RANDWICK CITY COUNCIL

Coogee Bay Flood Study

Final Report | April 2013
 Coogee Bay
Flood Study
Final Report

Offices
Brisbane
Denver
Karratha
Melbourne
Prepared For: Randwick City Council Morwell
Newcastle
Perth
Prepared By: BMT WBM Pty Ltd (Member of the BMT group of companies) Sydney
Vancouver

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
 DOCUMENT CONTROL SHEET

BMT WBM Pty Ltd

BMT WBM Pty Ltd
126 Belford Street
Document : R.N1924.001.03.docx
BROADMEADOW NSW 2292 Project Manager : Darren Lyons
Australia
PO Box 266
Broadmeadow NSW 2292

Tel: +61 2 4940 8882

Fax: +61 2 4940 8887 Client : Randwick City Council
ABN 54 010 830 421 003
Client Contact: Mr. T. Kefalianos
www.wbmpl.com.au

Client Reference PROJ/10515/2009

Title : Coogee Bay Flood Study – Final Report

Author : Daniel Williams
Synopsis : Report for the Coogee Bay Flood Study covering the development and calibration of
computer models, establishment of design flood behaviour and flood mapping.

REVISION/CHECKING HISTORY

REVISION DATE OF ISSUE CHECKED BY ISSUED BY

NUMBER
0 10/05/11 DJL DJL
1 20/02/12 DJL DXW
2 10/01/13 DJL DXW
3 11/04/13 DJL DXW

DISTRIBUTION

DESTINATION REVISION
0 1 2 3
Randwick City Council 2p,1e 2p, 1e 1e 3p, 1e
BMT WBM File
BMT WBM Library 1 1 1 1

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
EXECUTIVE SUMMARY I

EXECUTIVE SUMMARY
Introduction

The Coogee Bay Flood Study has been prepared for Randwick City Council (Council) to define the
existing flood behaviour in the Coogee Bay catchment and establish the basis for subsequent
floodplain management activities.

The primary objective of the Flood Study is to define the flood behaviour of the local Coogee Bay
catchments through the establishment of appropriate numerical models. The study has produced
information on flood flows, velocities, levels and extents for a range of flood event magnitudes under
existing catchment and floodplain conditions. Specifically, the study incorporates:

 Compilation and review of existing information pertinent to the study and acquisition of additional
data including survey as required;

 Undertaking a community consultation and participation program to identify local flooding

concerns, collect information on historical flood behaviour and engage the community in the on-
going floodplain management process;

 Development and calibration of appropriate hydrologic and hydraulic models;

 Determination of design flood conditions for a range of design events including the 20% AEP, 5%
AEP, 1% AEP, 0.5% AEP, 0.2% AEP and extreme flood event; and

 Presentation of study methodology, results and findings in a comprehensive report incorporating

appropriate flood mapping.

Catchment Description

2
The study area catchments occupy a total area of 2.9km and incorporate the majority of Coogee and
parts of South Coogee and Randwick. The catchments drain to the east into Coogee Bay.

The natural creek systems have been heavily modified and the study area is now drained entirely by
a stormwater pipe network. When the capacity of this network is exceeded, overland flow will occur
along the alignments of the original creeks. Many of the old creek alignments are now located
through developed properties, which presents a significant flood risk.

Land use within the study area primarily consists of urban development (90%), open recreational
space (9%) and tree-covered land (1%). The urban development within the study area includes low,
medium and high density residential development and commercial uses, including the Coogee CBD
along the beachfront. Some of the developed areas would previously have been creek alignments.

Historical Flooding

There is a long history of flooding in Coogee, as it is an old suburb with development located on
natural creek lines. Floods reported in available newspaper articles include 1959, 1989 and 1998.
The October 1959 event is the largest recorded within the catchment, with a daily total of 265mm of

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
EXECUTIVE SUMMARY II

rainfall recorded at Randwick Bowling Club, however, the duration of the event is understood to have
been of the order of three and a half hours.

The most significant recent floods include the January 1999 and May 2009 events. The former is well
documented by both the Coogee Oval and Bowling Club Flooding Assessment (PBP, 1999) and
Assessment of Impacts from January 1999 Flooding (GBA, 1999). 74mm of rain were recorded on
th
24 January at Randwick Bowling Club, most of which is believed to have fallen in a 90 minute
period. This would be equivalent to around a 10% AEP event.

The May 2009 event is one of the largest recent local flooding events within the catchment This was
a localised storm event which primarily impacted the eastern suburbs of Sydney. Randwick Bowling
rd
Club recorded 77mm of rain on 3 May, which was the highest recorded daily total by any gauge in
the local area. Most of the rain fell within a 90 minute period, which would be equivalent to around a
10% AEP event.

Community Consultation

Community consultation undertaken during the study has aimed to collect information on historical
flooding and previous flood experience, and inform the community about the development of the flood
study and its likely outcome as a precursor to floodplain management activities to follow. The key
element of the consultation process involved the distribution of a questionnaire relating to historical
flooding.

Council mailed out the questionnaire to all residents and businesses located within the study area.
Council received back almost 1000 responses, of which around 250 had comments relating to
flooding. The comments relating to flooding that were received from the community provided valuable
data for the calibration process.

Model Development

Development of hydrologic and hydraulic models has been undertaken to simulate flood conditions in
the catchment. Hydrologic and hydraulic modelling has been undertaken using the TUFLOW two-
dimensional (2D) software developed by BMT WBM and utilising a direct rainfall approach to model
the catchment hydrology. The model simulates runoff routing, hydrological response, flood depths,
extents and velocities. The 2D modelling approach is suited to model the complex interaction
between channels and floodplains and converging and diverging of flows through structures and
urban environments.

The floodplain topography is defined using a high resolution digital elevation model (DEM) derived
from LiDAR survey for greater accuracy in predicting flows and water levels and the interaction of in-
channel and floodplain areas. The stormwater drainage system was modelled, using survey details of
pipe configuration, pipe sizes and invert levels. Land use surfaces wee derived largely from Council
GIS layers, including individual building footprint polygons.

Model Calibration and Validation

The selection of suitable historical events for calibration of computer models is largely dependent on
available historical flood information. Significant flooding in Coogee has occurred on numerous
occasions, with the most severe events in recent times including 1959, 1989, 1998, 1999 and 2009.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
EXECUTIVE SUMMARY III

The May 2009 event is considered the most suitable of the historical events for model calibration. The
vast majority of the community questionnaire responses related to the May 2009 event. The
availability of rainfall data and flood photographs provides a sound dataset to assist calibration of the
model.

The January 1999 event was also selected for model calibration. It is similar in magnitude to the May
2009 event and is the next most recent significant flood event in the catchment. The October 1959
event has been selected for model validation purposes as although available data is limited, it is the
largest event recorded within the study area.

A reasonable model calibration has been achieved given the available data for the catchment. The
developed model is considered to provide a sound representation of the flooding behaviour of the
catchment, as demonstrated through comparison of recorded peak water levels and known
inundation areas for the historical events simulated.

Design Event Modelling and Output

The developed model has been applied to derive design flood conditions within the Coogee Bay
catchments. Design rainfall depth is based on the generation of intensity-frequency-duration (IFD)
design rainfall curves utilising the procedures outlined in AR&R (2001). A range of storm durations
using standard AR&R temporal patterns were modelled in order to identify the critical storm duration
for design event flooding in the catchments.

The design events considered in this study include the 20% AEP (5-year ARI), 5% AEP (20-year
ARI), 1% AEP (100-year ARI), 0.5% AEP (200-year ARI), 0.2% AEP (500-year ARI) and PMF events.
The model results for the design events considered have been presented in a detailed flood mapping
series for the catchments. The flood data presented includes design flood inundation, peak flood
depths and peak flood velocities.

Provisional flood hazard categorisation in accordance with Figure L2 of the NSW Floodplain
Development Manual (2005) has been mapped for the events, in addition to the hydraulic categories
(floodway, flood fringe and flood storage) for flood affected areas.

Sensitivity Testing

A series of sensitivity tests have been undertaken on the modelled flood behaviour of the Coogee
Bay catchments. The tests provide a basis for determining the relative sensitivity of modelling results
to adopted parameter values. The tests undertaken include:

 Structure blockages – structure blockage due to flood debris can result in significant
increases to flood levels and redistributions of flood flows. A scenario of 100% blockage to the
stormwater drainage network has been applied to identify locations for which the blockage conditions
are significant;

 Design rainfall losses – a decrease in design rainfall losses has been simulated to adopt the
standard 15mm initial loss recommended by AR&R. This provides for an increase in effective rainfall
and therefore in increase in surface runoff for the design rainfall condition;

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
EXECUTIVE SUMMARY IV

 Increased sea-level – the downstream boundary condition in Coogee Bay was raised to an
extreme level, approximating a 0.5% AEP ocean flooding condition with a 0.9m climate change sea-
level rise allowance; and

 Increased rainfall intensities – the 0.5% AEP and 0.2% AEP event results were compared
with the 1% AEP results to assess the impact of a 10% and 25% increase to the 1% AEP design
rainfall intensities. This is similar to the recommended approach for considering increased rainfall
intensity as a potential impact of climate change.

Conclusions

The objective of the study was to undertake a detailed flood study of the local overland flow
catchments of Coogee Bay and establish models as necessary for design flood level prediction. In
simulating the design flood conditions for the local catchments in the study area, the following
locations were identified as potential problem areas in relation to flood inundation extent and property
affected:

 Alfreda Street and Coogee Oval;

 Brook Street;

 Coogee Bowling Club and Tennis Club;

 Coogee Street and Dolphin Street;

 Clyde Street;

 Oswald Street;

 Abbott Street;

 Bardon Park and Smithfield Avenue;

 Carr Street;

 Oberon Street; and

 Rainbow Street.

The flooding issues within the Coogee Bay study area are largely restricted to locations which were
naturally creek/gully lines, but are now occupied by urban development. Along these alignments
natural depressions in the topography and those created by man-made obstructions, such as roads
and other land-raising activities, fill to significant depths during major design flood events. This type of
flood behaviour is widespread throughout the study area.

