shown in Figure 3.17. This converts the data base to the new format.
Data for setting up DRAINS models can be imported directly from the 12d digital terrain
modelling program (formerly known as 4D). After design and analysis have been carried out
in DRAINS, the resulting system information can be returned to 12d for further analysis and
plotting. For details, please contact 4D Solutions (Robert Graham or Tony Ingold on (02)
9970 7117, [email protected] or [email protected]).
Data for entry into DRAINS can be imported directly from the Land Development Desktop
(LDD) digital terrain modelling program and the Advanced Road Design program, also
marketed by Autodesk. Files for both programs can be imported into DRAINS using the
options in the Import ► sub-menu in the File menu. After processing in DRAINS, the
results can be returned to these programs for further enhancement and plotting. Files for
Land development desktop are automatically generated as data base files, while those for
advanced Road Design can be generated using the option in the Advanced Road Design
File… option in the Export ► sub-menu opened from the File menu.
For details, please contact CADApps (Australia) Pty Ltd (Andrew English or Laurie
Comerford (03) 9568 0077, [email protected] or [email protected]).
New data bases can be established using the Default Data Base Option in the Project
menu. This opens the dialog box shown in Figure 3.18, from which a base can be selected
from the .db1 files stored in the C:/Program Files/Drains/Program folder. This is
followed by a warning indicating that the standard .db1 file, Drains.db1, will be overwritten.
Note that this can only be done with a DRAINS file that has an empty main window. Once
components are entered, the only way to add pipe or pit types is by hand. In addition, it is
not possible to delete pipe and pit types in this situation, though their characteristics can be
edited and changed. (.db1 files are related to the new pit data bases and .db files to the
older system described in Section 2.4.6(b).)
3.3.1 Introduction
DRAINS provides several options for viewing data on screen in addition to standard
Windows facilities such as scrolling bars. The options available before calculations are
performed are demonstrated in this section using the Toowoomba Estate.drn example.
This system has been set up in DRAINS using the procedures described in Chapter 1 and
saved as Toowoomba Estate.drn. The program is ready to run in Design mode, which
will define the pipe diameters and invert levels.
You can vary the way that the a drainage system is presented on screen using options that
are mainly included in the View menu;
If the system appears to be too small, you can enlarge it using the Zoom option, which is
also available through an icon on the Toolbar. When this is selected, the dialog box shown
in Figure 3.20 appears. You can nominate the magnification. If you accept the standard
value of 1.5, an enlarged presentation is obtained. Entering a factor less than 1 reduces the
size of the system, but the size of lettering remains the same.
The wheel on a mouse can also be used to zoom in and out of DRAINS models. There is no
pan facility; it is necessary to use the sliding bars on the margins to move the view, or the
Index Sheet facility described in the following section.
Selecting Index Sheet from the DRAINS View menu produces the view of the system shown
in Figure 3.21:
The rectangle represents the screen size. Placing this “mask” in a certain position and
clicking sets the screen to that position, as shown in Figure 3.22.
Note that the Main Window area is quite large and includes a title block in the lower right
corner. Text can be inserted into this block using the Description… option in the Project
menu, which opens the property sheet shown in Figure 3.23. Comments and lines for the
title block can be entered. If no block is required, the three TITLE BLOCK lines can be made
blank.
Facilities like the Status Bar at the bottom of the Main Window and components such as sub-
catchments can be removed from the window if desired, using the options in the central part
of the View menu. The Drawing Area can be extended at the four corners, cropped (reduced
selectively) or trimmed all round, using options in the View menu.
Figure 3.22 Toowoomba Pipe System selected using the Index Sheet
The Customise Text … option at the top of the View menu produces the dialog box shown
in Figure 3.24. By selecting options here, you can change the information provided, as
indicated in Figure 3.25. Many of the options are only available after a Design or Analysis
run has been made.
Figure 3.24 The Dialog Box for Customising or Changing the Text Displayed
This dialog box can also be opened by clicking on the name of any component, though not
the component itself.
The pop-up menus opened by right-clicking on an object are the main means of presenting
results of calculations on-screen. They also provide some information prior to calculations.
Two displays from the Toowoomba example are shown in Figure 3.26.
In the Run menu there are six options, four of which are for running DRAINS:
• The Standard Design (pipes only) option runs with the rainfall patterns chosen in the
Select Storms > Minor storms option in the Project menu, determining sizes and invert
levels for pipes according to the specifications in the Options property sheet called from
the Project menu. It also performs a simulation using the set of minor storms as a check
on the design, producing hydrographs and HGL traces.
• The Advanced Design (pits and pipes) option runs with the rainfall patterns chosen in
the Select Storms > Minor storms and Select Storms > Major storms options in the
Project menu It applies a complex procedure determining pit sizes and diameters and
The Standard Design process only applies to pipes. Pipes can be excluded from design
changes using the option buttons in the box labelled “During Design runs this pipe” in the
Pipe property sheet, shown in Figure 3.27. Invert levels must be specified for fixed pipes,
that are not intended to be changed.
If you choose the first option and also specify invert levels, these will most probably be
changed in a Design. (In calculations, the second option is treated exactly the same as the
third, except that when DRAINS calculates quantities of soil volumes for excavation, volumes
for pipes defined under Option 2 are included in the table of quantities along with those
defined under Option 1; volumes for Option 3 pipes are not.)
DRAINS does not specifically try to design around “existing” pipes with fixed invert levels, so
situations will be encountered where it is not possible to do this while obeying the restrictions
set in the Options property sheet opened from the Project menu. In these cases the invert
levels at the downstream end of designed pipes may be specified as being lower than the
existing pipe to which they connect.
Analysis runs treat all pipes as fixed, and do not alter the given pipe diameters and invert
levels. Complex situations, such as pits with the invert of the outgoing pipe being higher that
those of the incoming pipes, can usually be modelled.
The Advanced Design method, based on the Queensland Urban Drainage Manual (Neville
Jones & Associates et al., 1992) varies both pits and pipes to obtain an optimal result. It is
possible to set both the pit size and the pipe diameter and invert levels as fixed, using
options in the Pit and Pipe property sheets.
Both design procedures allow for intermediate levels between pits, along a pipeline route.
These are considered when pipe invert levels are determined, allowing for minimum cover
depths.
Following a run, DRAINS presents a log reporting on the results, as shown in Figure 3.27,
indicating problems and possible causes. The first example shows a trouble-free run and the
second one that has complex results.
The report must be closed (by clicking on the X at the top right of the window), but it can be
recalled using the Last Run Report option in the View menu. The information in the log is
also reproduced in the spreadsheet output for results.
DRAINS performs a number of checks as data is entered. One is to ensure that all data is
entered. In some instances DRAINS requires values to be within certain reasonable rages,
in others it queries values that appear to be unusual.
Warnings like those shown in Figure 3.28 also appear when a run is initiated, and after a run.
It is important to heed these, and to try to eliminate the causes.
• The Revise Pit Loss Coefficients option alters the pit pressure change coefficients
using an algorithm within DRAINS which is based on an approximate relationship
developed by Mills (see Section 6.6.3.). After this is done, you can again run in a design
or analysis mode
3.4.5 Quantities
The Quantities option in the Run menu displays or prints out a table of quantities for the
pipes in the current system, as shown in Figure 3.29. This complements the information
printed for each completely-defined pipe at the bottom of its property sheet, as shown in
Figure 3.30.
DRAINS calculates the volumes of excavation from the pipe lengths and invert levels,
assuming that the trench width is as shown in Table 3.1, with 200 mm being added for each
parallel pipe in addition to the first one, and that the bedding depth is 50 mm below the
outside of the pipe wall.
3.5.1 General
Options for the export and storage of data and results have developed progressively in
DRAINS. The first options reflected the ILSAX file outputs; now more useful forms of output
have been introduced.
Section 3.3.2 showed various screen displays that are provided by DRAINS prior to run
calculations. Additional displays become available once a run is made. These include
hydrographs, HGL level plots, tables of flowrates and HGLs, as shown in the previous
examples in this manual and in Figure 3.31.
Figure 3.32 Screen Capture of DRAINS Results showing Main Window and Hydrograph and
HGL Plot Windows
DRAINS has a facility for printing out the system displayed on a screen, either completely, or
as the view shown on the screen. This is implemented in the Print Diagram option in the
File menu, using the dialog box shown in Figure 3.33. Font sizes can be altered. The OK
button starts the printout, while the Setup... button opens a Printer Setup dialog box.
Note that this facility will not work with some printers, due to problems wuith the printer
drivers.
Until 2002, DRAINS contained an option to provide an ASCII or text file output similar to the
output file format for ILSAX. This has now been discontinued.
The process of importing data in DXF format was presented in Section 3.2.2. There are two
types of output via DXF file format, one of the most common formats used for drawings.
With the Toowoomba Estate.drn file, you can export a plan view to scale using the
Export DXF File… option in the File menu. This opens a Save As dialog box, and after a
file name and location are specified, opens the DXF File dialog box shown in Figure 3.34.
The resulting file can be opened in a drawing program, appearing as shown in Figure 3.35.
The background and pipes are supplied on different CAD layers.
A longitudinal section can be exported by nominating a path between neighbouring pits, and
then specifying drawing characteristics. The option File -> Export -> DXF Long Section…
opens the dialog boxes shown in Figure 3.36.
The smaller box is used to define a continuous pipe route. You need to specify the starting
and ending node names exactly, allowing for blanks and the case of words.
Figure 3.36 Dialog Boxes for setting Paths for Plotting Long Sections
Once a satisfactory layout is achieved, pressing the Save as DXF opens a window in which
the file name and location can be specified. This creates the DXF file, which can then be
viewed and manipulated in a CAD program, as shown in Figure 3.39 and printed from this.
The spreadsheet option provides a convenient way to view and store data and results, as
well as a medium for transferring information between DRAINS and other programs. It
effectively supersedes the text file output described in the previous section, although this is
retained for the convenience of users.
To exchange information with a spreadsheet program, say Excel, both programs must be
opened. Information is exchanged via the Windows Clipboard by selecting the copy and
paste options in the Edit menu. After selecting Copy Data to Spreadsheet in DRAINS, as
shown in Figure 3.40, transfer to Excel and select Paste from its Edit menu.
The information shown in Figure 3.41 appears. Almost all the information entered for
components is presented, organised by type of component - PIPE/NODE, SUB-
CATCHMENT, etc. This worksheet can be given a name such as “Data” by double-clicking
on the tag at the bottom of the sheet and writing in the name in the space that is highlighted.
X-Y coordinates are given for pits and nodes, referring to their positions in the Main Window.
If a base drawing is imported from a CAD or GIS file the coordinate system will be consistent
with this.
The results from a Design can then be transferred using the Copy Results to Spreadsheet
option from the Edit menu. This can be pasted into a second worksheet with the tag
“Design” or "Minor", as shown in Figure 3.42.
As with the data, results are organised by the type of component, in the same order.
Calculated flowrates, times, velocities and other information are presented. Where multiple
rainfall patterns are specified, the information presented is for the worst case result - the
greatest flowrate, highest HGL level, etc. among the results for the various storms.
The particular storm that causes this worst condition is noted in the last column for each
component. (DRAINS does not transfer the specific results for each individual storm. If you
wish to do this, you should use the Select Storms option in the Project menu to run
DRAINS with single storms and transfer the results one at a time.)
The velocities shown correspond to the peak flowrates and may be part-full or full pipe
velocities, depending on the conditions when the maximum flowrate occurred
A continuity check of inflow and outflow hydrograph volumes at each node (presented at the
bottom of the spreadsheet shown in Figure 3.43) applies to the storm that is most severe for
most components. It shows up differences in continuity due to factors such as:
• the absence of an overflow route when overflows occur,
• introduction of a baseflow or a user-provided inflow hydrograph.
Where there is a lack of continuity at a node, the cause can be explored by examining the
inputs and outputs to the relevant node using the View Hydrographs and View
Hydrographs as Tables options in the pop-up menus for pipes, channels, overflow routes
and sub-catchments.
The run log that appeared after the run that produced the results in the spreadsheet is also
presented, at the end of the output for results.
When a Design run is followed up by an Analysis run, the results can also be transferred,
pasting in the “Major” worksheet shown in Figure 3.44. These spreadsheets can be saved
and used to document a design or analysis. They can also be transferred from the
spreadsheet program to a word-processor for inclusion in a report.
Data for pipes, pits, nodes and sub-catchments can be transferred back into DRAINS using
the Paste Data from Spreadsheet option in the Edit menu. You must first make the
required changes and then copy the entire spreadsheet to the Clipboard using the Copy
option in the spreadsheet. (A quick way of selecting an entire Excel spreadsheet is to click
the cell top-left cell between the “1” and “A” cells.) The changes can then be pasted into
DRAINS using the Paste Data from Spreadsheet option in the Edit menu. Because the
transfers are made via the Clipboard, it is not necessary to have any direct connection
between the spreadsheet file and the DRAINS file.
It is first necessary to establish a system that is capable of being run, such as the example
shown in Figure 3.45.
If you continue, you will then need to nominate a filename for shapefiles in the following
dialog box:
You can see from the existing files in this example how six ESRI SHP files are established.
Another 12 SHX and DBF files will also be produced.
After a name is entered, the process is complete if there are no results. If results are
available, the dialog box shown in Figure 3.48 appears. A suitable name should be added
describing the results; here they are for a 2 year average recurrence interval storm. The
limited size is due to restrictions on the size of column headings in the database files used in
ArcView. After this is entered, the process is finished.
If a background is present in the DRAINS model, this will be transferred with the ESRI files.
The transferred files can now be viewed in ArcView and ArcMap, as shown in Figures 3.49
and 3.50.
A database table is associated with each theme, as shown in Figure 3.51. Note that most
values are specifies as strings of characters, and must be converted to numerical values
using procedures within ESRI programs if these are required to provide displays.
Note that this includes results with the "2Yr" added to headings as a suffix. If another run is
made and the process is repeated with one of the existing shapefiles nominated in the Save
As dialog box, additional results will be appended, as shown in Figures 3.52 and 3.53, where
100 year ARI flows and pipe velocities are added to the 2 year ARI results.
If data from a GIS data base can be assembled into this same format, less the results, the
File ► Import option ESRI Shapefiles… can be used to import data into DRAINS.
It is first necessary to establish a system that is capable of being run, such as the demo
example shown in Figure 3.45.
Selecting the MapInfo files… option from the File -> Export menu presents the following
message:
If you continue, you will then need to nominate a filename for MID/MIF files in the following
dialog box:
You can see from the existing files in this example how six MapInfo MIF files are established.
Another six MID files are also produced.
After a name is entered, the process is complete if there are no results. If results are
available, a dialog box appears such as that shown in Figure 3.48 appears. A suitable name
should be added describing the results; such as "2Yr" for a 2 year average recurrence
interval storm.
A background in the DRAINS model will be transferred with the MapInfo files. The
transferred files can now be viewed in MapInfo, as shown in Figure 3.56.
The merge options allow you to add DRAINS systems together. It is first necessary to export
a system as a merge file, before importing it into another system. The two systems are
linked the pits at each end of a common pipe, which is the lowest pipe in the system to be
added. The procedure is as follows:
(a) Edit the files so that both include two adjacent pits with the same names.
