RESEARCH & DEVELOPMENT

REPORT NO. RD 1051

REVIEW OF DSDS SCALE MODELING

EXPERIENCES AND PRACTICES

Research and Development Section

Land Drainage Division
Drainage Services Department
November 2009

Contents
1)

2)

3)

Page No.

Introduction
1.1 Background
1.2 Study Objectives
1.3 Scope of the Report
1.4 Forum
1.5 Acknowledgement

1
2
3
3
3

Scale Models in Hydraulic Engineering

2.1 Principle of Scaling
2.2 The Need of Scale Models
2.3 Scale Models versus Mathematical Models

4
5
6

Categorization of DSD Projects on Scale Modeling

3.1 Junction Design of Drainage Channels
3.2 Step Channel, Cascade & Stilling Basins
3.3 Tunnel Structures
- Intakes
- Outfalls
3.4 Pump Sumps for Stormwater & Wastewater Pumping Stations

9
10
13
14
17
18

4) Scale Modeling Practices

4.1 Procurement
4.2 Time and Cost
4.3 Instrumentation
4.4 Material and Inspection
4.5 Testing Record
4.6 Calibration and Verification with Field Measurements
4.7 Modeling Agencies

21
21
21
22
22
22
23

5) References of Previous Scale Modeling Experiences

25

6) Conclusions

26

References

28

1) Introduction
1.1

Background

Hydraulic structures are very often complex and in many cases their designs require
much attention so that the flow behaviour around hydraulic structures and the ir
influence on the environment can be predicted accurately. To predict the flow
behaviour around hydraulic structures, it is usually carried out by means of hydraulic
models of two distinguished types, viz. mathematical models and scale models.
Mathematical models can be used if the problem can be described mathematically
with sufficient details. In all cases, computers are used to solve these problems.
Scale models are used if the physical phenomena can be reproduced with sufficient
similarity by reducing the length dimensions of the real problem area even though the
hydraulic processes are not fully understood. 1
In general, these two types of models are used for different problems. In some cases,
both types of models can be used or a combination of the two model types is used to
predict the flow behaviour around hydraulic structures.
The purposes of hydraulic modeling studies can be briefly described as follows2 :a) to study general appearance and performance of structures;
b) to determine flow capacities and to calibrate flow measuring devices;
c) to find ways to improve the performance of hydraulic structures; and
d) to find a solution to hydraulic problems.
In recent years, mathematical models are used extensively to aid in the design of
hydraulic structures. However, most of the complicated hydraulic structures cannot
be modeled and analysed comprehensively by mathematical models alone using
common commercial 1-dimensional hydraulic softwares such as MIKE 11 (a),
InfoWorks(b) and SOBEK (c). Even if these structures can be modeled by the
softwares, the accuracies of the results may not satisfy the requirements for local
detailed design, particularly for large scale projects, due to:
(i)
the numerical assumptions and approximations necessary to construct the
generic model;
(ii)
ignored or approximated hydraulic phenomena because the model has
been set up to reproduce primarily flow routing only; and
(iii)
uncertainties in the user defined flow parameters.
(a)

MIKE 11
a DHI Software to simulate river network modeling
InfoWorks a Wallingford Software to simulate urban drain age and sewerage modeling
(c)
SOBEK
a WL | Delft Hydraulics Software for the integral simulation of processes in river and urban drainage networks
(b)

In these cases, scale models, in combination of mathematical models if necessary, can

be the solution. Nowadays, scale models are often called physical models because
other types of physical models (e.g. Hele Shaw model or electrical analog for ground
water flow) have become obsolete.
There are three interrelated phases in the modeling process: 1
(a)
(b)
(c)

Modeling implies the reproduction of the real problem area (the prototype)
in a suitable way in a model.
Solving of the technical problem in the model leads to a solution which in
principle only holds for the model.
Interpretation of the model solution is necessary to arrive at the solution
for the prototype.

The reliability of the predictions obtained by mean of the model is largely dependent
on its accuracy being tested in various events. Calibration and Verification are two
key processes in the tests. Hydrological measurements are essential to give the
necessary data for the calibrations and verifications.
Over the past two decades, Drainage Services Department (DSD) has conducted a
significant number of scale modeling tests under various projects including drainage
tunnel, pumping stations and river channel projects, etc. This Report discusses the
basic principles of scale models and summarizes DSDs experience on scale modeling
in order to enable design engineers to recognize the need of employing scale models
to aid or improve their hydraulic design.

1.2 Objectives of the Study

(a) To summarize previous experiences gained in the various scale modeling studies,
includ ing the three DSD stormwater tunnel projects that employed physical
modelling to select/design tunnel components (e.g. intake, aeration chamber,
audit tunnel, main tunnel and outfall) as well as various scale modeling studies
on pumping station design and junction design for drainage channel.
(b) To review the benefits of carrying out scale modeling tests and, where
appropriate, make recommendations for considering scale modeling tests in
future DSD projects.
(c) To summarize previous experiences on procurement of scale modeling tests and
to provide general information on the respective procurement in future DSD
projects.
2

1.3 Scope of the Report

(a) Desktop study on the scale models considered for junction design of drainage
channels, different types of drainage tunnel components including intake
structure, adit tunnel, main tunnel, stilling basin, and outfall.
(b) Desktop study on the scale models considered for pump sumps of pumping
stations. Review and summarise findings on application of scale modeling for
pump sump design.
(c) Desktop study on technical requirements of scale modeling tests to summarize
experience on preparation of technical specification.
(d) Review and make reference from past experiences on any need for additional
and/or staged model tests.
(e) Explore and examine alternative methods to scale modeling tests (e.g. advanced
mathematical models) and make recommendations on the applicable
circumstances.
(f)

Compile a list of service providers for scale modeling tests.

