COWI Report-Micro

MARCH 2013
WORLD BANK GROUP
Report on the Development

and Application of the
Urban Drainage Model
SELECTION NO. 1043824/SRI LANKA: UPDATE/UPGRADE OF THE HYDROLOGIC-
HYDRAULIC MODEL AND NEW URBAN MICRO-DRAINAGE SYSTEM MODEL FOR
GREATER COLOMBO BASIN
in association with
ADDRESS COWI A/S
Parallelvej 2
2800 Kongens Lyngby
Denmark
TEL +45 56 40 00 00
FAX +45 56 40 99 99
WWW cowi.com
MARCH 2013
WORLD BANK GROUP
Report on the Development

and Application of the
Urban Drainage Model
SELECTION NO. 1043824/SRI LANKA: UPDATE/UPGRADE OF THE HYDROLOGIC-
HYDRAULIC MODEL AND NEW URBAN MICRO-DRAINAGE SYSTEM MODEL FOR
GREATER COLOMBO BASIN
PROJECT NO. P-29106

DOCUMENT NO. 1
VERSION 2.1
DATE OF ISSUE March 29th, 2013
PREPARED NEJO, JUTT, Nandalal
CHECKED NEJO
APPROVED HEGL
REPORT ON THE DEVELOPMENT AND APPLICATION OF THE URBAN DRAINAGE MODEL 3
CONTENTS
1 Introduction 6
2 Objectives 7
3 Project Boundary 8
4 Time plan and resources 12
5 Data Background 14
5.1 Rainfall data 15
5.2 Catchment data 18
5.3 Land use map of existing condition 19
5.4 Land use zoning map for future condition 23
5.5 Soil data 27
5.6 Network data 27
5.7 Infiltration and other inflows 29
5.8 Water level boundary conditions 30
6 Design standards 32
7 LiDAR survey 33
8 Modelling 34
8.1 Hydrological Model 34
8.2 Pipe flow Model 36
8.3 Other utilities 41
9 Model Application 42
9.1 Design 42
C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

4 REPORT ON THE DEVELOPMENT AND APPLICATION OF THE URBAN DRAINAGE MODEL
9.2 Evaluation of O&M Policies 43

9.3 Urban Master planning 45
9.4 Flood and flood risk 45
9.5 Status on Application of the model 47
10 Production Environment 50
11 References 57
APPENDICES
Appendix A Example of staff members 58
Appendix B Importing cross-sectional data from

Excel 60
Appendix C LiDAR survey 63

C.1 Introduction 65
C.2 LiDAR digital terrain model 69
C.3 Details of the LiDAR scanner 81
Appendix D Curve Numbers 84
Appendix E Short cut to compute hydraulic length of

catchments 86
Appendix F Changes to DHIAPP.INI 88
Appendix G Merging of 2 models 89
Appendix H Rainfall study 100

Abbreviations and acronyms
CMC Colombo Municipality Council
DLMC Dehiwala Mt. Lavinia Municipal Council
DTM Digital Terrain Model
GIS Geographical Information System
HSG Hydrological Soil Groups
IDF Intensity Duration Frequency
KAMC Kaduwela Municipal Council
KOMC Kote Municipal Council
KPS Kotikawatta Pradesiya Saba
KUC Kollonawa Urban Council
LIDAR Light Detection And Ranging
LKR Sri Lanka Rupee
PPS25 Planning Policy Statement 25
UDA Urban Development Authority
UHM Unit Hydrograph Method
USD US Dollars
SLLRDC Sri Lanka Land and Reclamation Development Corporation
SWM Solid Waste Management

1 Introduction
This report constitutes the Urban Drainage Design Report for the project
'Update/upgrade of the Hydrologic-Hydraulic Model and New Micro-Drainage
System Model for Greater Colombo Basin'.
The report describes the following main activities:
1. Development, calibration and validation of the urban drainage model.
2. The LIDAR survey.
3. The application of the model.
4. The required production environment.
The project is carried out by the Consultant consisting of COWI A/S (Denmark)
and Lanka Hydraulic Institute, LHI (Sri Lanka) in corporation with SLLRDC and
CMC in Sri Lanka.
COWI has visited CMC and assisted training the staff during three visits conducted
respectively on the 25th to the 30th June, the 23th to the 29th June and the 21st
September to the 4th October 2012.

2 Objectives
The study is structured along a set of objectives focused on lowering the flood risk
in the Greater Colombo Basin.
COWI has assisted CMC in creating and deploying an urban drainage model based
on Mike Urban software, for the assessment of the urban drainage system. The
assistance has been divided over a set of visits.
COWI's assistance has mainly focused on supporting CMC to become skilled with
the use and application of the Mike Urban software and the process of designing
storm water utilities.
The outlined set of objectives is to use the model for:
1 Propose, test and validate design options.
2 Assess and predict flood risk.
3 Master planning and assessment of urban drainage performance.
4 Investigate design policies and standards.
5 Test operation and maintenance initiatives
6 Evacuation procedures
For CMC, applying a simulation model for planning of the urban drainage system
is new and the need for support in the initial phases of development is required in
order to meet the above listed objectives.

3 Project Boundary
The storm drainage systems have been divided into two segments. First segment
that defines the primarily larger natural drainage canals termed the "Macro
Drainage System" and a second segment that primarily defines the man-made
channels and conduits forming part of the urban drainage system termed the
"Micro Drainage System".
The definition can be imprecise as parts of the natural canal system have been
upgraded with concrete wall embankments replacing the original natural
embankments.
The responsible authority for the Macro Drainage System is SLLRDC under the
Ministry of Defence. The responsible authority for the Micro Drainage System is
CMC.
The project area consists of the greater Colombo basin as shown in Figure 3-1.
The management of the micro systems for greater Colombo basin is divided
between six authorities each covering specific catchments. The location of the
catchment for each of the six authorities is shown in Figure 3-2.

Figure 3-1 Overview of Greater Colombo Basin

Figure 3-2 Overview of Greater Colombo Basin.
The authorities are:
› Colombo Municipal Council (CMC)
› Dehiwala Mt. Lavinia Municipal Council (DLMC)
› Kote Municipal Council (KOMC)
› Kollonawa Urban Council (KUC)
› Kotikawatta Pradesiya Saba (KPS)
› Kaduwela Municipal Council (KAMC)
The current study is focused on the micro drainage system that lies within the
jurisdiction of CMC.
However the areas outside CMC's jurisdiction have grown significantly and
especially KOMC's and DLMC's catchments are densely urbanized. The
development of the infrastructure in those areas is also limited and flooding
frequently reported.
The need for capacity building of the surrounding authorities is equally required
and it would be natural to extend the current program to include these authorities
once the current project has obtained sufficient knowledge.

CMC's micro drainage area is further sub-divided into 15 intervention schemes.

The location of the intervention schemes is shown in Figure 3-3. The intervention
schemes have been prioritized based on the flood risk and the social and economic
impact from flooding.
Intervention scheme no. 15, the scheme with the highest priority, has been selected
as the "front-runner" project. COWI and CMC have jointly decided to focus on this
intervention scheme as an example that can be used for work on the other
intervention schemes.
Figure 3-3 Location of 15 intervention schemes in Colombo City.

4 Time plan and resources

The planned flood protection contracts have been divided into 9 infrastructure
packages, with an execution plan according to the priority of the intervention
schemes. The planned packages, estimated contract values, commencement and
completion dates are shown in Table 4-1 below.
Table 4-1 Overveiw of intervention schemes and the time plan.

Procurem
Intervention PMU ent Est. Commence- Duration
Priority Scheme No No. Name Pkg. Val ment Completion [months]
15 (SC 2,4 & Package
1 5) 7 Marine Drive 01 200 15-Jan-13 15-Nov-13 10
15 (SC 1,3, 6 Package
2 & 7) 21 Marine Drive 02 400 1-Apr-13 1-Jan-15 20
Package
3 9 26 Horton Place 03 190 5-Nov-13 15-Sep-14 10
George R De
Silva
Armour Street Package
4 2,4 & 5 29 Green Lane 04 410 1-Oct-13 1-May-15 20
Maligawatta
Housing
Scheme Package
5 7&8 37 Norries Canal 05 650 1-Jul-13 1-May-15 20
Nagalagam Package
6 1 38 Street 06 70 15-Sep-13 15-Apr-15 7
Siridamma Package
7 10 & 11 39 Mawatha 07 435 15-Jan-14 15-Nov-15 22
Package
8 3 42 Kimbula 08 80 15-Jan-14 15-May-16 13
Saunders place
Poorwarama
Mawatha
6,12, 13 & Park Road Package
9 14 43 Thunmulla 09 810 15-Aug-14 15-Apr-16 20
The total construction sum in flood protection over the period is expected to be
30.5 mio USD equivalent to 3,245 mio. LKR. The investment will be funded
through a World Bank loan of 100 m USD to the Government of Sri Lanka. The

loan will also fund upgrade of the Macro-Drainage system as well as the required
data collection.
COWI's involvement has been limited to support of the organisation local

organisation. COWI has assisted in creation of the storm water model for the
Greater Colombo Basin and furthermore assisted in practical use of the model for
the purpose of design of intervention scheme 15 (Marine Drive Catchment),
covering the tender packages 1 and 2.
The design team appeared to be light in resources as the staff was limited to 11
engineers from which 4 engineers were dedicated to modelling. Preferable the staff
members in the design team should be increased from 11 to 15 staff members. The
full study including supervision, support functions and project management should
be between 40-50 engineers.
The 4 staff dedicated to the modelling activity is sufficient, but it is important that
the team can continue with the modelling without being interrupted by preliminary
or detailed design activities. A proposal of an organisation is given under Appendix
A. The proposed organisation must be seen as guidance considering the suggested
time plan.
Below is a description of the model work so far. The intention is that CMC, with
limited assistance during the initial phase, will be able to solely conduct the
required studies.

5 Data Background
The micro drainage model requires a range of input data in order to simulate water
levels and flow in the micro drainage network. The model consists of two
components, a hydrological model that describes the run-off in terms of surface
flow on catchments and a network model that describes the flow of water in the
drainage canals and conduits.
The data required for the hydrological model consist of:
› Rainfall data
› Catchment data
› Area, length and slope
› Land use data
› Soil data
The data required for the network model consist of:
› Network data
› Manholes
› Drainage channels and conduits
› Outlets, overflows, pumps and other hydraulic structures
› Boundary conditions in terms of water levels
Below is a more detailed description of the data background.

5.1 Rainfall data

Rainfall data close to Colombo City are available from three rain gauge stations.
These are daily rainfall data from the stations Ratmalana (at latitude 6.82N and
longitude 79.88E) and Katunayake (at latitude 7.17N and longitude 79.88E) and 15
min. rainfall data in Colombo (at latitude 6.90 N and longitude 79.87 E).
The location of the rain gauges and the project area for respectively the macro
drainage system and the micro drainage system is shown in Figure 5-1 below.
Figure 5-1 Location of rain gauges near Colombo City.
Monthly average rainfall distribution obtained based on daily rainfall data in

Colombo from 30 years of records covering a period 1981 to 2010 is shown in
Figure 5-2. The average precipitation per year is 2302 mm and the wet months are
May, October and November with a monthly rain depth above 300 mm. As stated

under reference Ref. 1 "Heavy rainstorms occur mainly in the Southwest Monsoon
period from May to September and occasionally in the two Inter-monsoon periods
(March-April and October-November) as well".
400
350
300
Rainfall (mm)
250
200
150
100
50
0
Jan Feb Mar April May June July Aug Sept Oct Nov Dec
Figure 5-2 Average Monthly Rainfall observed in Colombo (Source: Meteorological Office
in Colombo)
For assessment of flooding and design of the micro-drainage network observation

of short duration extreme rainfalls is essential. Accordingly short duration rainfall
data from the Colombo station has been analyzed. It should be noticed that the
Colombo gauge is located within the boundaries of the project area.
Based on the 30 years record a so-called Intensity-Duration-Frequency (IDF)

curves has been derived from a statistical analysis and best fit. The detailed
analysis is described under the reference (See 11Appendix H). The IDF curve
expresses the rain depth that can be expected for a set of durations and return
periods. The IDF-curve developed for Colombo is shown below in Figure 5-3:
1000.00
2 yr
5 yr
Rainfall Intensity (mm/h)
10 yr
25 yr
50 yr
100.00
100 yr
10.00
10 100 1000
Rainfall Duration (min)
Figure 5-3 Intensity-Duration-Frequency curves developed for 1-100 years return period
based on data from Colombo rain gauge (source Ref. 1, Section 11Appendix
H).

Most of the micro-drainage catchments are small in size and the time of
concentration, i.e. the time it takes for surface run-off to flow from the most distant
point on the catchment to the manhole, is short. For most catchments the time of
concentration is in the range of 5-10 minutes.
The IDF relation has been extrapolated to derive a 60 min hyetograph based on 5
min. alternating blocks for the design of the micro-drainage system. The IDF
relation can be expressed as:
Where
I = Intensity [mm/hr]
a,b,c = constants.
Duration [minutes]
For a 10 year return period a = 3473.1, b = -19.0520 and c = -0.8059
The hyetograph is derived for a 10 year return period rainfall, which is the design
frequency for the micro drainage system. The hyetograph for the design rainfall is
shown in Figure 5-4.
25
20
Rainfall (mm)
15
10
0
5 10 15 20 25 30 35 40 45 50 55 60
Time (min)
Figure 5-4 Design hyetograph 10 year storm [mm] (source Ref. 1)
The total rain depth for the 1 hour rainfall is 103 mm. The provided hyetograph
forms part of the design basis.

5.2 Catchment data
5.2.1 Digital Elevation Model

An initial digital terrain model (DTM) has been provided by UDA. The DTM is
supposed to have an accuracy of 30-40 cm. The DTM has been compared with
survey data available for the network and there are deviations between the survey
data and the DTM. It has been necessary to adjust levels as described under Section
5.6.
The DTM is using the standard Sri Lankan SLD-99 projection. The parameters for
the SLD-99 projection is shown in Table 5-1.
Table 5-1 Overview of the model set-up in SLD-99 coordinate system.
Projected Coordinate System SLD-99

False Easting 500000
False Northing 500000
Central Meridian 80.77171310
Scale factor 0.99992384
Latitude of origin 7.00047150
COWI has just completed a LIDAR survey and the data has been processed (see
section 7. The LIDAR survey will provide a better basis for the micro-drainage
model.
COWI recommends to compare the existing DTM with the LIDAR based DTM
prior to importing the DTM data into the model.
This can be done by the ArcGIS spatial analyst software (CMC will receive a
license for the spatial analyst). Larger deviations would indicate where impacts on
the modelling results and update of the network data will be required.
5.2.2 Catchment delineation

The catchments delineation has been based on a visual inspection of the DTM. For
the focus areas (Intervention Scheme 15) the catchment delineation has been
further updated based on local knowledge of the area.
During the current study a LIDAR survey of the Greater Colombo Catchment has
been completed resulting in a more precise DTM (see Section 7). In the 2012
version of Mike Urban catchment delineation can be done automatically using a
DTM so once CMC has received the 2012 version of the software it is
recommended to update the catchment delineation using the new DTM.