(b) In the file to be added, use the Export a Merge File... option in the File menu to create
and name a .mrg merge file.
(c) Then close the file to be added and open the file with the receiving system. Using the
Import a Merge File… option in the File menu, read in the .mrg merge file created in
Step (b).
(d) The merged system will appear. The orientation of pipes will be that of the receiving
system. You can then tidy this up, save the combined file and make runs as required.
Figure 3.57 The File that is to be Added using the Merge Options
As part of the dedicated links from Autodesk Land Development Desktop, Advanced Road
Design and 12d to DRAINS, results are transferred back to these applications via database
files and the spreadsheet outputs. These processes have been set up to run automatically,
with the other applications interrogating files created by DRAINS.
From Help topics, underlined links can open additional Help topics. The contents and the
index can also be used to find particular topics.
With well over 200 topics, the DRAINS Help system provides a comprehensive guide to the
program, and a glossary of urban stormwater drainage terms and concepts.
It complements the material in this manual, and provides timely advice on enhancements to
DRAINS.
4.2 Integration
Chapter 1 describes how DRAINS can be set up and run. However, it is unlikely that you will
apply DRAINS alone. While it is self-contained, it is more effective when used in conjunction
with other programs.
The general trend in engineering software is towards integration of models on two fronts.
The first is within models such as DRAINS, providing many functions within a single
package, and avoiding the need to transfer data between different programs, which can be
difficult and time consuming. For example, the ILSAX and storage routing hydrological
models within DRAINS can be used together, and piped and open channel drainage systems
can be modelled together.
The second type of integration covers all the programs that can be used to document, model
and present information on urban stormwater systems. These include technical design and
analysis models such as DRAINS, spreadsheets, digital terrain models (DTMs), CAD and
GIS programs, and databases. Just as Microsoft and other software suppliers provide
integrated office suites, allowing data to be transferred easily from one application to another,
a similar arrangement can be made with programs for stormwater systems. What is needed
is a common data description. Information from DRAINS can be transferred to other
programs in spreadsheet and DXF formats in files and via the clipboard, or in database and
GIS formats. Possible linkages between programs are shown in Figure 4.1.
Spreadsheet
Stormwater
System Asset Clipboard
Data Base DRAINS
There is usually some give and take in design, so that the road and allotment layout can be
altered to suit drainage requirements. However, the initial layout made by an experienced
subdivision designer should anticipate potential conflicts.
The products or "deliverables" of the design will be a drainage layer in the drawings with all
drains and channels detailed, together with design calculations. Plans, specifications, tables
of quantities and estimated costs can be derived from these.
The main aims in designing pipe networks with DRAINS are to develop a file that describes
the proposed system, and to produce the deliverables - plans and documentation. The
single .drn file can be run for both Design and checking by Analysis, and can quickly be re-
run, with data and results being transferred to a spreadsheet or report. It forms the basis for
the design variations and checks that may be required.
For a small system, data can be entered from the keyboard into property sheets, as
described in Chapter 1. For larger systems, it is likely that information will be transferred by
CAD file, as described in Section 3.2.2, or from DTMs such as 12d and Land Development
Desktop. Imported data can be augmented with data entered directly into the property
sheets for components. The information for a component is retained when it is copied and
pasted using the Copy Shape option in the pop-up menu for a component and the
associated Paste Shape option. It is often easier to copy and paste an existing component
and to modify its data, rather than to enter all the data each time.
For large systems, the spreadsheet outputs and inputs described in Section 3.5.6 can
provide an efficient means of entering repetitive data. Components can be entered with
nominal values and can then be edited in the Data spreadsheet, before transferring the
information back to DRAINS. The process is shown diagrammatically in Figure 4.2.
When DRAINS starts, it checks for the existence of the file StandardStorms.drn in the
same folder as Drains.exe. You can set up one or more of these in another folder and
transfer them as required.
Using CAD programs such as AutoCAD and DTM programs, catchment areas can be
defined as polygons and the areas directly measured, so that planimeters are not needed.
The lengths of flowpaths, and in some models, their slopes, can also be determined. In
some models, the automatic definition of impervious and pervious areas will be possible
where suitable overlays are available.
For convenience in design the parts of the system can be separated into small sub-systems.
For example, where several branched pipe systems in steep terrain flow to a common open
channel, the pipe systems can be analysed independently, as long as there are no
backwater effects. In flat terrain, with backwater influences, the system will have to be
designed as a whole. Individual systems can be joined using the merging procedures
described in Section 3.5.8.
DRAINS produces information on pipe sizes, invert levels and locations that can be
transferred to drawing programs to produce detailed plans and longitudinal sections. The
spreadsheet tables act as documentation that can be printed or supplied in electronic form to
the consent authority, together with the DRAINS datafiles. Paper print-outs can be obtained
if required.
Since DRAINS runs rapidly and its results are quite apparent, it is easy for consent
authorities such as municipal councils to run files and inspect the results. As discussed in
Section 4.5, the files should be retained and incorporated into the authority's DRAINS model
of its overall drainage system.
The results provided by an appropriate ILSAX Hydrological Model will be superior to those
obtained using the Rational Method, since allowance can be made for detention storages,
and major system modelling is more accurate. Likewise, the Advanced Design Method will
produce superior results to the Standard Method, automatically performing many checks and
repeated calculations that would take a lot of time by trial and error.
The Advanced Method is dependent on having good information on pit inlet capacity
relationships. The best data is available from Queensland where overflows are larger than in
Southern States, and more attention has been given to controlling them. If good quality pit
capacity data is unavailable, the Advanced Method cannot be fully employed.
DRAINS allows hydrological models to be swapped easily, so that it is not difficult to convert
Rational Method models to ones using ILSAX hydrology. Only the hydrological model and
rainfall data need to be changed, and the impervious areas for each sub-catchment to be
split into paved and supplementary areas. The reverse change, from an ILSAX Model to the
Rational Method, can be done even more easily. This might be done to compare the results
give by the models, or to check an older design using Rational Method hydrology.
It is unlikely that designs of pipe systems can be carried out automatically, as in a new
subdivision. Analysis capabilities are required when exploring solutions. Users will probably
have to vary some features by hand to develop a trial and error solution. Fortunately,
DRAINS can be easily edited and re-runs can be carried out rapidly.
While DRAINS allows users to mix pipes that have fixed inverts with pipes with positions that
can be varied, it may not be able to come up with a suitable design in some cases. When
dealing with a complex situation, a strategy would be to see whether DRAINS can come up
with a satisfactory design first, and then make modifications by hand to cope with problems
such as conflicting services and the inability of a pipe system to match the inverts of the
downstream pipe to which it must connect, while carrying the required design flows.
DRAINS can "design around" existing services or utilities, and can allow for surface levels all
the way along a pipe if suitable survey data is provided in the Pipe property sheet. However,
the solution provided may set pipe inverts too deep, so that it will be necessary to make
adjustments by hand. It may be necessary to use stronger pipe classes (with greater wall
thicknesses), multiple pipes or box-section conduits to reduce the cover requirements. In
complex cases, relocation of existing stormwater pipes or other services may be the best
solution.
Because re-developments usually involve an increase in the density of development and the
percentage of impervious area, several drainage authorities have imposed on-site
stormwater detention (OSD) requirements. These have become an important and often
complicated issue for designers. The Upper Parramatta River Catchment Trust has been the
most influential developer of OSD design procedures in New South Wales, introducing
requirements such as a permissible site discharge (PSD) in L/s/ha of catchment, and site
storage requirement (SSR) in m3/ha.
The detention basin routing has to be explored by trial and error, but the ability to quickly edit
the data and re-run the model makes this a fast process.
The processes in creating DRAINS files for Analysis are almost the same as for Design.
Invert levels of all conduits must be defined.
Since an existing system has to be modelled in some detail, it will be necessary to draw
information from GIS and data base sources. If these are unavailable or inadequate, it will
be necessary to carry out topographic surveys to determine the exact positions and levels of
system components, including:
• surface levels of pits,
• invert levels of pipes, including if possible, those in sealed pits and junctions,
• lengths of pipes,
• floor levels of houses and businesses, and driveway levels where flows may enter
properties and yard levels where ponding may occur.
Inspections are needed to define many aspects of drainage systems, such as low points on
roadways and likely overflow paths. It is likely that the same areas may have to be inspected
two or three times during modelling to define drainage components and paths exactly.
Information can also be sought from residents about their experiences of flooding during
these visits. Closed circuit television (CCTV) investigations can provide detailed information
on pipes and defects such as erosion, cracks and faulty joints.
Gauging for a period of at least 3 months will probably be necessary. When the model is run
with the recorded data, the times of flow can be varied by altering values in the spreadsheet
output and re-inserting these into DRAINS. The DRAINS model can be calibrated or “tuned”
so that the times of the calculated hydrograph peaks match those of the recorded ones. A
similar process can be carried out by varying the Hydrological Model parameters and
percentages of land use, to make the calculated flow peaks or volumes match the recorded
ones. This is more difficult because pervious areas may only contribute flows in larger
storms, and the gauging period may be too short or dry to record significant runoff-producing
storms.
From the available mapping and the survey data, DXF or DWG drawings can be prepared,
and a suitable file prepared with the three layers containing pits (as circles), pipes and a
background. This can be then imported into DRAINS to provide the initial file for the data
entry and modelling processes.
In modelling existing systems, a difficult issue will be the definition of the flow paths taken by
flows from paved and grassed surfaces, as shown in Figure 4.3. These will be greatly
influenced by the size and arrangement of allotments and the buildings on them, and
especially by the style of fencing along allotment boundaries.
Another difficult issue will be the ponding of stormwater on streets and in backyards. This
occurs where development has occurred in the natural floodway areas, and various barriers
to flow have been erected, including road crowns, road embankments, walls and fences. It is
fairly easy to see where stormwater will run into properties. Usually those on the
downstream side of a road at a low point will be affected, as shown in Figure 4.4.
Storages within allotments can be very complicated, with flows being blocked by gates and
fences, so that several instances of ponding may occur on the one property. Some
situations can be modelled readily, such as flow under a fence represented as a sluice gate,
or flow over a low wall as a weir flow. However, in many Australian situations, the barrier
may be a metal Colourbond fence extending to the ground. Such fences can probably hold
back stormwater to a depth of 1 m or more. When failure occurs, there may be catastrophic
effects from the resulting rush of water and debris.
Modelling such events is difficult. Our knowledge of how they are initiated is poor, and
DRAINS does not model “dambreaks” of this type.
Existing systems that have been augmented can have two pipes with different characteristics
running more or less parallel. These can be modelled as multi-channels if there are no
The spreadsheet documentation provided by DRAINS is very useful for recording results,
which can be separated into worksheets and suitably tagged.
When analysing very large systems (say 500 pipes and over) computation times can be quite
long with multiple storms. It is therefore necessary to plan the analysis work, starting with
storm events likely produce the highest flowrates. Once the system has been refined, final
runs can be made with a wider range of rainfall patterns of different average recurrence
intervals and durations.
Once a basic working model has been established, the likely flowrates, heights of storages,
flooding impacts and resulting damages can be assessed. “Trouble spots” can be identified,
and remedial works can be considered. The basic DRAINS model can then be varied to
produce a number of models for assessing different remedies. In some cases the remedies
will interact with each other, some reinforcing the beneficial effects of other remedies, others
diminishing these. This makes the consideration of options quite complex.
The Rational Method Analysis procedure should not be used to simulate the behaviour of
existing systems, since the various flow peaks calculated can occur at different times, and
the flowrates obtained from combining peak flows are approximate. This procedure should
only be used to check newly-designed systems.
When the drainage system is constructed, it is likely that some details will have been
changed during construction. The model should be updated to reflect the work-as-executed
information. It will then require further modification as, whenever:
• additional drainage systems are connected,
• rezonings and re-developments create more impervious areas and increase runoff
volumes and rates,
• possible flow diversions occur within the catchment, and between it and other
catchments,
• compensatory detention storages are provided,
• additional information and experience about the drainage system accumulates,
• design rainfalls are revised, and climatic change effects occur,
• the system deteriorates and defects due to damage and ageing of assets become
apparent,
• remedial works are constructed, and
• design standards change.
Many municipalities and stormwater authorities do not have full information on their systems.
Nevertheless, they can start with this incomplete data, setting up data bases and models with
the available information, and then refining these. Experience during storm events special
surveys to determine pit and pipe invert levels, CCTV inspections, preparation of lists of
trouble spots and asset registers will provide added information, so the records can be
gradually expanded and the modelling improved.
As shown in Figure 4.5, DRAINS provides transfers to GIS programs in the form of ArcView
shapefiles and Mapinfo MID/MIF files, the most commonly-used GIS formats. The
connection of DRAINS to the GISs of drainage authorities allows the results of DRAINS
analyses to be included in the GIS. These can include flowrates and hydraulic grade line
levels for average recurrence intervals of 1, 2, 5, 10, 20, 50 and 100 years, plus probable
maximum precipitation (PMP) storms. These results can be mapped and displayed in many
ways, using colour-coded symbols and lines. The GIS can also act as a means of querying
the underlying database, so that flows or HGL levels at particular locations can be checked
on-screen.
Ultimately, drainage system managers can develop systems where revised DRAINS models
can be created from information on previous DRAINS models in their GIS. As new
developments and re-developments occur, it will be possible to include these. Results from
various models can be retained in the GIS system. The combination of DRAINS with GIS
allows managers to maintain and ongoing record of their drainage systems that included
records of performance and flooding risk.
Sub-catchment areas, channel reach lengths and other characteristics are then measured.
The number of sub-areas should reflect the detail of the information required and the
important features of the catchment, such as reservoirs and changes in the type of channel.
Streamflows and other data suitable for calibration of the model are then assembled. Ideally,
there should be at least three recorded flood events. Loss parameters and initial values of
the parameters (kc in RORB, BX in RAFTS and C in WBNM) are established.
The program is then run and the outflows are compared with the calibration data, or rural
catchment flood estimates developed by methods in Chapter 5 of Australian Rainfall and
Runoff (Institution of Engineers, Australia, 1987). The parameters are then adjusted and the
final calibration flowrates determined. If more than one storm event is available for
calibration, it may require different parameters to obtain exact matches to recorded peak
flows. A compromise set of parameters must then be selected.
With the parameters established, the model can be used to estimate the flows from large
floods such as a 100 year average recurrence interval flood. The effects of detention basins
and stream break-outs or diversions can be assessed.
If you wish to combine the storage routing model results with the open channel hydraulic
calculations available in DRAINS and/or an ILSAX model, this can be done to obtain more
detailed results. The open channel and ILSAX models can be set up in the usual way. This
integration of models should be useful in situations where there is interaction between a
large watershed and a smaller urban catchment.
5.2.1 Programming
DRAINS is written in C++ and works on PCs with the Microsoft Windows, NT and XP
operating systems. The calculation procedures from the PIPES program are used to model
pressurised flow situations. It inputs and outputs data in binary, spreadsheet CLV, DXF,
DWG, ESRI shapefile, MapInfo MID/MIF and data base formats.