1.4 Forum
A DSD Forum on Scale Modeling was conducted on 31.8.2009. Experiences on
DSDs projects were presented by the professors of the University of Hong Kong
(HKU) and the Hong Kong Polytechnic University (HKPolyU) and views were
exchanged with DSDs colleagues.

1.5 Acknowledgement
Special thanks are extended to Professor Joseph H.W. LEE of HKU, Professor C.W.
LI & Professor Onyx W.H. WAI of HKPolyU and Mr. Adri VERWEY of Deltares for
their expert advices on this R&D Report

2)
2.1

Scale Models in Hydraulic Engineering

Principle of Scaling

The principle of the use of scale models consists of the possibility to reproduce the
real problem (the prototype) on a smaller scale in such a way that the phenomena in
the scale model are similar to the prototype.
The similarity can be divided into various aspects: (i) geometric similarity (ii)
kinematic similarity (iii) dynamic similarity, etc.
The relation between the model and the prototype is found by defining for each
parameter a scale (or scale factor). For complete similarity between the prototype
and the model, the ratio of each variable of the model and the correspond ing one for
the prototype must conform to the specific principles.
It is important to achieve dynamic similarity between model and the prototype. The
two most important parameters for dynamic similarity are Froude number and
Reynolds number.

Fr =

V
Dg

(Froude Number)

where Fr = Froude number, V = velocity of flow, D = flow depth, g = gravitational

acceleration.

Re =

VD
v

(Reynolds Number)

where Re = Reynolds number, V = velocity of flow, D = linear dimension, ? =

kinematic viscosity.
The Froude Law
This is applicable when the gravitational forces are predominant in the flow condition.
Similitude (or prototype/model similarity) can be obtained if the ratios of gravitational
forces to the inertia forces in the prototype and model are equal. Other forces may
be neglected. This law implies that there must be equality of Froude number
between the prototype and the model in the case where gravitational forces
predominate.
4

Froude number can be regarded as a ratio of the inertia force of a water particle to the
gravitational force acting on it. Hence, for a high Froude number, inertia dominates
and gravity has little influence on the motion of the particle. Conversely, for a low
Froude number, gravitational force becomes the major factor controlling water
motion.

Fm
To satisfy the Froude law, ____
=1

, where

Fp
FFm
m ==

Vm
Vm
, and
Dm
gm
Dm gm

Fp ==
Fp

Vp
Vp
,
Dpgp
Dp gp

where the subscripts m and p represent model and prototype respectively.

The Reynolds Law
A similar interpretation applies to the Reynolds number, which represents the ratio of
the inertia force of the water particle to the viscous force acting upon it. When
viscous forces predominate, the similitude is based on the equality of the ratio of the
inertia force and the viscous force between the prototype and the model.
It is impossible to make the pair of Froude number and Reynolds number identical in
both model and prototype if the same fluid is used. For open channel flow, equality
of Froude numbers in modeling is commonly used. Otherwise, water surface
elevations in the model bear no relationship to the prototype. Therefore, a model
with a free water surface is almost invariably a Froude-scale model while at the same
time allowance is made for influence due to the non-scaled Reynolds number because
Reynolds number is used to classify flow type (laminar to turbulent etc). In open
channel flow, both prototype and model are usually in the turbulent state.

2.2 The Need of Scale Models

Scale modeling tests are commonly conducted for detailed and complicated hydraulic
structures within large-scale drainage projects, where application of general
mathematical modeling for specific design is found limiting/insufficient due to the
assumptions and approximations made. To optimize the hydraulic design of
complicated structures, it is therefore common to conduct scale modeling tests.
It is known that assumptions, such as linear relationship, one-dimensional (1D)
characteristics, isotropic behaviour and constant boundary conditions etc., are made in
common hydraulic formulae. Some of the equations are even empirical and only
applicable under certain specific circumstances only. Flow patterns in pump sumps
5

and the other complex structures, on the other hand, are three-dimensional (3D) and
complicated by the presence of turbulence, vortex, and sediment deposition. 1D
numerical models such as MIKE 11 and InfoWorks, which average velocity over the
whole cross-sectional flow area and assume plane surface for the transverse directions,
are no longer valid.
Scale modeling is therefore useful for identifying unusual hydraulic performance and
obtaining design data and information for complicated structures like vortex-intake
structure, pumping stations, cascades, portal structures and large-size channels with
abrupt bends etc., where two to three-dimensional (2D to 3D) flows dominate.
Whenever an engineer foresees that there is a high probability for unusual hydraulic
performance and review of other types of modeling indicate that the performance
cannot be accurately assessed, and the stakes at risk are high, it is recommended to
consider conducting scale modeling tests.
Some of the model scale for DSDs intake structures of drainage tunnel ranged from
1:9.5 to 1:24.5. Model scale for pumping station was about 1:10 while that for main
tunnel was about 1:40. Engineers are always reminded to check the proposed scale
for each test. It is not uncommon for some service providers to establish 2 model
scales under the same contract to verify the sensitivity of scaling. (Ref. 5 for Hong
Kong West Drainage Tunnel Intake Structure)
Scale models can help design engineers to obtain information from the measurement
of hydraulic quantities, observe physical phenomena, understand the underlying
physics and reduce the uncertainties. Furthermore, technology transfer may be
easier as the scale model can be used as a physical demonstration explaining to the
public, District Councils or concerned parties.
However, time and cost considerations are important factors when planning to
conduct scale models for detailing complicated hydraulic structures within large-scale
drainage projects.
2.3 Scale Models versus Mathematical Models
There is a large similarity between the use of scale models and mathematical models.
The same process of modeling, solving and interpretation is required for both types of
models.
Scale models are used if the physical phenomena can be reproduced with sufficient
similarity by reducing the length dimensions of the real problem area. Mathematical
6