5.2.3 Catchment area, length and slope

The catchment area is automatically computed by Mike Urban based on the
catchment boundary.
For the Intervention Scheme 15, the slope was determined for 3 zones. A flat zone
upstream with 0.2 % slope, an area with steep slope towards Marine Drive with a
slope of 1.0 % and last zone at the coast line with a slope of 0.2 %.
5.3 Land use map of existing condition

Land use maps from year 2000 have been received from UDA. These maps have
been updated to year 2010 to represent the 'existing conditions'. The land use maps
from UDA were received in AutoCAD format and were converted to GIS shape
files. The UDA land use maps were combined into a fewer categories, see Table
6.2.
Table 5-2Land use categories by UDA and the Consultant
UDA land use maps Lumped categories by Consultant

Bank
Commercial
Defense Use
Education
Hotels &
Guesthouse
Industrial
Public Building
Religious
Residential
Shanties
Stores &
Warehouses
Utilities
Vacant Buildings
Agricultural Uses
Marsh Green Areas
Paddy(abandoned)
Minor Road
Other Road Road
Main Road
Water Area
Water Bodies
Park & Playgrounds
Vacant Land
Balance (Total catchment-sum of
Waste Disposable
above layers)
Site
Home Garden

Figure 5-5 Areas covered by the GeoEye satellite image.
The land use maps were updated to year 2010 by overlaying the 2000 land use data
on a recent 'GeoEye' satellite image provided by the National Water Supply &
Drainage Board. After that the categories in the right column of Table 6.2 were
updated manually. The GeoEye image only covered part of the basin see Figure
5-5.
The remaining area, not covered by GeoEye, was covered by Google satellite
images which were less good in resolution. However, it was judged that the Google
images were sufficient accurate for the present purpose.
The shape files of the land use maps from year 2000 and year 2010 are provided in
electronic form as part of the Consultant's delivery.
The figure below (Figure 5-6) shows the division of the basin into a number of sub-
basins. These sub-basins are mainly used for providing rainfall-runoff input for the
hydraulic model. However, the sub-catchment unit size is useful to present overall
numbers of land-use characteristics. Figure 5-6 shows the distribution of
catchments which have imperviousness less or larger than 50%. The figure is
overlaid with the extent of flooding during a 40-yr flood event.
An analysis of the data from 2000 and 2010 reveal that there has been an overall
increase in impervious areas. The table below shows the percentage of

imperviousness for each of the sub-catchments, both for the 2000 land use
conditions and for the 2010 condition. Figure 6.7 show in graphical form the
results presented in Table 6.3.
Figure 5-6Greater Colombo Basin with sub-catchments. Imperviousness less and larger than
50% is shown together with a flood map corresponding to a 40 yr return
period.
The percentage of imperviousness in Table 5-3 has been computed based on the
area for each category (roads, buildings, green areas, water bodies) and the
estimated imperviousness as indicated in Table 5-4.
There are impervious surfaces directly connected to the storm water system such as
parking lots, roads and sidewalks, while other impervious surfaces are
disconnected and hence the storm water may flow onto pervious surfaces rather
than flowing into the storm water system. Roofs disconnected from the storm water
system are a typical example. Runoff from the roof may fall onto the lawn rather
than in the storm water system, reducing the impact on the storm water system.
There has not, in this stage of the project, been made a distinction between direct
and indirect imperviousness. Instead percentages of direct imperviousness have
been estimated as per table Table 5-4.

Table 5-3Percentage of impervious area for each sub-catchment in the Colombo Basin
Impervious
Impervious Land
Land use
use percentage %
Catchment percentage %
(2000)
Name (2010)
C_01 40,24 42,75
C_02 30,84 44,02
C_03 33,86 51,56
C_04 35,30 38,05
C_05 47,12 49,65
C_06 56,07 57,13
C_07 49,77 54,63
C_08 43,98 57,33
C_09 46,61 47,79
C_10 52,90 55,70
C_11 53,48 50,19
C_12 60,31 55,89
C_13 58,29 61,38
C_14 56,22 73,82
C_15 60,99 74,51
C_16 76,21 82,39
C_17 75,25 68,06
C_18 50,75 60,89
C_19 34,27 48,55
C_20 53,82 61,95
C_21 52,58 58,54
C_22 50,99 53,66
C_23 45,67 48,22
C_24 50,73 60,90
C_25 45,73 53,27
C_26 43,81 51,57
C_27 53,36 58,89
C_28 49,17 52,97
C_29 56,97 62,51
C_30 69,53 79,80
C_31 47,71 55,05
C_32 47,08 52,26
C_33 50,01 52,11
Madiwela 35,55 40,28
Total 44,79 48,24
Table 5-4 Overview of percentage of direct imperviousness for different landuse types
Land use categories Impervious

surface
Roads 90%
Houses 75%
Parks and green areas 10%
Water bodies (Rivers/Lakes) 0%*)
Other 40%
*) Water bodies such as lakes or rivers act as water level boundary conditions to
the micro model and does not contribute with runoff to the model.

COMPARISION OF PRACENTAGE OF IMPERVIOUS AREAS BETWEEN YEAR 2000 & YEAR 2010
90,00
80,00
70,00
60,00
PERCENTAGES (%)
50,00
40,00
30,00
20,00
10,00
0,00
CATCHMENT NUMBER
IMPERVIOUS LAND USE PERCENTAGE % (2000) IMPERVIOUS LAND USE PERCENTAGE % (2010)
Figure 5-7Percentage of imperviousness for each subcatchment in the basin in year 2000 and
2010.
During the model building process the 2010 land use map was not completed.
Therefore the present model includes data from the 2000 land use map instead.
The main parameter from the land use maps which the Urban Drainage model uses
is to determine the Curve Number (See selection of Curve Number Section
11Appendix D).
5.4 Land use zoning map for future condition

A future land use map projected for year 2040 (30 year horizon) can not be made to
the same level of detail as the land use maps from year 2000 and the updated map
from 2010. For such a detailed map to be made the planned development 30 years
ahead needs to be known in detail and needs to be available at the responsible
authorities in Greater Colombo Basin. Since this is not the case, a coarser zoning
map approach has been applied.
The following is a suggested way forward on how to use the relevant spatial
information in order to create a future zoning map for urban development. A
simple approach has been adopted and implies the following:
› The development of zoning maps only considers drainage aspects, and aims at
not to worsening the flooding situation in the future.

› The level of urbanisation is assessed through the degree of imperviousness on

sub-catchment level.
› The drainage capacity is represented through the present flood inundation

pattern.
› A cut-off value for the percentage of imperviousness of 50% has been applied.
The cut-off value is used to determine sub-catchments for which a further
increase in imperviousness (urbanisation) can take place and sub-catchments
where further increase can not take place dependent on the drainage
(represented through flooding) conditions.
› If the percentage of imperviousness in a sub-catchment is less than 50% (see

Table 5-3), a further increase in the imperviousness can take place regardless
of the state of flooding in a particular sub-catchment. However, in these sub-
catchments a careful evaluation shall be made to point out areas which can be
used for retention of flood waters.
› If the percentage of imperviousness in a particular sub-catchment is larger

than 50% (see Table 5-3) then a further increase hereof is conditional to the
degree of flooding. Hence if the percentage of flooded area within the same
sub-catchment is larger than 20% (in e.g. a 40-year flood event), then it is
judged that a further increase in imperviousness should be avoided as it may
increase the severity of the flooded area to an unacceptable level. However, if
the flooded area is less than 20% then from a drainage perspective the
percentage of imperviousness can be further increased.
› For sub-catchments in which no further increase in imperviousness should

take place it is recommended that future changes in the land use are
documented with respect to proving that the overall percentage of
imperviousness has not increased further.
The maps shown in Figure 5-8 to Figure 5-9 show the resulting zoning maps using
the above principles using 20% and 0% flooding criteria as examples. It is
recommended that the principles behind the maps are discussed and perhaps
refined further by the responsible authorities.
More precise flood maps can be produced as outcome of the combined micro
drainage and macro drainage modelling activity. It is recommended, once those
flood maps are available, to make influence the planning process, avoiding further
development in zones where the flood risk is high or where flood protection
measures will be expensive.

Figure 5-8 Zoning map showing the areas with allowed increase in
imperviousness (green), and area where increase can not take place (using 20%
flooding criteria).
Figure 5-9 Zoning map showing the areas with allowed increase in
imperviousness (green),and area where increase can not take place (using 0%
flooding criteria).

Figure 5-10 As Figure 5-8 but with 40-year flood inundation map overlaid.
Figure 5-11 As Figure 5-9 but with 40-year flood inundation map overlaid.

5.5 Soil data

Geological settings of the study area are available. Generally the study area
consists of coastal sands, marshes (peats and organic clays) or laterites and lateritic
gravels.
The different type of soils maps to hydrological soil groups (HSG) when
determining the curve numbers for the SCS model. Below is an overview of the
HSG and type of soil textures:
Table 5-5 Overview of hydrological soil groups (HSG) and soil textures.
HSG Soil textures

A Sand, loamy sand or sandy loam
B Silt loam or loam
C Sandy clay loam
D Clay loam, silty clay, loam, sandy clay, silty clay or clay
5.6 Network data

A shape-file of the entire drainage network in Colombo city was available. This file
only shows the location of nodes and links and does not include any information on
e.g. manhole invert levels or diameters. This network consists of 10000 nodes and
9000 links.
Most of the network data available is based on a detailed survey from 1986-1987.
The survey was conducted by Norplan A/S Norway. The survey data consists of
app. 180 maps and 137 manhole charts covering the drainage area which CMC is
responsible for. These data have been digitalized by CMC which has resulted in a
database of 2600 links and 4000 manholes with data such as invert/ground levels,
diameter etc. During the current study sub-catchments 2, 4 and 5 in Intervention
Scheme 15 was surveyed as well as a few drains along Galle Road. Additionally
As-Built-Drawings are available for some of the areas. Some areas in the
remaining 14 intervention areas were also surveyed in 2012. The data was
processed into the same format as the 1987 data which makes it easy to update the
model with the data. At the time of writing not all of the 2012 data has been
processed and the model is therefore only updated with 2012 data in the
intervention areas Horton Place (open drains) and along Galle Road in the Marine
Drive areas.
During the current study parts of the network outside the intervention areas have
also been surveyed in order to confirm the data from 1987.
The staff at CMC was taught how to update the model with new data so they are
able to do this when the survey data is fully processed. A description of how to
import open drains into the model is given in Appendix B.

5.6.1 Manhole data

The manhole charts contained overview of incoming and outgoing pipes, top and
bottom level of manhole and invert levels of incoming and outgoing pipes.
However several man holes had not been surveyed and the cover level as well as
the invert level was estimated based on interpolation from neighbouring manholes
or the DTM.
In order to keep track of which data are original and which are estimated values,
the status field of node record was changed when modifications of the data (cover
level, invert level, diameter etc.) were made. In below table (see Table 5-6) the
status field corresponding to a specific modification is listed for nodes.
Table 5-6 Status field and corresponding modification/assumption of the node data.
Status Element_S Invert_level_S Ground_level_S Diameter_S

Model Interpolated
GIS Assigned from
DTM
Modified =Ground level - =Invert level +
0.5m 0.3m
Inserted =1
Imported Data from 2012
surveys
5.6.2 Link data

It was decided that all half round drains with a diameter less than 300mm should be
excluded from the model and be simulated as surface flow by the hydrological
model.
Many of the pipes in the system do not have information on the dimensions.
Missing pipe diameters are therefore assumed to be equal to the upstream diameter.
If the upstream pipe is rectangular the diameter is assumed to equal the width of
the upstream pipe. If the pipe is already upstream the diameter is assumed to be
300mm. The same assumption was made for rectangular pipes where missing
widths and heights are assumed to be equal to the upstream width and height.
In the dataset of open drains (both 1987 data and 2012 survey data) the cross
sections of the open drains are given for a single point along the drain. In Mike
Urban a cross section should be given for each link in the drain. Therefore it was
assumed that the cross section given for a specific point applies for the whole
upstream link.

Similar for nodes, to keep track of which data are original and which are estimated
values, the status field of link record was changed when modifications of the data
(diameter, height and width) were made. In below table (see Table 5-7) the status
field corresponding to a specific modification is listed for links.
Table 5-7 Status field and corresponding modification/assumption of the link data.
Status Element_S Diameter_S Height_S Width_S

Model = Upstream = Upstream = Upstream
dimension dimension dimension
Inserted = 0.3m = 0.3m = 0.3m
Imported Data from 2012
surveys
5.6.3 Recommendations for further surveys

In some areas of the model there is not sufficient data available to interpolate or
estimate the invert levels or pipe diameters without a great uncertainty. If these
areas have not yet been surveyed it is recommended to do so in order to improve
the data background of the model. The areas of special concern are:
› Kandewatha
› Ward Place/Kinsey Road/Rosmead Place
› Danister De Silva Mawatha
› St. Joseph Street
› Aluthmawatha Road
› De La Salle Street/Elie House Road
As mentioned in Section 3 KOMC's and DLMC's catchments are densely

populated and flooding is frequently reported. If it is decided to extent the current
program to include these authorities surveys of the network in these catchments
will be necessary as well.
5.6.4 Pump stations

The current system is purely gravity based and there were no pump stations
identified in the network.
5.7 Infiltration and other inflows

During the site visit to Intervention Scheme 15 it could be observed that inflow
from infiltration, dewatering from construction sites and inflow from illegal foul
systems was present, see Figure 5-12.

Figure 5-12 Foul flow in the storm water network, Intervention Scheme 15.
For a few catchments additional constant inflow was assumed. E.g. for the foul
flow catchments additional flow of 0.002 m3/s was added.
It is worth noticing that it was decided to use the SCS method. Infiltration is
determined based on the curve number.
5.8 Water level boundary conditions

The water level at Colombo port is available and has been used to derive estimates
for the water levels at the outlets.
The boundary conditions of the water level have been set up conservatively. In
intervention area 1-6 the water level intervals are given in the conceptual design
reports. The maximum water level given in these reports are used in the model.
For intervention scheme 15 the outlet water levels has been divided into 3 areas:
outlets to the sea, coastal near outlets and outlets in the canals. For the sea outlets
the water level has been set to 0.80 m, 0.40 m contributing from a high tide level
(spring tide) and 0.40 m estimated to be contributed from a surge. For the outlets
along the river, but close to the coast the water level was given as 1.00 m and
further upstream given as 1.20 m.
More detailed data of the expected water levels will become available once
SLLRDC has finalized the modelling activity of the macro drainage system.

SLLRDC shall deliver time series of water levels for 1/10 year design rain events
to CMC at the outlet locations of the micro drainage model(s).
For 2012-events the water levels can be obtained from SLLRDC, but until these
data are obtained the rest of the outlets are assumed to be free outlets.
5.8.1 Macro model and backwater effects
Exchange of boundary conditions between the Mike 11 model for the macro
drainage system and the Mike Urban model for the micro drainage system has been
proposed as a possible improvement of the modelling. A detailed description of
this procedure is given in Ref. 3.

6 Design standards
The initial design has been based on the rational formula and Excel spreadsheets.
The following standards have been applied by CMC:
Guide line from Irrigation Department of Sri Lanka.

Eng.AJP Ponrajah (1984)
Design of Irrigation Head works for small catchment
Irrigation Department – Sri-Lanka
Applied Hydrology (1988)

Prof. Ven.Te. Chow
Maidment and May
Hydraulic Engineering circular No-22

US Department of Transportation Federal Highway Administration Publication
FHWA-HHI-10-009 September 2009
Third Edition Urban Drainage Design Manual
Urban Storm Water Management Manuals of Asian Countries.

Performance criteria of drains and structures

7 LiDAR survey
As mentioned in section 5.2.1 COWI has completed a LIDAR survey during the
current study. The survey was conducted during September - October 2012 of the
area shown in Figure 7-1.
Figure 7-1 Area of interest for which the LiDAR survey has been conducted.
Detailed description of planning of the survey and generation of the DTM (Digital
Terrain Model) is given in Appendix C.