In order to run, DRAINS requires information on run specifications, rainfall data and a pipe or
channel system. This data is stored temporarily in the computer’s memory and, more
permanently, in a binary data file with a .drn suffix, such as the sample files that have been
described in this manual.
After a data file has been saved, you can re-open it in DRAINS and modify the data. Since it
is saved in binary format, it cannot be viewed or changed using a text editor. The binary file
formats change as DRAINS is updated, but will always be back-compatible. That is, with the
current version of DRAINS, you will be able to open and use files created in previous
versions. However, you will probably not be able to open files created with a later version of
DRAINS than the one you are using - it is not forward-compatible.
Each DRAINS .drn file is effectively a data base describing a drainage system and its
components, together with reference data bases for pipes, pits and overflow routes. Most of
the data on the drainage system can be readily accessed in ASCII or text form using the
spreadsheet output option described in Section 3.5.6. Data on rainfall patterns, hydrological
models and run specifications are not transferred to spreadsheet.
As well as the reference sets of pipe, pit and overflow route types and associated information
contained in the .drn file, a more general set called Drains.db1 is contained in the
C:/Program Files/Drains/Program folder along with the Drains.exe file. (The
older pit type format described in Section 2.3.2(b) was stored in .db files.) A new set of data
information, such as the regional sets for New South Wales, Queensland and other states,
can be installed by using the Default Data Base option in the Project menu. It is important
that you determine what you require before starting a project, as it may be difficult to change
the available options later.
DRAINS uses metric units throughout. Where possible, it follows SI conventions for these,
but in many displays and outputs it is not possible to show superscripts. Thus “cu.m” and
“cu.m/s” are frequently used in place of “m3” and “m3/s”.
5.2.4 Operations
As shown in Figure 5.1, you can operate DRAINS through a number of processes, such as:
(a) Data entry and file storage,
(b) Performing calculations,
(c) Inspection and possible storage of results,
(d) Changes or corrections to data, and re-running calculations,
(e) Transfers of data and results to files and other programs.
Import Export
data Spreadsheet file results
DRAINS model in
computer memory from Run DRAINS using Results stored in
keyboard and other calculation engine in computer memory,
sources, displaying this Design or Analysis to be inspected
on screen modes
Export Import
Save Open file file
file file
Performing calculations in Action (b) is a batch process. Once started it continues without
intervention by the user, unless it is aborted by pressing the Esc key. On the other hand,
Actions (a), (c) and (d) are event-driven. They can be carried out in many different ways,
depending on your preferences. The programming style follows Microsoft Windows
conventions, so that it will be familiar to most users.
Applications using hydrological storage routing models may only go through the first of these
procedures.
Figure 5.2 shows the calculation procedures employed with the ILSAX hydrological model.
The standard design method involves a Design process that determines pipe sizes and invert
levels, followed automatically by an Analysis that checks the design using the minor storms
and presents results. This needs to be followed by an additional check by analysing major
storms. The advanced method performs calculations involving both minor and major storms,
without presenting results. Additional analyses with minor and major storms are required to
check the design.
For Rational Method calculations, the processes are simpler. The hydrological calculations
produce peak flowrates, to be used in design, and in a simplified analysis procedure. Details
of the processes in Figure 5.2 are given in the following sections.
The various run options are described in Section 3.4. Before these become available in the
Run menu, the program performs checks to confirm that:
• a hydrological model and rainfall patterns have been specified,
• the components of a system are joined correctly,
• the drainage system components have been fully specified.
The run options in the menu are greyed out if these conditions are not met.
Once a run begins, DRAINS sorts through the various components to define linkages
throughout the system and the order in which calculations should occur. Using the
coordinates of the objects, it identifies connections between pits or nodes and links,
↓ ↓ ↓
Figure 5.2 The Core Calculation Procedures for Pipe System in DRAINS
(using ILSAX Hydrograph Model)
such as pipes or channels, at locations where the positions of the ends of a link are within
the symbol of a node in the Main Window. The connections between pits and sub-
catchments are established where the symbols for these overlap. The pits or nodes at the
extremities of drainage systems are identified as those having no incoming links.
With the ILSAX model, hydrological calculations involve the computation of the hydrographs
from the paved and grassed surfaces of each sub-catchment using the Horton loss model
and the time-area routing methods described in Section 6.3.3. They are carried out in the
same way for both Design and Analysis runs.
In the calculated hydrographs, flowrates are defined at times that are multiples of the
calculation time step defined in the Options property sheet called from the Project menu or
automatically defined by DRAINS using a criterion that unpressurised flows should take at
least one time step to travel through any pipe in the system. Hydraulic modelling
considerations are more critical than hydrological factors, and these determine the time step
required.
The hydrographs in all links begin at the same time, the start of the storm rainfall. Any
baseflows and user-provided inflow hydrographs introduced at pits or nodes are assumed to
begin at this same starting time. User-provided hydrographs can have any time step, but will
be converted to the calculation time step by linear interpolation.
All hydrographs produced for paved and grassed sub-catchment surfaces, total sub-
catchments, pipes, channels, overflow links and detention basins can be viewed as graphs or
tables using the pop-up menus for individual components, and can be transferred to the
Clipboard, as shown in Figure 5.3. They also can be printed out in the Print Data and
Results… option in the File menu. Calculated paved and grassed hydrographs are stored
temporarily as part of each sub-catchment “object”.
With the Rational Method model, peak flows are calculated and stored with the data for each
component.
The procedures for the storage routing models emulating the RORB, RAFTS and WBNM
models, are simpler than those from ILSAX, involving the non-linear relationship S = k.Qm
between storage S and discharge, Q. Results are presented in the same way as ILSAX
hydrograph outputs in Figure 5.3. Details of the three types of storage routing model are
provided in Section 6.5.
Once hydraulic calculations start, it is necessary to determine the inflow into the pipe and
channel system at each time step. At each node, the following flows are first combined:
• the flow off paved and grassed areas on the local sub-catchment,
• any overflows from upstream pits or detention basins that are directed to this destination,
• any baseflows or flows from user-provided inflow hydrographs that are applied at the
surface for a pit or simple node.
This surface flow is assumed to enter the system without restriction at an ILLUDAS type pit,
a simple node, detention basin, culvert or bridge. For an on-grade or sag pit, the pit capacity
relationship defined in the Pit property sheet is applied to estimate the inflow rate (see
Section 2.3.2).
For Design calculations, a pipe system will be sized to carry all flows that enter the system.
The only overflows will be the bypasses caused by restrictions on inlet capacities. In
Analysis, there may be upwelling of flows from the pit due to the capacity of the downstream
Inflow
Upwelling
Upstream
Pipe Flows Downstream
Pipe Flows
DRAINS calculates upwelling flows using hydraulic analyses. No outlet restrictions are
placed on upwelling flows unless the pit has a bolt-down lid. Otherwise, it is assumed that
any flowrate can escape from a pit.
No overflows can occur at ILLUDAS type pits, or from a simple node in DRAINS. Situations
where overflows for breakouts occur from channels might be modelled by substituting a
detention basin, as noted in Table 2.2.
With these calculations determining flowrates at each time step, full hydrographs of inflows
and overflows can be built up. With the Rational Method, only peak flow conditions are
considered. The methods applied to this are the same as those applied to any of the
instantaneous flows occurring at the various times during a hydrograph generated by ILSAX
or storage routing models.
In the Standard Design mode, DRAINS makes one pass down the drainage system from pits
at the tops of lines down to the main outlet. At each pit it determines the maximum pipe
outflow, allowing for inlet flows, flows in upstream pipes, and any baseflows or user-provided
hydrograph flows. It then determines suitable pipe sizes and invert levels, taking account of:
• the roughness and the allowable cover depth associated with the chosen pipe type,
• the values set of minimum pipe slope, pit freeboard and fall in the Options property sheet
opened from the File menu,
• a restriction preventing pipes decreasing in diameter as the calculations move
downstream,
• likely pit pressure changes at pits in full or part-full pipe flows and
• the hydraulic capacities of pipes with various diameters and slopes.
The selection of diameter is mainly based on allowable cover depths. The aim is to keep the
pipe as shallow as possible, and pipe sizes are increased where necessary to achieve this.
(In cases where pipes are to be set deep enough to pass under other services, such as
The selection of invert levels is mainly dictated by cover depths and slope restrictions.
Allowance is made for cover depths at intermediate levels along a pipeline, as defined in the
Survey Data property sheet called from a Pipe property sheet (see Section 2.3.4). A result is
shown graphically in Figure 5.5.
This output also shows a pit with a significant drop, which can be a consequence of aiming to
keep the pipe system as shallow as possible. If you wish to grade the upstream pipe down
to the pit, it will be necessary to adjust the invert levels and run the model in Analysis mode,
or make the pipe inverts fixed.
DRAINS can automatically design to avoid fixed services, where possible, using service
location information entered in the Survey Data property sheet and a minimum design
clearance to services set in the Options property sheet called from the Project menu. One
such service is shown in Figure 5.5 and also in Figure 5.6.
If a Design run is made with the positions of some of the pipes fixed, the results must be
carefully checked, especially if these are located in the middle of lines. In some complex
Design cases, pipes entering pits might be lower than a fixed pipe flowing out. The fallback
in this case is to run in Analysis mode and to vary pipe sizes and invert levels individually to
achieve a satisfactory design.
In carrying out Standard Design calculations, DRAINS moves down the system, setting pipe
sizes and invert levels. The procedure used produces similar pipe sizes to Rational Method
design procedures such as the one in Chapter 14 of Australian Rainfall and Runoff (1987).
Where multiple storm patterns are specified, the program repeats the downwards pass for
each storm and selects the pipe diameters and invert levels that convey the most critical
flows.
With the Standard Design Method, once the pipe details are established, DRAINS performs
an Analysis run using the same storm or storms that were used for Design. It checks
whether the hydraulic grade line (HGL) rises above the freeboard limit set below the gutter
level (commonly 150 mm) and reports on this, and on the presence of bypass flows, when
the run is completed. Thus Analysis provides an independent check on the adequacy of the
design using a comprehensive, full hydrograph analysis with the same design rainfall data.
This method is based on pipe drainage system design procedures in the Queensland Urban
Drainage Manual (Neville Jones & Associates et al., 1992).
The method takes as its starting point the flows along overflow routes. It sets appropriate
safety levels for these, in terms of tolerable depths in the minor and major storms and a
maximum velocity x depth product. A point along each flow path must be nominated, by
specifying a cross-section from the Overflow Route data base as shown in Figure 5.7, a
percentage of downstream catchment contributing to the flow, and a longitudinal slope
The basis for selecting the percentage of the downstream sub-catchment is explained in
Section 2.3.6.
Having established safe flows for each flow path, the method then determines the pit and
pipe sizes needed to restrict the surface overflows to the safe limits.
This process involves the organisation of pit types into families and sizes, in a similar way to
the classification of pipe types and diameters. With the user having defined the pit and pipe
types required, DRAINS automatically searches through the available sizes to determine the
required ones at each overflow location. It also selects pipe sizes so that at a major storm
level, such as the 100 year ARI storm event, HGL levels at pits are still below the ground
surface. This ensures that the drainage system does not completely fill with water, and the
pipe flows will be maintained even when stormwater ponds over pits.
Because the method involves consideration of both minor and major storms, DRAINS does
not automatically perform an analysis of the minor storm patterns, as it does in the Standard
Design method. You must make follow-up Analysis runs using the options in the Run menu
for minor and major storms.
In some cases DRAINS cannot arrive at a solution that meets the safety requirements, most
obviously when the flow from a sub-catchment is much larger than the capacity of any of the
pits that can be selected. DRAINS returns the notice in Figure 5.8, advising that the pipe
system must be changed or that different pits are required.
The advantage of this method is that quite complex calculations, often involving many runs
and alterations to a proposed drainage system, can be performed in a single run. The
results can be checked by Analysis runs to ensure that the design conditions are met. By
taking full advantage of allowable surface flow capacities, the sizes and costs of pipe
systems can be minimised.
(a) Introduction
Originally, the hydraulic procedures in DRAINS were intended to be rather simple, based on
projections of HGLs upwards through pipe and channel systems from a set tailwater level for
each time step during a storm in which flows and water levels increased and decreased.
Each time step was to be effectively independent, and no attempt was to be made it solve
the equations of mass and momentum conservation. Even with this simplification, it became
apparent that a different method was required for pressurised flow, when pipes run full under
pressure. The "engine" from the PIPES program was used for this purpose, modelling
pressurised flows accurately and producing complex hydrographs and HGL patterns. It is
desirable that users have some understanding of why these occur.
Drainage systems are analysed by making downwards and upwards passes through the pipe
or channel network at each time step, going from each pit or node to the next one
downstream or upstream.
The first pass moves downwards from the top of each line in the system, establishing the
surface flows arriving at each node by adding flows from the local catchment, overflows from
upstream and user-provided flows. Using the pit inlet capacity relationship, bypass flows are
determined. The flow entering the pit is then added to any flows through upstream pipes and
possible user-provided hydrographs to define provisional pipe flows.
When the calculations reach the system outlet or outlets, DRAINS makes the upwards pass,
starting from the tailwater level at the outlet. Allowing for pipe friction and pressure changes
at pits, it defines the position of the HGLs at pits and nodes, and if necessary, modifies the
flowrates in the pipes and the corresponding overflows. For part-full pipe flow, this process is
carried out by projecting HGLs upwards and allowing for pressure changes at pits. If a pipe
flows full, a pressurised flow calculation procedure is used to define HGLs at pits and
flowrates in pipes.
Whenever it encounters a junction, DRAINS projects HGLs up both branches from the pit
water level. If the calculated water level in a drop pit is determined to be below the invert
level of an incoming pipe, the tailwater is set as the higher of the normal and critical depths in
this pipe, and upwards HGL projections are continued.
The calculation time step defined by DRAINS is determined so that it will take at least one
time step for water to travel through each conduit. The minimum time step is 0.025 minutes
or 1.5 seconds.
DRAINS calculates HGLs at nodes and inside pipes for subcritical part-full flows, but it only
presents the results at nodes. It defines flowrates in links such as pipes or channels, and
provides a continuity check in the spreadsheet output summing the inflows and outflows at
each node. The flowrates presented for pipes are those calculated at their upper ends, so
that the flows displayed in DRAINS outputs at a particular time will probably differ from the
flowrates emerging from the pipe at that time. If the pipe is unpressurised, these outflows will
equal the flows that entered a conduit a certain number of time steps previously (depending
on the pipe length and flow velocity). If it is pressurised, there is no time delay. DRAINS
manages the transfers between part-full and full-pipe flow so that there are only small
continuity errors.
Within pipes flowing part-full, the water level is taken to be the normal depth for supercritical
flow, while for subcritical flow water surface profiles are calculated by the standard step
method using the Colebrook-White or Manning’s Equations.
Within pipes the open channel flow is always controlled from downstream, so that the
backwards projection of HGLs applied in DRAINS is valid. While this assumption may
overestimate HGL levels somewhat, it prevents calculations from being unstable or involving
extensive iterative calculations to locate hydraulic jumps.
For part-full flows, the pressure change at each pit can be defined by three methods:
(i) The change can be a constant, with the amount increasing for greater changes of
horizontal direction. (The default value is 35 mm.)