or numerical models can be used if the problem can be described mathematically with
sufficient details.
Nowadays, computers are used to solve these problems.
However, not for all hydraulic problems such a large selection of possibilities is
present. The computational methods may not be powerful enough to solve all
problems. The accuracy of numerically predicted results relies on the initial and
boundary conditions that are sometimes difficult to determine, especially for new
designs and setups.
Proprietary 2D and 3D numerical model may be alternatives to scale modeling.
Examples include the use of InfoWorks 2D models of HR Wallingford Software,
MIKE 21 and MIKE 3 of Danish Hydraulic Institute (DHI), Delft3D modeling series
of WL | Delft Hydraulics (Deltares). These computer programmes are state-of-the
-art softwares for hydraulic modeling. They are becoming increasingly popular and
widely used in recent years. Nonetheless, some complicated phenomena such as
heavy splashing at the outfall outlet and air trapped in the chamber, etc. may not
be satisfactorily modeled by some numerical modeling codes like Delft3D and
MIKE3. Other software such as FLUENT may be a solution to study detailed flow
processes for complex geometries, with turbulence and density flow processes.
2D and 3D numerical models have already been adopted for use in Hong Kong
though their use is currently mostly for sewerage projects such as those by the
Harbour Area Treatment Scheme (HATS) Division of DSD and those for marine water
quality impact assessments by Environmental Protection Department (EPD).
Furthermore, Computational Fluid Dynamics (CFD) Models, used in the design of
wastewater treatment plant clarifiers and bioreactors in DSD, may be able to handle
complicated structures though CFD codes are still at research and development stage.
2D and 3D numerical models have yet to be widely applied to drainage projects in
Hong Kong. However, they have already been used widely overseas, which include
modeling for flood control projects, river flow simulations and morphology
assessments. If these complex numerical model studies have to be done by
expertises abroad, the cost may be substantially higher than when the stud ies can be
done in Hong Kong.
With the advancement of computational technology in recent years, numerical
modeling may be a trend in the future. However, expertise in the effective
application of the softwares and in interpretation of the model results will be most
demanding on the hydraulic modeling staff. Having said that, with the improving
capability of 2D/3D numerical models and accumulation of relevant experiences,
useful data and information for design can be derived for characteristics of relatively
7

less complicated structures such as for flows in main tunnels, stilling basins and
simple weirs as well as that for flow splitting etc. Nevertheless, detailed comparison
of results from scale models and those from the 2D and 3D models is recommended.
For complex nature of flow, scale modeling is still a more reliable and practical
modeling technique to resolve flow problems of pump sumps, intake structures and
other complex structures. Also, air entrainment problems such as step channels are
not easily solvable in 3-D computational hydraulic model.

3) Categorization of DSD Projects on Scale Modeling

In recent years, DSD has conducted a numbers of scale modeling tests under various
drainage and sewerage projects. Some of the DSD projects that have conducted
scale modeling tests in the following categories are as follows:3.1 Junction Design of Drainage Channels
One of the examples for modeling the junction design of drainage channels is Yuen
Long Bypass Floodway (YLBF). The YLBF is designed to divert flood flows from
Sham Chung River, the San Hui Channel, and part of the flow from the designed to
divert flood flows from Sham Chung River, the San Hui Channel, and part of the flow
from the Yuen Long Main Nullah into the Kam Tin Floodway. Under a 200-year
return period design cond ition, the YLBF is designed to divert approximately 15% of
the main nullah flow and convey a total combined inflow of 278m3 /s to downstream
within acceptable flood levels.

Complicated junction flows cannot be resolved by 1D drainage models

Tight land availability
Figure 3-1 Success of Flood Control Scheme Depends on the Junction Design

The key hydraulic elements for the YLBF are the three junction designs: Main
Nullah-Bypass Floodway (MN-BF) junction, the Sham Chung-Bypass Floodway
(SC-BF) junction, and the San Hui-Bypass Floodway (SH-BF) junction. The flood
levels and hydraulic design depend greatly on the interaction of the flows at the
junction.
These complicated junction flow interactions could not be satisfactorily resolved by
1D mathematical models for their unsteady free surface flows. A scale model was
therefore developed under the project to (i) observe water levels and flow
characteristics at critical points in the model, (ii) predict actual water levels in the
YLBF and (iii) propose design to improve flow conditions.

Figure 3-2 Main Nullah-Bypass Floodway

Figure 3-3 San Hui-Bypass Floodway

(MN-BF) Junction

(SH-BF) Junction

A 1:50 undistorted model based on Froude similitude is designed to simulate the

predominantly gravity driven flow and to model the crucial three-dimensional flows
at the channel junctions. The scale model has been an invaluable tool in gaining
insights into the flow characteristics, and exploring possibilities which led to the final
design.
3.2 Stepped Channels, Cascades & Stilling Basins
On 24 June 2005, extensive overshooting of stormwater runoff from the drainage
channel on the cut slope of Smithfield was observed. This flooding resulted in a
huge amount of overland flow down the slope and onto the street. The Smithfield
drainage channel is located between Smithfield and Pokfulam Road in Hong Kong
Island (Figure 3-4).