Most of the study area drains to two large depressions – Coogee Oval and Rainbow Street. At
Coogee Oval the higher ground of Arden Street and Goldstein Reserve is situated some 2m above
the bottom of the Oval. During major flood events the available storage of the Oval will be exceeded
and flood waters will spill across Arden Street and the reserve to the beach. The Rainbow Street
depression is some 10m deep and as such the storage capacity will never be exceeded. In extreme
flood conditions such as the PMF event or under a blocked stormwater drainage scenario, a
significant flood risk to this area is posed, with possible flood depths of several metres.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
CONTENTS V

CONTENTS

Executive Summary i
Contents v
List of Figures vii
List of Tables viii

GLOSSARY I

1 INTRODUCTION 1
1.1 Study Location 1
1.2 Study Background 1
1.3 The Need for Floodplain Management at Coogee Bay 1
1.4 The Floodplain Management Process 3
1.5 Study Objectives 3
1.6 About This Report 4

2 STUDY APPROACH 5
2.1 The Study Area 5
2.1.1 Catchment Description 5
2.1.2 History of Flooding 5
2.1.3 Previous Investigations 7
2.2 Compilation and Review of Available Data 7
2.2.1 Previous Studies 7
2.2.1.1 Assessment of Impacts from January 1999 Flooding (GBA, 1994) 7
2.2.1.2 Coogee Oval and Bowling Club Flooding Assessment – 24 January 1999
(PBP, 1999) 8
2.2.2 Historical Flood Levels 8
2.2.3 Rainfall Data 9
2.2.4 Council Data 9
2.3 Site Inspections 11
2.4 Community Consultation 12
2.5 Development of Computer Models 12
2.5.1 Hydrological Model 12
2.5.2 Hydraulic Model 12

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
CONTENTS VI

2.6 Calibration and Sensitivity Testing of Models 12

2.7 Establishing Design Flood Conditions 13
2.8 Flood Result Presentation 13

3 COMMUNITY CONSULTATION 14
3.1 The Community Consultation Process 14
3.2 Community Questionnaire 14
3.3 Public Exhibition 16

4 MODEL DEVELOPMENT 17
4.1 Hydrological Model 18
4.1.1 Flow Path Mapping 18
4.1.2 Rainfall Data 18
4.1.3 Surface Type Hydrologic Properties 20
4.2 Hydraulic Model 20
4.2.1 Extents and Layout 20
4.2.2 Topography 21
4.2.3 Hydraulic Roughness 21
4.2.4 Drainage Layer 23
4.2.5 Boundary Conditions 24

5 MODEL CALIBRATION 25
5.1 Selection of Calibration Events 25
5.2 May 2009 Model Calibration 25
5.2.1 Rainfall Data 25
5.2.2 Rainfall Losses 29
5.2.3 Downstream Boundary Condition 30
5.2.4 Adopted Model Parameters 30
5.2.5 Observed and Simulated Flood Behaviour 32
5.2.5.1 Flood Photographs 34
5.2.5.2 Rainbow Street Calibration 39
5.3 January 1999 Model Calibration 39
5.3.1 Rainfall Data 40
5.3.2 Observed and Simulated Flood Behaviour 41
5.4 October 1959 Model Validation 42
5.4.1 Rainfall Data 42
5.4.2 Observed and Simulated Flood Behaviour 43

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
LIST OF FIGURES VII

6 DESIGN FLOOD CONDITIONS 44

6.1 Design Rainfall 44
6.1.1 Rainfall Depths 44
6.1.2 Temporal Patterns 45
6.1.3 Rainfall Losses 46
6.2 Design Flood Results 46
6.2.1 Peak Flood Levels, Depths and Velocities 46
6.2.2 Hydraulic Categorisation 52
6.2.3 Provisional Hazard 53
6.3 Sensitivity Tests 54
6.3.1 Stormwater Drainage Blockage 55
6.3.2 Sea Level 57
6.3.3 Rainfall Losses 59
6.3.4 Climate Change 61

7 CONCLUSIONS 66

8 REFERENCES 68

APPENDIX A: DESIGN FLOOD MAPPING A-1

APPENDIX B: COMMUNITY QUESTIONNAIRE B-1

APPENDIX C: CALIBRATION DATA C-1

APPENDIX D: HISTORIC NEWSPAPER ARTICLES D-1

APPENDIX E: SUMMARY OF PUBLIC EXHIBITION SUBMISSIONS E-1

LIST OF FIGURES

Figure 1-1 Study Locality 2

Figure 2-1 Topography of the Coogee Bay Catchments 6
Figure 2-2 Rainfall Gauges in the Vicinity of Coogee Bay 10
Figure 3-1 Distribution of Responses to the Questionnaire 15
Figure 4-1 Coogee Bay Catchment Boundary and Overland Flow Paths 19
Figure 4-2 Linked 1D/2D Model Layout 22
Figure 4-3 Sample Drainage Line Longsection 23

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
LIST OF TABLES VIII

Figure 5-1 Coogee Bay Rainfall Radar Coverage 26

Figure 5-2 Rainfall Intensity Profiles for the May 2009 Event 28
Figure 5-3 Comparison of Derived May 2009 Rainfall with IFD Relationships 29
Figure 5-4 Modelled Land Use Map 31
Figure 5-5 Example Location of Solid Wall Modelling 33
Figure 5-6 May 2009 Flood Photograph Locations 35
Figure 5-7 May 2009 Flood Photograph: Location A 36
Figure 5-8 May 2009 Flood Photograph: Location B 36
Figure 5-9 May 2009 Flood Photographs: Location C 37
Figure 5-10 May 2009 Flood Photograph: Location D 37
Figure 5-11 May 2009 Flood Photograph: Location E 38
Figure 5-12 May 2009 Flood Photographs: Location F 38
Figure 5-13 May 2009 Flood Photograph: Location G 39
Figure 5-14 Comparison of Derived January 1999 Rainfall with IFD Relationships 41
Figure 5-15 January 1999 Flood Photograph of Coogee Oval 42
Figure 5-16 Comparison of Derived October 1959 Rainfall with IFD Relationships 43
Figure 6-1 Critical Durations for the 1% AEP Event 47
Figure 6-2 Distribution of Design Result Reporting Locations 48
Figure 6-3 Design Flood Level Long Section 49
Figure 6-4 Comparison of Design Flood Results with Historic Map 51
Figure 6-5 Provisional Flood Hazard Categorisation 54
Figure 6-6 Impact of Stormwater Pipe Blockage on 1% AEP Event 56
Figure 6-7 Impact of Increased Sea-level on 1% AEP Event 58
Figure 6-8 Impact of Decreased Initial Rainfall Loss on 1% AEP Event 60
Figure 6-9 Impact of 10% Rainfall Increase on 1% AEP Event 62
Figure 6-10 Impact of 25% Rainfall Increase on 1% AEP Event 63

LIST OF TABLES

Table 2-1 Summary of Rainfall Gauges in the Coogee Bay Locality 11

Table 5-1 Conversion of Radar Reflectivity to Rainfall Intensity 27
Table 5-2 Daily Rainfall Totals for Each Radar Cell 28
Table 5-3 Adopted Hydraulic Roughness Coefficients Based on Land Use 32
Table 5-4 Rainfall Data for the January 1999 Event 40
Table 6-1 Design Flood Terminology 44
Table 6-2 Average Design Rainfall Intensities (mm/hr) 45
Table 6-3 Summary of Design Flood Levels 49
Table 6-4 Hydraulic categories 53
Table 6-5 Summary of Blockage Sensitivity Results 55

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
LIST OF TABLES IX

Table 6-6 Summary of Sea-level Sensitivity Results 57

Table 6-7 Summary of Initial Loss Sensitivity Results 59
Table 6-8 Summary of 10% Increased Rainfall Assessment 64
Table 6-9 Summary of 25% Increased Rainfall Assessment 64

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
GLOSSARY I

GLOSSARY

annual exceedance AEP (measured as a percentage) is a term used to describe flood

probability (AEP) size. It is a means of describing how likely a flood is to occur in a
given year. For example, a 1% AEP flood is a flood that has a 1%
chance of occurring, or being exceeded, in any one year. It is also
referred to as the ‘100 year ARI flood’ or ‘1 in 100 year flood’. The
term 100 year ARI flood has been used in this study. See also
average recurrence interval (ARI).

Australian Height Datum National survey datum corresponding approximately to mean sea
(AHD) level.

attenuation Weakening in force or intensity

average recurrence interval ARI (measured in years) is a term used to describe flood size. It is
(ARI) the long-term average number of years between floods of a
certain magnitude. For example, a 100 year ARI flood is a flood
that occurs or is exceeded on average once every 100 years. The
term 100 year ARI flood has been used in this study. See also
annual exceedance probability (AEP).

catchment The catchment at a particular point is the area of land that drains
to that point.

design flood A hypothetical flood representing a specific likelihood of

occurrence (for example the 100yr ARI or 1% AEP flood).

development Existing or proposed works that may or may not impact upon
flooding. Typical works are filling of land, and the construction of
roads, floodways and buildings.

discharge The rate of flow of water measured in tems of vollume per unit
3
time, for example, cubic metres per second (m /s). Discharge is
different from the speed or velocity of flow, which is a measure of
how fast the water is moving for example, metres per second
(m/s).

flood A relatively high stream flow that overtops the natural or artificial
banks in any part of a stream, river, estuary, lake or dam, and/or
local overland flooding associated with major drainage before
entering a watercourse, and/or coastal inundation resulting from
super-elevated sea levels and/or waves overtopping coastline
defences excluding tsunami.

flood behaviour The pattern / characteristics / nature of a flood.

flood fringe Land that may be affected by flooding but is not designated as
floodway or flood storage.

flood hazard The potential for damage to property or risk to persons during a
flood. Flood hazard is a key tool used to determine flood severity
and is used for assessing the suitability of future types of land
use.The degree of flood hazard varies with circumstances across
the full range of floods.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
GLOSSARY II

flood level The height of the flood described either as a depth of water above
a particular location (eg. 1m above a floor, yard or road) or as a
depth of water related to a standard level such as Australian
Height Datum (eg the flood level was 7.8 mAHD). Terms also
used include flood stage and water level.

flood liable land see flood prone land

floodplain Land susceptible to flooding up to the probable maximum flood

(PMF). Also called flood prone land. Note that the term flood liable
land now covers the whole of the floodplain, not just that part
below the flood planning level.

floodplain risk management Studies carried out in accordance with the Floodplain
study Development Manual (NSW Government, 2005) that assesses
options for minimising the danger to life and property during
floods. These measures, referred to as ‘floodplain management
measures / options’, aim to achieve an equitable balance between
environmental, social, economic, financial and engineering
considerations. The outcome of a Floodplain Risk Management
Study is a Floodplain Risk Management Plan.

floodplain risk management The outcome of a Floodplain Risk Management Study.

plan

flood planning levels (FPL) The combination of flood levels and freeboards selected for
planning purposes, as determined in Floodplain Risk Management
Studies and incorporated in Floodplain Risk Management Plans.
The concept of flood planning levels supersedes the designated
flood or the flood standard used in earlier studies..

flood prone land Land susceptible to inundation by the probable maximum flood
(PMF) event. Under the merit policy, the flood prone definition
should not be seen as necessarily precluding development.
Floodplain Risk Management Plans should encompass all flood
prone land (i.e. the entire floodplain).

flood stage See flood level.