(ii) The special Queensland Urban Drainage Manual procedure described at the end of
Section 6.7.3 can be applied. (This carries out a check on levels to decide whether the
HGL is above the obvert of the pipe leaving a pit.)
(iii) A pressure change coefficient applicable to part-full flows can be specified, which is
multiplied by the velocity head of the downstream part-full flow.
There is little information available about pit pressure changes and energy losses in pits with
part-full flow. It is not possible to be definite about this, and it is suggested that results be
inspected to endure that one of the above assumptions does not lead to improbably high
increases in HGLs. If this is the case, alternative methods may be applied.
When the pit becomes submerged, a pit pressure change relation is applied, using the kU
factor (see Figure 5.9) specified by the user in the pit’s property sheet.
The change from part-full to full pipe flow often results in a large increase in the pit pressure
change, setting up a surge that may appear as a sudden jump or "spike" in hydrographs and
HGL plots. This raises the HGL level and transmits a wave upwards through the system.
Thus a jump in pressure that occurs at a pit may be due to another pit somewhere
downstream filling suddenly. A similar drop in HGL and pressure may occur when a full-
flowing pipe changes to part-full flow.
While these changes are modelled in mathematical form in DRAINS, using the relationships
for part-full pit pressure changes and the different relationships applying to full-pipe flows, the
2
k LV0
2g TEL
2
k uV0
HGL
2g
Pit
V0
pressure rise phenomenon does actually occur, so that DRAINS presents generally realistic
pipe system behaviour.
If the water surface rises above a pipe obvert, DRAINS applies the pressurised system
hydraulics model.
If the HGL projects above the surface of an open pit, DRAINS will specify a pit overflow or
upwelling, caused by downstream backwater effects or inadequate downstream pipe
capacity. During upwelling of an open pit, the highest position of the HGL is fixed by
DRAINS at the surface level for an on-grade pit, or at the maximum specified depth of
ponding for a sag pit. If there is scope for the HGL in the downstream pit to go higher, the
slope of the HGL in the pipe is reduced, since it is “squeezed” between the upstream pit’s
fixed level and the downstream pit HGL level. The full-pipe flow slows down and the flowrate
decreases, producing a hollow-topped hydrograph of the type shown in Figure 5.10.
If a sag or on-grade pit is defined as being sealed using the check box labelled “Pit has bolt
down lid” in the Pit property sheet, the HGL can rise above the surface without any upwelling
occurring.
The analysis in the ILSAX program would have simply cut the top of the hydrograph at an
assumed constant pipe capacity, as shown in Figure 5.11. The DRAINS result is more
accurate, and in the case of Figure 5.10, more conservative.
If a pipe system fills completely, the flows through it will ultimately be determined by the
surface levels. If a pit at the downstream end of a pipe has a surface level that is the same
as the upstream pit, the HGL will be horizontal when the system is fully submerged. If the
downstream pit’s surface level is higher than the upstream pit's, a flow reversal will occur
when both pits fill to their surface levels. DRAINS will record negative flows.
Once the peak of a hydrograph is past, and pipe flowrates reduce, a reverse process takes
place. HGLs are lowered and the system goes from full-pipe to part-full pipe flow. If the
rainfall pattern is complex, with several peaks, then levels may move up and down in
response to this, and flows may switch from part-full to full-pipe flow many times.
Along with these backwater effects, there can be sharp jumps in water levels and flowrates
when pipes change from part-full to full pipe flow and vice versa. The main cause is that,
with part full flow, the head loss or pressure change at this pit is usually assumed to be quite
small, say 35 mm. When the downstream pipe runs full, the pressure change becomes
kU.V2/2g, which may be much larger. The water level increases significantly within one time
step in response to this.
It may be difficult to trace the causes of jumps of water level in a large pipe system, as
effects can travel through the system rapidly and affect many other pipes.
DRAINS results appear to be consistent with hydraulic principles and with results from other
detailed hydraulics programs.
At system outlets, DRAINS sets a tailwater level, depending on the entries in the Outlet
property sheet. If a free outfall is specified, it determines the higher of the normal and critical
Where a drop pit is so deep that the pit water surface is below the invert of the upstream
pipe, the starting level for upstream HGL projections will be set in the same way as for a free
outlet. It will be the higher of the normal and critical depths in the upstream pipe. In effect,
the open channel flow calculations start again at this pit.
(e) Overflows
Overflows follow the defined overland flow path to a destination, with flows being lagged by
the specified time delay, which must be at least one calculation time step.
Although a slope and cross-section must normally be specified for flow paths, the network
calculations do not allow for differences in levels, so that the path can go from one pit or
node to another at a higher surface level. DRAINS provides warnings of such situations, and
you are advised to resolve such anomalies where possible.
In calculating open channel flows it is important to note that DRAINS does not model
supercritical flows. If flows do become supercritical, the water surface level is set to the
critical depth. Full modelling of supercritical depth would involve modelling of hydraulic
jumps and search procedures for their location. The selection of critical depth instead of the
lower normal depth for supercritical flows will overestimate water depths at some locations,
but since supercritical flows can be unstable and revert to higher depths if they become
subcritical due to an obstruction, this procedure is considered to be reasonable.
Hydraulic calculations in open channels (prismatic, irregular and multi-channels) are more
complex than in non-pressurised pipes. In such channels, DRAINS determines intermediate
cross-sections by interpolation and carries out complex backwater calculations. In-non-
pressurised pipes DRAINS typically assumes normal depth (or critical depth if the flow is
supercritical).
The calculations for detention basins in DRAINS can be complex because the elevation-
discharge relationship will change if the downstream tailwater level submerges the outlet.
This can happen at many time steps during a DRAINS run, so that the relationship changes.
By contrast, ILSAX and most other models for trunk drainage systems assume that the
relationship is fixed. Thus DRAINS can model interconnected basins.
DRAINS provides several warning messages for detention basins. These should be heeded
and changes made where possible to eliminate conflicts.
Culverts and bridges are treated like detention basins, in that they obstruct flows and can
have low level outlets (the channel under the roadway) and high level outlets (overflows over
the road). They do not have any associated storage. Where this may be significant, they
can be modelled as a detention basin. DRAINS returns similar information in the Main
Window for detention basins, culverts and bridges - the upstream and downstream water
levels.
All of these relate to physical quantities that are easily understandable; so that values that
are estimated, as is usually the case, will not be greatly wide of the mark.
Where rainfall and runoff data for storms is available, the hydrological modelling in DRAINS
can be improved by calibration, though not to a large extent (O’Loughlin, Haig, Attwater and
Clare, 1991). Times of entry and travel through a drainage system can be defined more
accurately.
Less accurate calibrations can also be carried out on the basis of ponded volumes. If rainfall
is available for a storm, DRAINS can estimate the stored volume at a location where depths
have been observed. The volume from DRAINS can be compared with that corresponding to
the maximum depth observed.
In hydraulics calculations for piped systems, DRAINS results can be significantly different
from those of ILSAX when a system is pressurised and upwelling flows occur. The DRAINS
hydraulic calculations are more accurate, as they allow for HGLs with pit pressure changes
and tailwater levels, and for both pressurised and unpressurised flows. DRAINS tightens up
many loose procedures in ILSAX.
Detention basin routing results match those in ILSAX and checks on culvert and bridge
calculations show that these produce results consistent with the sources for the methods
used.
In comparisons with data from gauged urban catchments, ILSAX has been shown to provide
results that are at least as good as other urban hydrology programs such as SWMM (see
Vale, Attwater and O’Loughlin,1986; O’Loughlin et al, 1991; and Dayaratne, 1997, 2000).
Table 5.1 ILLUDAS-SA and Observed Results (Mein and O'Loughlin, 1985)
Table 5.3 and Figure 5.11 present more recent comparisons between ILSAX, DRAINS and
observed data for 25 storms recorded at the University of Technology, Sydney gauging
station at Hewitt, Penrith. This and other comparisons with data recorded at Penrith (Pereira,
1998; Tran, 1998) have shown that ILSAX and DRAINS produce similar hydrographs at
catchment outlets, except in large storms where backwater effects influence the pipe system
hydraulics.
Open channel hydraulics procedures were tested using examples from ACADS (1981) and
other material. Culvert and bridge calculations have been checked for agreement with
culvert equations and procedures for bridges in the AUSTROADS (1994) Waterway Design
manual.
Results for storage routing models have been checked against examples in manuals for
other programs and against projects carried out using RORB, RAFTS and WBNM.
Table 5.3 Comparisons between ILSAX and DRAINS Calculated Flows and Observed
Flows at the Hewitt Gauging Station, Penrith, Sydney
(O'Loughlin, Stack and Wilkinson, 1998; Shek and Lao, 1998; Chan, 1998)
5
Calculated Flowrates (m 3/s)
4 ILSAX
DRAINS
Linear (ILSAX)
Linear (DRAINS)
3
0
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5
3
Recorded Flowrates (m /s)
6.2.1 General
DRAINS is the result of a the following chain of development that originated with the U.K.
Transport and Road Research Laboratory (TRRL) Method in the early 1960s:
The TRRL method was developed by Watkins (1962) following extensive studies by the U.K.
Transport and Road Research Laboratory. Storm rainfalls and runoff were recorded over
several years on twelve catchments. The Rational Method and other hydrological models
were applied to the catchments, and their flow estimates were compared to this recorded
data. From this research, a general design procedure was developed (U.K. Transport and
Road Research Laboratory, 1976). This was released as a FORTRAN program in 1963,
replacing the Rational Method as the most widely-used drainage system design method in
the U.K.
In the TRRL Method, a time-area diagram for each sub-catchment is combined with a design
rainfall pattern to produce a flow hydrograph. After entering the pipe system, this joins with
other hydrographs of flows through upstream pipes, and is routed through the available
storage in the downstream pipe. Pipes were sized to carry the peak flowrate of the routed
hydrograph. This procedure avoided the assumptions about peak flows and times required
in the Rational Method. Since total storm hydrographs were generated and combined,
The research of Watkins indicated that the greatest runoff peaks were produced by
thunderstorms in summer when soils were dry. Thus it appeared that runoff from pervious
areas, and impervious areas draining onto these, could be ignored in the U.K. This
assumption was incorporated into the TRRL Method. When the Method was tested in the
U.S. by Stall and Terstriep (1969) and in Australia by Heeps and Mein (1973) and Aitken
(1975), all found that the neglect of pervious area runoff was a serious limitation.
To overcome this, a "tropical" version named HYDRAN was developed using African data
(Watkins and Fiddes, 1984). This allowed for pervious area runoff, routing it through linear
storages. It was not widely used. The TRRL Method's handling of storage on catchment
surfaces and in the pipe system was criticised in the U.K., and it was replaced by the
WASSP suite in the 1980s. Although the TRRL Method is now superseded by other
methods, its originators deserve kudos for developing the first practical computer rainfall-
runoff model for urban drainage systems.
The Illinois Urban Drainage Area Simulator, ILLUDAS, was developed and extensively tested
by Terstriep and Stall (1974), who adapted the TRRL Method to cope with pervious area
runoff and added other features. Although it was popular among researchers, it has not
been widely used by designers in North America. In most design applications ILLUDAS and
the SWMM (Stormwater Management Model) have been overshadowed by the U.S. Soil
Conservation Service programs and other, simpler methods.
ILLUDAS is versatile, and many variants with special features have been developed.
A number of tests have been made by Terstreip and Stall (1974) and by Watson (1981) who
tested the model on two South African gauged catchments.
Besides testing ILLUDAS, Watson (1981) produced a new metric version named ILLUDAS-
SA, with many additional features. This was the basis for the ILSAX program. In Australia,
ILLUDAS-SA and various development versions of ILSAX were applied to data from gauged
urban catchments in Sydney and Melbourne by Cartwright (1983), Mein and O'Loughlin
(1985), Vale, Attwater and O'Loughlin (1986), and others. The first practical application was
on a large-scale drainage study of Keswick and Brownhill Creeks in suburban Adelaide in
1982-83. Carleton (1983) then applied ILLUDAS and the EXTRAN block of SWMM to
drainage problems in Wooloomooloo, Sydney.
6.2.4 ILSAX
ILSAX was developed by Geoffrey O’Loughlin between 1982 and 1986, with the aim of
producing a better stormwater drainage design program than the Rational Method. The later
part of this development was integrated with the preparation of the chapter on urban
stormwater drainage in Australian Rainfall and Runoff, 1987,. ILLUDAS-SA was adapted to
include, among other additions, facilities for routing overflows from pits elsewhere in a
drainage system, so that it could satisfactorily model major event flows as part of the
major/minor system of design recommended in Australian Rainfall and Runoff.
The name ILSAX stood for "ILLUDAS-SA with something extra". Although some versions
were distributed on tape for mainframe computers, the program only started to be used
In 1990 O’Loughlin produced a separate data entry program for ILSAX, named ILSIN. This
QuickBASIC program provided screen data entry and checking for ILSAX Input files and a
Help System.
The early testing described above showed that ILSAX was at least as accurate as other
urban hydrology models, such as SWMM. The hydrological model has proved to be robust
in many applications. The simplicity of its parameters has made it easy to understand and to
use, although its pervious area calculations are relatively complex.
In the 1990s ILSAX received a new role as a tool for analysis of on-site stormwater detention
systems. It is still being used for many applications in 2003, 17 years after its first release,
but has been superseded by DRAINS. ILSAX and its manual can be obtained from the
website www.eng.uts.edu.au/~simonb.
6.2.5 DRAINS
This new program was developed as an extension of ILSAX, providing a modern Windows
interface and overcoming many of the flaws of ILSAX, particularly the limited modelling of
pipe hydraulics. However, it is intended to be much more, a general, integrated model of
gravity drainage systems.
Users who have the Rational Method version can switch between the ILSAX model and
Rational Method models by selecting an appropriate default model in the Hydrological
Models dialog box. The spreadsheet output has provided an efficient means of connecting
with other programs, as well as an organised method of presenting data and results.
6.3.1 General
Simulation models such as ILSAX and DRAINS require a model to transform rainfall patterns
to runoff hydrographs. These models deal with the part of the hydrological cycle known as
the rainfall-runoff process, shown in Figure 6.2.
Precipitation
Evapo-
transpiration Snow Rainfall
SNOWPACK
STORAGE
Snowmelt
(delayed)
Surface Runoff
DEPRESSION
STORAGE C
(fast) H
Ponding A
Infiltration N
N
E
Interflow
SOIL MOISTURE L
STORE (slow) S
T
Percolation O
R
Groundwater A
Flow G
GROUNDWATER
STORE E
(very slow)
Many such models and variations on models are available. The decision on which to use
depends on the capabilities of the model and also on the extent to which engineers and other
users of hydrology are accustomed to it. In most parts of the world, certain models have
become established and it is customary to use these until new models with clear advantages
become available. Endorsement by important organisations or manuals is sometimes an
important factor in encouraging use of models. Australian Rainfall and Runoff (Institution of
Engineers, Australia, 1987) is an example of such influence.