10

Figure 3-4 Drainage Channel on Smithfield and Model Coverage

Figure 3-5 Smithfield Drainage Channel

The objective of this scale modeling study is to determine the causes of the
overshooting of the stormwater runoff and to study the effectiveness of drainage
improvement measures to prevent the recurrence of flooding. The flow conditions in
the drainage tunnel are studied in a 1:16 undistorted Froude scale model for a range of
flows.
The study was carried out in three phases: Straight channel experiment, Stepped
Cascade Smithfield drainage system and modified drainage channel. The purpose of
the straight channel experiments was to study the characteristics of the stepped
channel flow (with a view of determining the bottom roughness), and to isolate the
effect of the bend on the flooding problem.
In the second phase, scale modeling tests were performed on a 1:16 undistorted
Froude scale model of the prototype Smithfield drainage system. To facilitate flow
visualization the model is made of perspex and includes the full length of the steep
cascade, and detailed steps in the channel.
11

In the third phase, the Smithfield drainage channel is modified to improve the
drainage of the stormwater flow. The modification consists of three parts: (i)
covering the upstream portion of the lower channel with a tapered roof; (ii) raising the
side walls of the uncovered section of the lower channel; and (iii) replacing the
original stepped invert of the inlet to the steep cascade with a smooth curved spillway
entrance.
Straight Stepped
Channel

Figure 3-6 Smithfield Drainage Channel

Figure 3-7 Overview of Smithfield model

(view of covered cascade)

It was observed from the scale model that the Smithfield stepped channel has a
skimming flow result, with the main flow skimming above a pseudo-bottom joining
the corners of the steps; the re-circulating flow in the lee of the steps also serves to
dissipate energy (Figure 3-8).

Figure 3-8 Skimming Flow over Stepped Channel

12

Figure 3-9 Overshooting at Trash Grille

Spiral flow downstream of bend

Figure 3-10 Ove rs hoo ting at Smithfie ld Drainag e Channe l

From the scale modeling test, it was concluded that the overshooting flooding flow
was caused by both the channel bend and more importantly the trash grills located just
at the entrance to the steep cascade. The improved Smithfield channel design
includes:
(i)
reposition of trash grilles to upper channel,
(ii) raising side walls in uncovered section of low channel,
(iii) cover extension, and
(iv) streamlined smooth inlet to steep cascade.
The scale modeling results on the modified Smithfield channel showed that the
proposed drainage improvement works could effectively prevent the overflow and
also confirmed that the presence of the two trash grilles in the lower channel was the
dominant cause of the overshooting on the Smithfield channel.
3.3 Tunnel Structures
The design of 3D non-uniform flow hydraulic structures such as vortex shaft, outfall
from the tunnel, junction between adits and main tunnel entails much uncertainty.
These complicated structures cannot be easily identified and analyzed by simple 1D
hydraulic computational model. Scale modeling tests serve to give an insight on
how to cope with hydraulic jump, vortices, flow re-circulation and abrasion from
13

sedimentation on structures, etc. Without identifying and resolving these problems

beforehand, the subsequent maintenance cost on the tunnel and other facilities would
continue to soar. It is therefore important to consider any need for conducting scale
modeling tests prior to construction of large scale complex stormwater tunnels.
Intakes
For the intakes structures of the three drainage tunnel, vortex structures have been
widely used. Some of the general problems occurred in the vortex structures are
summarized below:
Problems due to Vortex Structures

Solution by Scale Modeling

1. Unstable and fluctuating air

core, especially during high
flow5
2. Shock wave in inlet to the
vortex structure5

a. Installation of inner guide wall in vortex

structure5
b. Use of warped invert screw type intake5
c. Reduction of approach channel length5
d. As items a and b above 5

3. Undesirable interception flow e. Lowering of the weir crest8

ratio from existing system to f. Decreasing the sluice gate opening8
the vortex structure8
g. Installation of intake weir25
h. Increasing length of the approach channel
weir26
i. Installation of convex-curved weir26
4. Overtopping in the approach j. Increasing height of wing wall25
channel25
5. Backup of flow from vortex k. As item g above 25
chamber25
l. Increasing the slope of the
channel25
6. Hydraulic jump16

approach

m. Installation of baffle walls on the cascade

structure16

Intake structures also include direct link tunnel and stilling basin. An example is the
intake and tunnel structure of the Eastern Portal of Hong Kong West Drainage Tunnel3
and the intake weir/ sluice gate of Intake-1 of Tsuen Wan Drainage Tunnel under
contract 38 . Vortex was found in the original design of the Eastern Portal. This
could be prevented by eliminating the abrupt change in flow direction in the transition
from the stilling basin to the link tunnel. For the intake weir/sluice gate of Intake-1,
main drawback of the original design was the relatively low interception flow ratio
from existing system to tunnel. This was solved by lowering the interception weir
and enlarging the sluice gate opening.
14

Scale Model

Site Condition

Conditions at Upstream of Intake (Existing condition)

Flow intercepted by
the vortex intake
system

Flow bypass to
natural stream

Approach
channel

vortex drop
shaft

vortex
intake

Figure 3-11 Scale model of the vortex drop intake system in Tsuen Wan Drainage Tunnel8

15

Figure 3-12 Typical flow pattern in the intake structure in Tsuen Wan Tunnel8

Figure 3-13 Intake structure layout in Tsuen Wan Drainage Tunnel25

Turbulence

Figure 3-14 Turbulent flows in vortex structures in Tsuen Wan Tunnel25

16

Figure 3-15 Intake and main drain tunnel in Hong Kong West Drainage Tunnel3

Sharp change of
channel

Figure 3-16 Intake and main drain tunnel in Hong Kong West Drainage Tunnel3

Outfalls
Some of the problems in the outfall design revealed and subsequently resolved by
scale modelling tests are listed below:
Problem
Solution by Scale Modeling
1. Heavy splashing at the outfall a. Installation of baffle wall on stepped
outlet30
channel30
2. Flow not well distributed30

b. Installation of triangular baffle block30

3. Hydraulic jump in box culvert30

c. Installation of baffle wall30

d. Installation of stilling basin30
e. Increasing the height of baffle wall30

4. Rip-raps on apron progressively

scoured by outflow30

f. Additional differential landing step30

g. Sandy gravel layer on apron30

17

Figure 3-17 Layout of outfall structure in Tsuen Wan Drainage Tunnel30

Splashing at outfall outlet

Figure 3-18 Physical behaviour of outfall structure in Tsuen Wan Drainage Tunnel30