flood storage Floodplain area that is important for the temporary storage of
floodwaters during a flood.

flood study A study that investigates flood behaviour, including identification

of flood extents, flood levels and flood velocities for a range of
flood sizes.

floodway Those areas of the floodplain where a significant discharge of

water occurs during floods. Floodways are often aligned with
naturally defined channels. Floodways are areas that, even if only
partially blocked, would cause a significant redistribution of flood
flow, or a significant increase in flood levels.

freeboard A factor of safety usually expressed as a height above the

adopted flood level thus determing the flood planning level.
Freeboard tends to compensate for factors such as wave action,
localised hydraulic effects and uncertainties in the design flood
levels.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
GLOSSARY III

high flood hazard For a particular size flood, there would be a possible danger to
personal safety, able-bodied adults would have difficulty wading to
safety, evacuation by trucks would be difficult and there would be
a potential for significant structural damage to buildings.

hydraulics The term given to the study of water flow in rivers, estuaries and
coastal systems.

hydrology The term given to the study of the rainfall-runoff process in

catchments.

low flood hazard For a particular size flood, able-bodied adults would generally
have little difficulty wading and trucks could be used to evacuate
people and their possessions should it be necessary.

m AHD metres Australian Height Datum (AHD).

m/s metres per second. Unit used to describe the velocity of

floodwaters.
3
m /s Cubic metres per second or ‘cumecs’. A unit of measurement for
creek or river flows or discharges. It is the rate of flow of water
measured in terms of volume per unit time.

overland flow path The path that floodwaters can follow if they leave the confines of
the main flow channel. Overland flow paths can occur through
private property or along roads. Floodwaters travelling along
overland flow paths, often referred to as ‘overland flows’, may or
may not re-enter the main channel from which they left; they may
be diverted to another water course.

peak flood level, flow or The maximum flood level, flow or velocity that occurs during a
velocity flood event.

probable maximum flood The largest flood likely to ever occur. The PMF defines the extent
(PMF) of flood prone land or flood liable land, that is, the floodplain. The
extent, nature and potential consequences of flooding associated
with the PMF event are addressed in the current study.

probability A statistical measure of the likely frequency or occurrence of

flooding.

risk Chance of something happening that will have an impact. It is

measured in terms of consequences and likelihood. In the context
of this study, it is the likelihood of consequences arising from the
interaction of floods, communities and the environment.

runoff The amount of rainfall from a catchment that actually ends up as

flowing water in the river or creek.

stage See flood level.

topography The shape of the surface features of land

velocity The term used to describe the speed of floodwaters, usually in

m/s.

water level See flood level.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
INTRODUCTION 1

1 INTRODUCTION
The Coogee Bay Flood Study has been prepared for Randwick City Council (Council) to define the
existing flood behaviour in the Coogee Bay catchment and establish the basis for subsequent
floodplain management activities.

1.1 Study Location

2
The study area drains east to Coogee Bay, with a total area of some 2.9km as shown in Figure 1-1.
Around 80% of the study area forms a single catchment, which is drained via the stormwater pipe
network and discharges to the sea at the northern end of Coogee Beach. Another 10% of the study
area drains to a topographic depression located on Rainbow Street, which is drained by the
stormwater pipe network. The remainder of the study area drains to the sea via smaller local pipe
networks. The study area is predominantly urban and the natural creek systems have been heavily
modified.

1.2 Study Background

Previous studies within the Coogee Bay catchments have focussed on response to the January 1999
flood event, which is one of the more significant recent floods in the study area. Hydrological models
have been used to assess flooding in the Coogee Oval catchment, but a detailed hydraulic modelling
investigation of the entire study area had not been carried out prior to the undertaking of the current
study.

It is recognised that the runoff from local catchments can pose a significant flood risk to parts of the
Coogee Bay catchments. The majority of the study area is serviced by an underground stormwater
drainage system. However, these systems have a finite capacity and are generally designed to
convey runoff for events of the order of 20% AEP to 10% AEP at best. For events of a larger
magnitude that exceed the drainage system capacity, overland flows are generally conveyed along
road networks or designated overland flow paths. In Coogee Bay there are a number of overland flow
paths that are not aligned with the road system which flow through private property.

1.3 The Need for Floodplain Management at Coogee

Bay
As evidenced in the recent May 2009 and January 1999 events, a significant flood risk from overland
flow is posed to residents in parts of the Coogee Bay area. The existing development situated within
historic creek and gully alignments are particularly at risk from flooding. This existing flood risk may
be exacerbated by potential climate change impacts through increased storm intensities and
therefore more catchment runoff. This may result in more frequent and more severe flooding in some
locations within the study area.

There is likely to be a future increase in development pressures across the wider Randwick LGA,
including Coogee to accommodate general population growth expectations. Whilst the majority of the
study catchments are largely developed, infill development may see an intensification of the existing
urban areas. This in time will increase the number of people potentially exposed to flood risk, many

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
INTRODUCTION 2

Figure 1-1 Study Locality

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
INTRODUCTION 3

of whom would be oblivious to existing flood risk given no previous experience of flooding in the
catchment

Floodplain risk management considers the consequences of flooding on the community and aims to
develop appropriate floodplain management measures to minimise and mitigate the impact of
flooding. This incorporates the existing flood risk associated with current development, and future
flood risk associated with future development and changes in land use.

Accordingly, Council desires to approach local floodplain management in a considered and

systematic manner. This study comprises the initial stages of that systematic approach, as outlined
in the Floodplain Development Manual (NSW Government, 2005). The approach will allow for more
informed planning decisions within Coogee.

1.4 The Floodplain Management Process

The State Government’s Flood Prone Land Policy is directed towards providing solutions to existing
flooding problems in developed areas and ensuring that new development is compatible with the
flood hazard and does not create additional flooding problems in other areas. Policy and practice are
defined in the Government’s Floodplain Development Manual (2005).

Under the Policy the management of flood liable land remains the responsibility of Local Government.
The State Government subsidises flood mitigation works to alleviate existing problems and provides
specialist technical advice to assist Councils in the discharge of their floodplain management
responsibilities.

The Policy provides for technical and financial support by the State Government through the following
four sequential stages:

Stages of Floodplain Management

Stage Description
1 Flood Study Determines the nature and extent of the flood problem.
2 Floodplain Risk Management Evaluates management options for the floodplain in
Study respect of both existing and proposed developments.
3 Floodplain Risk Management Involves formal adoption by Council of a plan of
Plan management for the floodplain.
4 Implementation of the Construction of flood mitigation works to protect existing
Floodplain Risk Management development. Use of environmental plans to ensure
Plan new development is compatible with the flood hazard.

This study represents Stage 1 of the above process and aims to provide an understanding of local
catchment flood behaviour within the Coogee Bay catchments.

1.5 Study Objectives

The primary objective of the Flood Study is to define the flood behaviour of the local Coogee Bay
catchments through the establishment of appropriate numerical models. The study has produced
information on flood flows, velocities, levels and extents for a range of flood event magnitudes under
existing catchment and floodplain conditions. Specifically, the study incorporates:

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
INTRODUCTION 4

 Compilation and review of existing information pertinent to the study and acquisition of additional
data including survey as required;

 Undertaking a community consultation and participation program to identify local flooding

concerns, collect information on historical flood behaviour and engage the community in the on-
going floodplain management process;

 Development and calibration of appropriate hydrologic and hydraulic models;

 Determination of design flood conditions for a range of design events including the 20% AEP, 5%
AEP, 1% AEP, 0.5% AEP, 0.2% AEP and extreme flood event; and

 Presentation of study methodology, results and findings in a comprehensive report incorporating

appropriate flood mapping.

The principal outcome of the flood study is the understanding of flood behaviour in the catchments
and in particular design flood information that will underpin subsequent floodplain management
activities.

1.6 About This Report

This report documents the Study’s objectives, results and recommendations.

Section 1 introduces the study.

Section 2 provides an overview of the approach adopted to complete the study.

Section 3 outlines the community consultation program undertaken.

Section 4 details the development of the computer model.

Section 5 details the model calibration and validation process.

Section 6 presents the design flood conditions and sensitivity tests.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 5

2 STUDY APPROACH
2.1 The Study Area
2.1.1 Catchment Description
2
The study area catchments occupy a total area of 2.9km and incorporate the majority of Coogee and
parts of South Coogee and Randwick. The catchments generally drain to the east into Coogee Bay.

The topography of the study area is shown in Figure 2-1. The northern 80% of the study area forms a
single catchment, with the alignment of the natural gully line being similar to that of Dolphin Street.
The southern 20% of the study area forms a number a smaller, less well-defined catchments, all of
which drain east to the sea. The upper catchments are largely elevated above a level of 60m AHD,
peaking at above 80m AHD in some locations. The topography is mostly steep, with slopes typically
in the order of 5% to 10%.

The natural creek systems have been heavily modified and the study area is now drained entirely by
a stormwater pipe network. When the capacity of this network is exceeded, overland flow will occur
along the alignments of the original creeks. Many of the old creek alignments are now located
through developed properties, which presents a significant flood risk.

There are a number of localised depressions in the catchment topography, which will be liable to fill
with water during flood events. Deep flood waters in these locations will not be uncommon once any
local drainage capacity is exceeded. One such depression centred around Rainbow Street is a
significant feature of the catchment topography. It is situated at the southern edge of the study area
and has a catchment area of around 26ha (see Figure 2-1). The topography of this depression
provides no natural outlet and is around 10m deep from the bottom of Rainbow Street to the lowest
point along the catchment boundary. Drainage from the depression will be largely restricted to the
capacity of the trunk drainage line and sub-surface infiltration.

Land use within the study area primarily consists of urban development (90%), open recreational
space (9%) and tree-covered land (1%). The urban development within the study area includes low,
medium and high density residential development and commercial uses, including the Coogee CBD
along the beachfront. Some of the developed areas would previously have been creek alignments.

The study area is traversed by a number of roads, some of which run perpendicular to overland flow
routes through the catchments. These include Alison Road, Coogee Street, Carrington Road, Mount
Street, Brook Street and Arden Street. In certain locations these routes incorporate significant
embankments across the overland flow routes that are evident in the catchment topography shown in
Figure 2-1.

2.1.2 History of Flooding

There is a long history of flooding in Coogee, as it is an old suburb with development located on
natural creek lines. Floods reported in available newspaper articles include 1959, 1989 and 1998.
The articles typically reference flooding of Coogee Oval, including photographs and reported flood
depths. Coogee Oval is situated in a natural depression of over 2m depth, located behind the higher

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 6

Figure 2-1 Topography of the Coogee Bay Catchments

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 7

ground of Arden Street and the Coogee Bay foreshore area. As such it is prone to flooding and these
occurrences are generally well documented.