8 Modelling
A model with the whole network was set up by CMC under the close guidance of
COWI DK. Furthermore, CMC have worked intensively in parallel, updating and
refining the Marine Drive model. Therefore the modelling activity focused on the
intervention scheme 15 only (the Marine Drive Model) and that part of the network
will later be imported into the model covering the remaining part of the network.
The below section describes mainly the work executed on the Marine Drive Model.
8.1 Hydrological Model

CMC specifically requested to apply the Unit Hydrograph Method (UHM) based
on the US. Soil Conservative Service (SCS) curve number method. The method
simulates single storms and is useful for designing storm water systems. The
method has been applied by CMC and SLLRDC over long time and the two
organisations are familiar with this approach. Accordingly it was decided to move
ahead with the UHM hydrological model.
The use of the UHM hydrological model was illustrated by COWI and is described
below. A general update of the whole hydrological model shall be conducted once
the updated LIDAR DTM has been received.
The principle was developed by Mockus (1949). The curve number run-off can be
computed as:
For P < Ia the depth of runoff is 0.
Where
Q = depth of runoff [mm]
P = depth of rainfall [mm]
Ia = initial abstraction [mm]

S = maximum potential retention [mm]
The maximum potential retention can be expressed based on a so-called curve

number. The curve number is given as:
1000
10
25.4 !
Where CN is the curve number. As can be seen no potential retention (S=0) derives
a CN of 100 and very high potential retention derives low curve numbers.
The curve number can be derived based on a combination of soil type and land use.
The land use is dense urban. The soil type is sandy (coastal sands) as per Figure
11-1 under Appendix D and for the soil type category please refer to Table 5-5,
where the coastal sand is categorized as belonging to hydrological soil type group
A. An overview of curve numbers based on land use and soil types is shown in
Appendix D. The imperviousness is 49%, lower than the 65%. It was chosen to use
a curve number of 75 for the Marine Drive model. By interpolating in Table 11-1
based on the degree of future imperviousness (60% for Marine Drive) derives a
curve number of 74.2, which is rounded up to 75.
The curve number was assumed to reflect the expected urban development in year
2040.
The UHM model operates with three different antecedent moisture conditions
(AMC) defining the level of moisture in the ground before rain starts. Level 1
specifies dry conditions, level 2 average moisture conditions and level 3 moist
conditions. As conservative estimate CMC decided to use level 3 for moist
conditions.
The time lag between the peak in the rain and the peak in the hydrograph is
computed based on the so-called hydraulic length. For the UHM a hydraulic length
is to be specified. The hydraulic length determines the lag-time. The lag-time
specifies the time it takes from the peak in hyetograph to the peak in the
hydrograph.
The lag time is computed as:
'.,
1000
" # 3.28 10& '.(
) 9+ / 1900 . './
Where:
tl is the lag-time in hours
CN is the curve number
Y is the slope in %

L is the Hydraulic Length in km
A detailed description of the UHM model and how the routing model works can be
found under Ref. 4.
The hydraulic length depends on the size and shape of the catchment. As a "short-
cut" to compute the hydraulic length for the hundreds of catchments a process was
implemented. The process use the process tool available for Model A.
Initially the process tool to compute the time of concentration for Model A is used.
A constant velocity of 0.2 m/s is inserted in the process and the time of
concentration is automatically computed.
At last the hydraulic length is derived and updated by SQL statements in the
database. The procedure is described under Appendix E.
The model was simulated with a 60 s time step and the rainfall chosen was the 10
year design rainfall as described under Section 5.1.
As stated earlier the hydrological model could not be calibrated as there are no
detailed measurements of water level or flow.
8.2 Pipe flow Model

As for the Run-off model no actual measurements of water levels, flows or flood
levels are available. CMC has only an overview of expected areas where flooding
has been reported.
Accordingly a calibration and validation process could not take place. Under
section 9.4.2 COWI propose a "flood chaser" team that can be dispatched to
locations where frequent flooding occurs during extreme rain events. Such a team
could provide valuable information that can be used for calibration of the model.
8.2.1 Roughness and Manning numbers

The pipe materials of the drainage system were categorised into 3 main categories
in the database: PVC, Clay and Other. Initially these materials were assumed to
correspond to the predefined materials in Mike Urban: Plastic (Manning's number
= 80 m1/3/s), Ceramics (Manning's number = 70 m1/3/s) and Concrete (Rough)
(Manning's number = 68 m1/3/s), respectively. The large model is based on those
Manning numbers.
However for intervention 15 the area was visited and it was clear that the channels
were in poor condition. The problems noticed were:
› Litter and debris
› Sedimentation

› General deterioration
› Growth of plants along the channels
Examples illustrating some of the more severe locations are shown in Figure 8-1
and Figure 8-2.
All the pipes in the intervention scheme 15 consisted of rough concrete only. For
Intervention scheme 15 it was decided to lower the Manning number for rough
concrete to describe the state of the channels, so Mike Urban's default value of 68
m1/3/s was reduced to 60 m1/3/s to reflect the reduced smoothness.
At some locations downstream it may be required to lower the Manning numbers

even further as debris and sediments were detected at several locations in the
system.
Figure 8-1 Litter and debris

Figure 8-2 Sedimentation close to outlet.
The water level boundaries were set as specified under Section 5.8. The run-off
came as a result from the hydrological model as described in details under Section
8.1.
8.2.2 Simulation with open cross-sections

It should be noticed that some of the links consisted of the side drains as well as the
roads. Initially the cross-sections were defined as "open-HW" cross-sections,
however due to a defect in the Mike Urban software all HW open cross-sections
were computed as closed.
As a work around cross-sections were replaced with "open X,Z" cross-sections.

DHI informed COWI that they will work on a solution to the defect, but the
timeline does not permit to have the defect resolved in the Mike Urban 2012
release. The process of converting open H-W cross-sections for the other areas to
open X,Z cross-sections is to be conducted by CMC.
The DHIAPP.INI file was updated as the water level quickly went above the height
of the open cross-sections. The change in the DHIAPP.INI file allowed 90 degree
vertical extrapolation of the cross-section ensuring that flooded cross-sections

would not stop the simulations. The required change to the DHIAPP.INI file is
shown in Appendix F.
Simulations were run for a period of 6 hours to ensure that all run-off was
conveyed through the network. No hot-start file was used.
Initially several lines had negative slope and appeared unrealistic when looking at
the profiles. Survey data was investigated and several of the links were updated.
Several simulations were conducted in order to stabilize the model until the model
results looked realistic.
8.2.3 Refinement of the model

The current model stores artificially the flooded water on top of the manholes. The
model represents an artificial area of 1000 times the manhole area above the
manhole. The model may be able to indicate flood locations, but not actual flood
levels. The current limitations in the model could be reduced or eliminated by
refining the model.
It was decided not to refine the model further. Importing the new LIDAR DTM
would require an update of the model and network and accordingly time devoted to
enhancing the models should be done after the LIDAR DTM has been received and
imported. Furthermore the time did not permit enhancing the model further.
Once the new DTM arrives the following enhancements would be proposed:
› To include the major roads in the system as part of the conveyance system.
During heavy rain the side drains quickly overflows and the water is conveyed
along the streets (see Figure 8-3).
› To describe the inundated areas with basin and weirs. During overflow the
inundated areas are filled and slowly emptied as capacity becomes available to
convey the water downstream. This process can show actual flood levels.
› At locations with sub-surface storm water drainage, include a description of

the gully capacity with a weir description.
› Coupling with macro-drainage system and simulation of outlet boundary

conditions.
The above mentioned enhancements would make the model better to predict actual
flooding and also be more precise in determining actual flood protection measures.

Figure 8-3 Water flowing along the streets, Colombo.
8.2.4 Simulation of alternatives

The staff was divided into two teams; Team 1 consisting of Mr. Chandrapale and
Mr. Rohana and Team 2 consisting of Mr. Kumara and Mr. Chalita.
Team 1 concentrated on sub-catchment 1, 2, 3 and 4 while Team 2 considered sub-

catchment 5, 6 and 7 of the Marine Drive model.
Both teams assessed 3 alternative solutions for each sub-catchment. The

assessment was based on value engineering principles where the following aspects
were considered.
› Efficiency, i.e. how much did the solution reduce the flooding at the different
locations.
› Constructability, could the suggested measures be done within the rights of

way, was it easy to construct or would it be complex etc.
› Cost. CMC had estimates for most of the typical pipes and channels used in
Colombo. However other relevant cost was included in the calculation such as
land acquisition, traffic diversion, diversion of existing utilities etc.
› Positive and negative social and environmental impacts
› Expected time required for implementation

› Requirements with respect to approvals. Especially approvals of additional

outfalls were complex.
The preferred solution for a given sub-catchment was chosen as the solution with
the best score. The process was repeated for each sub-catchment.
The solutions will need to be refined and verified against all design criteria such as
ensuring self cleaning velocities to avoid siltation, ensure designs that can be easily
operated and maintained etc. The applied standards are listed under Section 6.
8.3 Other utilities

Normally so-called typical cross-sections are developed for planning of major
infrastructure installations. The typical cross-sections define the location (utility
corridors) of the utilities for different road cross-sections. However such typical
cross-sections do not exist for Colombo. This makes it more complex to plan
utilities as there are no well-defined utility corridors.
Furthermore overview of existing utilities does not currently exist in digital format.
This could potentially be a major challenge planning the location of the new storm
water infrastructure as the proposed storm water infrastructure may clash with
other existing or planned utilities.
COWI propose CMC to work closely with other utility departments and entities
such as waste water, water supply, telecom and electrical installations ensuring
clashes are avoided during design stage so necessary diversion of other utilities can
be planned and construction work can go ahead unhindered. Otherwise such
clashed will be discovered in the field and can be a source of significant delays,
construction problems and interruptions in utility services provided to the clients
during implementation.

9 Model Application
Modelling is a new discipline to CMC. The Mike Urban model is a one-
dimensional computational tool for simulations of unsteady flows in pipe networks
with alternating free surface and pressurized flow conditions.
A one-dimensional simulation will naturally have its limitations. Eventually

coupling of the macro-system, micro-system and 2-D overland flow description
using a Mike Flood model would be the preferred long term solution. However
there are several reasons for sticking to a 1-D model for now:
› Mike Flood is more complex to operate and the current organization does not
possess the skill set to operate the model.
› Computations in Mike Flood takes longer time
› The license is more expensive
The one-dimensional model can be utilised for a wealth of applications and with
further refinements the model would be able to simulate the same physical
processes as Mike Flood.
Below is a short description of the possible model applications CMC could

consider; however some of the applications may require further refinements to the
existing model.
9.1 Design
CMC previously used Excel spreadsheets and the rational method to design storm
water utilities. Deploying an urban drainage model allow CMC to provide better
design and evaluate the impact from different solutions before implementation.
The current Mike Urban model is suitable for evaluation of designs and flood
protection measures. The performance of the design ideas can be evaluated and the
most efficient solutions can be chosen.

The integration with ArcGIS allows importing GIS layers of land use and sensitive
areas with respect to flooding can be identified.
The new designs can be tested and the impact on the existing and new proposed
network can be fully evaluated.
During COWI's visit examples of how to design the storm water system applying
the Mike Urban model was shown.
9.2 Evaluation of O&M Policies

Proper O&M of the network is paramount. During site inspection of the Marine
Drive catchment it was evident that proper O&M of the network is a "low-
hanging" fruit for increasing the performance of the current Micro Drainage
system. According to CMC only 60 % of the system is well functioning.
Better O&M policies should be a focus area by the management.
CMC is also responsible for Solid Waste Management (SWM). This provides an
opportunity to solve the problem with debris and litter which can be observed from
several sites through cross-department initiatives. Better SWM policies can prevent
the litter from entering the storm water system in the first place. It is expensive to
remove litter from underground systems compared to removing litter at ground
level and it is even cheaper to empty bins than collecting litter from the ground.
With proper SWM policies debris and litter can be reduced.
In general the O&M department should first conduct a major clean up of the storm
water system. Such a major clean up should start from the down stream areas and
move upstream in the system. Removing litter from upstream to downstream could
potentially worsen the problem with flooding downstream.
It is possible to evaluate O&M policies with Mike Urban. COWI propose, once the
major clean up has been conducted and the most urgent needs for the design has
been addressed, to extend the use of the model to the O&M team within CMC.
The following options can be considered to evaluate O&M policies:
› Increase roughness to illustrate the effect of poor maintenance
› Conduct sensitivity analysis and identify focus areas
› Deploy a MOUSE TRAP model to simulate effects of sediment transport in

the system
Below is a short description of the options.

9.2.1 Increase roughness

A common method to illustrate the effect of different levels of O&M is through
reduction of the Manning numbers in the network.
Different levels of O&M can be tested by changing the Manning numbers

accordingly. Lower Manning numbers will increase the resistance in the pipe
network and flood levels.
Well maintained systems have lower resistance. As an example COWI has used
reduced Manning numbers (Rough concrete was reduced from 68 to 60 m1/3/s to
reflect the current state of the system).
9.2.2 Sensitivity analysis

A sensitivity analysis of the system can be conducted by blocking or reducing the
hydraulic conveyance of the system at different locations. This can be done in the
model by e.g. introduce artificial valves and close them or decrease the Manning
numbers to reflect partial blockage.
This allows the user to visualize the effect of poor maintenance at different
locations. For locations depending on single routes for drainage, blocking or
reduction in the conveyance can have severe impact during extreme rains. Other
may have redundancy and the effect from poor maintenance can be less severe.
The sensitivity analysis can point to locations in the city where the O&M effort
should be intensified.
9.2.3 Deploy a MOUSE Trap model

The MOUSE TRAP model can simulate the sediment transport in the urban
drainage system.
During extreme rainfall sediments are washed off and end up in the drainage
system. It is important to design so it is self-cleansing, however this may not
always be possible due to topographical constraints. Sediments can build up in the
system and reduce the capacity.
Knowing the locations of built up of sediments is important when maintaining the

storm water system.
As an example, during inspection of the Marine Drive sediments appeared to be a

problem in the down stream areas close to the coast, but the problem was not
significantly pronounced in the upstream areas. Due to topographical restrictions
the area downstream is designed with reduced slope and the sediments will be
naturally deposited.
CMC does not have a license to the MOUSE TRAP model. Deploying the MOUSE
TRAP model requires additional expertise and COWI does not recommend to
purchase MOUSE TRAP before the team is fully familiar with Mike Urban.

9.3 Urban Master planning

The model can be used by the city planning departments (UDA). Storm water
management is often neglected during city planning and often with severe
consequences.
Location of buildings, important infrastructure and protection of sensitive sites

against flood risk is important.
Urban development such as construction of roads or changes in grading due to land

fill can change the flow pattern and increase the flood risk. The following effects
could be observed.
› Increase in flow due to increased imperviousness
› Change in flow patterns caused by land fill or redirection of storm sewers
› Change in conveyance due to change of the land surface
The changes can be evaluated by the Mike Urban model and the consequences of
proposed urban development can be simulated before it is deployed. This can
ensure, that not only the flood risk of the new development is evaluated, but also
the change in flood risk on surrounding existing buildings and infrastructure.
The model can also point to locations where further development should be
avoided due to the flood risk as developing the required infrastructure would
exceed the value of the proposed development.
The Planning Policy Statement 25 (PPS 25) (See Ref. 2) describes how flood risk
management can be taken into account in the planning process to ensure
sustainable urban development and COWI recommends CMC to study the policies
outlined in PPS 25.
9.4 Flood and flood risk

The current model is limited in the sense that it has been created with the focus on
design. However with a few updates the model could potentially be applied to
predict flood levels and flood risk.
9.4.1 Model detailing

For the purpose of flood modelling COWI propose to include roads as conveyance
channels and describe inundated areas with basins which will allow Mike Urban to
simulate actual flood levels. The different level of detail is described in under
Section 8.2.3.
The level of detail may not be deployed over the entire model, but can be limited to
the flooded areas. COWI propose to conduct an initial assessment of flood prone
areas. The assessment can be based purely on the topography (DTM) where local

low points are identified. Once identified the model description is detailed further
around the flood prone zones.
The model should be calibrated to ensure the model predicts the flood levels
correctly.
9.4.2 Calibration of flood model

Once a flood model has been established it should be calibrated against measured
flood levels. Current data of flood levels are not available. Accordingly COWI
propose that CMC deploy a "flood chaser" team that during extreme rainfall and
with short notice can be dispatched to known flood locations. The team should
conduct the following tasks:
1 Measure the depth of the flooding as a function of time
2 Log the exact location of the measurement
3 Take photos and measure the extend of the flooding
It is important the team has synchronized the time, have a GPS device to log the
position, a camera for taking pictures, yard stick to measure flood levels and tape
measure to measure the extend of the flooding. For larger flooded areas it may be
required to have a laser measure.
The data should be documented and actual rainfall data should be collected from
the Department of Meteorology, Colombo.
Based on the collected data the micro drainage model should be calibrated against
the flood levels.
It should be noticed that the rainfall is only represented by a single rain gauge and
variations in rainfall over such a large catchment can be significant. Preferable data
from several extreme rain events should be collected to increase the basis for
calibration.
9.4.3 Application of the flood model

Once calibrated the flood model can be used to predict the frequency of flooding at
different locations as well as predict the extent, duration and depth of the flood.
It is important to know how the city is impacted when the design criteria are
exceeded.
Flood maps of the city can be produced based on the model results. The maps can
be overlaid with the land use indicating what part of the city that will be impacted
by different severity of rainfall.