3.4 Pump Sumps for Stormwater & Wastewater Pumping Stations

The standard protocol for pump sump modeling is given by Ref. 35 for BHRA &
CIRIA (July 1977), The Hydraulic design of Pump Sumps and Intakes (MJ Prosser).
Stormwater Pumping Stations
Layout and hydraulic behaviour of stormwater pumping station (e.g. Sheung Wan
Stormwater Pumping Station9 ) are shown in Figures 3-21 to 3-22 respectively.
Several major problems in the original design revealed and subsequently resolved by
scale modeling tests are listed below:
Problem
Solution by Scale Modeling
1. Sharp change of angle9
a. Installation of curved channel wall9
2.

Drag force at columns9

3.

Turbulence
region9

4.

Curved flow and jet-like d. Increasing side wall height in pump sump
phenomenon in the pump sump
region42
42
region
e. Reduction of side slope in pump sump
region42

5.

High velocity zone near a guide f. Reduction of length of guide wall42

wall42
Air trapped in the chamber42
g. Installation of slab in a chamber42

6.

in

pump

b. Concrete block surrounding the column9

sump c. Installation of weir at the outlet of the
conveying channel9

18

Figure 3-21 Layout of Sheung Wan Stormwater Pumping Station9

Turbulence

Figure 3-22 Flow pattern of Sheung Wan Stormwater Pumping Station9

Wastewater Pumping Stations

Some of the major problems in the original consultant design revealed and
subsequently resolved by scale modeling tests are listed below:
Problem
Solution by Scale Modeling
37
1. Recirculation and dead zone
a. Installation chamfer at base of wet
well37
2. Uneven flow distribution37

b. Relocation of pump to centreline of

wet well37

3. Air bubbles generated in wet well38

c. Installation of baffle wall in the inlet

of wet well38

The efficiency of the pumping stations was enhanced by introducing the

recommendation from the scale modeling test. The maintenance cost on the
structures was also reduced.

19

Figure 3-23 Typical layout of sewage pumping station in Ting Kok Road Sewage
Pumping Station36

Figure 3-24 Hydraulic behaviour of sewage pumping station in Ting Kok Road Sewage
Pumping Station36

20

4) Scale Modeling Practices

4.1 Procurement
Tendering/procurement of scale modeling test shall be based on the Stores and
Procurement Regulations (SPR). The cost of a scale modeling test normally does
not exceed the financial limit ($1.43 million in 2009) stated in SPR 220(c). In
accordance with SPR 222(b), departments shall follow the procedures set out in SPR
280 under the heading Direct Procurement of Services by Departments. It
provides guidelines on officers authorized to approve the selection of service
providers for quotation invitation, issue of invitations and the acceptance of offers
under different values of the purchase. Guidelines are also given for circumstances
when the number of invitation of quotation and quotations received fall below the
minimum requirement (five numbers) as stipulated in SPR 280(c), (i), and (ii).
The scale modeling tests of most of the projects, e.g. stormwater drainage tunnels
were employed as part of the consultancy. The consultants assisted DSD to call
quotations/tenders according to the requirements specified in the project Briefs.
Tender Reports with recommendations were submitted after checking and assessing
the quotations. Recommendations of consultants were reviewed accordingly. For
pumping station projects, which are delivered by in- house staff, all tendering
processes for scale modeling tests were carried out by DSD in- house.
4.2 Time and Cost
The test results are normally used in the design and construction of the project. The
duration of scale modeling studies is, therefore, limited by the design and construction
schedule. The contract periods of completed scale modeling tests were generally 6
to 12 months. All DSD contracts for scale modeling tests were lump sum contracts.
The contract sums varied from about HK$ 0.25M to HK$ 1.25M. Payments were
made after the service providers achieved designated milestones.
4.3 Instrumentation
Instruments are required to obtain the necessary information from scale models. The
accuracy of the model results depends, to a large extent, on the quality of the
instrumentation.
Generally, ultrasonic level meter, pressure transducer and
ultrasonic flow meter are used to measure the water level, pressure and discharge of
the flow. For some cases of which flow concentration is a controlling factor, the use
of sediment to increase the salinity is required.

21

4.4 Material and Inspection

Materials used in scale modeling tests are generally either Perspex; polycarbonates or
wood. They are durable and can be rigidly supported without distortion. Clear
materials should be used for the main body of the hydraulic structure to provide clear
visual observation on flow pattern and water profile. On-site inspection of the scale
model is necessary. Dosing dye is normally used for visualizing the flow patterns of
the scale model.
4.5 Testing Record
The progress report should be submitted every two weeks. Regular progress
meeting should also be arranged to monitor the progress of the modeling services. A
Methodology Report should be included for preliminary assessments of the details of
model design and construction. The service provider should consult the consultant/
employer prior to any necessary modifications. Depending on the parties involved,
draft reports should normally be submitted with at least 4 copies, while 6 reports are
required for the final stage. And video records on the testing should be included in
the reports.
To capture the dynamic hydraulic performance of the hydraulic structures, both video
clips and colour photographs are useful tools. Their records should be included in
the report for easy reference. For example, a hydraulic jump in a cascade structure
may be so unstable that it continues to move down the stream channel. A video clip,
rather than a thousand words, can easily demonstrate the movement of the hydraulic
jump. It should be noted that not many reports have allowed for the inclusion of
video clips to capture the testing performance.
Test models are normally stored for 12 months after submission of the final report.
The disposal of the model is the responsibility of the service providers.
4.6 Calibrations and Verifications with Field Measurements
Calibration of the model implies adjusting the model by means of prototype
measurements in such a way that the model data fit the prototype data sufficiently.
The model is then reproducing a specific, known situation in the prototype.
Verification of the model implies hindcasting of another known situation without
adjusting the model anymore. In fact, verification is a must because calibration
alone is not a sufficient guarantee for reliability. It follows that model studies cannot
replace prototype measurements. On the contrary, model studies require prototype
data 1 . General speaking, post-project monitoring is a part of verification process of
the modeling work by making use the prototype data.