The October 1959 event is the largest recorded within the catchment, with 265mm of daily rainfall
recorded at Randwick Bowling Club. Mayoral records provided by Council related to this event
suggest the total rainfall depth fell within a period of three and a half hours. This equates to around
twice the rainfall of a 1% AEP event of similar duration.

The most significant recent floods include the January 1999 and May 2009 events. The former is well
documented by both the Coogee Oval and Bowling Club Flooding Assessment (PBP, 1999) and
Assessment of Impacts from January 1999 Flooding (GBA, 1999). 74mm of rain were recorded on
th
24 January at Randwick Bowling Club, most of which is believed to have fallen in a 90 minute
period. This would be equivalent to around a 10% AEP event.

The May 2009 event is one of the largest recent local flooding events within the catchment This was
a localised storm event which primarily impacted the eastern suburbs of Sydney. Randwick Bowling
rd
Club recorded 77mm of rain on 3 May, which was the highest recorded daily total by any gauge in
the local area. Most of the rain fell within a 90 minute period, which would be equivalent to around a
10% AEP event.

2.1.3 Previous Investigations

Previous investigations of the flooding characteristics of the study area were undertaken following the
January 1999 flood event. These studies focused on specific locations within the catchment where
flood damage had occurred during the event.

An assessment of flooding impacts at seven properties within Randwick LGA was undertaken by
Gary Blumberg & Associates (1999). Of these, two properties are located within the study area. The
report details the damage and likely flood mechanisms at each location.

A detailed investigation of flooding at Coogee Oval and Bowling Club was undertaken by Patterson
Britton & Partners (1999). It included a hydrological assessment of the main catchment of the Coogee
Bay study area. A local inspection of the drainage network and a hydraulic assessment were carried
out for the area of concern.

Further details of these previous investigations and their relevance in the context of the current flood
study are presented in Section 2.2.1.

2.2 Compilation and Review of Available Data

2.2.1 Previous Studies

2.2.1.1 Assessment of Impacts from January 1999 Flooding (GBA, 1994)

th
Following the flood event of 24 January 1999 and associated complaints from local residents,
Council engaged Gary Blumberg & Associates to provide an engineering assessment of flooding
impacts which occurred at seven properties. The properties were located across the wider Randwick
LGA. Of these, two are located within the study area:

 Albi Place, Randwick; and

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 8

 Clyde Street, Randwick.

The study provides a description of the January 1999 storm event, including a 5-min interval recorded
rainfall series at Little Bay (BoM station), in which 93mm of rain fell in one hour. It is stated that this
may statistically be described as a 1 in 85 year ARI (or close to a 1% AEP event), as advised by
Council at the time. However, the study found the storm recurrence interval to be lower for some local
sub-catchments, based on the assessments undertaken.

Comments regarding the flooding at Albi Place suggest that runoff from the roadways to the north-
west resulted in a scour hole being created behind the retaining wall at the rear of the property. At
Clyde Street runoff from Pitt Street exceeded the gutter capacity, proceeding to flow through
properties and into Clyde Street. Runoff from the northern end of Judge Street probably also
contributed. A flood depth of 150mm to the rear of the property at Clyde Street is quoted.

2.2.1.2 Coogee Oval and Bowling Club Flooding Assessment – 24 January

1999 (PBP, 1999)

The Patterson Britton & Partners study focussed on flooding of Coogee Oval and the nearby bowling
club. It included hydrologic and hydraulic investigations for the catchment using the RAFTS and
RatHGL software packages. The modelling was done to assess the capacity of the stormwater
network in the vicinity of Coogee Oval. Runoff exceeding the capacity of the stormwater drainage was
routed to a detention basin representing Coogee Oval. Survey data of some of the pipes was
collected as part of the study.

The January 1999 event was modelled using the rainfall data from Little Bay, which totalled 114mm
over a two hour period. It is stated that this represents around a 1 in 60 year ARI storm event. Little
Bay is located some 7km to the south of the Coogee Bay catchment. Inspection of the Randwick
Bowling Club rain gauge, which is less than 1km away, shows a daily rainfall total of 74mm. The
majority of the rainfall fell in a 90 minute period. A rainfall depth of 74mm over a 90 minute period
would be closer to a 10% AEP event than the previously suggested 1 in 60 year ARI storm.

An observed peak flood level of around 5.4m AHD within Coogee Oval is specified within the report.
The modelled peak level was over-estimated at 6.1m AHD. It is not apparent in the report why the
Little Bay rainfall depth was adopted over the more local Randwick Bowling Club total, but this may
explain the over-estimation of peak flood level, in addition to the modelling limitations cited in the
report.

Recommendations were made regarding short-term and long-term measures to reduce flooding
within the Coogee Bay catchment.

2.2.2 Historical Flood Levels

Available flood level records in the catchment are limited. Coogee Oval is the one location for which
information relating to flood levels exists for a number of events. During significant storm events,
excess runoff from the catchment collects in the Oval, which is situated in a natural depression of
over 2m depth. Although no official flood level records were available at this location, the area can be
inundated for several hours. As a result a number of flood photographs and additional anecdotal
evidence is readily available, enabling estimation of peak flood levels in the Oval.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 9

Flood photographs of Coogee Oval were available and identified for the following events:
th
 6 January 1989;
th
 24 January 1999; and
nd
 2 May 2009.

For each of these events the peak flood level within Coogee Oval can consistently be determined to
be around 5.4m AHD (approximately 1m depth). Daily rainfall totals recorded for the relevant dates at
Randwick Bowling Club are 74mm, 74mm and 77mm respectively. It is likely that these events were
in the same order of magnitude in terms of catchment runoff volume, and have resulted in similar
peak flood levels within Coogee Oval being attained.

th
References within the Mayoral records and a newspaper article relating to a flood event on 29
October 1959 mention flood depths in Coogee Oval of around 10 feet. This indicates that a flood level
of 7.5m AHD or more was reached within the Oval, which would also have involved substantial
overtopping of Arden Street and the Coogee Bay foreshore area.

The records available for Coogee Oval were further supplemented by observed flood levels and
photographs (largely relating to the May 2009 event) obtained through the community consultation
process, as discussed in Section 3.2. Data obtained from historic records and the community
consultation process is presented in Section 5.2.5, for the purposes of calibration.

2.2.3 Rainfall Data

There is an extensive network of rainfall gauges across the Sydney area, many of which are operated
by the Bureau of Meteorology (BoM) and Sydney Water Corporation (SWC). There are no gauges
located within the study area. The closest gauge to the Coogee Bay catchments is a BoM operated
daily read gauge, located at Randwick Bowling Club. This gauge has a long period of record, from
1917, and is still operational. There are a further 16 rainfall gauges located within 5km of the study
area, four of which are daily read gauges operated by BoM. The remainder are continuous gauges
and are operated by SWC. The closest BoM-operated continuous gauge is located around 6km from
the study area at Little Bay. A list of these rainfall stations with their respective period of record,
including closed stations, is shown in Table 2-1. The location of the gauges is shown in Figure 2-2.

The May 2009 is the largest recent local catchment event in the study area. For this event RADAR
rainfall data has also been acquired from BoM. A more detailed discussion of the rainfall data
available for this and other events is discussed in Section 5.2.1.

2.2.4 Council Data

Digitally available information such as aerial photography, cadastral boundaries, topography,

watercourses, drainage networks, land zoning, vegetation communities and soil landscapes were
provided by Council in the form of GIS datasets.

LiDAR (Light Detection and Ranging) land survey data covering the entire study area was acquired in
2005. LiDAR data is of good vertical accuracy (generally ~ +/- 0.1m) and provides data at around a
2m interval, providing excellent coverage over an extensive area. Flood behaviour is inherently
dependent on the ground topography.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 10

Figure 2-2 Rainfall Gauges in the Vicinity of Coogee Bay

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 11

Advanced GIS analysis also allows the LiDAR imagery to be assessed in concert with spatial 2-D
flood model data, facilitating mapping, categorisation, and overall flood management.

Table 2-1 Summary of Rainfall Gauges in the Coogee Bay Locality

Station Start End
Name Operator Type
No. Year Year
66051 Little Bay (The Coast Golf Club) BoM Pluvio 1925 current
566009 Rushcutters Bay Tennis Club SWC Pluvio 1998 current
566010 Cranbrook School at Bellvue Hill SWC Pluvio 1998 current
566028 Mascot Bowling Club SWC Pluvio 1973 current
566032 Paddington (Composite Site) SWC Pluvio 1961 current
566034 Pagewood SWC Pluvio 1959 1973
566043 Randwick (Army) SWC Pluvio 1956 1970
566077 Bondi (Dickson Park) SWC Pluvio 1989 2001
566088 Malabar STP SWC Pluvio 1990 current
566099 Randwick Racecourse SWC Pluvio 1991 current
566114 Waverley Bowling Club SWC Pluvio 1995 current
566115 Bondi Golf Club SWC Pluvio 1994 1995
566123 Maroubra Bowling Club SWC Pluvio 1995 1998
66052 Randwick Bowling Club BoM Daily 1917 current
66073 Randwick Racecourse BoM Daily 1937 current
66098 Rose Bay (Royal Sydney Golf Club) BoM Daily 1928 current
66160 Centennial Park BoM Daily 1900 current
66209 Dover Heights (Portland St) BoM Daily 2007 current

Details of stormwater drainage were provided in a GIS database format for the entire study area. The
dataset included full survey details of the pipes and pits, which were collected over the previous five
years. Details include pipe sizes, invert levels and pit inlet configuration and dimensions.

Flood information collated from a Council file and library search was also made available.

2.3 Site Inspections

A number of site inspections were undertaken during the course of the study to gain an appreciation
of local features influencing flooding behaviour. Some of the key observations to be accounted for
during the site inspections included:

 Presence of local structural hydraulic controls such as walls and kerbs that may have an impact
on overland flooding behaviour;

 Confirmation of the location and configuration of the stormwater drainage pits and outlets;

 Location of existing development and infrastructure on the floodplain.

This visual assessment was useful for defining hydraulic properties within the hydraulic model and
ground-truthing of topographic features identified from survey.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 12

2.4 Community Consultation

The success of a floodplain management plan hinges on its acceptance by the community, residents
within the study area, and other stake-holders. This can be achieved by involving the local community
at all stages of the decision-making process. This includes the collection of their ideas and knowledge
on flood behaviour in the study area, together with discussing the issues and outcomes of the study
with them.

The key elements of the consultation process in undertaking the flood study have included:

 Issue of a questionnaire to obtain historical flood data and community perspective on flooding
issues;

 Public exhibition of Draft Report and community information session (to be undertaken).

These elements are discussed in further detail in Section 3.