An assessment of the existing flood risk can be carried out and sensitive sites to
flood can be located. Relocation of sensitive infrastructure could be initiated based
on the flood maps, so adverse effects of flooding can be prevented.
The model can also be used for evacuation procedures. Extreme rainfall can in
severe cases be life threatening. The civil defence can use the flood model to
simulate the nature of extreme rainfalls and prepare evacuation plans and conduct
drills to reduce loss of life and impacts on property during extreme rain events.
9.5 Status on Application of the model

A short summary of the modelling activity is provided below. The status describes
the state of the project when COWI demobilized from Sri Lanka on the 4th October
2012.
In general COWI focused on training CMC to a level where CMC could create,
calibrate, simulate and design using the model without assistance.
9.5.1 The greater Colombo model

As mentioned in Section 8 a model of the whole network was established. The
level of details in this model is very low compared to the Marine Drive model due
to the time limit of this study and the limited data available. Especially the
hydrological model is very simplified and the same imperviousness was defined for
all catchments due to the incomplete landuse data. The model includes all the
necessary information to run a simulation but it is required to update the model
before application. The Marine Drive model was used as an example of this during
the training of the CMC staff. The update can initially be limited to the intervention
areas and the catchments affecting these areas.
› To make sure the naming convention as defined in Section 10.1.1 is followed

for all Mike Urban Elements.
› In the areas Park Road, Andarawatha, Sidratha, Somadevi, Robert and

Kendewatha the sub-catchments have been created using Thiessen Polygons
method resulting in simple catchments. These sub-catchments should be
recreated based on the actual topography applying the LIDAR DTM as in the
rest of the model.
› Sub-catchments outside the intervention areas are relatively large compared to

the catchments inside the intervention areas. Those catchments should be
further subdivided once detailed data is available.
› To establish the UHM hydrological model and determine the catchment

slopes, curve numbers and imperviousness.
› To replace all cross-sections defined as "open-HW" with "open-XY" in order

to avoid the defect in the Mike Urban Software, as described in Section 8.2.2.

› To update the model with the remaining survey data from 2012. As mentioned
in Section 5.6 some of the 2012 survey data have imported to the model.
› To verify atypical profiles against available data. Atypical profiles may still
occur and if necessary deviations should be updated.
› To determine the external water levels. As mentioned in Section 5.8 only

external water levels in intervention areas 1-6 were included into the model.
External water levels of the remaining outlets should be estimated using the
same principle as described in Section 5.8.
› To reconsider the Mannings number in areas with high density of litter and
sediment etc., see Section 8.2.1.
› To add additional inflow to areas where infiltration and other inflows are
assumed to take place (see Section 5.7).
Some improvements may not be necessary to apply on the whole model but can be
restricted to flood prone areas.
› Include major roads in the system as part of the conveyance system.
› To describe the inundated areas with basin and weirs.
Finally the model should be calibrated to the extent that is possible to ensure the
model predicts the flood levels correctly.
9.5.2 Intervention scheme 15

For demonstration of the MikeUrban model capabilities and training of the staff
COWI and CMC focused on the Marine Drive model, intervention scheme 15.
The design hyetograph was derived and used as basis for all simulations as
described in details under Section 5.1.
The UHM hydrological model was fully established and curve numbers were
determined based on a best estimate of land use and soil types. The team could
conduct hydrological simulations and analyze the results.
For the network model profiles were visually inspected and all atypical profiles
were verified against available data (survey data or as-built drawings). Any
deviations were updated. The team was capable of conducting QA checks on the
network using the project check tool in Mike Urban as well as executing queries to
the Mike Urban database.

The network consisted of roads, side drains and conduits. The roads were only
introduced where needed and as artificial cross-sections.
The process of executing simulations and stabilizing the model numerically

required COWI assistance. The team fully understood the importance of analyzing
the continuity balance. The team was informed about the numerical concepts and
acquired skills to root-cause numerical instabilities.
The team conducted simulations with the model and followed a typical
development process as described under Section 10.1.5.
Simulations were conducted with alternative scenarios as described under Section

8.2.4.
The design process was not fully completed when COWI left Colombo, but it was
clear that the team consisting of Mr. Chandrapale, Mr. Rohana, Mr. Kumara and
Mr. Chalita was capable of completing the design process using the Mike Urban
model.
The need for communication and coordination with road and other utility teams as
well as evaluating the solutions with respect to cost, social and environmental
impacts were fully understood by the team and the communication channels to the
other teams was established.

10 Production Environment
The production environment is important to ensure the teams productivity. The
following aspects are to be considered:
› Name conventions of Mike Urban elements
› Name conventions of files and directories
› Overview of the models to maintain
› Quality Assurance
› How to keep track of changes and different versions of the model
› Single user vs. multi-user environment
› Reporting and document control
Below is a short overview of the discussions and the decisions taken during
COWI's two weeks visit to CMC in September/October 2012.
CMC must, over time, develop the required processes for a productive working
environment.
10.1.1 Name conventions of Mike Urban elements

It is important that the elements in Mike Urban are proper named and that the
elements are following the same naming convention. The following main elements
are defined in Mike Urban:
› Links
› Manholes or structures
› Catchments

Colombo is split up in quadrants. Each quadrant has a unique identification. For

manholes it has been decided to use the quadrant as first identifier followed by the
number automatically generated by Mike Urban.
So a manhole will be named <Quadrant>_<MikeUrban_ID>
Example of a manhole name: 66_13_23C_2152
Where "66_13_23C" is the quadrant and 2152 is the number generated by Mike
Urban. Typically Mike Urban generates a name "Node_Number", so the "Node"
string is replaced by the quadrant number.
Some of the manholes have been artificially created and interpolated. This is
typically the case when the survey indicates two or three lines meet, but the
connecting manhole has not been surveyed. In this case the manhole is appended
with a 'T', so the name becomes <Quadrant>_<MikeUrban_ID>T, example
66_13_22D_1282T.
For links the name is combined by the from node ID and the to node ID, so the link
name will be:
<Upstream_Node_ID>_<Downstream_Node_ID>
Example of a link ID is:
66_13_23A_2981T_66_13_23A_2853 where the upstream Node ID is

66_13_23A_2981T and the downstream ID is 66_13_23A_2853.
The catchments are named according to the name of the intervention scheme and
the
<Name_Of_Catchment>_<SubCatchment>_<MikeUrbanID>
Example:
Marine_Sub_Catch3_359
It is important that the name convention is fully maintained. Any deviations must
be resolved.
It could be seen that deviations exist in the larger model. CMC needs to clean up
all the deviations from the decided naming convention.
10.1.2 Name conventions for files and directories

A master machine was appointed. In the network the master machine had the
hostname "ACER 2". Appointing a master machine ensured that only one machine
could contain the basis of all the models and any development must be maintained
on the master machine only.

It was decided to create a file directory structure on the master machine that
reflects different levels of the model.
Furthermore some of the data was global data and would be shared between all the
models. Such data was GIS data, DTM, rainfall data etc. Those data was kept
separate under the root.
So the directory structure decided on the Master Machine was:
MikeUrban
Time Series
GIS Data
Model
Existing
Global Model
Intervention_Scheme_1
Intervention Scheme 15
Subcatchment_1_2_3_4
Subcatchment_5_6_7
Design
Intervention_Scheme_1
Intervention Scheme 15
Subcatchment_1_2_3_4
Subcatchment_5_6_7
The above structure will need to be expanded once more models become available.
It could be flood model, O&M model, further sub divisions of catchments etc.
It was also agreed that back-up should be taken regularly and that a backup should
be placed at another location than at the building.

On file level it was decided that the file name of sub-models should reflect the
purpose of the model, the intervention scheme number and the sub catchment
number. Example of filename:
Design_I15_SC_1_2_3_4.mdb
10.1.3 Models
At the current stage it was decided to maintain an existing model and keep the
design models separate.
However once the proposed solutions are locked and the packages are tendered a
model reflecting the future designs is to be maintained in parallel with the existing
model. The existing model will then be updated once the design has been
implemented in the field.
The process requires that it is possible to merge models together. The process of
merging models is described under Appendix G.
Furthermore future models for the purpose of evaluating O&M, flood assessment
or urban planning as described under Section 9 may be added.
It should be pointed out that too many models should be avoided as it will require
extra manpower to maintain the models, so if possible, maintain as few models as
possible.
10.1.4 Quality Assurance

A person must be dedicated to ensure the quality of the modelling system. COWI
propose to appoint a Quality Controller that reviews the changes to the models
before it is included.
10.1.5 Versioning and Scenario Manager

MikeUrban has an option to use the Scenario Manager. The Scenario Manager is a
tool in MikeUrban that can keep track of changes between base model and the
alternative models.
The alternative model can be created as a Child model to the Parent base model.
Such an approach is preferable when changes are relatively few. It is possible to
create children of children. This option will be used as described below.
However it is not recommended to use the Scenario Manager for describing

significant changes in a large model. As an example; deletion of a large amount of
pipes and nodes from the global model to create alternative solutions for a small
sub-catchment should be avoided. It is better to create a sub-catchment model as a
separate project file/database and create alternative solutions for the specific sub-
catchment. Otherwise every time the simulation starts it first takes the base model
as the starting point and then conducts all the required operations to come to the

child model. Here it is better to have the sub-catchment represented as the base
model and let the alternative models be children hereof.
For the Marine Drive sub-catchment the task was divided between two teams and
accordingly two separate Mike Urban models (projects). The first team considered
sub-catchment 1 to 4 and the second team sub-catchment 5 to 7. Two smaller and
separate models were created each reflecting the smaller project areas. However
within each model, each sub-catchment could be considered as a separate scenario.
So e.g. for the first model covering sub-catchment 1 to 4 three alternatives were
tried for sub-catchment 1. The preferred solution was then chosen and a new child
was created for sub-catchment 2 based on the preferred solution from sub-
catchment 1 and so forth. Finally a copy of the model was taken so the process
could later be documented. The preferred child containing all the preferred
solutions were the "set as root", so all the changes done to the model became the
new base model. The principal is shown in Figure 10-1 below.
Figure 10-1 Overview of the Scenario Manager process.
Once the two teams were fully completed a preferred model could be merged
together. The method for merging the models together is described in Appendix G.

The intention is to finally end up with a preferred proposal for each intervention
scheme that can be handed over to the design team for detailing.
10.1.6 Single user vs. multiple user

One of the issues is that the team consist of 4 members (shared between 3
MikeUrban licenses). However each team member does not have access to each
other databases and accordingly cannot concurrently work on the same intervention
scheme without doing manual merges between the models and keep track of the
changes.
Such problems could be avoided if the team could share the same database.
MikeUrban uses a Geo Database to store data. As standalone versions this is based
on Microsoft access and each database is a personal geodatabase where data cannot
be directly shared. However there is the option to allow a full networked version
where multiple users can connect to the same geo database allowing more users to
share the data and concurrently work on the same model.
Figure 10-2 Possible solution for multi-user environment sharing same database.
At the time of writing a small network had been installed for office. Having a
network allow implementing the above shown solution, where MikeUrban,
together with other ESRI client software can share the same database. In this way
data can be kept centralised and multiple users can work concurrently on the same
model.

COWI and CMC requested to receive a trial version of the ArcGIS server to test if
the solution will indeed be preferred over the existing single user approach.
The solution can have additional perspective as external access to the database
could be provided in the future. This would enable other departments, stakeholders
and authorities to access the data available.

11 References
Ref. 1 Design Rainfall for Greater Colombo Basin, Prof. K.D.W. Nandalal,University
of Peradeniya, Peradeniya, Sri Lanka, August 2012.
Ref. 2 Planning Policy Statement 25, Development and Flood Risk Practice Guide,
Department for Communities and Local Government, December 2009.
Ref. 3 Report on the Upgrading and Application of the Hydraulic Model, COWI,
October 2012.
Ref. 4 Mouse Runoff Reference Manual, (See Section 5), Unit Hydrograph Surface
Runoff Model (UHM).Mike By DHI, 2012. Danish Hydraulic Inst.

Appendix A Example of staff members

Below is an overview of the staff members proposed for design, tendering and
supervision of the proposed upgrade of the Micro Drainage System. Some of the
roles may be outsourced outside the organisation.
Table A-1 Overview of staff members
PM Unit Staff No.

Project Manager 1
Dep. PM 1
Utility Engineer 2
Technical Officer 2
Contracts specialist 1
Secretary 1
Total PM Unit 8
Support team
GIS specialist 1
IT specialist 1
Quantity surveyors 2
Survey team 2
Social specialist 1
Environmental specialist 1
Traffic specialist 1
Total support team 9
Modelling team
Urban drainage experts/Modeling specialists 4
Total Modeling team 4
Design team
Utility engineers (piping) 2
Civil engineer (structural) 2

Mechanical engineer (if and when required) 1
Electrical engineer (if and when required) 1
Road engineer 1
CAD Technicians 3
Secretary/Document Controller 1
Total Design Team 11
Supervision team
Residential Engineer (RE) 1

Assistant Residential Engineer (ARE) 1

Site Engineer 4
Civil Inspectors 4
QS 2
Surveyor 2
HSE 1
Secretary/Doc controller 1
Total staff Supervision 16

Appendix B Importing cross-sectional data

from Excel
1) Open the model in Mike Urban
2) Go to File Import/Export
3) Choose "Use Advanced" and "Create New Import/Export" Click "Next"
4) Right-click in the menu to the left choose "Create New Job" Name it
"CRS"
5) Right-click the new job you just created Choose "Create New Table Config."
Name it "CRS_MH"
5a) Right-click the new job you just created Choose "Create New Table
Config." Name it "CRS_Links"
6) Click the new job ("CRS") In Source-tab click (…) double-click "ODBC"
In Target-tab click (…) double-click "ESRI Geodatabase Storage"
In Basic-tab click (…) in Source-field Click the "Machine Data Source"-

tab select "Excel Files" Select the excel file with the CRS data
In Basic-tab click (…) in Target-field Click "Yes" when asked to connect to

actual project.
7) Click "Append and update existing data"
8) Click the new table config. called "CRS_MH" In Basic-tab click (…) in
source field select the sheet with CRS-node data in the excel-file.
In Basic-tab click (…) in Target field select the table "msm_Node".
Click "Append and update existing data"
In the field "Source ID" type the name of the unique ID of

nodes (for example Manholes_ID)
In the field "Target ID" type "MUID"
9) In the Assignment-tab keep only MUID, Invert level, Ground level, Diameter,
_X and _Y, If 2012-data: Status =3 (imported)

10) Click the new table config. called "Pipes" Do the same as in step (8) only
with pipes instead of nodes in the Assignment-tab keep only MUID, FromNode,
ToNode, Length, TypeNo = 2, CRS ID, If 2012-data: Status=3 (imported).
11) Click "OK" nodes and links of open channels are imported!
12) Once more go to File Import/Export
13) Choose "Use Advanced" and "Create New Import/Export" Click "Next"
14) Right-click in the menu to the left choose "Create New Job" Name it
"CRS_ID"
15) Right-click the new job you just created Choose "Create New Table
Config." Name it "CRS"
16) Click the new job ("CRS_ID") In Source-tab click (…) double-click
"ODBC"
In Target-tab click (…) double-click "ESRI Geodatabase Storage"
In Basic-tab click (…) in Source-field Click the "Machine Data Source"-

tab select "Excel Files" Select the excel file with the CRS data
In Basic-tab click (…) in Target-field Click "Yes" when asked to connect to

actual project.
17) Click "Only append to existing data"
18) Click the new table config. called "CRS" In Basic-tab click (…) in
source field select the sheet with the CRS-node data in the excel-file.
In Basic-tab click (…) in Target field select the table "ms_CRS".
Click "Only append to existing data"
In the field "Source ID" type the name of the CRS-ID as given in the excel-
sheet
In the field "Target ID" type "MUID"
19) In the Assignment-tab: MUID = the CRS-ID as given in the excel-sheet,

TypeNo = 2 (open H/W), Description = the Link-ID as given in the excel-sheet.
20) Click OK
21) Open the model in Access and open the table ms_CRSD

22) Copy the columns: CrsID, Sqn, HX, WZ to an excel-sheet
23) Add the CRS-data from the survey (CRS_ID, SqnNo, H, W)
24) Copy all into Access.
25) Check that the right CRS is connected to the correct link.