Calculations for urban stormwater drainage systems must be carried out for many
components such as sub-catchments and pipes. Without computers, it is not possible to
perform sets of complex, repetitive calculations, so prior to the 1960s designers had to rely
on simple procedures, notably the Rational Method. These used statistical rainfall
information, on the form of rainfall intensity-duration-frequency (I-D-F) curves, to calculate
design flowrates for various components of drainage systems. As shown in the first part of
Figure 6.3, these models only produced peak flowrates.
Rainfall
Intensity, I Flowrate Peak Q = C I
(mm/h) (m3 /s) A
ARI
I
t
c
SIMPLE
tc Time (h) tc Time (h)
MODEL
Rainfall
Intensity, I
(mm/h) Full Hydrograph
Hyetograph Flowrate
(m3 /s) Ordinates
Rainfall
"blocks"
COMPLEX
∆t Time (h) Time (h)
MODEL
With computers it became possible to create models which could take complex rainfall
patterns and convert these to hydrographs of flows, as shown in the second part of the
diagram above. The models involved could be split into loss models and routing models, as
shown in Figure 6.4.
Rainfall Hyetograph
Losses
Time
Rainfall Rainfall
Excess
Time
Runoff Flowrate
Time
Routing models allow for the distribution of rainfall across a catchment surface, with some
rainfall inputs being closer to the outlet than others, and so spreading out the pattern of flow
or hydrograph at the outlet. They also account for storage effects on the catchment. The
main types are:
• time-area routing,
• synthetic unit hydrographs
• artificial storage routing, and
• kinematic wave routing.
These models are described in many hydrology texts and in manuals such as Australian
Rainfall and Runoff. O’Loughlin, Chocat and Huber (1996) give a description of their
development and assess our current state of knowledge about them.
Using these methods, the flows occurring during a particular rainfall event can be
determined. With the loss model describing how the infiltration properties of a catchment
change as it becomes wetter, rainfall excess can be determined and routed through the
model. For most design and flood applications, such event models are adequate, although
there is often a problem in determining the antecedent wetness of a catchment just before a
storm.
However, most models cannot model the continuous generation of runoff from a catchment
from groundwater sources, which occurs when no rain is falling. To do this, a more
elaborate, soil moisture accounting model is needed. This includes soil storages
representing the unsaturated and saturated zones and parts thereof. By storing water in
these and releasing it, the drying out of the catchment soils can be modelled, as well as their
ILSAX and DRAINS use a medium-level rainfall-runoff model which combines a Horton loss
model with time-area routing in a complex way. It is adaptable to many situations, but does
not perform continuous simulation. The storage routing models that operate like the RORB,
RAFTS and WBNM models commonly used in Australia are also "event models", designed to
produce hydrographs for flood estimation, but not capable of modelling long periods of runoff
under wet and dry conditions.
The description here and in subsequent sections will relate particularly to the way that this
model is implemented in DRAINS rather than in ILSAX. There are minor differences, but
with the same parameters, the resulting hydrographs are identical.
This model relates to an urban or semi-urban catchment, subdivided into several sub-
catchments associated with a drainage system of pipe and channel sections as shown in
Figure 6.5. Sub-catchments are divided into three surface types - paved (impervious areas
directly connected to the pipe system), supplementary (impervious areas not directly
connected) and grassed (pervious areas connected to the pipe system).
HYETOGRAPH
Rainfall
Time
INLET
PAVED AREA BYPASS
(directly-connected impervious) FLOW
SUPPLEMENTARY GRASSED
AREA AREA
(impervious area not (pervious area PIPE
directly-connected) directly-connected)
SUB-CATCHMENT
Flowrate
HYDROGRAPH PIPE or
CHANNEL
Time DETENTION
STORAGE
OUTFALL
Since it is an event model, conditions at the start of each storm event must be established by
defining a value of the antecedent moisture condition, AMC, for the soil underlying the
pervious portions of the catchment.
The loss model subtracts depression storages from all surfaces and calculates additional
losses for grassed or pervious areas using Horton’s infiltration model. The soil type and
AMC parameters are easily understandable and can be related to identifiable soils and
rainfall depths preceding a storm. Results are quite sensitive to the AMC and users must
consider the effects of their choices using sensitivity studies. Despite this, few problems with
employing this model have been reported.
The model relies on times of travel as the main parameters used in routing. These can be
determined to an acceptable level of accuracy (or inaccuracy) for urban catchments, but are
very variable for rural catchments. Thus, while the ILSAX model can be applied to pervious
sub-catchments of a drainage system, it is not strictly applicable to rural catchments. This
reflects the lack of suitable studies to calibrate the model in rural conditions, rather than a
defect in the model itself.
The basis of the ILSAX model’s hydrograph generation is the time-area method, illustrated in
Figure 6.6. This “convolves” the rainfall hyetograph with a time-area diagram, in a similar
manner to unit hydrograph calculations. A time of entry (or time of concentration) must be
determined for a drained area using methods discussed later in Section 6.3.5.
For this explanation, assume that the rainfall hyetograph has had losses removed and so
represents rainfall excess. The hyetograph is divided into time steps of ∆t. So is the time-
area diagram, a plot of the catchment area contributing after a given number of time steps.
This diagram can be visualised by drawing isochrones, or lines of equal time of travel to the
catchment outlet. For times greater than the time of concentration, the area contributing
equals the total area of the catchment.
When a storm commences on a catchment that has a time of entry of 5∆t, the initial flow Q0
is zero. After one time step ∆t, only sub-area A1 contributes to the flow at the outlet. Any
runoff from other sub-areas is still in transit to the outlet. Thus the flowrate at the end of the
first time step can be approximated by Q1 = c.A1 . I1, where c represents the conversion
factor from mm/h to m3/s units, and I1 is the average rainfall intensity during the first time
step.
At the end of the second time step, there are two contributions to the outlet flow, Q2, due to
the second block of rainfall falling on the sub-area nearest to the outlet, c. A1 .I2, and to runoff
from the first rainfall block on the second sub-area, c. A2 . I1. At the end of the third time
step, there are three contributions, Q3 = c. (A1 . I3 + A2 . I2 + A3 . I1), and so the process
continues, as shown in Figure 6.6. The hydrograph builds up to a peak and then recedes
once rainfall stops and the catchment drains.
Losses can be subtracted from the rainfalls and flows before or after these time-area
calculations are made. The latter choice is recommended for grassed or paved areas, as
this allows infiltration to occur from flows moving across a sub-catchment after rainfall has
stopped. In this case, the hydrograph of Q values represents a "supply rate", from which
losses must be subtracted later.
This process is simple to apply in a computer program, and effectively models the behaviour
of runoff from distributed areas. It does not specifically allow for storage effects, as do
alternative models such as linear or non-linear reservoirs or kinematic wave calculations.
However, in practice it gives similar accuracy to these models because of the difficulties
involved in describing flow paths and times. Storage routing effects might be simulated by
lengthening the time of concentration. In DRAINS, as in ILSAX, it is assumed that all time-
area diagrams are straight-lines. It is conceivable that they could be concave or convex,
depending on catchment shape on other factors, however, investigations conducted in the
U.K. with the TRRL Method concluded that this degree of accuracy was not necessary.
In the ILSAX Type Hydrological Model used in DRAINS, each sub-catchment must be
divided into the sub-areas shown in Figure 6.7, with the following surface and drainage
characteristics:
• paved areas, impervious areas directly connected to the pipe system, including road
surfaces, driveways, roofs connected to street gutters, etc.,
• supplementary areas, impervious areas not directly connected to the pipe system, but
draining onto pervious surfaces which connect to this system (These may include tennis
courts surrounded by lawns, house roofs draining onto pervious ground, etc.), and
• grassed areas, pervious areas directly connected to the pipe system, including bare
ground and porous pavements as well as lawns.
There may also be undrained areas, which do not connect at all to the pipe system, and can
be ignored in a DRAINS model.
In DRAINS, the total sub-catchment area and the percentages of paved, supplementary and
grassed areas must be specified for each sub-catchment.
Generally, fully-developed medium density residential catchments will have areas impervious
between 30 and 70%. Dayaratne (2000) has obtained the following relationships from
modelling of storms on 16 gauged residential catchments in four Victorian municipalities:
Directly connected impervious area (or paved area) percentage, DCIA (%)
= -0.85 hhd2 + 23.38 hhd - 101.19 (r2 = 0.90) … (6.1)
Times of entry must be specified or calculated for the paved and grassed areas (and also for
the supplementary area in DRAINS). These are effectively the same as the times of
concentration or times of travel used in the Rational Method. They set the base lengths of
the time-area diagrams used to create hydrographs.
The DRAINS property sheet for a sub-catchment is shown in Figure 6.8. Information on
surface types is arranged in three columns. The length of these varies according to the level
of detail selected in the Use box. For many applications, fixed times can be entered.
However it is also possible to calculate a time by the steady-state “ kinematic wave” equation
for overland flows (Ragan and Duru, 1972):
toverland = 6.94 .
(L ⋅ n)
0.6
.... (6.3)
I ⋅S
0.4 0.3
The surface roughness n is the same as the coefficient n in Manning's Formula for open
channel flows. It typically takes the values set out in Table 6.2.
Values for lawns and grassed surfaces show considerable variation, depending on the depth
of flow relative to the height of grass blades. Values from 0.05 to 1.0 have been obtained by
various researchers, as described by Engman (1986).
In DRAINS, intensity I is taken as the mean intensity of the rainfall pattern supplied. This
should be satisfactory for design rainfall bursts such as those supplied in Australian Rainfall
and Runoff, 1987, but may be erroneous for some more variable or patchy patterns that
occur naturally.
For paved areas, it is also possible to have a gutter flow time calculated using the equation
for flows in gutters, recommended by the U.S. Federal Highway Administration (1984):
where Q (m3/s) is the total flowrate, estimated by dividing the section as shown and applying
Izzard's equation for a triangular channel with a single crossfall or cross-slope:
Z 8/3 0.5
Q = 0.375.F. .d .S0 .... (6.5)
n
F is a flow correction factor, ZG and ZP are the reciprocals of the gutter and pavement cross-
slopes (m/m), nG and nP are the corresponding Manning's roughness coefficients, dG and
dP are the greatest gutter and pavement depths (m), and S0 is the longitudinal slope (m/m).
Clarke, Strods and Argue (1981) estimated values for F of about 0.9 for simple triangular
channels and 0.8 for sections of the type shown in Part (a) of Figure 6.6. These values may
be used in the absence of more precise information. Typical values of n would be 0.012 for
concrete, 0.014 for hotmix, 0.018 for flush seal and 0.025 for stone pitchers (Dowd et al.,
1980).
Where the face of a kerb is steeply sloping, it can be considered as vertical. For "lay-back"
kerbs with sloping kerb faces, Equation 6.4 can be applied, assuming ZG to be equal to w ,
dG
as defined in Part (b) of the above figure.
Gutter ⋅ Length
The equation given in DRAINS is Time = + 5 .... (6.5)
GUT ⋅ S0 ⋅ 60
where GUT is a factor based on Equation 6.4 and the continuity equation, V = Q , equal to
A
( 8 8
) 8
0.375 . F . ZG ⋅ dG3 − dP 3 + ZP ⋅ dP 3 divided by area,
[(d
2
G − dP2 ) ⋅ ZG + dP2 ⋅ ZP ]
nG nP 2
Although this is an awkward expression, it will be a constant for a specific gutter and
roadway combination, for a given depth of flow. Thus it can be used when a typical depth
can be set.
In the Sub-Catchment property sheet shown earlier, various combinations of flows can be
given for paved areas. These are added as follows:
This allows you to have constant time alone, overland flow time alone, gutter flow time alone
or various combinations of these times. Only the constant time is entered directly. The
others are calculated from the values of length, slope and roughness entered.
The arrangement is similar for the supplementary area, except that a gutter time is not
involved. The grassed area arrangement is much the same, with a constant time and
overland flow segment.
The property drainage time is that required for all water to contribute to flow at the boundary
outlet. There is conflicting evidence on property drainage times (Stephens and Kuczera,
1999, Goyen & O’Loughlin, 1999, Dayaratne, 2000), with some pointing to short times, 1 or 2
minutes, and some to longer times, 5 to 10 minutes). 2 to 5 minutes is probably a
reasonable compromise.
The Queensland Urban Drainage Manual (Neville Jones & Associates et al., 1992)(QUDM)
recommends a simplified procedure for setting inlet times, using the following values:
Should the calculated tc be less than 5 minutes, this minimum value is customarily adopted as
the tc.
Table 6.3 Recommended Standard Inlet Times in Queensland Urban Drainage Manual
A lag time for grassed area flows can be applied where flows from such areas must flow over
paved surfaces before reaching a pit, as shown in Figure 6.10.
This can be given as a constant time or as a proportion of the calculated paved area
(overland + gutter flow time). It is the former with the abbreviated data or more detailed
data options in the Sub-Catchment Data property sheet, and the latter with the
comprehensive data option. This provision allows considerable flexibility in dealing with
different distributions of paved and grassed surfaces in a sub-catchment.
Losses from paved and supplementary areas are calculated simply. Depression storages
are considered as initial losses and are subtracted from rainfall hyetographs prior to time-
area calculations.
The general range of depression storages is from 0 to 2 mm for impervious surfaces and 2 to
10 mm for pervious surfaces. Commonly-used values for paved, supplementary and
grassed areas are 1, 1 and 5 mm, respectively. Dayaratne (2000) recommends values of 0
to 1 mm for impervious areas.
The procedures for grassed areas are much more complex. They are based on the general
equation developed by Horton in the 1930s:
This describes the curves shown in Figure 6.11. These only apply when there is sufficient
rainfall to completely satisfy the infiltration capacities, and accumulated infiltration is
increasing at its full rate.
The curves represent soil types which follow the classification used by Terstriep and Stall
(1974), based on the system developed by the U.S. Department of Agriculture, and
described in references such as Chow (1964). These are used in widely-used procedures in
the U.S. and Canada, such as Technical Release 55 of the U.S. Soil Conservation Service
(1975). There are four main soil classifications, designated A, B, C and D (corresponding to
1, 2, 3 and 4 in the ILSAX type model). These are described as:
1 (or A) - low runoff potential, high infiltration rates (consists of sand and gravel);
2 (or B) - moderate infiltration rates and moderately well-drained;
3 (or C) - slow infiltration rates (may have layers that impede downward movement of
water);
4 (or D) - high runoff potential, very slow infiltration rates (consists of clays with a
permanent high water table and a high swelling potential).
These soil types are used in conjunction with antecedent moisture conditions (AMCs) which
fix the points on the infiltration curves at which calculations commence. This is specified, not
by an initial infiltration rate in mm/h, but by an antecedent depth of moisture, which
corresponds to the area under the curve to the left of the starting point. On each curve in the
above figure, four starting points (numbered 1, 2, 3 and 4) are shown, representing possible
AMCs.
AMCs can be estimated from Table 6.4. Both soil types and AMCs can be interpolated
between the levels of 1, 2, 3 and 4.
For the curve and AMC selected, the model calculates an infiltration loss in each time step.
This is subtracted from the rainfall inputs to the pervious area.
Values of parameters involved with various combinations of soil types and AMCs are set out
in Table 6.5. Users also can also provide their own values.