2.5 Development of Computer Models

2.5.1 Hydrological Model

Traditionally, for the purpose of the Flood Study, a hydrologic model is developed to simulate the rate
of storm runoff from the catchment. The output from the hydrologic model is a series of flow
hydrographs at selected locations such as at stormwater drainage pit inlets, which form the inflow
boundaries to the hydraulic model.

In recent years the advancement in computer technology has enabled the use of the direct rainfall
approach as a viable alternative. With the direct rainfall method the design rainfall is applied directly
to the individual cells of the 2D hydraulic model. This is particularly useful for overland flow studies
where model results are desired in areas with very small contributing catchments. This study has
adopted the direct rainfall approach for modelling hydrology, details of which are discussed in Section
4.1.

2.5.2 Hydraulic Model

The TUFLOW hydraulic model (discussed in Section 4.2) developed for this study includes:
2
 two-dimensional (2D) representation of Coogee Bay covering an area of approximately 2.8 km
(complete coverage of the total catchment area); and

 one-dimensional (1D) representation of the stormwater pipe network.

The hydraulic model is applied to determine flood levels, velocities and depths across the study area
for historical and design events.

2.6 Calibration and Sensitivity Testing of Models

The hydrodynamic model was primarily calibrated to the May 2009 flood event to establish the values
of key model parameters and confirm that the models were capable of adequately simulating real
flood events.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
STUDY APPROACH 13

The following criteria are generally used to determine the suitability of historical events to use for
calibration or validation:

 The availability, completeness and quality of rainfall and flood level event data;

 The amount of reliable data collected during the historical flood information survey; and

 The variability of events – preferably events would cover a range of flood sizes.

The available historical information highlighted only one flood with sufficient data to potentially
support a calibration process – the May 2009 event. Flood information relating to Coogee Oval for the
January 1999 and October 1959 events has also been used to aid the model calibration and
validation process.

The calibration and validation of the model is presented in Section 5. A series of sensitivity tests were
also carried out to evaluate the model. These tests were conducted to examine the performance of
the models and determine the relative importance of different hydrological and hydrodynamic factors.
The sensitivity testing of the model is detailed in Section 6.3.

2.7 Establishing Design Flood Conditions

Design floods are statistical-based events which have a particular probability of occurrence. For
example, the 1% Annual Exceedance Probability (AEP) event, which is sometimes referred to as the
1 in 100 year Average Recurrence Interval (ARI) flood, is the best estimate of a flood with a peak
discharge that has a 1% (i.e. 1 in 100) chance of occurring in any one year. For the Coogee Bay
catchments, design floods were based on design rainfall estimates according to Australian Rainfall
and Runoff (IEAust, 2001).

The design flood conditions form the basis for floodplain management in the catchment and in
particular design planning levels for future development controls. The predicted design flood
conditions are presented in Section 6.2.

2.8 Flood Result Presentation

Design flood result presentation is undertaken using output from the hydrodynamic model. Figures
are produced showing water depth and velocity for each of the design events. The figures present the
peak value of each parameter. Provisional flood hazard categories and hydraulic categories derived
from the hydrodynamic model results are also presented. The flood model outputs are described in
Section 6.2 and presented in Appendix A.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
COMMUNITY CONSULTATION 14

3 COMMUNITY CONSULTATION
3.1 The Community Consultation Process
The consultation has aimed to inform the community about the development of the flood study and its
likely outcome as a precursor to subsequent floodplain management activities. It has provided an
opportunity to collect information on their flood experience, in particular historical flood data related to
overland flooding.

The key elements of the consultation process have been as follows:

 Distribution of a questionnaire to all landowners, residents and businesses within the study area;
and

 Public exhibition of the draft Flood Study (to be undertaken).

These elements are discussed in detail below.

3.2 Community Questionnaire

A questionnaire was distributed to residents within the study area to collect information on their
previous flood experience and flooding issues. The focus of the questionnaire was historical flooding
information that may be useful for correlating with predicted flooding behaviour from the modelling.

Council mailed out the questionnaire to all residents and businesses located within the study area.
Council received back almost 1000 responses, of which around 250 had comments relating to
flooding. The responses were compiled into a GIS layer by Council. A copy of the questionnaire is
included in Appendix B.

The focus of the questionnaire was to gather relevant flood information from the community, including
photographs, observed flood depths and descriptions of flood behaviour within the catchment.
Thirteen responses included photographs showing flooding, flood damage or flood marks showing
high water levels that occurred during the flood event.

Comments relating to flood behaviour contained within the responses were extracted where useful for
model calibration purposes. Around 200 such comments were extracted, many of which included
indicative flood depths. Most of these comments appeared to relate to the recent May 2009 flood
event, with only a few relating to older events.

The distribution of questionnaire responses is presented in Figure 3-1. It can be seen that there is a
fairly comprehensive coverage of responses across the study area. The locations of responses with
comments relating to flooding have been highlighted. The two main flowpath alignments within the
study area can be discerned, as can a cluster of comments from the local depression catchment
centred on Rainbow Street.

The comments relating to flooding that were received from the community were an important part of
the calibration process, which is discussed in Section 5.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
COMMUNITY CONSULTATION 15

Figure 3-1 Distribution of Responses to the Questionnaire

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
COMMUNITY CONSULTATION 16

Many comments received related to local scour issues. These typically occur in areas with steep
slopes over 10% grade. The high flood velocities coupled with sandy soils result in local scour and
downstream deposition problems, even in areas of sheet flow where no substantial overland flow
path has been generated. Comments of this nature were prominent in the south-east of the study
area, such as Cairo Street.

A number of newspaper articles relating to storms in 1912, 1914, 1922 and 1933 were also provided
through the community consultation process and are included in Appendix D. They contain some
useful anecdotal evidence relating to flooding in the catchment. The 1922 storm in particular appears
to have caused severe damage. It was an intense hailstorm, lasting only around 30 minutes, but
caused extensive flooding. Unfortunately there is insufficient data available to use these events for
model calibration purposes. Also, topographic modifications within the catchment over such
timeframes may also have changed local flood behaviour.

3.3 Public Exhibition

The Draft Coogee Bay Flood Study was paced on public exhibition from Tuesday 19 February 2013
to Tuesday 26 March 2013.

Public displays were placed at the following locations:

 Bowen Library, 669-673 Anzac Parade, Maroubra;

 Randwick Library, Level 1 Royal Randwick Shopping Centre, Randwick; and

 Council’s administration centre, 30 Frances Street, Randwick.

Exhibition material at the public displays included:

 Copies of the draft reports;

 Fact Sheets;

 Comment Sheets; and

 Comment Box.

Newspaper advertisements were placed in the Southern Courier on 19 February and 5 March
providing details of the public exhibition. The public exhibition was also advertised on Council’s
website and included a copy of the draft Flood Study.

A letter was sent to all property owners identified as being below the 1% AEP flood plus freeboard or
below the Probable Maximum Flood. A total of 7317 letters were sent to property owners providing
details of the public exhibition and the community drop in session. A community drop in session was
held at Bowen Library, 669-673 Anzac Parade, Maroubra on Monday 11 March between 6pm and
8pm. Staff from Council, BMT WBM and the Office of Environment and Heritage were available for
the community to come along and find out about the study or ask questions.

A total of five written submissions were received during the public exhibition period. A summary of
the feedback from residents during the public exhibition period is provided in Appendix E

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 17

4 MODEL DEVELOPMENT
Computer models are the most accurate, cost-effective and efficient tools to assess a catchment’s
flood behaviour. Traditionally, for the purpose of the Flood Study, a hydrologic model and a hydraulic
model are developed.

The hydrologic model simulates the catchment rainfall-runoff processes, producing the stormwater
flows which are used in the hydraulic model.

The hydraulic model simulates the flow behaviour of the drainage network and overland flow paths,
producing flood levels, flow discharges and flow velocities.

In recent years the advancement in computer technology has enabled the use of the direct rainfall
approach as a viable alternative. With the direct rainfall method the design rainfall is applied directly
to the individual cells of the 2D hydraulic model. This is particularly useful for overland flow studies
where model results are desired in areas with very small contributing catchments. This study has
adopted the direct rainfall approach for modelling hydrology and therefore only a single TUFLOW
model has been developed.

Information on the topography and characteristics of the catchments, drainage network and
floodplains are built into the model. Recorded historical flood data, including rainfall and flood levels,
are used to simulate and validate (calibrate and verify) the model. The model produces as output,
flood levels, flows (discharges) and flow velocities.

Development of a hydraulic model follows a relatively standard procedure:

1. Discretisation of the catchment, drainage network, floodplain, etc.

2. Incorporation of physical characteristics (stormwater pipe details, floodplain levels, structures

etc).

3. Establishment of hydrographic databases (rainfall, flood flows, flood levels) for historic events.

4. Calibration to one or more historic floods (calibration is the adjustment of parameters within
acceptable limits to reach agreement between modelled and measured values).

5. Verification to one or more other historic floods (verification is a check on the model’s
performance without further adjustment of parameters).

6. Sensitivity analysis of parameters to measure dependence of the results upon model

assumptions.

Once model development is complete it may then be used for:

 establishing design flood conditions;

 determining levels for planning control; and

 modelling development or management options to assess the hydraulic impacts.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 18

4.1 Hydrological Model

The hydrologic model simulates the rate at which rainfall runs off the catchment. The amount of
rainfall runoff from the catchment is dependent on:

 the catchment slope, area, vegetation and other characteristics;

 variations in the distribution, intensity and amount of rainfall; and

 the antecedent conditions (dryness/wetness) of the catchment.

Hydrological modelling is undertaken to establish inflow boundaries to the TUFLOW hydraulic model
(flow hydrographs from external catchments and local rainfall directly on to the flood-prone area). A
direct rainfall approach has been adopted for the study using the TUFLOW software. The runoff
routing and hydrological response of the catchment within the model is driven by the surface type and
underlying topography. Where appropriate, runoff is diverted into 1D pipe domains of the 2D/1D
model (more detail is provided in Section 4.2). The general modelling approach and adopted
parameters is discussed in the following sections.

4.1.1 Flow Path Mapping

2
The study catchments drain an area of approximately 2.9km to their outlets in Coogee Bay. The
extent of the study area hydrologic catchment is shown in Figure 4-1.

Flow path mapping and catchment delineation has been undertaken using the CatchmentSIM
software. The generated DEM was imported into the software and following hydrologic conditioning
(removal of flats and pits), flow paths and catchment boundaries were generated.

The delineation of the hydrologic catchment boundary was important for defining the limits of the
hydraulic model extent and the associated direct rainfall input.

4.1.2 Rainfall Data

Rainfall information is the primary input and driver of the hydrological model which simulates the
catchments response in generating surface run-off. Rainfall characteristics for both historical and
design events are described by:

 Rainfall depth – the depth of rainfall occurring across a catchment surface over a defined period
(e.g. 270mm in 36hours or average intensity 7.5mm/hr); and

 Temporal pattern – describes the distribution of rainfall depth at a certain time interval over the
duration of the rainfall event.