Appendix C LiDAR survey
C.1 Introduction 61
C.2 LiDAR digital terrain model 65
C.3 Details of the LiDAR scanner 77

C.1 Introduction
This report is composed as conclusion and documentation for planning of the
project "Colombo LiDAR survey" September - October 2012.
The project involves aerial LiDAR Survey and generation of terrain model of the
Area Of Interest (AOI). The AOI is in total approximately 248 sqkm (Figure 11
below).
Figure 11 Area of interest marked by red line. The image used for this figure is taken from
Google earth.
The aerial LiDAR survey is basis for the production of a Digital Terrain Model
(DTM).
The LiDAR survey has been done in a helicopter Bell Jet Ranger 206B3 with a IGI
LiteMapper 5600.

Figur 1 - Installation of a LiteMapper installed in a Bell Jet Range 206 reg. 4R-DLK photo from
Deccan Aviations facebook
The aerial surveys started and ended at the Colombo airport, Ratmalana.
In order to support to aerial surveys, 3 GPS base station named BIYA, A88 and
A90 was placed in the project area (see figure below). The GPS receivers were set
to collect 1 Hz. The stations are makred with yellow and the Ratmalana is marked
with a blue aircraft

Figur 2- Base stations from survey markred with yellow and the Ratmalana airport marked with
a blue aircraft.
C.1.1 Project organisation

Client/Beneficiary World Bank/SLLRDC
Sri Lanka Land Reclamation &

Development Corporation (SLLRDC)
No.03, Sri Jayawardenapura Mw.
Welikada, Rajagiriya
Sri Lanka
End client/Beneficiary: General

Manager Ms Thushari
Client/Donor: Program manager Ms
Rosanne Nitty
Client/Donor: Local program manager

Ms Samantha Wijesundera
Consultant COWI A/S

Parallelvej 2
DK-2800 Kongens Lyngby
Denmark
Team leader: Mr Henrik Garsdal

LiDAR survey project Manager: Ms
Yvonne Morville Petersen
Sub-Consultant to COWI IMAO Aerial Survey
Aéroport International de Limoges
81, Avenue de l'Aéroport
87100 - Limoges
France
Operator: Ms Perrine Thebault

Contact person: Mr Damien Giolito
Sub-Consultant to COWI Prasad Surveys Private Limited

No. 43B, Bandarawatta
Katuwawala, Boralesgamuwa
Sri Lanka
Flight service operator (provided by Deccan Aviation (Lanka) Pvt Limited
SLLRDC) The Land Mark, Level 6
No 385, Galle Road
Colombo 3
Sri Lanka
C.1.2 Requirements for the LiDAR survey

The requirement for the Colombo Aerial LiDAR survey is the following:
› LiDAR aerial survey covering the area of 248 sqkm (priority 1).
› Coordinate system is projection - TM. 2 different Coordinate systems - WGS

84 and SLD99. Datum Kandawala.
› The elevation value will refer to WGS 84 ellipsoid.
› The LiDAR point cloud must have a vertical accuracy of 10 cm RMSE and
was planned for 3,5 pts/m2.
C.1.3 Deliverables
The services include the following deliveries:
› A DTM and DEM of an area of app. 248 km2 covering Greater Colombo
Basin with a horizontal resolution of 1x1 m and a vertical accuracy of < 0.1 m

› The Datum is based on Kandawella, the projection is TM, and data will be
delivered in two different coordinate systems, WGS 84 and SLD99. The data
file formats are ASCII and USGS/DEM.
In addition the following will be delivered:
› All flight trajectories of data capture POS files (IGI format) provided with
aircraft position (easting, northing, elevation), altitude (heading, pitch, roll)
and GPS time recorded at regular intervals of 1 second or less, in ASCII text.
› Quality checked point clouds in .las-format on harddiscs
› Progress reporting as judged relevant by the Consultant.
C.2 LiDAR digital terrain model

The following goes through the course of the LiDAR project starting from the
planning of the laser scanning, the data acquisition and the data handling, analysis
of the accuracy and the data deliverables.
C.2.1 LiDAR organisation

The following persons have been responsible for carrying out the LiDAR project:
Technical Responsible - LiDAR LiDAR Specialist;

Johnny Koust Rasmussen
Phone: +45 5640 5715

E-mail: jyr@cowi.dk
Quality Responsible Head of Production Section;
Søren Vosgerau Jespersen
Phone: +45 5640 1654

E-mail: srj@cowi.dk
C.2.2 Planning and data acquisition

In the preliminary stage the data acquisition was planned in IGIplan (v.1.4). The
planning was carried out with the LM5600/IMU through the CCNS and AEC
airborne FMS (Flight Management System). Subsequently the flight plan was
transferred to the helicopter and displayed to the pilot via the CCNS (Computer
Controlled Navigation System) which is connected to the navigation system.

Figur 3 - The image shows the planned 67 flight lines (Image from Google Earth)

Data acquisition facts
Figur 4 - Shows the coverage of the flight lines. Layout Google Earth
Category Plan
Time of scanning 25-28 of September 2012
Helicopter Bell 206 Jet Ranger reg. 4R-DLK
Flying height 1,362 Ft (415m) Above Ground Level (AGL).
Speed 80 knots
Flight lines in project 67
LiDAR scanner parameters (see section C.3):

Category Result
Type Lite Mapper 5600-240
IMU 256 Hz
FOV 40º
Scan rate 100 Hz
Pulse rate 50 Hz
Nadir Point Density 4.4 points / m2

Average Point Density 4.4 points / m2

Illuminated footprint diameter 0.49 m at 1/e2
The data acquisition was carried out in cloud and fog free conditions and the result
was satisfactory.
C.2.3 Ground survey and system calibration- control points

(PRASAD surveyor)
The calibration of the LiteMapper was done - the area for the calibration is
visualized below.
The ground control points (GCPs) used in the project is measured using GPS RTK
fast static system on 4 areas located in the project. Each area contains 20-25 GCPs.
The GCPs were used to verify the vertical accuracy of the LiDAR data.

Figur 5- The picture shows the location of the calibration area and the 4 areas with ground
control points. (image from Google Earth)

Figur 6 - Shows the calibration area in detail. (image from Google Earth)
Figur 7 - The image shows the ground control point area 1 in detail. (image from Google Earth)

Figur 10 - The image shows the ground control point area 4 in detail. (image from Google
Earth)
C.2.4 LiDAR data processing
GPS/INS calculation
The GPS data is processed with the GRAFNAV Software from Trimble. Hereafter
the GPS/INS data was integrated with AERO OFFICE.
120925_COLOMBO [Combined] - Difference DGPS-IMU
0.2
0.18
0.16
0.14
0.12
0.1
0.08
0.06
0.04
[m ]
0.02
0
-0.02
-0.04
-0.06
-0.08
-0.1
-0.12
-0.14
-0.16
203 000 203 500 204 000 204 500 205 000 205 500 206 000 206 500 207 000 207 500 208 000 208 500 209 000 209 500 210 000 210 500
[sec]
north east up

120926_MATIN_COLOMBO [Combined] - Difference DGPS-IMU
0.22
0.21
0.2
0.19
0.18
0.17
0.16
0.15
0.14
0.13
0.12
0.11
0.1
0.09
0.08
[m]
0.07
0.06
0.05
0.04
0.03
0.02
0.01
0
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-0.07
271 000 272 000 273 000 274 000 275 000 276 000 277 000 278 000 279 000
[sec]
north east up
120926_MIDI_COLOMBO [Combined] - Difference DGPS-IMU
0.14
0.12
0.1
0.08
0.06
0.04
[m]
0.02
0
-0.02
-0.04
-0.06
-0.08
-0.1
288 000 289 000 290 000 291 000 292 000 293 000 294 000 295 000
[sec]
north east up
0.14
0.12
0.1
0.08
0.06
0.04
[m]
0.02
-0.02
-0.04
-0.06
-0.08
360 000 361 000 362 000 363 000 364 000 365 000 366 000 367 000
[sec]
north east up
120927_MIDI_COLOMBO [Combined] - Difference DGPS-IMU

0.18
0.16
0.14
0.12
0.1
0.08
0.06
[m]
0.04
0.02
0
-0.02
-0.04
-0.06
-0.08
-0.1
373 500 374 000 374 500 375 000 375 500 376 000 376 500 377 000 377 500 378 000 378 500 379 000
[sec]
north east up
0.25
0.2
0.15
0.1
[m]
0.05
-0.05
-0.1
446 000 446 500 447 000 447 500 448 000 448 500 449 000 449 500 450 000 450 500 451 000 451 500 452 000 452 500 453 000
[sec]
north east up
The conclusion is that the GPS/INS calculation has been successful and provides
basis for a good accuracy according to the specifications.
Processing of point cloud

The laser range measurements were combined with the GPS/INS data (trajectory)
with RI-WORLD Post Processor software from RIEGL. The output is a point cloud
in UTM zone 44 Ellipsoidal heights.
Classification of point cloud

After the calculation of the point cloud an automatic classification of the point
cloud (into terrain and non-terrain points) takes place using the Terrascan software
from Terra Solid.
Figur 11 - Cross section of city centre in Colombo.

Figur 12 - Cross section of a group of buildings in Colombo.

Figur 13 - A colour graduated DTM ASCII grid seen from above.
Quality control of point cloud and classification

After the automatic classification in Terra Scan the data was edited manually. In
the manual editing a visual inspection of the point cloud was performed and
classification errors and blunders were corrected.
It is checked that the automatically produced DTM meets the quality targets. In
case of faulty classification, e.g. in connection with slopes, woods or bridges, the
data is corrected manually.
It is checked that the data does not contain any serious errors. Serious errors (birds,
skylights, erroneous reflections in water areas) are moved to an error layer and thus
removed from the models (DSM and DTM).
The check is made in MicroStation, using the TerraScan and TerraModeler

applications. Cross sections are created by means of the "draw section" function
and colored TIN. The most likely sources of errors are checked as well as the
whole blocks are checked as a colored TIN.
The editors are in particular considering the following:
Water areas: Points under the water surface are moved to the error layer.
Skylights: Points surveyed through skylights are moved to the error layer.

Birds: Points very high above terrain are moved to the error layer.
Slopes: Slopes/dikes are often classified as DSM by automatic DTM generation.

Slopes classified as DSM are moved to the DTM layer.
Woods: In dense woods or woods with dense underwood, wrong classifications of

the automatic DTM production may have been made. The editor makes a manual
re-classification of the area in question. Different kinds of woods have to be
investigated carefully.
Bridges: Points on bridges are moved to the DSM layer.
Large temporary objects: Large temporary objects (trains, ships etc.) must be
moved to the error layer. Small temporary objects (cars, trucks etc.) should remain
in DSM layer.
The geometric accuracy was checked by measuring some control points in the area.
A comparison of the control points and laser data can be seen below.
C.2.5 Density of the LiDAR data
Point cloud
The point cloud consists of up to 4 echo's from each submitted pulse. To visualize
the density of the point cloud only 1 point from each pulse is considered.
The density is calculated as points / m2. Minimum 95% of cells of 10 x 10 m shall

consist of minimum 2 pts/m2.
The point density - excluding water areas and other parts with low reflection - shall
satisfy the expected point density of 2 points pr. square meter. The point density
varies by scanning angle and actual flight conditions no more than expected. Some
areas have higher density than others - this is due to overlapping flight lines.
C.2.6 Quality
Quality assessment
The final quality of the LiDAR is a complex process. To ensure fulfilment of
the height accuracy of the LiDAR data of 10 cm RMSE (in z) the following
must be observed:
› GPS/INS calculation (position and orientation/trajectory of the helicopter):

The Surveyor provided two GPS base stations within the original AOI. Data
from the GPS antennas is based on GPS service from the Sri Lankan Survey
Department. The location of the base stations was based at the edges of the
original AOI to ensure a low distance impact.
› Range measurement (distance from the helicopter to the ground)

The flying height was 1,362 feet. This low flying height ensured a low impact
on the error from range and orientation measurements.
› Leverarm measurement
The distance between the IMU (scanner unit) and the GPS receiver fixed on
the helicopter was measured with high accuracy.
› Calibration of the sensors (Boresight angles) is done onsite on a dedicated area

with a specific flight pattern. This was done twice during the missions.
› Quality check of the survey output was implemented to ensure a thorough

check on the accuracy of the data. Therefore the error contribution above in
the LiDAR data was checked with measurement of five control surfaces
within the AOI.
By comparing these control surfaces with the survey point cloud the difference
calculated stipulates the accuracy of the LiDAR survey. The difference between the
LiDAR data and the ground measurement was expected to be less than 10 cm
RMSE. The control report for the 97 ground control points (GCPs) had a RMSE
value of 7.4 cm.
The data collecting, processing was carried out successfully. The RMSE is within
limit of the expected 10 cm.
C.3 Details of the LiDAR scanner

Below is shown the specifications of the LiDAR scanner Lite Mapper.



Appendix D Curve Numbers

Table 11-1 Overview of curve numbers as a function of soil group and land use.
Curve Number by
Hydrologic Soil
Average %
Description Group Typical Land Uses
Impervious
A B C D
Multi-family,
Residential (High
65 77 85 90 92 Apartments, Condos,
Density) Trailer Parks
Residential (Med. Single-Family, Lot Size

30 57 72 81 86 ¼ to 1 acre
Density)
Residential (Low Single-Family, Lot Size 1

15 48 66 78 83 acre and Greater
Density)
Strip Commercial,
Commercial 85 89 92 94 95 Shopping Ctrs,
Convenience Stores
Light Industrial, Schools,

Industrial 72 81 88 91 93 Prisons, Treatment Plants
Gravel Parking, Quarries,

Disturbed/Transitional 5 76 85 89 91 Land Under
Development
Cultivated Land, Row

Agricultural 5 67 77 83 87 crops, Broadcast
Legumes
Parks, Golf Courses,

Open Land – Good 5 39 61 74 80 Greenways, Grazed
Pasture
Hay Fields, Tall Grass,

Meadow 5 30 58 71 78 Ungrazed Pasture
Forest Litter and Brush

Woods (Thick Cover) 5 30 55 70 77 adequately cover soil
Light Woods, Woods-

Woods (Thin Cover) 5 43 65 76 82 Grass combination, Tree
Farms
Paved Parking, Shopping

Impervious 95 98 98 98 98 Malls, Major Roadways
Water Bodies, Lakes,

Water 100 100 100 100 100 Ponds, Wetlands

Figure 11-1 Soil map for Greater Colombo.

Appendix E Short cut to compute hydraulic

length of catchments
1) Open the model in Mike Urban and set the model in Edit Mode
2) Select MOUSE Catchment Tools Catchment Processing
3) The Catchment Processing Wizard will start
4) Press Next
5) Select Model Type to "MOUSE T-A Curve (A)
6) Tick the "Calculate Hydrological Parameters"
7) Press Next
8) Specify a mean surface velocity to 0.2 m/s and keep the other default values.
8) Press Next
9) Select "Choose between default Curves. This will allow MikeUrban to choose a
relation based on the catchment shape.
10) Press Next
11) Press Finish and let Mike Urban process the data.
12) Select Mouse Catchment & Catchment Parameters Time-Area (A) …
13) Press Edit for the field Parameter set
14) Choose Select by attribute
15) Select records where the time of concentration is less than 4 minutes by
choosing Advance Show Selected
16) Update time of concentration to a minimum time of concentration of 4 minutes

using the field calculator.
17) Select Tools General SQL Command …
18) Execute the below 4 SQL statements:
UPDATE msm_HModUHM u set u.MethodNo=1, u.LossModelNo=3, u.LagTimeMethodNo=2;
UPDATE msm_HModUHM u INNER JOIN (SELECT parA.MUID, parA.ConcTime,

parA.TACurveID, modA.CatchID, modA.ParAid FROM msm_HParA as parA, msm_HModa

modA WHERE (modA.ParAID=parA.MUID) AND (parA.TACurveID="TACurve1")) AS s ON

u.CatchID=s.CatchID set u.HydraulicLength=1.0*parA.ConcTime*60*0.1;


19) Select MOUSE Catchments & Catchment Parameters UHM and verify
the hydraulic length has been recomputed.