This classification involving soil type and AMC has been found to give good fits to recorded
storm hydrographs from gauged catchments in Australia, and the soil types have been
accepted by ILSAX and DRAINS users. Siriwardena, Cheung and Perera (2003) compared
the infiltration rates in Table 6.4 with those measured with infiltrometers at eight urban
gauged catchments in Victoria. They found that the f0 and fc values measured were
generally higher than those for the same soil classification in Table 6.4. They obtained f0
Antecedent Rainfall
Depths (mm) for AMCs:
1 0 0 0 0
2 50 38 25 18
3 100 75 50 38
4 150 100 75 50
Initial Infiltration
Rates (mm/h) for AMCs:
1 250 200 125 75
2 162.3 130.1 78.0 40.9
3 83.6 66.3 33.7 7.4
4 33.1 30.7 6.6 3.0
values of 28 to 503 mm/h compared to 13 mm/h for a Type B soil, and fc values of 4 to 135
mm/h, compared to values of 31 to 200 mm/h. They also obtained a shape factor, k, of 0.85
h-1 compared to 2 h-1 in the table.
The Siriwardena, Cheung and Perera paper does not explore the implications of changing
these parameters in modelling hydrographs from the test catchments. It is not possible to
assess the effects of this at present, but Victorian users of DRAINS and similar programs
should take the above results into consideration when setting parameters. DRAINS s allows
user-provided parameter values to be specified in the hydrological model inputs.
In ILLUDAS-SA, the following form of Horton's equation was used to determine the infiltration
rate from the accumulated depth of infiltration. This allows for variable rainfall intensities
which might be less than the infiltration capacities at some times.
f
f = f + fc . 1 − … (6.8)
f + (f − f ).e(f0 − kF − f f
)/
c 0 c c
The infiltration rate calculated from this is subtracted from the hyetograph or supply rate, and
should any water remain, depression storage is subtracted. Once the depression storage
has been fully satisfied, any excess over infiltration is assumed to be runoff. The
accumulated infiltration depth is increased by the amount assumed to be infiltrated. For
porous soils and light rainfalls, it is quite possible that there will be zero runoff from pervious
surfaces.
Malcolm Watson, the developer of ILLUDAS-SA, suggested that an alternative method could
be used which would not involve iterative calculations, and this was incorporated into ILSAX.
The alternative procedure, described by Watson (1981b), involved the division of the
infiltration curve described in Equation 6.7 into diminishing and constant components:
The actual depth of infiltration, ∆F, over time step ∆t is the lesser of I . ∆t , where I is rainfall
intensity, and
( − )
Fcap = (1 - e-k∆t) . f 0 f c − Fd + fc . ∆t .... (6.9)
k
∆F
Fd = Fd - . (∆Fcap - fc. ∆t) .... (6.10)
∆Fcap
which apportions actual infiltration depths between diminishing and constant components, as
shown in Figure 6.12.
∆Fd
.∆t
∆Fcap - fc ∆ t
Infiltration
Rate ∆F
(mm/h) ∆Fd = (∆F - fc . ∆ t)
∆Fcap cap
∆Fcap
f=
∆t
fc
Time
∆t ∆F
∆Fc = .f .∆t
∆Fcap c
The time-area method is applied separately to the paved, supplementary and grassed area
portions of the catchment. DRAINS differs from ILSAX in the treatment of the supplementary
area runoff. In ILSAX, no losses are applied to rainfall on the supplementary area and it is
transferred instantly to the grassed area. DRAINS allows for a supplementary area
depression storage and time of travel. (These must both be set to zero if you wish to exactly
reproduce ILSAX hydrographs.) The DRAINS procedures can be applied to situations where
there is significant supplementary area, such as airfields.
The process for paved and supplementary areas is shown in Figure 6.13. Hyetograph
values are scaled (by area/360) to convert intensities to flowrates in m3/s.
Full Area
The more complex process for grassed area runoff is shown in Figure 6.14.
Runoff
... convolved ... produces ...
with ...
Rainfall
Runoff
from the
Supplementary
Area
Runoff
Infiltration
• that supplementary area runoff is added to the grassed area flows, and
• that depression storage is actually calculated after the infiltration is calculated.
All hydrographs in the program are linked to the same time base and are synchronised, and
combination of hydrographs is a straightforward addition process. As shown in Figure 6.15,
the supplementary area hydrograph is incorporated in the grassed area hydrograph. This is
added to the paved area hydrograph and possible user-provided hydrographs or baseflow, to
obtain the total runoff hydrograph coming off the local sub-catchment. Overflows from
upstream pits, if present, are then added to this to obtain the total approach flow to a pit,
simple node or detention basin.
It can be assumed that all flows can be accepted into the node, or that a pit entry relationship
applies, and bypass flows can occur. Modelling of bypass flows involves the separation of
hydrographs.
Once inside the pit, the admitted flows are combined with flows in upstream pipes and
possible user-provided hydrographs or baseflows, as shown in Figure 6.16. These are then
checked against the calculated capacity of the pipe, and if part of the hydrograph exceeds
this capacity, overflows are assumed to occur. These may be held at the pit until the
hydrograph recedes and there is sufficient capacity for them to re-enter the system, or they
may be directed to a downstream pit or out of the system altogether.
Rainfall Hyetograph
"Above Ground"
Combined User-Provided
Runoff Hydrograph Hydrograph
or Baseflow
Overflows from
Upstream
Total Local
Approach Hydrograph
Runoff
to Entry Point to
Hydrograph
Drainage System
The first option available in the Rational Method Model property sheet that is called from the
Hydrological Model Specifications box is a general Rational Method procedure. It is
necessary to specify four runoff coefficients - an impervious and a pervious area coefficient
for design, and another set of these for analysis.
DRAINS performs a search between the time specified for the impervious area and the
(usually longer) time specified for the pervious area, to find the time that provides the
greatest value of Q = C.I.A.
Rational Method times of concentration are specified in exactly the same way as ILSAX
model times of entry in the property sheet for sub-catchments.
This general method is a “plain” implementation that has no special fixed features, and can
be applied outside Australia, or within Australia if the user wants to depart from the Australian
Rainfall and Runoff, 1987 method.
The method from Australian Rainfall and Runoff (institution of Engineers, Australia, 1987) is
the second option provided by DRAINS. The method is fully explained in that publication,
and a detailed example is provided.
Here it is necessary to enter C10 values for the impervious and pervious areas. These are 10
year average recurrence interval (ARI) runoff coefficients that are adjusted for other
recurrence intervals. The value for the impervious area is always 0.9.
C10 = 0.1 + 0.0133 (10I1 - 25) with upper and lower values of 0.1 and 0.7 … (6.12)
10
I1 is the 10 year ARI, 1 hour rainfall intensity, used as an index of the rainfall climate.
The search for the time of concentration that provides the greatest value of Q = C.I.A is also
carried out with this method.
The third method is taken from the Australian / New Zealand Standard 3500.3.2. This gives
different procedures for Australia and New Zealand, and only the Australian procedure is
implemented in DRAINS at present.
The Rational Method formula is expanded to allow for the three surface types:
where Ar, Ai and Ap are the areas of roofs, impervious areas and pervious areas in the sub-
catchment being considered. Only a property site itself is considered in these calculations.
The runoff coefficient for roofs is assumed to be 0.1 and that for impervious surfaces at
ground level to be 0.9. In calculations, the time of concentration is fixed, at 5 minutes.
All methods require intensity-frequency-duration (I-F-D) data that is entered in the Rational
Method Rainfall Data property sheet opened from the Rainfall Data… option in the Project
menu.
Previous models, notably synthetic unit hydrograph procedures, provided a flow estimate at
the outlet to a catchment. By dividing the catchment into sub-areas, the storage routing
models provided flood estimates at several points throughout the stream system. They also
allowed hydrological losses to be varied across the catchment area, reflecting various soil
types. Since these models are essentially networks of storages, detention basins and
reservoirs can be easily incorporated.
RORB, RAFTS and WBNM.belong to a class of models termed runoff routing models, which
also includes models based on unit hydrograph and kinematic wave calculations. Runoff
routing models can "route" a hydrograph from one geographical location to another, allowing
for changes such as translation and attenuation of the hydrograph.
The basic non-linear model used in Australian storage routing models was developed by Eric
Laurenson. RORB, a practical computer application of this model, was developed by Eric
Laurenson, Russell Mein and Tom McMahon at Monash University, Melbourne, and released
in the mid 1970s. At the same time the RSWM (Regional Stormwater Model) was developed
Initially, these models were run on mainframe and mini computers. RAFTS (Runoff Analysis
and Flow Training Simulation), an enhanced version of RSWM, was released in 1983 and
sold commercially by Willing & Partners (later WP Software and XP-Software). This includes
continuous modeling processes as well as the storage routing model discussed here. A
version for IBM-compatible PCs was released in 1987. A PC version of RORB was released
in 1988.
WBNM was revised in 1987 and a new version was produced by Michael Boyd, Ted Rigby of
Forbes Rigby Associates and Rudy VanDrie of Balance Research and Development in 1994.
This modeled urban catchments. WBNM2000, introduced in 1999, used a different structure
to earlier models and added many features.
A number of other storage routing models using non-linear routing have been developed, but
have not had the wide use or long development history of the RORB, RAFTS and WBNM
models. Runoff routing methods have been covered in Australian Rainfall and Runoff
(Chapter 9, 1987 and or Book 5, Section 3, 1998), where regional estimates of parameters
are provided.
Storage routing calculations are carried out over a series of time steps, with the information
obtained from solving equations at one time step being used as an input to the next step.
Each of the models available in DRAINS has been developed on different principles. RORB
performs calculations based on the equation:
S = k . Qm … (6.14)
The routing factor k is defined as kc . kr, where kc is a general routing factor applicable to all
sub-catchments, and kr is a factor for a particular sub-catchment, being the ratio of the
stream length running through that sub-catchment and the average flow distance from sub-
catchments to the catchment outlet. kc acts as a calibration parameter, enabling the model's
results to be varied and fitted to recorded hydrographs. A kc of 0.0 will perform no routing, so
that values of rainfall excess and flows from upstream storages will pass through a sub-
catchment unchanged. A kc that is very large will delay flows considerable, so that flowrates
will be very low. By adjusting kc, the peak of a calculated hydrograph can be varied over the
range from the peak rate of rainfall excess to zero. Decreasing kc increases flowrates.
Allowance is made for different channel conditions by multiplying the routing factors by the
values in Table 6.6, in which Sc is the reach slope (%).
For routing along stream reaches, RAFTS applies a translation over a nominated time, or
performs Muskingum-Cunge routing based on the stream cross-section and roughness.
Values of the WBNM lag parameter are typically between 1.3 and 1.8. This can be used to
calibrate the model in a similar way to the RORB parameter, kc. WBNM2001 also allows for
translation and Muskingum routing in stream reaches.
S = 0.6.60.LP.A0.57.Q0.77 … (6.17)
with the 0.6 allowing for the routing effects in the reach, the length of which is related to the
area of the catchment through which it runs, A. A stream lag factor can be applied to allow
for different types of channel. Indicative values are shown below:
Modelling facilities based on RORB, RAFTS and WBNM have been included in DRAINS.
The three models have different structures, as shown in Figure 6.18: RORB has a well-
defined structure, with nodes located close to sub-catchment centroids. Routing is only
carried out in the stream reaches. There is modelling of losses at nodes but no routing.
By contrast, RAFTS can carry out routing at nodes representing sub-catchments, and also in
stream reaches, where flows can be translated or routed using the Muskingum-Cunge
method, based on the reach cross-section and roughness. The layout of RAFTS models is
flexible, and positions of sub-catchment nodes and lengths of reaches can be set in ways
that do not exactly reflect the model structure. The routing within sub-catchments differs
from RORB and WBNM in that flows are commonly routed through 10 successive non-linear
storages, as indicated in one of the sub-catchments in Figure 6.18.
In WBNM, routing occurs at the sub-catchment nodes and in stream reaches that convey
runoff from upstream sub-catchments through the local sub-catchment. Like RAFTS, it is
flexible, and can be set out in different configurations.
To fit these different structures into the DRAINS framework, it has been necessary to apply
different property sheets and relationships between model sub-catchments and stream
routing reaches. These are described in Chapter 2.
For stream channels, routing can also be undertaken by methods such as the Muskingum
Method, lag and route methods, Muskingum-Cunge routing and hydraulic routing using
methods such as kinematic wave calculations. DRAINS employs the latter in RAFTS-style
stream routing reaches, following a method given in Chapter 9 of Open Channel Hydraulics
by F.M. Henderson (Macmillan 1966).
The ILLUDAS type pit is retained in DRAINS because there may be some situations where
an overflow is impossible, and all runoff coming to the pit will be held there. In the great
majority of cases, it will be better to model this situation as a sag pit with a high overflow limit
and some restriction on the inlet capacity.
On-grade pit inlet capacities are defined as a relationship between the inlet flow or capture
and the approaching gutter (or channel) flow. As there is no direct theoretical relationship
between these, an empirical relationship must be established between these as a result of
laboratory tests or field observations.
Approach
Flow
Bypass Flow
Kerb and
Gutter
On-Grade Pit
Sag Pit
Figure 6.20 shows the relationships for kerb inlets on grade, as measured in hydraulic model
studies published by the N.S.W. Department of Main Roads (1979). It is evident that at low
flows all of the approach flow will be captured, but that at some level of approach flow, some
bypass will start to occur. As the magnitude of the approach flow increases, the percentage
of the flow captured will decrease. This is represented by the curved line becoming gradually
flatter and crossing the dotted lines indicating various percentages of capture.
Various tests have been conducted in Australia by the NSW Public Works Department for the
NSW Department of Main Roads (now the Roads and Traffic Authority) and NSW
Department of Housing, by the University of South Australia for various Queensland bodies,
and by the Victorian Country Roads Board (now Vic Roads). In addition, there have been
tests by overseas organisations such as the US Federal Highway Administration. These
have produced a rather confusing array of results, from which it is difficult to generalise.
The relationship used in ILSAX and in the past in DRAINS to describe on-grade pit inlet
capacities is:
As an example, in Section 12.3.2 of the superseded 1977 Australian Rainfall and Runoff, the
capacity of a double grating with an extended kerb inlet was quoted as "0.08 m3/s plus half of
the flow in excess of 0.08 m3/s ". This can be described by the above equation, with CAP1 =
0.04, CAP2 = 0.5, CAP3 = 0 and CAP4 = 0.
If the relationships in Figure 6.20 are plotted on log-log paper, a line of best fit can be made
to the sections of the curves which depart from the 100% capture line, and the following
relationships can be derived:
giving CAP1 = 0.0, CAP2 = 0.0, CAP3 = 0.37 and CAP4 = 0.7.
giving CAP1 = 0.0, CAP2 = 0.0, CAP3 = 0.07 and CAP4 = 0.52.
Relationships for a range of New South Wales RTA pits were developed by O'Loughlin,
Darlington and House (1992) and were included in some of the DRAINS demonstration
examples.
From April 2001, DRAINS now applies another approach, in which inlet capacity
relationships for on-grade pits are defined as a series of points, as shown in Section 2.4.6,
rather than by an equation. This allows you to set a maximum capacity, or a relationship that
approaches an upper limit of inflow capacity asymptotically. It is better suited to the
Advanced Design method than the ILSAX relationship in Equation 6.14. (DRAINS can still
apply this relationship with older data files, but cannot mix the two formulations, as they set
out information differently in the drains.db file that stores the pit data base in the
C:/Program Files/Drains/Program folder.)