Both of these properties may vary spatially across the catchment.

The procedure for defining these properties is different for historical and design events. For historical
events, the recorded hyetographs at continuous rainfall gauges provide the observed rainfall depth
and temporal pattern. Where only daily read gauges are available within a catchment, assumptions
regarding the temporal pattern may need to be made.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 19

Figure 4-1 Coogee Bay Catchment Boundary and Overland Flow Paths

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 20

For design events, rainfall depths are most commonly determined by the estimation of intensity-
frequency-duration (IFD) design rainfall curves for the catchment. Standard procedures for derivation
of these curves are defined in AR&R (2001). Similarly AR&R (2001) defines standard temporal
patterns for use in design flood estimation.

The rainfall inputs for the historical calibration/validation events are discussed in further detail in
Section 5 and design events discussed in Section 6.

4.1.3 Surface Type Hydrologic Properties

The response of the catchment to the input rainfall data is dependent on the spatial distribution and
hydrologic properties of the land use surface types. The properties assigned to each surface type (or
material) within TUFLOW that influence the hydrologic response of the model are:

 Initial and continuing losses determine how much rainfall is lost to surface and soil storage etc.
and therefore the effective rainfall contributing to surface runoff;

 Roughness parameters for sheet flow govern the speed with which the runoff will travel,
influencing the hydrologic response of the model.

The material layers input to the model define these properties for each land use surface type within
the catchment. Each material has initial loss, continuing loss and roughness parameters assigned to
it. Along with the model topography, it is these parameters which determine the runoff routing and
hydrological response of the model.

4.2 Hydraulic Model

The overland flow regime in urban environments is characterised by large and shallow inundation of
urban development with interconnecting and varying flowpaths. Road networks often convey a
considerable proportion of floodwaters due to the hydraulic efficiency of the road surface compared to
developed areas (eg. blocked by fences and buildings), in addition to the underground pipe network
draining mainly to open channels. Given this complex flooding environment, a 2D modelling approach
is warranted for the overland flooding areas.

BMT WBM has applied the fully 2D software modelling package TUFLOW. TUFLOW was developed
in-house at BMT WBM and has been used extensively for over fifteen years on a commercial basis
by BMT WBM. TUFLOW has the capability to simulate the dynamic interaction of in-bank flows in
open channels, major underground drainage systems, and overland flows through complex overland
flowpaths using a linked 2D / 1D flood modelling approach.

4.2.1 Extents and Layout

Consideration needs to be given to the following elements in constructing the model:

 topographical data coverage and resolution;
 location of recorded data (eg. levels/flows for calibration);
 location of controlling features (eg. dams, levees, bridges);
 desired accuracy to meet the study’s objectives;
 computational limitations.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 21

With consideration to the available survey information and local topographical and hydraulic controls,
a linked 1D/2D model was developed extending from the catchment outlets in Coogee Bay at the
downstream limit, to the head of the catchments. The stormwater drainage network has been
modelled as 1D branches underlying the 2D (floodplain) domain. This approach enables the
hydraulic capacity of the pipe drainage to be accurately defined by true pipe dimensions, whilst
enabling the overland flow to be represented in 2D. The model layout is presented in Figure 4-2.

2
The floodplain area modelled within the 2D domain comprises a total area of some 2.9km (up to
approximately 80m AHD) which includes the entire of the study catchments and the Coogee Beach
area. A high resolution DEM was derived for the study area from the LiDAR data provided by Council.
The ground surface elevation for the TUFLOW model grid points are sampled directly from the DEM.

A TUFLOW 2D domain model resolution of 2m was adopted for study area. It should be noted that
TUFLOW samples elevation points at the cell centres, mid-sides and corners, so a 2m cell size
results in DEM elevations being sampled every 1m. This resolution was selected to give necessary
detail required for accurate representation of floodplain topography and its influence on overland
flows.

4.2.2 Topography

A high resolution DEM has been derived for the study area from the LiDAR data provided by Council.
The ground surface elevation for the TUFLOW model grid points are sampled directly from the DEM.
It is a representation of the ground surface and does not include features such as buildings or
vegetation.

In the context of the overland flow path study, a high resolution DEM is important to suitably represent
available flow paths, such as roadway/gutter flows that are expected to provide significant flood
conveyance within the study area. Experience has proved this to be a successful approach and
enables detailed simulation of flooding from overland flow paths.

The ability of the model to provide an accurate representation of the overland flow distribution on the
floodplain ultimately depends upon the quality of the underlying topographic model. For the Coogee
Bay catchments, a high resolution DEM (0.5m grid) was derived from LiDAR survey provided by
Council.

4.2.3 Hydraulic Roughness

The development of the TUFLOW model requires the assignment of different hydraulic roughness
zones. These zones are delineated from aerial photography and cadastral data identifying different
land-uses (eg. forest, cleared land, roads, urban areas, etc) for modelling the variation in flow
resistance.

The hydraulic roughness is one of the principal calibration parameters within the hydraulic model and
has a major influence on flow routing and flood levels. The roughness values adopted from the
calibration process is discussed in Section 5.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 22

Figure 4-2 Linked 1D/2D Model Layout

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 23
4.2.4 Drainage Layer

The study requires the modelling of the drainage system in each catchment. Council provided
information where available on the existing drainage system. This data comprised a GIS layer of
pit/pipe locations, together with survey details including pipe sizes, invert levels and pit inlet
structures. The review of the available stormwater drainage system found the data to be largely
complete with only local gaps where survey access had not been possible.

In areas where no pipe survey was available pipe size details were assumed from upstream and
downstream configurations. The invert levels were interpolated between known locations, maintaining
the upstream and downstream pipe gradients where appropriate. These were then cross-checked
against the DEM elevations to take account of any local topographic features and to maintain
minimum cover levels.

For this study the entire trunk drainage network indicated by the council GIS data was modelled. The
study area contains a number of locations that would drain poorly without the inclusion of the pipe
network. Modelling all pipes ensures that the drainage of these areas is well represented.

A sample longsection of a modelled drainage line is shown in Figure 4-3. The figure shows the invert
levels and obvert according to culvert dimension, the ground surface level as derived from the DEM,
and a minimum cover level of 600mm.

Figure 4-3 Sample Drainage Line Longsection

The pipe network, represented as a 1D layer in the model, is dynamically linked to the 2D domains at
specified pit locations for inflow and surcharging. Pit inlet capacities have been modelled using
dimensions contained within the GIS database. Pit inlet curves have been developed for sag pit

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL DEVELOPMENT 24

configurations. The modelled pipe network, which consists of around 1400 pipes with a combined run
length of approximately 25km, is shown in Figure 4-2.

For the magnitude of events under consideration in the study, the pipe drainage system capacity is
expected to be well exceeded with the major proportion of flow conveyed in overland flow paths. For
this study the pipe network data was of a high quality, providing for a good representation of the
drainage system in the model. Nevertheless, any limitations in the available data or model
representation of the drainage system may not have a significant affect on flooded area for the major
flood events considered.

4.2.5 Boundary Conditions

The catchment runoff is determined through the hydrological component of the model and is applied
directly to the TUFLOW model 2D domain, where it is routed as sheet flow until the runoff contribution
is substantial enough to generate an overland flow path. Flow is automatically transferred to the 1D
domain where sufficient pipe and inlet capacity is available. Surcharging will then occur from the 1D
to the 2D domain once the pipe capacity becomes exceeded.

The downstream model limit corresponds to the water level in Coogee Bay. This has been set to a
conservative level of 1m AHD but is insignificant in its influence on upstream flood levels. The
adopted sea level boundary is discussed further in Section 6.3.2. Additional model boundaries have
been included at a few locations where runoff will spill over the catchment boundary and exit the
study area. In these instances constant water level boundaries have been applied in the 1D domain
and QH relationships applied in the 2D domain. The impact of these boundaries is not significant in
determining flood levels within the study area.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 25

5 MODEL CALIBRATION
5.1 Selection of Calibration Events
The selection of suitable historical events for calibration of computer models is largely dependent on
available historical flood information. Ideally the calibration and validation process should cover a
range of flood magnitudes to demonstrate the suitability of a model for the range of design event
magnitudes to be considered.

Significant flooding in Coogee has occurred on numerous occasions, with the most severe events in
recent times including 1959, 1989, 1998, 1999 and 2009. The May 2009 event is considered the
most suitable of the historical events for model calibration. The vast majority of the community
questionnaire responses related to the May 2009 event. The availability of rainfall data and flood
photographs provides a sound dataset to assist calibration of the model.

The January 1999 event was also selected for model calibration. It is similar in magnitude to the May
2009 event and is the next most recent significant flood event in the catchment. The October 1959
event has been selected for model validation purposes as although available data is limited, it is the
largest event recorded within the study area.

The model calibration therefore is based on the historical data available for the three events. The
available data, modelling approach and model results for each of these events are discussed in
further detail in the following sections.

5.2 May 2009 Model Calibration

5.2.1 Rainfall Data

The distribution of rainfall gauge locations in the vicinity of the Coogee Bay catchments was shown in
Figure 2-2 with their respective periods of record shown in Table 2-1. The closest gauge to the study
area is located at Randwick Bowling Club, which records daily rainfall totals. It recorded a total rainfall
rd
depth of 76.6mm on 3 May 2009.

The May 2009 storm was localised and intense and so rainfall depths and temporal patterns would
have exhibited significant spatial variation. The best data source available to estimate the rainfall that
fell on the catchments during the event is the rainfall radar data from the Sydney radar station,
located at Terry Hills and operated by BoM. Data was acquired from this station for the May 2009
event.

A total of seven cells of the rainfall radar dataset intersected with the study area. These have been
referred to as cells: N, E, SE, S, W, NW and NNW for the purposes of this study. The coverage of
these cells in relation to the study area is shown in Figure 5-1. The radar data provides signal
strength (dBZ) returns at 10 minute intervals, classified into 16 bands. These signal strengths can be
converted to rainfall intensities in mm/h using the following equation:

( )
( ) ( )

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 26

Figure 5-1 Coogee Bay Rainfall Radar Coverage

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 27

Table 5-1 shows the 16 radar intensity bands and their corresponding rainfall intensities.