Appendix F Changes to DHIAPP.INI

The following changes to the DHIAPP.INI file is required for allowing the cross-
sections to extrapolate when the water level reaches the top of the cross-section.
; When modelling open channels the model will normally stop if flooding
occurs.
; Here it can be chosen globally to extrapolate all the cross-sections.
; CRS_EXTRAPOLATION=1 gives no extrapolation, i.e. equal to older MOUSE
versions (<2002)
; CRS_EXTRAPOLATION=2 gives an axtrapolation according to the angle give
in:
; CRS_EXTRAPOLATION_ANGLE=90.0 is the angle between a horizontal level and
the extrapolated CRS
; CRS_EXTRAPOLATION_HEIGHT_FACTOR=3.0 is the maximum height of the
extrapolated CRS given
; as this factor times the maximum height of the CRS
CRS_EXTRAPOLATION=1
CRS_EXTRAPOLATION_ANGLE=90.0
CRS_EXTRAPOLATION_HEIGHT_FACTOR=1.0

Appendix G Merging of 2 models

Copy the text in the box below to a txt-file (e.g. via Notepad). Change the path
marked in red to the path leading to the location of the model you want to merge
into the total model. Save the txt-file and substitute the extension "txt" with "xml".
<?xml version="1.0" standalone="yes"?>

<muIO>
<Task>
<a Name="Import_Node_Link_Catchments" />
<a Enabled="True" />
<a WriteMode="Update" />
<a ShowDetails="True" />
<a SelectedOnly="False" />
<a Source="C:\Documents and Settings\jutt\Colombo\Urban
Drainage\Models_CMC\model.mdb" />
<a Target="RunTime Workspace" />
<SourceStorage>
<a Name="ESRI Geodatabase Storage" />
<Property Name="Read_OID" Value="No" />
<Property Name="Read_ESRIGeometry" Value="Yes" />
<Property Name="Read_AlternativeGeometry" Value="All" />
<Property Name="Compute_FromToNode" Value="Yes" />
<Property Name="ESRICompatible" Value="Yes" />
<Property Name="ESRITopologyCompatible" Value="Yes" />
<Property Name="WriteMode" Value="Update" />
<Property Name="Enable_Fill_Default" Value="Yes" />
</SourceStorage>
<TargetStorage>
<a Name="ESRI Geodatabase Storage" />
<Property Name="Read_OID" Value="No" />
<Property Name="Read_ESRIGeometry" Value="Yes" />
<Property Name="Read_AlternativeGeometry" Value="All" />
<Property Name="Compute_FromToNode" Value="Yes" />
<Property Name="ESRICompatible" Value="Yes" />
<Property Name="ESRITopologyCompatible" Value="Yes" />
<Property Name="WriteMode" Value="Update" />
<Property Name="Enable_Fill_Default" Value="Yes" />
</TargetStorage>
<Bridge>
<a Name="Standard MU DM Bridge" />
</Bridge>
<Section>
<a Name="Node" />
<a Source="msm_Node" />
<a Target="msm_Node" />
<a AdvancedMode="True" />
<a SectionMode="Update" />

<a SourceID="MUID" />

<a TargetID="MUID" />
<Commands>
<C C="OBJECTID = OBJECTID" />
<C C="SHAPE = SHAPE" />
<C C="MUID = MUID" />
<C C="TypeNo = TypeNo" />
<C C="InvertLevel = InvertLevel" />
<C C="GroundLevel = GroundLevel" />
<C C="Diameter = Diameter" />
<C C="GeometryID = GeometryID" />
<C C="CriticalLevel = CriticalLevel" />
<C C="LossParID = LossParID" />
<C C="LossParNo = LossParNo" />
<C C="OutletShapeNo = OutletShapeNo" />
<C C="LossTypeNo = LossTypeNo" />
<C C="LossCoeff = LossCoeff" />
<C C="EffAreaNo = EffAreaNo" />
<C C="PMTypeNo = PMTypeNo" />
<C C="PMLevel = PMLevel" />
<C C="CoverTypeNo = CoverTypeNo" />
<C C="BufferPressure = BufferPressure" />
<C C="SpillCoef = SpillCoef" />
<C C="QHTypeNo = QHTypeNo" />
<C C="OutletQHID = OutletQHID" />
<C C="InletControlNo = InletControlNo" />
<C C="MaxInlet = MaxInlet" />
<C C="AssetName = AssetName" />
<C C="DataSource = DataSource" />
<C C="Description = Description" />
<C C="SubModelNo = SubModelNo" />
<C C="NetTypeNo = NetTypeNo" />
<C C="Element_S = Element_S" />
<C C="InvertLevel_S = InvertLevel_S" />
<C C="GroundLevel_S = GroundLevel_S" />
<C C="Diameter_S = Diameter_S" />
<C C="Geometry_S = Geometry_S" />
<C C="Coupling_2D = Coupling_2D" />
<C C="CouplingMethod_2D = CouplingMethod_2D" />
<C C="MaxFlow_2D = MaxFlow_2D" />
<C C="InletArea_2D = InletArea_2D" />
<C C="QdH_2D = QdH_2D" />
<C C="OrificeCoef_2D = OrificeCoef_2D" />
<C C="CrestWidth_2D = CrestWidth_2D" />
<C C="ScaleFactor_2D = ScaleFactor_2D" />
<C C="Exponent_2D = Exponent_2D" />
<C C="_X = _X" />
<C C="_Y = _Y" />
<C C="_Z = _Z" />

</Commands>
</Section>
<Section>
<a Name="Link" />
<a Source="msm_Link" />
<a Target="msm_Link" />
<Commands>
<C C="UpLevel = UpLevel" />
<C C="DwLevel = DwLevel" />
<C C="Length = Length" />
<C C="UpLevel_C = UpLevel_C" />
<C C="DwLevel_C = DwLevel_C" />
<C C="Length_C = Length_C" />
<C C="Slope_C = Slope_C" />
<C C="Diameter = Diameter" />
<C C="Width = Width" />
<C C="Height = Height" />
<C C="CrsID = CrsID" />
<C C="TopographyID = TopographyID" />
<C C="MaxDx = MaxDx" />
<C C="MaterialID = MaterialID" />
<C C="FricTypeNo = FricTypeNo" />
<C C="FricNo = FricNo" />
<C C="Manning = Manning" />
<C C="Rough = Rough" />
<C C="HWCoef = HWCoef" />
<C C="PMApprNo = PMApprNo" />
<C C="NonReturnNo = NonReturnNo" />
<C C="DataSource = DataSource" />
<C C="UpLevel_S = UpLevel_S" />
<C C="DwLevel_S = DwLevel_S" />
<C C="Length_S = Length_S" />
<C C="Diameter_S = Diameter_S" />
<C C="Width_S = Width_S" />
<C C="Height_S = Height_S" />
<C C="CrsID_S = CrsID_S" />

<C C="TopographyID_S = TopographyID_S" />

<C C="Manning_S = Manning_S" />
<C C="Rough_S = Rough_S" />
<C C="_FromNodeID = _FromNodeID" />
<C C="_ToNodeID = _ToNodeID" />
<C C="_Length = _Length" />
<C C="_FromX = _FromX" />
<C C="_FromY = _FromY" />
<C C="_FromZ = _FromZ" />
<C C="_ToX = _ToX" />
<C C="_ToY = _ToY" />
<C C="_ToZ = _ToZ" />
</Commands>
</Section>
<Section>
<a Name="Catchment" />
<a Source="ms_Catchment" />
<a Target="ms_Catchment" />
<Commands>
<C C="Area = Area" />
<C C="X = X" />
<C C="Y = Y" />
<C C="Area_C = Area_C" />
<C C="X_C = X_C" />
<C C="Y_C = Y_C" />
<C C="MaxLevel = MaxLevel" />
<C C="MinLevel = MinLevel" />
<C C="SystemID = SystemID" />
<C C="SubSystemID = SubSystemID" />
<C C="PEsNo = PEsNo" />
<C C="Persons = Persons" />
<C C="_X = _X" />
<C C="_Y = _Y" />
<C C="_Z = _Z" />
<C C="_Area = _Area" />
</Commands>
</Section>

<Section>
<a Name="Catchment_Geometry" />
<a Enabled="False" />
<a Source="ms_Catchment_Geometry" />
<a Target="ms_Catchment_Geometry" />
<a SourceID="GID" />
<a TargetID="GID" />
<Commands>
<C C="GID = GID" />
<C C="Sqn = Sqn" />
<C C="X = X" />
<C C="Y = Y" />
<C C="Z = Z" />
</Commands>
</Section>
<Section>
<a Name="Catchment_Connection" />
<a Source="msm_CatchCon" />
<a Target="msm_CatchCon" />
<Commands>
<C C="CatchID = CatchID" />
<C C="NodeID = NodeID" />
<C C="LinkID = LinkID" />
<C C="Distance = Distance" />
<C C="Fraction = Fraction" />
</Commands>
</Section>
<Section>
<a Name="Catchment_Link" />
<a Source="msm_CatchConLink" />
<a Target="msm_CatchConLink" />
<Commands>
<C C="CatchConID = CatchConID" />

<C C="_FromNodeID = _FromNodeID" />

<C C="_ToNodeID = _ToNodeID" />
<C C="_Length = _Length" />
<C C="_FromX = _FromX" />
<C C="_FromY = _FromY" />
<C C="_FromZ = _FromZ" />
<C C="_ToX = _ToX" />
<C C="_ToY = _ToY" />
<C C="_ToZ = _ToZ" />
</Commands>
</Section>
<Section>
<a Name="ModA" />
<a Source="msm_HModA" />
<a Target="msm_HModA" />
<a SourceID="CatchID" />
<a TargetID="CatchID" />
<Commands>
<C C="ImpArea = ImpArea" />
<C C="ParAID = ParAID" />
<C C="LocalNo = LocalNo" />
<C C="ConcTime = ConcTime" />
<C C="RFactor = RFactor" />
<C C="ILoss = ILoss" />
<C C="CoeffNo = CoeffNo" />
<C C="TACoeff = TACoeff" />
<C C="TACurveID = TACurveID" />
</Commands>
</Section>
<Section>
<a Name="ModB" />
<a Source="msm_HModB" />
<a Target="msm_HModB" />
<Commands>
<C C="Slope = Slope" />
<C C="ParBID = ParBID" />

<C C="AISteep = AISteep" />

<C C="AIFlat = AIFlat" />
<C C="APSmall = APSmall" />
<C C="APMedium = APMedium" />
<C C="APLarge = APLarge" />
<C C="MISteep = MISteep" />
<C C="MIFlat = MIFlat" />
<C C="MPSmall = MPSmall" />
<C C="MPMedium = MPMedium" />
<C C="MPLarge = MPLarge" />
</Commands>
</Section>
<Section>
<a Name="ModC" />
<a Source="msm_HModC" />
<a Target="msm_HModC" />
<Commands>
<C C="C1Area = C1Area" />
<C C="C2Area = C2Area" />
<C C="ParCID = ParCID" />
<C C="RedFactor = RedFactor" />
<C C="LagTime = LagTime" />
<C C="TimeConst = TimeConst" />
</Commands>
</Section>
<Section>
<a Name="UHM" />
<a Source="msm_HModUHM" />
<a Target="msm_HModUHM" />
<a SectionMode="Append" />
<Commands>

<C C="AreaFactor = AreaFactor" />

<C C="MethodNo = MethodNo" />
<C C="Cp = Cp" />
<C C="SuhSlope = SuhSlope" />
<C C="LossModelNo = LossModelNo" />
<C C="InitLoss = InitLoss" />
<C C="ConstLoss = ConstLoss" />
<C C="RunoffCoeff = RunoffCoeff" />
<C C="CurveNum = CurveNum" />
<C C="AMC = AMC" />
<C C="InitAbstractDepth = InitAbstractDepth" />
<C C="LagTimeMethodNo = LagTimeMethodNo" />
<C C="HydraulicLength = HydraulicLength" />
<C C="SuhCt = SuhCt" />
<C C="SuhL = SuhL" />
<C C="SuhLc = SuhLc" />
<C C="BasFactor = BasFactor" />
<C C="StreamSlope = StreamSlope" />
<C C="LagCurveNum = LagCurveNum" />
</Commands>
</Section>
<Section>
<a Name="HParA" />
<a Source="msm_HParA" />
<a Target="msm_HParA" />
<Commands>
<C C="RedFactor = RedFactor" />
<C C="InitLoss = InitLoss" />
<C C="ConcTime = ConcTime" />
<C C="TAMethodNo = TAMethodNo" />
<C C="TACurveID = TACurveID" />
<C C="TACoeff = TACoeff" />
</Commands>
</Section>
<Section>
<a Name="HParB" />
<a Source="msm_HParB" />
<a Target="msm_HParB" />


<Commands>
<C C="WetSteep = WetSteep" />
<C C="WetFlat = WetFlat" />
<C C="WetSmall = WetSmall" />
<C C="WetMedium = WetMedium" />
<C C="WetLarge = WetLarge" />
<C C="StorageFlat = StorageFlat" />
<C C="StorageSmall = StorageSmall" />
<C C="StorageMedium = StorageMedium" />
<C C="StorageLarge = StorageLarge" />
<C C="InfMaxSmall = InfMaxSmall" />
<C C="InfMaxMedium = InfMaxMedium" />
<C C="InfMaxLarge = InfMaxLarge" />
<C C="InfMinSmall = InfMinSmall" />
<C C="InfMinMedium = InfMinMedium" />
<C C="InfMinLarge = InfMinLarge" />
<C C="InfExpWetSmall = InfExpWetSmall" />
<C C="InfExpWetMedium = InfExpWetMedium" />
<C C="InfExpWetLarge = InfExpWetLarge" />
<C C="InfExpDrySmall = InfExpDrySmall" />
<C C="InfExpDryMedium = InfExpDryMedium" />
<C C="InfExpDryLarge = InfExpDryLarge" />
<C C="ManningSteep = ManningSteep" />
<C C="ManningFlat = ManningFlat" />
<C C="ManningSmall = ManningSmall" />
<C C="ManningMedium = ManningMedium" />
<C C="ManningLarge = ManningLarge" />
</Commands>
</Section>
<Section>
<a Name="HParC" />
<a Source="msm_HParC" />
<a Target="msm_HParC" />
<Commands>
<C C="RFactor = RFactor" />
<C C="CTime = CTime" />
<C C="InfiltrNo = InfiltrNo" />
<C C="MaxCap = MaxCap" />

<C C="MinCap = MinCap" />

<C C="WetCond = WetCond" />
<C C="DryCond = DryCond" />
</Commands>
</Section>
<Section>
<a Name="BoundaryItems" />
<a Source="msm_BBoundary" />
<a Target="msm_BBoundary" />
<Commands>
<C C="ApplyBoundaryNo = ApplyBoundaryNo" />
<C C="GroupNo = GroupNo" />
<C C="ConnectionTypeNo = ConnectionTypeNo" />
<C C="SourceLocationNo = SourceLocationNo" />
<C C="SourceXCoor = SourceXCoor" />
<C C="SourceYCoor = SourceYCoor" />
<C C="IndividualConnectionNo = IndividualConnectionNo" />
<C C="NodeID = NodeID" />
<C C="NodeLoadTypeNo = NodeLoadTypeNo" />
<C C="LinkID = LinkID" />
<C C="ListName = ListName" />
<C C="CatchLoadNo = CatchLoadNo" />
<C C="CatchLoad = CatchLoad" />
<C C="OpenBoundaryNo = OpenBoundaryNo" />
<C C="Kmix = Kmix" />
<C C="DistributeNo = DistributeNo" />
<C C="LoadCategoryNo = LoadCategoryNo" />
<C C="GridTypeNo = GridTypeNo" />
<C C="Distance = Distance" />
</Commands>
</Section>
</Task>
</muIO>
Open the model of the entire network in Mike Urban.