Sag pits can be modelled more easily, as the theory of weir and orifice flow can be applied.
Experimental investigations have confirmed the following weir equation given in Australian
Rainfall and Runoff (Institution of Engineers, Australia, 1987, page 303).
Orifice flow can occur above 0.12 m for a grate and 1.4 times the slot height for a kerb inlet,
though there is a large transition zone for grates, in which either flow mechanism may occur.
Most cases of interest to designers, including major flows, are described by the weir
equation.
In ILSAX, depths of ponding were not calculated, and the following relationship was
developed with the volume of ponded water:
In some cases, depths can be related to volumes, and hence to inflows, as in the following
method presented by O'Loughlin, Darlington and House (1992).
If the volumes of ponding on the straight section of roadway shown in Figure 6.21 are
assumed to be irregular tetrahedrons, the total volume V (m3) can be given as:
d
3 1 1
V = ⋅ − = c.d
3
... (6.21)
6 Sc S
L1 SL2
For example, if Sc is 0.03 (3%), and SL1 and SL2 are 0.01 and 0.015 respectively, c is 926,
and V = 926 d3.
V 1/
Combining these equations, with d = 3, the following equation is obtained:
c
V ( /3)
1.5 V 0.5
Qcap = 1.66 . L . = 1.66 . L . ... (6.22)
c c
V 0.5
Thus for a 2 m wide kerb inlet, Qcap = 1.66 . 2 . = 0.109 V0.5. … (6.23)
926
In the form of Equation 6.20, VCAP1 = 0, VCAP2 = 0.109 and VCAP3 = 0.5.
An inlet capacity wizard for sag pits was included in DRAINS. If you clicked the Inlet
Capacity Wizard button in the Drainage Pit property sheet, the dialog box shown in Figure
6.22 appears, requesting road slopes and the perimeter of the pit.
“Right” and “Left” referred to the two sides from which gutter flows can approach a sag pit. If
you entered values into the boxes shown and the click the Calculation button, suggested
values for the parameters VCAP1, VCAP2 and VCAP3 were generated by the above
equations and presented. You could elect to choose these by clicking the Save and Finish
button.
A similar analysis can be carried out on the corner ponding shown at the bottom of the figure
showing ponding above. In this case, volume is related to depth by:
d
3
1 1 2
V = . + + ... (6.24)
6 Sc1. SL1 Sc2 . SL2 Sc1. Sc2
Other expressions can be derived for cases where a corner has an angle other than a right
angle.
These points are obtained from relationships in published sources, most of which are
smoothed graphs fitted to experimental data from the types of rigs operated by the University
of South Australia (www.unisa.edu.au/uwrc/rig.htm) and the New South Wales Department of
Public Works and Services Manly Hydraulics Laboratory
(http://marlin.mhl.nsw.gov.au/www/kivt.htmlx).
These rigs and occasional field tests have been used to define pit inlet capacities for many
common types of pits. DRAINS provides sets of relationships for various states of Australia.
The relationships provided in DRAINS have been extrapolated well beyond the published
relationships using hydraulic principles, allowing for approach flows up to 1 m3/s for on-grade
pits and depths of ponding of up to 0.6 m for sag pits. None of these relationships have been
approved by the originating authorities. It is up to each user of DRAINS to determine
whether these relationships are suitable for their purposes. Users can readily modify the
relationships.
Published relationships appear in different forms, and it is not unusual for pits of similar forms
and dimensions to have quite different inlet capacities.
For New South Wales, data has been provided for the pits shown in Table 6.8.
Figure 6.23 Type SA1, SA2 and SA5 Pits tested for the NSW Department of Main Roads
(now the RTA) in 1979
For Queensland, a number of pits have been tested to provide the most comprehensive data
available in any Australian State. Further tests have recently been carried out and some
relationships will eventually be modified. The Queensland pit capacity relationships provided
with DRAINS have been derived from information in the Queensland Main Roads
Department Road Drainage Design Manual (1999) (available for download using internet
search engines). Characteristics of pits are shown in Table 6.9.
Victorian relationships have been obtained by extrapolating the curves given in the VicRoads
Road Design Guidelines Part 7, Drainage, 1995. Types are shown in Table 6.10.
For each sag pit, you must define a ponded volume and an allowable depth of ponding
corresponding to this maximum volume. Overflows are assumed to occur if this is exceeded.
This models the situation where water ponds at a location such as the corner of an
intersection until it becomes high enough to cross the crown of the road and escape.
For both on-grade and sag pits, a choke factor can be applied to simulate blockage of the pit.
This is 0.0 for no blockage and 1.0 for complete blockage. Typical values might be 0.2 for an
on-grade pit and 0.5 for a sag. These are multiplied by the capacity defined by the inlet
capacity relationships, whatever magnitude this may be. While this is acceptable for the type
of blockage that might occur for sag pits, it may not be realistic for on-grade pits. If you have
doubts about this, it would be better to define the required inlet capacity relationship in the pit
data base, and to employ this with a blocking factor of 0.
Additional information on pit entry capacities is given in sources from the U.S. such as
Circular No. 22 of the US Federal Highway Administration (1996) (available from
www.fhwa.dot.gov/bridge/hyd.htm), Guo (1996) and Section 5.5 of Mays (2001). The former
contains the only general theoretical methodology available for defining inlet capacities for all
1 m grated Transverse
VicRoads none
inlet pit grates A & B -
1.0 m wide x
0.75, 1.0, 1.5,
2.0 or 2.5 m
long
6.7.1 General
The hydraulic models used in design and analysis of urban stormwater drainage systems
can be considered to operate at three levels:
• open channel hydraulics assuming steady flow, described as “pipe full but not under
pressure” in Australia,
• part or full-pipe flow calculations determining hydraulic grade lines (HGLs) and water
surface profiles,
• full hydrodynamic modelling, usually employing finite difference solution of the partial
differential equations describing the conservation of mass and momentum, the St. Venant
Equations.
ILSAX calculations operate at the first level, so their hydraulics is quite limited. The HGL
calculations associated with Rational Method in the urban stormwater drainage chapter of
Australian Rainfall and Runoff, and those in steady state hydraulics programs such as HEC-2
and HEC-RAS, are at the second level. Several programs, mostly proprietary ones, offer full
hydrodynamic modelling options. These can give the most accurate results with an
experienced, hydraulically-knowledgeable operator, but are subject to stability problems due
to the complex finite difference calculations employed.
In DRAINS there is a middle-level method that is well in advance of methods commonly used
for drainage system design, but which is stable, quick and easy to apply. Hydraulic grade
lines are determined throughout the system from the tailwater levels at the outlet at each
calculation time step. In effect it performs the same types of water surface profile
calculations as HEC-RAS something like 50 to 200 times for each storm pattern. These are
“snapshots” assuming steady flows at a specific time.
At each time step, Analysis runs make a pass downwards through the a drainage
system, determining flows into pits, possible bypass flows and the flows along pipes.
They then retrace this path from a specified tailwater level at the system’s outfall,
determining hydraulic grade lines and water levels in pits. Allowance is made for pipe
friction and pit pressure changes, and both part-full and full-pipe flows are modelled.
The possibility of water upwelling from pits due to the flow capacity of the downstream
pipe system being exceeded is also considered.
DRAINS does use a complex hydraulic engine, taken from the PIPES program, to
model pressurised flows. It switches to this when pipes go from part-full to full pipe
flow, handling the complex timing and flow volumetric transitions that are required in
jumping between calculation methods.
In a Design run, DRAINS determines pipe sizes and invert level positions by
determining the peak flows of calculated hydrographs and designing for these in a
downwards pass followed by an upwards one. An Analysis run using the Design
storms is performed automatically following design calculations, and results are given
for the entire hydrograph.
Details of the operation of these methods are given in Section 5.3 in the previous chapter.
Here we will present some background details of the hydraulic models adopted.
For circular pipes, you have a choice of the Colebrook-White Equation or Manning's Formula.
The Colebrook-White Equation employs the formula:
k 2.51 ⋅ υ
V = -0.87 2g ⋅ D ⋅ S ⋅ loge + .... (6.25)
5.7 ⋅ D D ⋅ 2g ⋅ D ⋅ S
The pipe wall roughness values in Table 6.11 are recommended by Hydraulics Research
(1983) and the Standards Association of Australia (1978). Values for other materials are
also given in these publications.
Concrete
Precast, with "O"Ring Joints 0.06 0.15 0.6
Spun precast, with "O" Rings 0.06 0.15 0.3
Monolithic construction, against 0.3 0.6 0.15 0.03 to 0.15
steel forms
Monolithic construction, against 0.6 1.5 -
rough forms
UPVC
Chemically-cemented joints 0.03 0.003 to 0.015
Spigot and Socket Joint 0.06
Manning's Equation is
1 0.667 0.5
V = .R S .... (6.26)
n
(Note that as C decreases, the pipe is rougher and there is greater frictions loss.)
The head losses and changes to the energy grade line and hydraulic grade line at pits and
junctions are extremely important in determining pipe system behaviour accurately. Figure
6.24 shows how these are represented by two functions of the pit outlet velocity Vo, for full-
pipe flow.
Grate Flow
2
k LV0
2g TEL
2
k uV0
HGL
2g
Pit
V0
The TEL will drop by the amount of the head loss for the pit, which can be expressed as:
2
Vo
hL = kL ⋅ … (6.28)
2g
ku can be positive, with the HGL dropping down, or negative, with the line rising due to the
downstream pipe diameter being larger and having a lower velocity than the upper one. This
has been termed “static regain”.
It is assumed that head losses and pressure changes take place at the centreline of the pit.
The actual losses occur mainly in the outlet pipe just downstream of the pit. Where there is
significant turbulence in the pit, the water level may be higher than the incoming HGL. A
factor kw, greater than ku, might be used in place of ku to establish water levels, if information
on these factors is available.
The factors described above apply only to full pipe flow. As the next section describes, a
large proportion of DRAINS calculations deal with part-full flows. For these, a nominal
energy loss and pressure drop of 30 mm has been assumed. (These estimates will not be
critical for analysis or design applications.)
In the DRAINS Run menu, there is the option named Revise Pit Loss Coefficients. This
alters the coefficients using an adaptation of an approximate method devised by Mills (Mills
and O’Loughlin, 1982-98).
Q Q
Basically, this is ku = 0.5 + 2. m + 4. g … (6.30)
Qo Qo
where Qm is the inflow from upstream pipes that are misaligned,
Qa is the aligned flow,
Qg is the grate inflow, and
Qo is the outflow, equal to Qm + Qa + Qg
DRAINS assumes that pipes at angles of 45o or more to the outlet pipe are misaligned.
A value of 0.5 is subtracted if the outlet pipe diameter is larger than that of any of the inlet
pipes. For a drop pit, the incoming flow from a pipe may be classed as grate flow if the inlet
pipe’s invert is located above the pit water level. Mill’s assessment of misalignment of
incoming pipes cannot be judged automatically by DRAINS.
This procedure is implemented by performing a design run, and then choosing the option
Revise Pit Loss Coefficients in the Run menu. This changes the original coefficients,
using the flows determined in the previous design run, thus overcoming the difficulty of
having to estimate ku factors roughly in advance when exact flows are not known. The
procedure can also be implemented after an Analysis run. This may lead to somewhat
different coefficients because the relative values of Qm, Qo and Qg may be different.
The process can be repeated to refine the result. It must be noted that this procedure is
approximate and may give poor estimates for some situations. The estimated coefficients
need to be checked and corrected where necessary.
In the version of DRAINS released in April 2001, allowance is made for pit pressure changes
occurring for part-full flows. In previous versions various allowances had been made, with
the allowance for most versions being the drop across the pit (usually 20 mm) plus 5 mm.
You now have three choices. The first is to apply a constant amount (default value 35 mm).
For pits with a 90o bend, a higher value would be appropriate than for a straight-through pit
arrangement. It is difficult to give specific guidance on this because there is only limited
scientific literature on pressure changes in part-full pipe junctions.
Configuration ku
Straight through line 0.5
Change of direction 0o to 45o 0.75
Change of direction 46o to 90o 1.0
Multiple pipes 1.0
The effect of part-full pressure changes will generally be small, as most systems will run full
under major flow conditions, but there may be some cases where the part-full pressure
change at one or more pits may be critical. The new facility in DRAINS enables designers to
cover this possibility.
The third option is to apply a kU factor to the velocity head of the downstream flow (using the
actual, part-full velocity). This may differ from the kU factor for full pipe flow. At this stage,
there is little general information on part-full kU factors, except that they are about on-third of
the full flow factors in sewer manholes.
A review of pit pressure changes and head losses is given in the paper by O'Loughlin and
Stack (2002). This discusses possible algorithms that might be used to determine pressure
changes at each time step in a model such as DRAINS. At this stage, only the Mills equation
has been implemented.
DRAINS calculations proceed down through the system until the tailwater level is reached.
This can be a level specified by the user in the Outlet Node property sheet, or it will be
assumed to be the pipe’s normal depth (or critical depth for supercritical flow) if it discharges
freely to the atmosphere.
Setting an appropriate tailwater level can be difficult. The following cases provide some
guidance: For a pipe system discharging to a free water body such as a lake, large stream
or the sea, the tailwater will be the water level occurring in this body at the time that the
storm passing through the drainage system occurs. Using DRAINS, you must determine the
most likely level coinciding with the storm for normal design or analysis, and high values
such as high tide levels in marine waters for modelling of extreme conditions.
Where the drainage system catchment is significantly smaller than the catchment of the
larger, receiving water body, it is likely that the rainfalls over the two catchments will differ in
intensity and timing. The estimation of appropriate events to define critical conditions
requires some statistical skill and knowledge of local storms.
Where the system being analysed is a pipe system discharging into a larger pipe or trunk
drain, the level to be selected should be the higher of the receiving pipe’s HGL or receiving
open channel’s water surface level at the junction. Hydraulic calculations may well be
necessary to establish these levels, but valid results cannot be obtained unless appropriate
tailwater levels are used.
The calculation processes are similar in circular pipes and in open channels. A hydraulic
grade line is projected upstream from a tailwater level using the standard step method
described in many open channel hydraulics texts (e.g. Chow, 1959, Henderson, 1966,
Chaudhry, 1993). Open channel sections are divided into several sections and the HGL is
projected upstream in each, so that water surface calculated is a series of straight lines
rather than a single line.
If the channel is hydraulically mild (with a normal depth greater than the critical depth) flow is
subcritical, and this projection from downstream will define surface correctly. However, for
steep channels, backwater must be calculated in a downstream direction and it is difficult to
determine where hydraulic jumps might occur. In DRAINS where channels are steep and
flows are supercritical, the water level will be held at critical depth, which will overestimate
the actual water level. Thus the calculated depths of flow will be conservative. Some
velocities will be underestimated and care should be taken if there is a possibility of scour.
These procedures are described in greater detail in Chapter 5.