Table 5-1 Conversion of Radar Reflectivity to Rainfall Intensity

Intensity Band Signal Strength (dBZ) Rainfall Intensity (mm/h)

0 0 0.0
1 12 0.2
2 23 1.0
3 28 2.1
4 31 3.2
5 34 4.9
6 37 7.5
7 40 11.5
8 43 17.8
9 46 27.3
10 49 42.1
11 52 64.8
12 55 99.9
13 58 153.8
14 61 236.8
15 64 364.6

The rainfall radar measures the reflectivity of rain clouds. This is strongly dependant on the size of
raindrops in the cloud and not the amount of rain drops. Therefore, differences between rainfall totals
estimated from radar data and those recorded at gauge sites are often experienced. The radar data
gives a good indication of the temporal and spatial distribution of rainfall, but requires calibration to
recorded gauge totals to provide accurate rainfall depth estimations. For this study the radar data has
been used to provide the temporal pattern of the May 2009 storm and the spatial distribution of
rainfall intensities within the catchment, in relation to the Randwick Bowling Club rainfall gauge
location and recorded rainfall depth.

The calculated rainfall intensities for each cell and time interval were converted to rainfall depths and
rd
totalled for the 24 hours (09:00 – 09:00) of 3 May. The total rainfall depth for cell NW, as captured by
the radar data is 22.9mm. This is the cell in which the Randwick Bowling Club daily rainfall gauge is
situated. The total rainfall depth recorded by the gauge was 76.6mm, or approximately 3.3 times as
much as was indicated by the radar. The daily totals for each radar cell were therefore scaled by this
amount, to match with the recorded gauge data. The rainfall totals for each radar cell are presented in
Table 5-2. The data shows a trend of higher rainfall on the coast, decreasing to the west of the study
area. The data suggests that higher rainfall intensities were experienced within the study catchments
than at the Randwick Bowling Club gauge location. The May 2009 event was largely a coastal event,
with localised high intensities.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 28

Table 5-2 Daily Rainfall Totals for Each Radar Cell

Calculated Daily Scaled Daily Rainfall Area Within Study

Radar Cell
Rainfall Total (mm) Total (mm) Catchment (ha)
N 31.9 106.9 58.6
E 34.6 115.9 28.9
SE 35.0 117.1 16.0
S 34.2 114.5 86.0
W 28.7 96.1 69.3
NW 22.9 76.6 18.0
NNW 25.1 84.1 16.5
Catchment
31.3 104.9 293
Weighted

A representative rainfall profile for the catchment was then derived from the scaled radar rainfall
intensities using an areal weighted approach. The area of each radar cell that intersects with the
study catchment are also shown in Table 5-2 and were used to derive the areal weighted catchment
average rainfall. The rainfall intensities calculated for each radar cell and the resultant scaled
catchment weighted average intensity are presented in Figure 5-2. A total rainfall depth of 105mm
was determined to have fallen on the catchment in the 24 hour period. However, the main event
lasted around one hour, with a rainfall depth of around 62mm. Around 12mm fell in the three hours
preceding the event, with a further 27mm within the six hours following the event.

Figure 5-2 Rainfall Intensity Profiles for the May 2009 Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 29

To gain an appreciation of the relative intensity of the May 2009 event, the derived rainfall depths for
various storm durations is compared with the design IFD data for Coogee as shown in Figure 5-3.

The derived depth vs. duration profile for the May 2009 event from the scaled catchment averaged
radar data shows it generally tracking around the design 10% AEP (10-year ARI) rainfall for a one
hour duration event.

Figure 5-3 Comparison of Derived May 2009 Rainfall with IFD Relationships

5.2.2 Rainfall Losses

The initial loss-continuing loss model has been adopted in the TUFLOW model developed for the
Coogee Bay catchments. The initial loss component represents a depth of rainfall effectively lost from
the system and not contributing to runoff and simulates the wetting up of the catchment to a saturated
condition. The continuing loss represents the rainfall lost through soil infiltration once the catchment is
saturated and is applied as a constant rate (mm/hr) for the duration of the runoff event.

Typical design loss rates applicable for NSW catchments east of the western slopes are initial loss of
10 to 35 mm and continuing loss of 2.5mm/hr (AR&R, 2001). However, losses for the Coogee Bay
catchments are likely to be higher due to the sandy nature of the soils. The Coogee Oval and Bowling
Club Flooding Assessment – 24 January 1999 (PBP, 1999) adopted initial and continuing losses of
35mm and 5mm/h respectively. The flood level within Coogee Oval was substantially overestimated
by the modelling and one of the reasons given for this was that the losses may be considerably
higher. The initial loss was set at 35mm to remain within the recommendations given in AR&R.

Given the availability of flood records for Coogee Oval it was possible to assess the likely losses for
the catchment. Being a flood storage area, the peak water levels within Coogee Oval are

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 30

predominantly driven by the volume of runoff generated during an event. The volume of water
entering the oval is dependent on the rainfall depth and the rainfall losses. The volume of water
exiting the oval is restricted to the capacity of the stormwater drainage. Given the detailed survey
information and hydraulic modelling of the stormwater drainage system, it is assumed that the outlet
capacity is well represented. Provided that the adopted event rainfall is close to that of the actual
event then the modelled flood level in the oval can be used to calibrate the rainfall losses.

Modelled flood levels in Coogee Oval for this event and the January 1999 event were used to
iteratively determine appropriate initial and continuing loss parameters. These were found to be
50mm and 5mm/h respectively for pervious areas and 5mm and 0mm/h for impervious areas. These
values are representative of the whole catchment, but may vary locally. Steep, rocky areas will likely
have reduced soil infiltration and corresponding loss rates. Areas where the soils contain a higher
proportion of loose sand may have a higher loss rate. The adopted rainfall loss parameters are higher
than the standards in AR&R but are appropriate for Coogee Bay, which experiences a high rate of
infiltration. However, despite this high infiltration the study area is still subject to flooding from high
intensity rainfall, as evidenced by the design results in Section 6.2.

5.2.3 Downstream Boundary Condition

The downstream model limit corresponds to the water level in Coogee Bay. This has been set to a
conservative level of 1m AHD but is insignificant in its influence on upstream flood levels. A lower
boundary condition of 0m AHD was tested and found to have no impact on the modelled flood levels.
Additional model boundaries have been included at a few locations where runoff will spill over the
catchment boundary and exit the study area. In these instances constant water level boundaries have
been applied in the 1D domain and QH relationships applied in the 2D domain. The impact of these
boundaries is not significant in determining flood levels within the study area and so no event specific
boundary conditions have been applied.

5.2.4 Adopted Model Parameters

The development of the TUFLOW model requires the assignment of different hydraulic roughness
zones. These zones are delineated from aerial photography and cadastral data identifying different
land-uses (eg. cleared land, scrub, roads, urban areas, etc) for modelling the variation in flow
resistance. Council provided GIS layers representing different land use types including paved
surfaces and building polygons.

The adopted hydraulic roughness (Manning’s ‘n’) applied in the model according to land use type
(material) is shown in Table 5-3. A roughness map for the study area is shown in Figure 5-4
illustrating the subdivision of the model area by land use type. A higher roughness value has been
applied to the materials at shallow depths (<30mm) to represent sheet flow. This will provide a more
realistic hydrologic response of the model to the direct rainfall inputs. For overland flow paths (depths
>150mm) standard roughness values have been applied. Between these two depths a linear
interpolation of the roughness value is applied.

The high Manning’s value for residential/industrial buildings is adopted to account for inundation
within buildings (accounting for storage) but not simulating significant flow through the building. A

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 31

Figure 5-4 Modelled Land Use Map

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 32

lower roughness value is adopted for sheet flow on buildings to allow the rainfall to quickly “runoff” the
building and on to the surrounding land.

Other obstructions to flow paths have been identified from site inspections undertaken during the
model building phase. Preliminary flow path mapping was used to identify potential flow paths and
impedance by existing on ground works such as fences. There are a variety of construction types
whose structural integrity and subsequent flow impedance perform differently in flood events. Further
complication is added by the presence of gaps at the bottom of fences allowing some through flow,
albeit controlled. The general approach has been to only include solid walls as obstructions to flow,
where located along flow paths and deemed to have a significant impact on local flood behaviour.
The locations at which solid wall obstructions have been modelled have been guided by the model
calibration process and includes both the Bowling and Tennis Clubs. The walls have been modelled
at their correct heights, as observed on site and openings in the walls are accounted for where
present. Figure 5-5 shows an example location where solid walls have been modelled and is
presented with the modelled flood depths for the May 2009 calibration event.

Table 5-3 Adopted Hydraulic Roughness Coefficients Based on Land Use

‘n’ for Sheet Flow ‘n’ for Overland Flow

Material
(depths <30mm) (depths >150mm)
Urban Lots 0.1 0.05
Buildings 0.05 1.00
Road Reserve 0.05 0.03
Paved Areas 0.02 0.02
Maintained Grass 0.05 0.03
Unmaintained Grass 0.1 0.05
Light Vegetation 0.15 0.06
Medium Vegetation 0.25 0.08
Dense Vegetation 0.40 0.10
Sand 0.05 0.03

The model parameters that were adopted were shown to provide a reasonable calibration to
observed data and so were not modified. Modifications to the model through the calibration process
were restricted to the rainfall loss parameters and local modifications to the model topography to
correctly represent significant hydraulic controls.

5.2.5 Observed and Simulated Flood Behaviour

There are no official water level records available for calibration within the study area. Alternatively,
calibration data was derived through relevant comments and photographs from community
questionnaire responses and other available resources. Comments relating to flood behaviour were
compiled and compared with modelled outputs for the May 2009 event. These have been presented
in Appendix C. Although most comments received relate to the May 2009 event, some do refer to
other flood events, but have been included for completeness. The reliability of individual flood depth
observations is highly variable. Some observers will be able to better assess flood depths than
others. Also, there is typically no indication as to what the depth is referenced to, i.e. the gutter,

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 33

Figure 5-5 Example Location of Solid Wall Modelling

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 34

pavement, garden, floor, etc. The general pattern and magnitude of flooding indicated by the model
results provides a good match with the comments received from the community. Specific calibration
data for particular flooding hotspots is presented in more detail below.

5.2.5.1 Flood Photographs

For locations where flood photographs are available a more detailed assessment of model calibration
is possible. Several such photographs were received from the community and other sources relating
to the May 2009 event, the locations of which are indicated on Figure 5-6.

With the May 2009 event occurring during the evening the availability of flood photos for Coogee Oval
was not as substantial as for other events. However, some flood photos were available, one of which
is presented in Figure 5-7 (Location A on Figure 5-6). A flood level of around 5.4m AHD has been
estimated from the photograph. However, it is not known whether the photograph was taken at the
peak of the flood and so the peak level in the Oval may have been higher than this. The peak
modelled flood level in Coogee Oval for the May 2009 flood event is around 5.7m AHD.

Another photo is available showing a flood mark on the door of the Senior Citizens Centre, located
just to the west of the Oval on Brook Street and presented in Figure 5-8 (Location B). A flood level of
around 6.8m AHD has been estimated from the photograph, based on the height of the flood mark
above the local kerb level. A similar peak level was modelled by the May 2009 calibration event.