22

Calibrations and validations of physical models with field measurements are

important. For some projects in which hydraulics structures are to be constructed,
field data will not be available in the course of scale modeling tests. As substitutes
for field measurements, tests can be run with two or more scale models with different
sizes. The model of larger size can be considered as the prototype of the model
with smaller size. If the results from the two models are inconsistent, the engineers
and the modelers will be required to investigate the problems more thoroughly.
Testing scenarios can be varied for different cases. The baseline test is used to verify
the model. The baseline test consists of the original structures without any
improvement works. The initial model is usually tested with 8 test scenarios such as
amber rainfall conditions, 2, 5, 10, 20, 50, 100 and 200 years of rainfall and sea level
combinations.
Generally, monitoring devices are recommended to record flow values and water
levels of the actual prototypes and to compare with the data obtained from the scale
modeling test. The information is important for verifying the accuracy of the scale
model and provides feedbacks for fine-tuning future scale modeling test.
4.7 Modeling Agencies
There are a number of Hong Kong registered organizations capable of conducting
scale model tests. Most of them are based in Hong Kong and Mainland China while
others are based in Europe. They include amongst others:
1. PolyU Technology and Consultancy Company (PTeC)
2. The University of Hong Kong
3. The Hong Kong Polytechnic University
4. China Institute of Water Resources and Hydropower Research
5. BHR Group Limited
6. Hydrotec Consultants Limited
7. HR Wallingford Limited
8. Department of Hydraulic Engineering, Tsinghua University
9. Hong Kong Productivity Council
10. Deltares (Delft Hydraulics)
11. Danish Hydraulic Institute (DHI)
In general, scale modeling tests have mostly been undertaken up by the following four
service providers:
1. PolyU Technology and Consultancy Company
2. The University of Ho ng Kong
23

3.
4.

China Institute of Water Resources and Hydropower Research

BHR Group Limited

The majority of scale modeling test contracts was awarded to local tertiary institutes.
This is mainly because of their comparatively lower bidding prices, their local
knowledge and understanding as well as their recognized track record on hydraulic
competence. Indeed, their performance is up to standard. Notwithstanding this,
there was one physical model test carried out by overseas service providers in
Europe8,16,17 . It should be noted, however, that for scale modeling tests carried out
overseas, it may be difficult for DSD staff and the consultants design team to visit the
laboratories and give on-site comments on the proposals and findings.
One of our modeling test contracts30 was awarded to an organization from the
Mainland China. Their performance was also competent and comparable to their
local counterparts even though they had a tight budget. Indeed, in their tests, a very
comprehens ive modeling schedule was prepared for a number of relatively large-scale
modelling tests. By varying the modeling details, the optimal design for the
hydraulic structure was achieved to the satisfaction of the client.

24

5) References of Previous Scale Modeling Experiences

This desk top study on previous scale modeling tests has highlighted the presence of
3D non-uniform flow in complex hydraulic structures such as stormwater drainage
tunnel, intake shafts, cascade structures in the outfall and pumping stations etc. The
contract documents and test reports of the previous physical model test have provided
useful reference and direction on how to conduct physical model tests and design
complicated hydraulic structures of similar nature during the detailed design stage.
A library keeping the aforementioned DSD physical model test reports has been setup.
The reports can be borrowed from the Lands Drainage Division (LDD) of DSD. The
list of previous DSD physical model test reports is given in the Reference section.
A summary for the different types of scale models carried out is shown in Table 1
below.
Table 1

Types of Scale Models

No.

Junction of Drainage Channel

Step Channel/Cascade

Tunnel

Stormwater Pumping Station/Storage Tank

Wastewater Pumping Station

Total

16

25

6) Conclusions

Scale modeling tests are useful tools for identifying and locating problematic areas
induced by 3D non-uniform unsteady flow in complex hydraulic structures such as
intake and outfall structures, stormwater drainage tunnels and pumping stations. The
problems can be mitigated step by step refining the geometry of the scale models of
the hydraulic structures in question. Indeed, good specifications of scale modeling
tests should define the main objective of identifying the 3D non-uniform flow
problems to the service providers who will work out the most economical and
effective mitigation measures. For the scale modeling tests completed by DSD,
mitigation measures were successfully introduced in the hydraulic design, including
inlet weirs on intakes structures, baffles on stepped channel to alleviate sharp change
in water level, curved channel and baffles in the wet well of sewage pumping stations
etc.
Scaling is important in scale modeling test but calibration and verification are also
two key processes in modeling work. Post-project monitoring is a recommended
process to verify the hydraulic design based on a scale model and complete the project
with prototype measurements, and possibly to optimize hydraulic performance (e.g.
adjusting weir levels).
When the actual prototypes are constructed and put into operation, it is recommended
to install monitoring devices to record the discharges and water levels. These
measurements can be used to verify the performance of the prototype. The
comparison can verify the accuracy of the scale models and provide feedback
information and direction for fine-tuning the prototype and improving future scale
modeling tests.
Computational modeling for 3D non-uniform flows is still at the research and
development stage. Expertise in the effective application of the softwares and in
interpretation of the model results will be most demanding on the hydraulic modeling
staff. Numerical simulation has its limitation in identifying the non-uniform flow
problems. At this stage, where stakes are high, conducting scale modeling test for
large-scale and complex 3D non-uniform hydraulic structures prior to construction is
still mandatory. It serves to physically verify the design and provides basis for
introducing effective measures to overcome hydraulic problems. It should also be
noted that, as more and more study reports of previous scale modeling tests have been
collated, they provide good references for solving the hydraulic problems of future
complex hydraulic structures with effective and economical solutions.
26

In procurement of scale modeling tests, time and cost considerations are important
factors when planning to conduct scale models for detailing complicated hydraulic
structures within large-scale drainage projects.
In the long run, with the continued development of more complex hydraulic structures
resulting in more experiences in scale modeling tests, good designs can be achieved
not only to serve with the desired hydraulic performance during operation as well as
less maintenance efforts in the future.