Go to:
File Import/Export
Click "Next"

Select "Load Import/Export config file" and navigate to the xml-file you just
created containing the text in the box above Click "Next" and answer "No" to
the question Click "Finish" and the 2 models will be merged.

Appendix H Rainfall study
Objective of the rainfall study
Objective of the study is to develop Intensity-Duration-Frequency curves and

hyetographs for Colombo. The design rainfalls will be used for the design of
Colombo's storm water system.
Data
• 15 min rainfall data of Colombo for the period from 1981 to 2010 for all the
days with daily total rainfall more than 10 mm. The 15 min data have been
read from the daily charts of a syphon type automatic rainfall recorder
installed at the premises of the Meteorological Office in Colombo
approximately at latitude 6˚54ˊ17˝N and longitude 79˚52ˊ19˝E.
• Daily data collected over the same period in a standard rain gauge installed at
the same location in Colombo.
Catchment description
The Greater Colombo Water Basin comprises some 100 km2 of land around
Colombo. The basin is bordered by the Kelani Ganga in the North, the Weras
Ganga in the South-East and the sea (Indian Ocean) in the West. The layout of the
basin including the Parliament Lake and surrounding storage areas is shown in
Figure 1. The basin slopes from south-east to north-west, with elevations of 10-20
m MSL in the upper reaches to about 1 m MSL downstream. Figure 1 shows
rainfall gauging stations at Colombo, Ratmalana (at latitude 6.82N and longitude
79.88E) and Katunayake (at latitude 7.17N and longitude 79.88E). Only daily data
were available at Ratmalana and Katnayake stations.

Figure 1. Project area
Climate
The climate of Sri Lanka is classified as tropical monsoon, having a wet season and
a short dry season. The study area is located in the South-Western quarter of the
Island, which is classified as a wet zone. The annual climate is characterized by the
Northeast Monsoon from December through February and the Southwest
Monsoon from May through September. The dry season is brought about by the
Northeast Monsoon and the wet season by the Southwest Monsoon. Heavy
rainstorms occur mainly in the Southwest Monsoon period from May to

September and occasionally in the two Inter-monsoon periods (March-April and

October-November) as well.
Rainfall
Monthly average rainfall distribution obtained based on daily rainfall data in

Colombo for 30 years from 1981 to 2010 is shown in Figure 1. Rainfalls during the
months, May, October and November are observed to be high while January and
February seem to be very dry months.
The project area is highly urbanized and consists of a considerable amount of built
up areas in towns where runoff is high. The groundwater table is also high in the
area because of the coastal and lake environments hence the soils are less water
penetrative.
400
350
300
Rainfall (mm)
250
200
150
100
50
0
Jan Feb Mar April May June July Aug Sept Oct Nov Dec
Figure 1. Monthly average rainfall in Colombo
Temperature
Average annual temperature in the project area is in the range of 25.5-27.5°C. The
variation of monthly average temperature is shown in Table 1. These data are
based on the meteorological station in Colombo.
Table 1. Monthly Average Temperature

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Temperature
26.6 26.9 27.7 28.2 28.3 28.0 27.6 27.5 27.5 27.0 26.7 26.6
(°C)
Relative Humidity
Relative humidity data are available for Ratmalana meteorological station, which
is close to the project area. Location of the Ratmalana meteorological station is
shown in Figure 1. Relative humidity is from 66% to 75% during the day-time, and
it varies from 84% to 90% during the night-time. Details are presented in Table 3.

REPORT ON THE DEVELOPMENT AND APPLICATION
N OF THE URBAN DRAINAGE
DRAIN MODEL 103
Table 3. Relative humidity data of Ratmalana meteorological station

Month Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec
Relative Humidity-
67 68 66 71 75 75 75 74 74 75 73 75
Day (%)
Relative Humidity-
87 88 89 89 86 84 85 84 85 87 90 88
Night (%)
Wind speed
Average maximum wind speed in the study area is in the range of 8-138 km/h
according to the wind data recorded at the Colombo meteorological station
during the period from 1993 to 2008. Figure 2 shows the average maximum wind
speed on monthly basis.
Figure 2. Average maximum monthly wind speed in Colombo
Drainage and Flood Management
Surface Drainage Pattern

The area of the Greater Colombo catchment is around 85 km2. It is located on the
left bank of the lower valley of the Kelani river. Main inflow into the area is
through rainfall and local runoff that reaches the area via various drainage lines,
the lowest parts of which are canals. The Colombo drainage system consists
con of a
network of canals and marshes functioning as retention basins. The Colombo
canal drainage network has six outlets, three discharging into the sea
(Wellawatta, Dehiwala, Mutwal tunnel), one is closed at present (Beira lake) and
the other two into the Kelani river.
Development of Intensity-Duration-Frequency
Intensity Frequency Relationships
Daily total rainfalls obtained from the gauge recorded 15 minute rainfalls were
compared with the daily rainfall data read in a check gauge (standard rain gauge
observations made
made at 0830 h). Records which deviated by more than 30% were
considered unreliable and eliminated from the study. In this process annual
maximum records of five years (1983, 1987, 1989, 2000 and 2010) had to be
discarded. Those were not included in the annual
annual exceedance series, too. Two
more rainfall records in the annual exceedance series (in 1985 and 2005) were
C:\Documents and Settings\NEJO\My Documents\Projects\Colombo

Colombo\Reports\A029106-1.0 ver.2.1b.docx
also discarded as they were unreliable based on the above criteria. Out of the
records that showed deviations, the percentages of data that deviate upto 10%,
10% to 20%, 20% to 30% and more than 30% were found to be, 81%, 7%, 4% and
8%. This shows that the number of rainfall records that have shown considerable
deviations in the total daily rainfalls is very small. Also, in general rainfall records
upto 20% deviation were found to be not affecting the annual maximum rainfalls.
Rainfall for 15, 30, 60, 90, 120, 180, and 240 min were compiled from the 15 min
rainfall data. Annual maximum and annual exceedance data series were extracted
from the validated records. Annual exceedance series refer to Peak Over
Threshold rainfalls by retaining the N largest values, where N is equal to the
number of years in the series. Table 4 provides the annual maximum rainfalls for
the years from 1981 to 2010.
Table 4. Annual maximum rainfall series for different durations in mm (1981-

2010)
Year 15 min 30 min 60 min 90 min 120 min 180 min 240 min
1981 38.82 56.07 94.35 94.35 95.43 96.50 111.06
1982 26.06 49.52 77.15 83.14 84.71 88.10 92.01
1983 21.74 40.96 54.36 56.38 56.38 57.14 57.14
1984 37.19 67.34 96.49 98.00 98.00 98.00 98.00
1985 49.89 65.35 80.07 81.81 83.56 94.97 109.67
1986 24.99 35.40 54.72 65.23 78.75 79.77 79.77
1987 19.70 33.37 53.70 56.85 59.02 62.75 75.25
1988 39.47 64.00 86.93 110.06 121.28 131.97 140.25
1989 22.12 40.97 54.74 62.85 65.64 69.44 72.23
1990 27.31 44.98 59.00 62.25 73.46 75.10 75.10
1991 22.62 36.61 46.94 48.98 62.13 76.85 77.10
1992 40.99 63.10 92.13 122.42 150.21 222.06 262.62
1993 29.04 47.71 65.91 76.65 111.92 133.51 152.60
1994 33.09 53.49 73.34 75.55 77.20 77.20 77.20
1995 28.75 43.67 62.04 69.25 70.50 70.50 70.50
1996 33.29 59.71 81.72 82.99 89.10 96.17 99.87
1997 27.66 43.36 77.06 87.96 103.82 116.70 117.20
1998 43.60 69.96 96.06 98.09 98.09 98.35 98.35
1999 50.81 62.46 82.42 99.74 105.42 125.13 129.01
2000 41.13 66.83 105.39 110.53 120.81 133.35 140.41
2001 33.03 56.44 69.50 69.50 69.50 69.50 71.79
2002 29.79 40.46 57.50 60.57 63.06 63.80 63.80
2003 25.00 45.00 46.50 60.67 67.24 72.04 74.82
2004 27.68 37.08 70.73 77.81 81.09 94.98 96.50
2005 44.59 54.16 71.17 86.34 94.51 108.80 109.09
2006 30.00 60.00 80.00 97.50 110.50 116.25 118.25
2007 35.96 56.22 99.77 111.16 119.77 128.10 132.38
2008 35.32 50.45 79.71 100.90 100.90 100.90 100.90
2009 26.19 46.34 70.51 96.70 116.34 148.58 197.93
2010 44.95 87.78 132.19 155.46 159.16 163.39 167.88

Then the above data were tested for suitability for using in a frequency analysis.
The data screening procedure adopted consisted of following steps.
• Plot the totals according to the chosen time step and note any trends or
discontinuities
• Test the time series for absence of trend with Spearman’s rank-
correlation method
• Apply the F-test for stability of variance and the t-test for stability of mean
to split, non-overlapping, sub-sets of the time series
• Test the series for absence of persistence by computing the first serial
correlation coefficient
Data were plotted for visual examination in order to identify any abrupt changes
in the time series. Annual maximum rainfall series for different durations are
presented in Figure 3. No appreciable trend is noticed from the plot.
300 15 min
30 min
Annual maximum rainfall (mm)
250
60 min
200
90 min
150
120
100 min
50
0
1980 1985 1990 1995 2000 2005 2010
Figure 3. Annual maximum rainfall from 1981 to 2010
Testing of high and low outliers was done using the equation 01 02 34 56 and
07 02 34 56 where yH and yL are high and low outlier thresholds in log and 02 is
the mean, n is the sample size, sy is the standard deviation and Kn is the
parameter given in Chow et al (1988).
The thresholds calculated for the period 1981 to 2010 are shown in Table 5.
Table 5. High and low thresholds for identifying outliers in mm

15 min 30 min 60 min 90 min 120 min 180 min 240 mi
High outlier threshold 61.96 93.56 139.97 163.17 180.16 219.88 251.42
Low outlier threshold 16.51 28.06 38.54 41.64 44.57 42.98 41.52
All the data series except two values in the 180 min and 240 min series were
found to be within these limits. For the 180 min and 240 min durations, the
rainfalls were found to be 222.06 mm and 262.62 mm respectively. Since these
are very marginally more than the threshold level and the comparison of rainfall
records at a gauging station located close to Colombo at Ratmalana, where daily

rainfall totals are known, indicated similar concurrent high rainfalls it was decided
to retain the observed rainfalls in the study.
Test for Absence of Trend
It is important to make sure that there is no correlation with the order in which
the data have been collected and with an increase or a decrease in the magnitude
of those data. For that Spearman’s rank correlation method was used to verify the
absence of trend at a significance level of 5%. The time series has no trend if
t{v,2.5%} <t1<t{v.97.5%}, where, t1 is the test statistic, v is the degrees of freedom
(Dahmen and Hall, 1990). The results of the calculations are given in Table 6.
Table 6. Rainfall data check for trend

Trend Check
Duration t{v,2.5%} <t1<t{v.97.5%}
t1 t{v,2.5%} t{v.97.5%}
15 min 1.391 -2.04 2.04 ok
30 min 0.853 -2.04 2.04 ok
60 min 0.887 -2.04 2.04 ok
90 min 2.014 -2.04 2.04 ok
120 min 2.052 -2.04 2.04 ok
180 min 2.238 -2.04 2.04 Not ok
240 min 1.646 -2.04 2.04 ok
All the data series except 180 min series have no trend and are thus stationary.
Though the 180 min data set does not satisfy the condition indicating that the
time series not stationary it was decided to check the data set further to study its
performances in other tests.
Test for Stability of Variance and Mean
The test for stability of variance is done first since instability of the variance
implies that the time series is not stationary and thus not suitable for further use.
The test statistic is the ratio of the variances of two split, non-overlapping, subsets
of the time series (Dahmen and Hall, 1990). The region for test statistic of Ft was
taken as F{V1,V2,2.5%} <Ft <F{V1,V2,97.5%} where, v1=n1-1 (number of degrees of freedom
for the numerator), v2=n2-1 (number of degrees of freedom for the denominator), n1, n2
equals the number of data in each sub set.
Then the t-test for stability of the mean was conducted using the same two subsets. The
test statistic was taken to be bounded as, t {v,2.5%} < tt < t{v,97.5%} where, v=(n1-1)+(n2-
1).
The checks carried out resulted in the results given in Table 7.

Table 7. Rainfall data check for stability of variance and mean

Stability check
Variance Mean
Duration
F {V1,V2,2.5%} <Ft <F{V1,V2,97.5%} t {v,2.5%} < tt < t{v,97.5%}
Ft F{V1,V2,2.5%} F{V1,V2,97.5%} tt t{v,2.5%} t{v,97.5%}
15 min 1.256 0.336 2.98 Ok -1.249 -2.04 2.04 ok
30 min 1.322 0.336 2.98 Ok -1.006 -2.04 2.04 ok
60 min 1.134 0.336 2.98 Ok -1.277 -2.04 2.04 ok
90 min 1.603 0.336 2.98 Ok -1.593 -2.04 2.04 ok
120 min 1.792 0.336 2.98 Ok -1.179 -2.04 2.04 ok
180 min 2.722 0.336 2.98 Ok -0.767 -2.04 2.04 ok
240 min 2.302 0.336 2.98 Ok -0.463 -2.04 2.04 ok
All the data series satisfy the checks and the means and variances are stable.
Therefore, the time series are stationary and could be used in the study.
Test for Absence of Persistence
The serial correlation coefficient verifies the independence of a time series which
in turn helps to ensure that the time series is completely random (Dahmen and Hall,
1990). It is sufficient to compute the first lag serial correlation coefficient for this
purpose and a confidence level of 95% was used in this study. The lag one serial
correlation coefficient, r1 should be within region at 5% level of significance as;
LCL(r) < r1 < UCL(r), where, UCL(r) = {-1,(-1-1.96(n-2)0.5)/(n-1)} and LCL(r) = {(-
1+1.96(n-2)0.5)/(n-1),+1 }
Table 8 presents the calculations and the results.
Table 8. Rainfall data check for absence of persistence

Absence of persistence
Duration LCL(r) < r1< UCL(r)
r1 LCL(r) UCL(r)
15 min 0.013 -0.392 0.823 ok
30 min -0.039 -0.392 0.323 ok
60 min 0.052 -0.392 0.323 ok
90 min 0.129 -0.392 0.323 ok
120 min 0.179 -0.392 0.323 ok
180 min 0.206 -0.392 0.323 ok
240 min 0.210 -0.392 0.323 ok
As Table 8 indicates there is no correlation between successive records in the data

series. The data are independent and there is no persistence in all the time series.
Thus it was decided to use the data series in the study as all the tests were
satisfied satisfactorily.
The Ranked Annual Exceedance Series extracted from the data are given in Table
9.