Suggested roughness values for channels are given in Table 6.13. More comprehensive
lists are given in texts and manuals on open channel flow and in Chapter 4 of Australian
Rainfall and Runoff, 1987.
Various energy losses can occur at changes or transitions in channel sections. These are
covered by contraction and expansion losses, typically 0.1 and 0.3, respectively. These
factors allow for energy losses due to changes in cross-sections and velocities through
these. If the velocity increases or decreases between two cross-sections, the HGL is
lowered by a coefficient multiplied by the difference in velocity heads at the two sections. For
example, if the upper and lower sections are labelled 1 and 2, the losses for the two cases
will be:
2 2
expansion coefficient . V1 − V 2 … (6.31)
2g 2g
Ii + Ii +1 -
Qi + Qi +1
= Si +1 − Si ... (6.32)
2 2 ∆t
Average of Inflow rates Average outflow Rate of change
at the start of a period, Ii rate over the of storage
and at the end, Ii+1 period
applied together with a relationship linking outflow rates with corresponding storages for
various water levels in a reservoir or basin.
DRAINS allows for separate height-outflow relationships for low- and high-level outlets of the
type shown in Figure 6.25. Routing is performed with a combined relationship, but outflows
via high-level outlets, such as diversion weirs, can be directed out of the system or to
reaches other than the one immediately downstream.
Low-level outlets from basins consist of culvert or drop pit pipe systems. The outflow rate for
these is dependent on the headwater and tailwater levels, and energy losses through the
pipe system. Height-outflow (or depth-discharge) relationships for various kinds of outlets
are given in standard textbooks and manuals on hydraulics. The equations shown in the box
on the next three pages are used for calculating height-outflow relationships in DRAINS for
detention basins, culverts and headwalls.
High-level outlets such as weirs and slots are usually governed by the weir equation:
These depend on the threshold level, TH, which is usually the invert level at the upstream
end of the outlet pipe or culvert (m AHD), and pipe diameter D (m).
For (HW - TH) < 0.8 D, the flowrate for Inlet Control is:
0.05 1.9 0.6
Qi = Nc . 1.50 . (Sc/40) . (HW - TH) .D
(inlet control, unsubmerged inlet, Henderson, 1966)
The outflow rate, Q (m3/s), corresponding to level in the basin or headwater level, HW (m
AHD), is the lesser of the calculated Qi and Qo values. Parameters used in the
equations are:
TW is the higher of :
(a) the tailwater level downstream of the outlet (m AHD), and
(b) a level half way between the outlet obvert level, equal to (TH - Sc . Lc + D)
and the level of critical depth of the flow at the pipe outlet, calculated from :
0.287
Q
dc = D . for dc >= 0.82, and
4.038 ⋅ D D
2.5
0.510
Q
dc = D . for dc < 0.82.
3 . 005 ⋅ 2.5
D D
(Since Q is not known exactly when the tailwater level is being established,
an iterative procedure is used.)
V ⋅D Qi D
= ⋅
υ π 2 υ
⋅D
4
V being velocity of flow (m/s), and υ being the kinematic viscosity of water
(1.14 x 10-6 m2/s at 15oC). Note that the flowrate is assumed to be to be the inlet flow
estimate and the pipe is assumed to be flowing full. This is not exact, and an iterative
procedure should be used. However, the value of f is insensitive to the NR used, and
this approximation should be adequate.)
These are based on the threshold level TH (m AHD), usually the invert at the upstream
end of the culvert, and the height of the culvert, H (m).
If (HW - TH) < 1.35 H, the flowrate for Inlet Control is:
1.50
Qi = Nc . 1.70 . (HW- TH) .B
(inlet control, submerged inlet, Boyd, 1986)
Q2 0.333
dc =
2
g⋅B
H⋅B
Area / Wetted Perimeter =
2(H + B)
Laurenson and Mein (1990) provide the weir coefficients shown in Figure 6.26.
Equation 6.32 can be rearranged so that the known terms are placed on the left hand side
(LHS):
2 Si 2 Si +1
Ii + Ii+1 - Qi + = Qi+1 + ... (6.34)
∆t ∆t
Inflow values Ii, Ii+1, etc. are known in advance, and outflow Qi and storage Si at the
beginning of each period are known. Routing procedures estimate values for Qi+1 and Si+1 by
various methods of graphical or numerical interpolation.
DRAINS applies this procedure at each time step, but also allows for tailwater effects that
might alter the elevation-discharge (or height-outflow) relationship. It can therefore project a
HGL backwards through a series of interconnected basins. This is a complicated process,
because of the difficulties of determining when an outlet is drowned. This can lead to
oscillations in flowrates and water levels. Small variations can be ignored, but cases where
large ones occur should be referred to Watercom Pty Ltd.
DRAINS is set up to model on-site stormwater detention (OSD) storages of the type shown in
Figure 6.27, including high-early discharge (HED) systems, as shown in Figure 6.28.
Pit Pipe
Orifice
Plate
Bund
Above-Ground Storage (weir)
Underground Tank
Screen
OSD storages are usually controlled by circular orifices with the discharge equation being:
π 2
Q = Cc . d . (2gh)0.5 … (6.35)
4
6.10.1 Introduction
Open channels in urban environments are commonly crossed by roads, requiring the
construction of a culvert or bridge. DRAINS provides methods that allow for the effect of
these on water surface profiles.
6.10.2 Culverts
There are two meanings to the word, culvert. The first is a long pipe; the second is a pipe,
usually short, constructed to allow flows in streams and artificial open channels to pass under
road and railway embankments. The culvert component in DRAINS models the latter case.
The first type of conduit should be modelled as a channel, or if storage and overflows are
important, as a detention basin.
HED
Pit
Main Storage
Flap Orifice
Valve Plate
Outlet
Inlet Pipe
Pipe
State 1: The HED Pit Fills State 2: The Larger Storage Fills
State 3: Both Storages are Full, and Act Together. State 4: The Storages Empty
Culverts convey flows in pipes or rectangular conduits that through road embankments,
usually obstructing flows by reducing the available waterway areas. Upstream water levels
are raised, creating a headwater level higher than the water levels occurring under
unobstructed flows. Downstream levels are lower, since flow emerges rapidly from the
culvert, creating supercritical flow conditions until a hydraulic jump occurs.
Several procedures are available for the design of culverts and analysis of their behaviour.
In DRAINS the sets of equations presented by Henderson (1966) and Boyd (1986) given in
Section 6.8 are used to determine the headwater levels occurring with a given flowrate and
downstream tailwater level at each calculation time step. These equations allow for inlet
control, where the constriction at the opening of the culvert is the determining factor, and for
outlet control, where a high tailwater level and significant head losses make the conduit flow
full.
In DRAINS the flowrate and the downstream water level at each time step are established,
and the corresponding headwater level is determined using the above equations. When
6.10.3 Bridges
The U.S. Federal Highway Administration has published methods by Bradley (1970) which
have been used in the AUSTROADS Waterway Design manual (1994). More extensive
procedures are incorporated in the HEC-RAS computer program (Hydrologic Engineering
Center, 1997). You are referred to these references for further details. DRAINS covers
relatively simple bridge layouts. Use of HEC-RAS is recommended for complex
arrangements involving multiple openings and broad channel cross-sections,.
DRAINS uses the AUSTROADS procedures to define the afflux or rise in upstream water
level caused by a bridge constriction. It does this at each calculation time step, for the
current flowrate and downstream water level. As with culverts, allowance is made for
possible overtopping and submergence of the bridge deck, treating this as a weir. Any
overflows are added to the flows through the bridge opening occurring at the same time.
This section provides some notes on file formats, as a guide to persons exchanging data
between DRAINS and other programs.
DRAINS can import and export graphical data in DXF format. As shown in Figure 6.29, this
is an ASCII format which can be edited on a text editor. Drawings can also be imported
using the binary DWG format, which cannot be read by an editor.
Geographic Information System (GIS) programs combine a mapping facility with a data base
of information on the spatial position of components, such as drainage pits and pipes, and on
their other attributes, such as pipe diameters. Objects displayed in different ways, according
to one or more of their attributes. Maps can be produced on paper or can be inspected
electronically.
To fully specify an object such as a pipe, it is necessary to establish a set of these three files.
The transfers to and from DRAINS involve files for up to six objects - pits, sub-catchments,
pipes, overflow routes, survey data on ground levels along pipe routes and positions of other
services, a total of 18 files, plus a DXF file containing the background to the drainage
system, which can be transferred at the same time.
For nodes, a table with the following 13 headers for columns or "fields' are required:
For pipes, a table with the following 12 headers for columns or "fields' are required:
For information on the other four components, you can refer to the formats of exported ESRI
files.
The MID file involved for transfer of information on nodes between MapInfo and DRAINS
includes a table with the following 13 headers:
For pipes, the MID file is a table with the following 12 headers:
For information on the other four components, you can refer to the formats of exported
MapInfo files.
DRAINS transfers data to spreadsheet programs in the space-delimited ASCII format shown
in Figure 6.30.
The datafiles are included in Manual Example Files folder on the CD-ROM supplied with
DRAINS, which is also obtainable from the www.watercom.com.au website.
7.2.1 General
When DRAINS is installed, the files for eight examples are placed in the folder: C:\Program
Files\Drains\DataFiles. All of these are within the limit of 5 pipes or pits. The
following descriptions are similar to those in the Guide that can be downloaded from
www.watercom.com.au.
7.2.2 Examples
The DRAINS main window shown in Figure 7.1 presents a system of five pits and six pipes
or nodes, against a background of a street. The land is assumed to slope from top to
bottom. A property sheet defining the inputs for one of the pits is also shown.
Two storms have been defined. A 2 year ARI, 25 minute design storm is to be used in a
Design run to establish pipe sizes and invert levels. Both the Standard design procedure
(that follows the major/minor design method in Australian Rainfall and Runoff (Institution of
Engineers, Australia, 1987)) and the Advanced design procedure (based on a method from
the Queensland Urban Drainage Manual (1992)) can be applied. These give slightly different
results, with the advanced procedure defining a smaller pit at Pit A.3 and a smaller diameter
for Pipe A.3. The methods are explained in the Help system and the DRAINS Manual.
The designed systems can then be used in an Analysis run employing a 100 year ARI, 25
minute storm to check the adequacy of this system during a major storm.
Initially, the names of pipes start with “??” because invert levels and slopes have not been
entered in the property sheet. This partial entry of data is permitted because a Design run
will determine appropriate pipe sizes and invert levels.
This is the same pipe and catchment system as in Example 1. The Rational Method
procedure from Australian Rainfall and Runoff (1987) has been used. The Hydrological
Models and Rainfall Data inputs in the Project menu are different, as are the data for the
sub-catchments. The fall of land is from north to south.
Outputs are similar in most respects to those from Example 1, although the calculated flows
are different, being peak values rather than the full hydrographs calculated by the ILSAX
model. You can export results to a spreadsheet and produce long-sections, such as that
shown below.
This example, shown in Figure 7.4, includes the top five pipes in the pipe drainage design
example from Chapter 14 of Australian Rainfall and Runoff (Institution of Engineers,
Australia, 1987). It is similar to Example 1, except that eight 2 year ARI design storms are
used in the Design run and eight 100 year ARI storms are used for Analysis. In addition,
overland flow paths in the sub-catchments are defined in more detail, using a constant time +
kinematic wave calculation, rather than a constant time of entry. We suggest that you
explore the drainage system and results in the same way as for the first example.
This example shows how DRAINS can be applied in the modelling of an established
stormwater drainage system, where overland flows are blocked by fences and other barriers.
As shown in Figure 7.5, a detention basin is used to model a runoff storage located behind
commercial buildings.
The run with major storms determines the division of flows as they escape from the trapped
ponding area along the two escape routes. Data and results can be exported to a
spreadsheet.
This example involves a property drainage system for a dual occupancy development on a
800 m2 house lot with an on-site stormwater detention (OSD) storage at the outlet.
The land falls from west to east. A new house is to be constructed in a backyard. The
detention basin is situated on and near the driveway running past the original dwelling,
located to the right. Flows in and out of the basin are through pipes. Invert levels must be
specified for the pipes entering or leaving the detention basin, but other pipes can be
designed by DRAINS. In this example UPVC pipes are selected from the Pipe Database.
You can add additional pipe and pit types to the data base from the Edit menu.
Two systems are presented – the pre-development situation, represented by a single sub-
catchment and node, and the more detailed, post-development situation. DRAINS allows
Note how DRAINS deals with high-level or overflow outlets from a detention basin. These
are controlled by information specified for the overflow path, rather than for the basin itself.
The example is set up with a high early discharge pit, as specified by the Upper Parramatta
River Catchment Trust (www.uprct.nsw.gov.au). You can run the model with and without this
to see the difference it makes.
Example 6 - Mixed Pipe and Open Channel System at Bendigo, Victoria, with a
Detention Basin (File Bendigo Trunk Drain.drn)
This system involves pipe and artificial channel links including a detention basin, as shown in
Figure 7.7. Note that pipes connect into open channels via nodes, for which surface levels
must be specified, so that Design can proceed.
This example shows how DRAINS can calculate the afflux occurring at culverts and bridges,
and determine water surface profiles along irregular channel sections.
The lowest channel reach is an irregular one containing four cross sections. These can be
viewed in the View Cross Sections option in the pop-up menu for this reach. After an
Analysis run is carried out, water levels are shown on the cross-sections. A longitudinal
profile of the water surface along this reach can then be seen using the View maximum
water level profile option.
This model is set up for a purely rural catchment using the RORB type of storage routing
model. The hydrological model and sub-catchment data entry are different to those for
ILSAX models. Some data is entered in the sub-catchments and some in the stream
reaches. Reaches do not have levels specified. They only route flows and hydraulic grade
levels are not provided.
Detention basins can be added, and the model can be linked to open channels that are
completely-defined.
When the model runs, it produces hydrographs and indicates clearly the degree of flow
routing that takes place in each reach.
Models of the same catchment using the RAFTS and WBNM models are also provided, as
Shepparton RAFTS Rural.drn and Shepparton WBNM Rural.drn.
Example 9 - A Model that mixes ILSAX and WBNM Storage Routing Models
(File Bowral Rural-Urban.drn)
This example shows the flexibility of the arrangements in DRAINS for running two
hydrological models side by side. A small urban area is modelled using ILSAX, while the
stream to which it drains is modelled using WBNM.
In the calculations, both hydrological models are operated at the same time step.
Here the stream levels do not define a tailwater level for the pipe system, but this can be
done by adding open channel sections into which the WBNM model flowrates can be
directed. Water levels will be calculated in these and will define a variable tailwater level for
the systems.
The equivalent DRAINS model in Figure 7.12 replicates almost all of these.
(Note that this example, and the one that follows, use the older form of pit inlet capacity
specification described in Section 2.3.2.)
A notable feature of this example is the treatment of rainfalls. There are three pluviographs
covering the catchment. Using a Thiessen polygon analysis, “zones of influence” were
established for the three pluviographs. The pattern for each pluviograph is applied to the
sub-catchments within its zone. One of the patterns is shown in Figure 7.15.
A Customise Storms dialog box shown in Figure 7.16 must be opened from each sub-
catchment to define the particular pluviograph pattern that applies to it.
Zoom 3.14