27

Reference
1. Scale Models in Hydraulic Engineering, Prof. dr. M de Vries, (1986- Handouts)
International Institute for Hydraulic and Environmental Engineering, Delft, the
Netherlands
2. Design of Low- head Hydraulic Structures, Water Resources Series No. 45,
United Nations, New York, 1973
3. PolyU Technology and Consultancy Company (PTeC), Final Report on Physical
Hydraulic Model Tests for Hong Kong West Drainage Tunnel in Northern Hong
Kong Island - Eastern Portal Structure and Main Drainage Tunnel, (LDD
Library Document No R&D 1051 Item No. 1)
4. Drainage Services Department, Contract Documents on Physical Hydraulic
Model Tests for Hong Kong West Drainage Tunnel in Northern Hong Kong Island
- Eastern Portal Structure and Main Drainage Tunnel, (LDD Library
Document No: R&D 1051 Item No. 2)
5. The University of Hong Kong, Final Report (Revised) on Physical Hydraulic
Model Tests for Hong Kong West Drainage Tunnel in Northern Hong Kong Island
- Intake Structures, (LDD Library Document No: R&D 1051 Item No. 3)
6. Drainage Services Department, Contract Documents on Physical Hydraulic
Model Tests for Hong Kong West Drainage Tunnel in Northern Hong Kong Island
- Western Portal Structure , (LDD Library Document No: R&D 1051 Item No.
4)
7. Drainage Services Department, Contract Documents on Physical Hydraulic
Model Tests for Hong Kong West Drainage Tunnel in Northern Hong Kong Island
- Intake Structures, (LDD Library Document No: R&D 1051 Item No. 5)
8. BHR Group Limited, Final Report on Drainage Improvement in Tsuen Wan and
Kwai Chung - Tsuen Wan Drainage Tunnel - Investigation, Physical Model Test
Contract 3, (LDD Library Document No: R&D 1051 Item No. 6)
9. PolyU Technology and Consultancy Company (PTeC), Final Report on
Modelling Test for Sheung Wan Stromwater Pumping Station, (LDD Library
Document No: R&D 1051 Item No. 7)

10. Drainage Services Department, Contract Documents on Physical Hydraulic

Model Tests for Lai Chi Kok Transfer Scheme - Channel, Stilling Basin, Siphon
& Outfall, (LDD Library Document No: R&D 1051 Item No. 8)
11. The University of Hong Kong, Variation Order No. 1 on Physical Hydraulic
Model Tests for Lai Chi Kok Transfer Scheme - Intake Structures, (LDD Library
Document No: R&D 1051 Item No. 9)
12. Hong Kong Polytechnic University, Study Brief on Modelling Tests for Wang
28

Chau Pumping Station, (LDD Library Document No: R&D 1051 Item No. 10)
13. Drainage Services Department, Letter (Tender) on Modelling Test for Sheung
Wan Stromwater Pumping Station, (LDD Library Document No: R&D 1051
Item No. 11)
14. The University of Hong Kong, Final Report (copy) on Physical Hydraulic Model
Tests for Hong Kong West Drainage Tunnel in Northern Hong Kong Island Western Portal Structure , (LDD Library Document No: R&D 1051 Item No.
12)
15. The University of Hong Kong, Final Report on Physical Hydraulic Model Tests
for Hong Kong West Drainage Tunnel in Northern Hong Kong Island - Western
Portal Structure , (LDD Library Document No: R&D 1051 Item No. 13)
16. BHR Group Limited, Draft Report on Tsuen Wan Drainage Tunnel, Physical
Model Study, Intake I-1, Model of Cascade Structure , (LDD Library Document
No: R&D 1051 Item No. 14)
17. BHR Group Limited, Final Report on Drainage Improvement in Tsuen Wan and
Kwai Chung - Tsuen Wan Drainage Tunnel - Investigation, Physical Model Test
Contract 3, (LDD Library Document No: R&D 1051 Item No. 15)
18. Drainage Services Department, Contract Documents on Drainage Improvement
in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage Tunnel-Investigation,
Physical Model Test Contract 1, 2, 3, (LDD Library Document No: R&D 1051
Item No 16
19. Drainage Services Department, Tender Documents and Brief on Modelling Test
for Sheung Wan Stromwater Pumping Station, (LDD Library Document No:
R&D 1051 Item No. 17)
20. Drainage Services Department, Contract Documents on Physical Hydraulic
Model Tests for Lai Chi Kok Transfer Scheme - Intake Structures, (LDD Library
Document No: R&D 1051 Item No. 18)
21. Drainage Services Department, Contract Documents on Physical Hydraulic
Model Tests for Lai Chi Kok Transfer Scheme - Channel, Stilling Basin, Siphon
& Outfall, (LDD Library Document No: R&D 1051 Item No. 19)
22. The University of Hong Kong, Final Report on Physical Hydraulic Model Tests
for Lai Chi Kok Transfer Scheme - Channel, Stilling Basin, Siphon & Outfall,
(LDD Library Document No: R&D 1051 Item No. 20)
23. The University of Hong Kong, Final Report on Physical Hydraulic Model Tests
for Lai Chi Kok Transfer Scheme - Intake Structures, (LDD Library Document
No: R&D 1051 Item No. 21)
24. Drainage Services Department, Requests, Approval Letters on Northern Hong
Kong Island, Lai Chi Kok, (LDD Library Document No: R&D 1051 Item No. 22)
25. PolyU Technology and Consultancy Company (PTeC), Final Report on Drainage
29

Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage

Tunnel-Investigation, Physical Model Test Contract 1, Intake I-2, (LDD Library
Document No: R&D 1051 Item No. 23)
26. PolyU Technology and Consultancy Company (PTeC), Final Report on Drainage
Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 1, Intake I-3, (LDD Library
Document No: R&D 1051 Item No. 24)
27. PolyU Technology and Consultancy Company (PTeC), Revised Report on
Drainage Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 1, Provisional Model Test for
Trash Grill of Intake I-3, (LDD Library Document No: R&D 1051 Item No. 25)
28. PolyU Technology and Consultancy Company (PTeC), Draft Test Report on
Drainage Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 1, Intake I-3, (LDD Library
Document No: R&D 1051 Item No. 26)
29. Drainage Services Department, Contract Documents on Drainage Improvement
in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage Tunnel-Investigation,
Physical Model Test Contract 1, (LDD Library Document No: R&D 1051 Item
No. 27)
30. China Institute of Water Resources and Hydropower Research, Final Report on
Drainage Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 2, Outfall O-1, Volume 1,
(LDD Library Document No: R&D 1051 Item No. 28)
31. China Institute of Water Resources and Hydropower Research, Final Report on
Drainage Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 2, Outfall O-1 Additional
Physical Model Test, Volume 2, (LDD Library Document No: R&D 1051 Item
No. 29)
32. Drainage Services Department, Contract Document on Drainage Improvement in
Tsuen Wan and Kwai Chung - Tsuen Wan Drainage Tunnel-Investigation, Physical
Model Test Contract 2, (LDD Library Document No: R&D 1051 Item No. 30)
33. Drainage Services Department, Contract Document on Drainage Improvement in
Tsuen Wan and Kwai Chung - Tsuen Wan Drainage Tunnel-Investigation, Physical
Model Test Contract 3, (LDD Library Document No: R&D 1051 Item No. 31)
34. PolyU Technology and Consultancy Company (PTeC), Final Report on Pump
Sump Modelling Tests for Sewage Pumping Stations at Ma On Shan, Sai Kung
Area 4 and Ting Kok Road, (LDD Library Document No: R&D 1051 Item No.
32)
35. BHRA & CIRIA (July 1977), The Hydraulic design of Pump Sumps and
Intakes, MJ Prosser.
36. PolyU Technology and Consultancy Company (PTeC), Final Report on Pump
30

Sump Modelling Tests for Sewage Pumping Stations at Ma On Shan, Sai Kung
Area 4 and Ting Kok Road, Final Report for Ting Kok Road Pumping Station
No.5, (LDD Library Document No: R&D 1051 Item No. 33)
37. PolyU Technology and Consultancy Company (PTeC), Final Report on Pump
Sump Modelling Tests for Sewage Pumping Stations at Ma On Shan, Sai Kung
Area 4 and Ting Kok Road, Ma On Shan Pumping Station, (LDD Library
Document No: R&D 1051 Item No. 34)
38. PolyU Technology and Consultancy Company (PTeC), Final Report on Nam
Sang Wai Sewage Pumping Station, (LDD Library Document No: R&D 1051
Item No. 35)
39. PolyU Technology and Consultancy Company (PTeC), Final Report on Kam Tin
Sewage Pumping Station, (LDD Library Document No: R&D 1051 Item No. 36)
40. PolyU Technology and Consultancy Company (PTeC), Final Report on Sha Po
Sewage Pumping Station, (LDD Library Document No: R&D 1051 Item No. 37)
41. PolyU Technology and Consultancy Company (PTeC), Revised Report on
Drainage Improvement in Tsuen Wan and Kwai Chung - Tsuen Wan Drainage
Tunnel-Investigation, Physical Model Test Contract 1, Intake I-2, (LDD Library
Document No: R&D 1051 Item No. 38)
42. The Hong Kong Polytechnic University, Final Report on Modelling Tests for
Wang Chau Pumping Station (LDD Library Document No: R&D 1051 Item No.
39)
43. BHR Group Limited, Final Report on Tsuen Wan Drainage Tunnel Physical
Model Study Intake I-1 Model of Cascade Structure (LDD Library Document No:
R&D 1051 Item No. 40)
44. The University of Hong Kong, Final Report on Tai Hang Tung Storage Scheme
Hydraulic Model Study (LDD Library Document No: R&D 1051 Item No. 41)
45. The University of Hong Kong, Final Report on Kai Tak Transfer Scheme
Hydraulic Model Study (LDD Library Document No: R&D 1051 Item No. 42)
46. Drainage Services Department, Agreement for Undertaking the Study on
Physical Model Study for Smithfield Road Flooding (LDD Library Document
No: R&D 1051 Item No. 43)
47. The University of Hong Kong, Final Report on Physical Model Study for
Smithfield Road Flooding (LDD Library Document No: R&D 1051 Item No.
44)
48. Drainage Services Department, Study Brief on Yuen Long Bypass Floodway
Hydraulic Model Study (LDD Library Document No: R&D 1051 Item No. 45)
48. The University of Hong Kong, Final Report on Yuen Long Bypass Floodway
Hydraulic Model Study (LDD Library Document No: R&D 1051 Item No. 46)
31