Table 9. Ranked Annual exceedance rainfall series for different durations in mm

(30 years)
15 min 30 min 60 min 90 min 120 min 180 min 240 min
1 50.81 87.78 132.19 155.46 159.16 222.06 262.62
2 49.89 69.96 105.39 122.42 150.21 163.39 197.93
3 44.95 67.34 99.77 111.16 123.44 148.58 167.88
4 44.59 66.83 96.49 110.69 121.28 133.51 152.60
5 43.60 65.35 96.06 110.53 120.81 133.35 140.41
6 41.13 64.00 94.35 110.06 119.77 131.97 140.25
7 40.99 63.10 92.13 100.9 116.34 128.10 139.26
8 39.47 62.46 86.93 99.74 111.92 125.13 137.40
9 38.82 60.00 83.05 98.09 110.5 123.70 132.38
10 37.81 59.71 82.42 98.00 105.42 116.70 129.01
11 37.19 56.44 81.72 97.60 103.82 116.25 123.70
12 35.96 56.22 80.07 97.50 101.07 110.42 118.25
13 35.32 56.07 80.00 96.70 100.9 108.80 117.20
14 33.29 55.04 79.71 94.35 98.09 101.60 111.06
15 33.09 54.99 77.28 90.25 98.00 100.90 109.67
16 33.03 54.16 77.15 88.44 95.43 98.35 109.09
17 31.50 53.75 77.06 87.96 94.51 98.00 101.60
18 31.06 53.49 76.01 86.34 90.25 96.50 100.90
19 30.99 53.08 73.34 83.90 89.52 96.17 99.87
20 30.12 52.50 72.66 83.14 89.10 94.98 99.67
21 30.04 51.82 71.33 82.99 89.04 94.97 98.35
22 30.00 50.45 71.17 81.81 88.13 92.95 98.00
23 29.86 49.52 70.73 79.26 86.66 92.49 97.63
24 29.79 49.27 70.51 77.81 84.71 92.25 96.50
25 29.35 47.71 70.51 76.65 83.56 92.05 94.70
26 29.19 46.93 69.50 76.01 82.67 91.14 93.05
27 29.04 46.34 65.91 75.55 81.33 90.63 93.00
28 28.87 45.00 64.95 73.66 81.09 90.51 92.22
29 28.85 45.00 64.75 73.58 78.75 88.10 92.01
30 28.75 44.98 63.88 73.54 77.79 85.80 91.04
For the more frequent events of return periods upto 10 years (2, 5 and 10 years)
annual exceedence series was used while for the return periods of 25 years and
above (25, 50 and 100 years) the annual maximum series was used in the
frequency analysis. Use of annual exceedance series for return periods upto 10
years and annual maximum series for higher return periods has been the practice.
The GEV distribution was observed to be the best fitting distribution for both
annual exceedence series and annual maximum series. Please note that many
distributions were tested and GEV was found to be the most suitable distribution.
The tested distributions were GEV, Weibull, Frechet, Gumbel and Figure 4 shows

these distributions with the observed data. Software “Easyfit” was used in the
fitting of probability distributions.
70
60
Annual Maximum Rainfall (mm)
50
40
30 Data
GEV
20 Frechet
Gumbel
10 Weibull
0
-2 -1 0 1 2 3 4 5
Gumbel Reduced Variate
Figure 4. Comparison of a few distributions
Then from the best fitted distribution (GEV distribution) rainfall intensities for
different return periods were calculated for all the rainfall durations. Then
corrections for the conversion of fixed/restricted 15 min data to
sliding/unrestricted data were introduced according to the guidelines developed
by Dwyer and Reed (1995). The correction factors used were 1.160, 1.085, 1.026,
1.008 and 1.003 for durations of 15, 30 60, 90 and 120 min rainfall data as
suggested in Dwyer and Reed (1995).
The different duration rainfalls for different return periods after the correction
introduced are given in Table 10. Figure 5 and Figure 6 present the relationships
of rainfall depth with rainfall duration and rainfall depth with return period.
Table 10. Different duration rainfall depths for different return periods (mm)
Duration (min) 2 yr 5 yr 10 yr 25 yr 50 yr 100 yr
15 38.84 45.54 50.78 58.92 63.68 68.17
30 59.10 67.97 74.18 84.01 89.67 94.84
60 79.20 91.83 101.65 117.35 126.15 134.31
90 90.01 105.39 116.57 133.05 147.65 160.62
120 96.48 113.95 127.20 147.18 161.18 175.59
180 102.99 123.51 142.16 179.05 203.90 230.31
240 109.43 135.18 159.04 205.31 243.43 287.15

350
2 yr
300 5 yr
10 yr
250 25 yr
50 yr
Rainfall (mm)
200 100 yr
150
100
50
0
0 30 60 90 120 150 180 210 240
Rainfall duration (min)
Figure 5. Rainfall against rainfall duration
300.00
15 min
30 min
250.00
60 min
90 min
200.00
120 min
Rainfall (mm)
180 min
150.00
240 min
100.00
50.00
0.00
1 10 100
Return Period (yr)
Figure 6. Rainfall against return period
These rainfalls were compared with the rainfall depths available in two previous
studies as presented in Table 11. One is the study by Dharmasena and Premasiri
(1990), who used the same raw data set, but for the period limited from 1960 to
1985. In their study Extreme Value Type I (Gumbel) Distribution (EVI) was fitted to
annual maximum rainfall series. The other is by JICA (2003), which used results of
a study reportedly carried out by H.Humphery. H. Humpheys listed all the high
intensity storm rainfall data for 15, 30, 45, 60, 75, and 90 minutes from 1921
through 1967. As the report mentions these data have been read directly from
the original daily charts of the automatic rainfall recorder at the Colombo
observatory. Based on these data, probable rainfalls have been estimated for the

given duration and frequencies. The rainfalls were observed to be a little increased
compared to the previous studies, but were very close to them.
Table 11. Comparison of different duration rainfalls derived in this study with
those values in two previous studies
Return Period
Study
Rainfall Duration 2 yr 5 yr 10 yr 25 yr 50 yr 100 yr
15 min Present 38.8 45.5 50.8 58.9 63.6 68.2
D&P 31.8 36.8 39.9 45.0 48.5 51.8
JICA 31.0 36.4 40.0 na na na
30 min Present 59.1 68.0 74.2 84.0 89.7 94.8
D&P 51.3 64.0 71.4 80.0 86.9 94.0
JICA 51.0 62.7 70.5 na na na
60 min Present 79.2 91.8 101.7 117.4 126.2 134.3
D&P 72.9 91.9 102.4 116.3 127.5 138.4
JICA 72.6 90.4 102.1 na na na
90 min Present 90.0 105.4 116.6 133.0 147.7 160.6
D&P 76.4 95.0 105.2 121.2 132.4 143.9
JICA na na na na na na
120 min Present 96.5 114.0 127.2 147.2 161.2 175.6
D&P 79.0 96.0 108.0 124.0 135.9 148.1
180 min Present 102.9 123.5 142.2 179.1 203.9 230.3
D&P 93.5 118.1 132.1 153.5 168.1 183.7
JICA 79.0 100.3 113.8 na na na
240 min Present 109.4 135.2 159.0 205.3 243.4 287.2
D&P 100.5 127.4 143.2 166.9 183 200.3
Rainfall intensities calculated from the rainfall depths are given in Table 12.
Table 12. Rainfall intensity at different return periods (sliding data) in mm/h
From Annual Exceedance From Annual Maximum
Series Series
Duration
2 yr 5 yr 10 yr 25 yr 50 yr 100 yr
(min)
15 155.35 182.15 203.11 235.68 254.73 272.67
30 118.21 135.94 148.35 168.01 179.34 189.67
60 79.20 91.83 101.65 117.35 126.15 134.31
90 60.01 70.26 77.71 88.70 98.43 107.08
120 48.24 56.98 63.60 73.59 80.59 87.79
180 34.33 41.17 47.39 59.68 67.97 76.77
240 27.36 33.80 39.76 51.33 60.86 71.79
These rainfall intensities were compared with the rainfall intensities estimated in
Ranatunga (2000) for Colombo as given in Table 13.

Table 13. Rainfall intensity at different return periods from the study by
Ranatunga (2000) in mm/h
Return period
15 127.82 150.01 162.81 182.15 198.02 212.00
30 96.33 115.85 126.48 143.14 155.83 167.81
60 65.97 81.19 89.36 102.32 111.60 120.93
90 50.91 63.36 70.13 80.79 88.24 95.94
120 41.79 52.36 58.21 67.31 73.60 80.20
180 31.18 39.36 44.05 51.18 56.04 61.24
240 25.12 31.85 35.81 41.73 45.75 50.08
The differences between the rainfall intensities derived in the present study with
the rainfall intensities estimated by Ranatunga (2000) are shown in Figure 7. The
differences are given as the percentage difference with respect to the previous
study.
50
2 yr
45
5 yr
Difference relative to previous
40
10 yr
35
estimation (%)
25 yr
30
50 yr
25
100 yr
20
15
10
5
0
15 30 60 90 120 180 240
Rainfall duration (min)
Figure 7. Difference of rainfall intensities with estimations by Ranatunga (2000)
The IDF curves for the data given in Table 12 are drawn and are presented in
Figure 8.

1000.00
2

10
25
100.00
10.00
10 100 1000
Figure 8. IDF Curves
For these rainfall intensities best fitting equations were fitted using “shifted
power equation”. The equation fitted for the different return periods is in the
following format.
8 9 :: ; 5 0 9 :: ;5
<
Where, a, b and c are constants. Table 14 gives a, b, c values for different return
periods. Rainfall duration is in minutes and the rainfall intensities are in mm/h.
Table 14. Constants in Power equation for rainfall intensity

Constant 2 yr 5 yr 10 yr 25 yr 50 yr 100 yr
a 6844.65 4383.95 2832.77 1579.57 1124.45 831.49
b -31.217 -22.872 -15.388 -5.818 -0.827 3.486
c -0.987 -0.875 -0.772 -0.626 -0.537 -0.455
Rainfall intensities were determined from the above equation and they are
presented in Table 15.
Figure 9 shows the final IDF curves developed using these results.
Table 15. Different duration rainfall intensities (mm/h) for different return periods
from the fitted shifted power equation
15 155.49 182.14 202.96 235.86 255.11 273.32
30 117.81 136.01 148.90 167.89 178.33 186.96
60 79.46 91.77 100.63 114.69 123.79 132.47
120 48.24 56.97 64.04 76.43 85.62 95.29

180 34.69 41.92 48.24 59.86 68.95 78.87

240 27.10 33.41 39.23 50.24 59.12 69.03
300 22.25 27.91 33.33 43.81 52.46 62.28
360 18.87 24.04 29.14 39.16 47.58 57.27
720 9.91 13.46 17.34 25.50 32.81 41.67
1080 6.73 9.52 12.75 19.81 26.39 34.62
1440 5.10 7.44 10.23 16.56 22.62 30.36
1000.00
2 yr
5 yr
10 yr
25 yr
50 yr
100.00
100 yr
10.00
10 100 1000
Figure 9. IDF Curve developed based on fitted power equation
Derivation of Design Rainfall
The available 15 min rainfall data record was used to identify the time distribution
of observed maximum rainfalls. Initially, largest three rainfalls were extracted
from the data record each year. Thus, totally 90 rainfalls were extracted and
analyzed in the study to identify hyetograph shapes.
The selected rainfall events were categorized into 5 levels with respect to rainfall
durations. That is, less than 1 h, 1h to 2h, 2h to 3h, 3h to 4h and more than 4h.
The numbers of rainfall events in each category are shown in Table 16.
Table 16. Number of rainfall event in each category considered in the study
Rainfall duration Number of events
<1h 9
1h – 2h 35
2h – 3h 23
3h – 4h 6
> 4h 17
Total 90

Plots of % cumulative rainfall against % time for the above different duration
rainfalls were drawn to investigate the shape of hydrographs and they are
presented in Figure 10. Approximate average lines are also drawn on the same
plots.
1 1
0.9 0 h To 1 h 0.9 1 h To 2 h
0.8 0.8
% Cumulative rainfall
% Cumultive rainfall
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3 0.3
0.2
0.2
0.1
0.1
0
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 % Time
% Time
1 1
0.9 2 h to 3 h 0.9 3 h To 4 h
0.8 0.8
0.7 0.7
0.6 0.6
0.5 0.5
0.4 0.4
0.3
0.3
0.2
0.2
0.1
0.1
0
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
% Time 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
% Time
1
0.9 Over 4 h
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
% Time
Figure 10. Plots of % cumulative rainfalls against % time for five different
durations

The average lines are drawn on one plot and presented in Figure 11. However,
please note that the average lines drawn are very subjective.
1
0h - 1h
0.9
1h - 2h
0.8 2h - 3h
0.7 3h - 4h
4h <
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1
% Time
Figure 11. Rainfall distribution for different duration rainfalls
Shapes of the Design Rainfall hyetographs for these different cases are shown in
the following Figure 12. Rainfall durations are given in the hyetographs. One unit
in the horizontal axis is one tenth of the rainfall duration. Vertical axis shows the
distribution of a unit rainfall.
0.4 0.4
0.35
< 1h 0.35
1h - 2h
0.3 0.3
0.25 0.25
0.2 0.2
0.15 0.15
0.1 0.1
0.05 0.05
0 0
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8

0.4 0.4
0.35
2h - 3h 0.35
3h - 4h
0.3 0.3
0.25 0.25
0.2 0.2
0.15 0.15
0.1 0.1
0.05 0.05
0 0
1 2 3 4 5 6 7 8 1 2 3 4 5 6 7 8
0.4
0.35
> 4h
0.3
0.25
0.2
0.15
0.1
0.05
0
1 2 3 4 5 6 7 8
Figure 12. Design rainfall hyetograph shapes for different duration rainfalls
References
Chow, V.T., Maidment, D.R. Mays, L.W. (1988) Applied Hydrology, McGraw Hill
Book Company, Singapore.
Dahman, E.R., Hall, M.J. (1990) Screening of Hydrological Data: Tests for
Stationarity and Relative Consistency, IILRI, The Nethrlands
Dharmasena, G.T., Premasiri, S.M. (1990) Rainfall Intensity Studies for Sri Lanka,
Engineer: Journal of the IESL, December, pp.38-52.
Dwyer,I.J., Reed, D.W. (1995) Allowance for discretization in hydrological and

environmental risk estimation (ADHERE), Report No.13, Institute of Hydrology,
Wallingford, UK.
Nippon Koei (2003) The Study on Storm Water Drainage Plan for the Colombo
Metropolitan Region in the Democratic Socialist Republic of Sri Lanka. Volume I:
Executive Summary. Volume II: Main Report. Volume III: Supporting Report 1:
Master Plan. Volume IV: Feasibility Study and Volume V: Data Book.
Ognik, H.M.J., (2012) Note on short duration rainfall statistics of Colombo, Period
1981-2010, Report prepared for World Bank

Ranatunga, D.G.L. (2000) Towards more efficient hydraulic and hydrological

design of cross drainage structures using new developed intensity duration
frequency equations, Hydrology Division, Irrigation Department.

COWI Report-Micro

Uploaded by

Copyright:

Available Formats

COWI Report-Micro

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

COWI Report-Micro

Uploaded by

Copyright:

Available Formats

MARCH 2013

WORLD BANK GROUP

Report on the Development

Report on the Development

PROJECT NO. P-29106

4 Time plan and resources 12

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

9.2 Evaluation of O&M Policies 43

Appendix B Importing cross-sectional data from

Appendix C LiDAR survey 63

Appendix D Curve Numbers 84

Appendix E Short cut to compute hydraulic length of

Appendix F Changes to DHIAPP.INI 88

Appendix G Merging of 2 models 89

Appendix H Rainfall study 100

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

Abbreviations and acronyms

CMC Colombo Municipality Council

DLMC Dehiwala Mt. Lavinia Municipal Council

DTM Digital Terrain Model

GIS Geographical Information System

HSG Hydrological Soil Groups

IDF Intensity Duration Frequency

KAMC Kaduwela Municipal Council

KOMC Kote Municipal Council

KPS Kotikawatta Pradesiya Saba

KUC Kollonawa Urban Council

LIDAR Light Detection And Ranging

LKR Sri Lanka Rupee

PPS25 Planning Policy Statement 25

UDA Urban Development Authority

UHM Unit Hydrograph Method

SLLRDC Sri Lanka Land and Reclamation Development Corporation

SWM Solid Waste Management

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

The report describes the following main activities:

1. Development, calibration and validation of the urban drainage model.

2. The LIDAR survey.

3. The application of the model.

4. The required production environment.

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

The outlined set of objectives is to use the model for:

1 Propose, test and validate design options.

2 Assess and predict flood risk.

3 Master planning and assessment of urban drainage performance.

4 Investigate design policies and standards.

5 Test operation and maintenance initiatives

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

Figure 3-1 Overview of Greater Colombo Basin

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx

Figure 3-2 Overview of Greater Colombo Basin.

The authorities are:

› Colombo Municipal Council (CMC)

› Dehiwala Mt. Lavinia Municipal Council (DLMC)

› Kote Municipal Council (KOMC)

› Kollonawa Urban Council (KUC)

› Kotikawatta Pradesiya Saba (KPS)

› Kaduwela Municipal Council (KAMC)

C:\Documents and Settings\NEJO\My Documents\Projects\Colombo\Reports\A029106-1.0 ver.2.1b.docx