Photographs showing flood debris on the Bowling Green and a flood mark in the Bowling Club
th
basement were included in a May 4 article by the Southern Courier and are presented in Figure 5-9
(Location C). The flood mark shows around a 1m peak flood depth within the basement, which based
from site observations could represent a flood level of around 8.8m AHD. The modelled flood level
here is close to this at around 9.0m AHD. A number of local hydraulic controls were incorporated into
the model at this location to properly represent the local flood behaviour. These were based on site
investigations and include wall structures between the bowling greens and around the tennis courts.
The floor levels of the Bowling Club and Tennis Club are set below ground level.

A photograph showing flood waters flowing from Dolphin Street, down Mount Street and through to
th
the Bowling Club was included in a May 4 article by the Sydney Morning Herald and is presented in
Figure 5-10 (Location D). The depth of the water flowing through the area at the deepest location
appears to be around the bonnet height of a car, or around 0.9m. This would require a flood level of
around 10.8m AHD, which is closely matched by the modelled flood level of around 10.7m AHD.

Further up the catchment, there is a depression located behind the eastern end of Oswald Street
which will fill with this excess runoff and result in significant flood depths. Figure 5-11 (Location E)
shows a flood mark at Oswald Street indicating a flood depth of around 1m. At this location the
ground level is around 22.1m AHD (based on the LiDAR data points), giving a flood level of around
23.1m AHD. The modelled flood level is higher, at around 23.8m AHD. This location is a topographic
depression with no local drainage included in the model. As it is an isolated depression with no
modelled connection to Council’s drainage network, the flood level will be highly sensitive to the
volume of water spilling into the depression. Council have indicated that a drain cover was removed
at this location to help drain the water away and this may account for the lower water level than that
which has been modelled.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 35

Figure 5-6 May 2009 Flood Photograph Locations

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 36

Figure 5-7 May 2009 Flood Photograph: Location A

Figure 5-8 May 2009 Flood Photograph: Location B

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 37

Figure 5-9 May 2009 Flood Photographs: Location C

Figure 5-10 May 2009 Flood Photograph: Location D

A property located behind Oswald Street, on Farnham Avenue, also reported flooding, providing flood
photographs of the front and rear of the property, as shown in Figure 5-12 (Location F). The flood
depths at the front and rear of the property can be judged to be around 0.1m and 0.5m respectively.
Using elevations from the LiDAR data, flood levels of around 24.7m AHD at the front and 24.1m AHD
at the rear of the property are appropriate. The model results indicate flood levels of 24.7m AHD and
23.8m AHD.

Another property, located on the Alison Road side of Farnham Avenue suffered from flood damage
during the May 2009 event, a photograph of which is provided in Figure 5-13 (Location G). It appears

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 38

from the photograph that local scouring has undermined the corner of the property, resulting in failure
of the wall. Runoff from Alison Road will flow down a steep embankment and into the rear of the
properties on Farnham Avenue. There are likely to be locally high velocities, which could result in
scouring such as this. The flood mark of around a 0.7m depth is not evident further along the walled
section of the building, or within internal photographs of the damaged room. The model has
generated an overland flow path in the vicinity of this property.

Figure 5-11 May 2009 Flood Photograph: Location E

Figure 5-12 May 2009 Flood Photographs: Location F

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 39

Figure 5-13 May 2009 Flood Photograph: Location G

5.2.5.2 Rainbow Street Calibration

The model calibration results generally match well with those indicated by comments from the
community and the available flood photographs. Rainbow Street was the only location for which the
performance of the model did not initially correspond to the available information.

The bottom of Rainbow Street is situated within a topographic depression, with a depth of 10m and
catchment area of 26ha. As such there is no natural outlet for catchment runoff and the stormwater
drainage provides the only means for transfer of water out of the catchment. Once the capacity of the
trunk drain is exceeded, the excess runoff will begin to fill the depression. This resulted in significant
flooding being modelled for the May 2009, with a peak level of around 42.1m AHD. Information
received a resident in this location suggested that the highest flood level experienced was closer to
41.6m AHD. It is also known that during the largest floods, the water drained away over the course of
an afternoon It is therefore likely that local drainage infrastructure in the Rainbow Street depression
connects into the Council stormwater network. A 300mm diameter pipe was added to the model to
provide drainage from the lowest-lying lot. This reduced the modelled flood level to around 41.7m
AHD.

5.3 January 1999 Model Calibration

The January 1999 event was used in conjunction with the May 2009 event to calibrate model
parameters such as roughness values and rainfall losses. These parameters are consistent with
those of the May 2009 calibration, as discussed in section 5.2.2 to section 5.2.4.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 40
5.3.1 Rainfall Data

The distribution of rainfall gauge locations in the vicinity of the Coogee Bay catchments was shown in
Figure 2-2 with their respective periods of record shown in Table 2-1. The closest gauge to the study
area is located at Randwick Bowling Club, which records daily rainfall totals. It recorded a total rainfall
th
depth of 73.8mm on 24 January 1999.

The temporal pattern for the January 1999 rainfall event has been derived from the recorded data at
the Little Bay gauge, located 6km to the south of the study area. This data was adopted for use in the
Coogee Oval and Bowling Club Flooding Assessment. The rainfall event recorded at Little Bay lasted
approximately two hours, with a total recorded depth of 114mm. For the purposes of this study the
data has been scaled to provide a total of 73.8mm, as recorded at Randwick Bowling Club. The
recorded rainfall data from Little Bay and the adopted rainfall for the model calibration are shown in
Table 5-4.

Table 5-4 Rainfall Data for the January 1999 Event

Recorded at Little Scaled for Randwick

Time
Bay (mm) Bowling Club (mm)
7:15 3 1.9
7:20 1 0.6
7:25 3 1.9
7:30 3 1.9
7:35 1 0.6
7:40 2 1.3
7:45 2 1.3
7:50 1 0.6
7:55 0 0
8:00 4 2.6
8:05 7 4.5
8:10 7 4.5
8:15 10 6.5
8:20 8 5.2
8:25 9 5.8
8:30 6 3.9
8:35 8 5.2
8:40 8 5.2
8:45 8 5.2
8:50 10 6.5
8:55 5 3.2
9:00 3 1.9

To gain an appreciation of the relative intensity of the May 2009 event, the derived rainfall depths for
various storm durations is compared with the design IFD data for Coogee as shown in Figure 5-14.

The derived depth vs. duration profile for the January 1999 event from the adopted catchment rainfall
shows it generally tracking between the design 20% AEP (5-year ARI) and 10% AEP (10-year ARI)
rainfall for a one to two hour duration.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 41

Figure 5-14 Comparison of Derived January 1999 Rainfall with IFD Relationships

5.3.2 Observed and Simulated Flood Behaviour

The January 1999 event was used in conjunction with the May 2009 event to assess appropriate
rainfall losses for the catchment. The Coogee Oval and Bowling Club Flooding Assessment
references a flood level of around 5.4m AHD in the Oval. This level is also supported by the available
flood photographs taken of the Oval during the flood. Figure 5-15 shows vehicles parked along the
southern side of Coogee Oval. At the deepest area of flooding the water level is at the level of the car
bonnets, or around 0.9m deep. The ground elevation here is around 4.5m AHD, giving an
approximate flood level of 5.4m AHD. The modelled flood level in Coogee Oval for the January 1999
calibration event is 5.5m AHD.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 42

Figure 5-15 January 1999 Flood Photograph of Coogee Oval

5.4 October 1959 Model Validation

The October 1959 event is the largest on record within the study area. There is little data available to
calibrate the model, but an indicative flood depth in Coogee Oval of 10 feet (referenced in the
Mayoral records) was used to validate the performance of the calibrated model.

5.4.1 Rainfall Data

The distribution of rainfall gauge locations in the vicinity of the Coogee Bay catchments was shown in
Figure 2-2 with their respective periods of record shown in Table 2-1. The closest gauge to the study
area is located at Randwick Bowling Club, which records daily rainfall totals. It recorded a total rainfall
th
depth of 265.4mm on 29 October 1959.

The Mayor’s minutes documenting flood damages from this event indicate a storm duration of around
3.5 hours. With no better available information, the three hour design event provided in AR&R has
been adopted as the temporal pattern for the October 1959 event.

To gain an appreciation of the relative intensity of the October 1959 event, the derived rainfall depths
for various storm durations is compared with the design IFD data for Coogee as shown in Figure
5-16.

The derived depth vs. duration profile for the October 1959 event from the adopted catchment rainfall
shows it far exceeding the 0.2% AEP (500-year ARI) rainfall, being almost twice as large as the 1%
AEP (100-year ARI) rainfall.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
MODEL CALIBRATION 43

Figure 5-16 Comparison of Derived October 1959 Rainfall with IFD Relationships

5.4.2 Observed and Simulated Flood Behaviour

The only calibration data available for the October 1959 event is a reference in the Mayor’s minutes
to a flood depth of 10 feet on Coogee Oval (also referenced as 12 feet within a newspaper). A depth
of this order would relate to a flood level of around 7.6m AHD. The model predicted flood level in
Coogee Oval is 7.3m AHD. There is much uncertainty regarding the water level record, adopted
rainfall pattern and changes to the catchment characteristics and so an exact match is not expected
in this instance. However, the similarity of the reported and modelled flood depths within Coogee
Oval suggests a reasonable model prediction.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 44

6 DESIGN FLOOD CONDITIONS

Design floods are hypothetical floods used for planning and floodplain management investigations.
They are based on having a probability of occurrence specified either as:
 Annual Exceedance Probability (AEP) expressed as a percentage; or
 Average Recurrence Interval (ARI) expressed in years.

This report uses the AEP terminology. Refer to Table 6-1 for a definition of AEP and the ARI
equivalent.

Table 6-1 Design Flood Terminology

1 2
ARI AEP Comments
500 years 0.2% A hypothetical flood or combination of floods which
represent the worst case scenario likely to occur on
average once every 500 years.
200 years 0.5% As for the 0.2% AEP flood but with a 0.5% probability or
200 year return period.
100 years 1% As for the 0.2% AEP flood but with a 1% probability or
100 year return period.
20 years 5% As for the 0.2% AEP flood but with a 5% probability or
20 year return period.
5 years 20% As for the 0.2% AEP flood but with a 20% probability or
5 year return period.
Extreme Flood / A hypothetical flood or combination of floods which
3
PMF represent an extreme scenario.
1 Average Recurrence Interval (years)
2 Annual Exceedance Probability (%)
3 A PMF (Probable Maximum Flood) is not necessarily the same as an Extreme Flood.

In determining the design floods it is necessary to take into account the critical storm duration of the
catchment. Small catchments are more prone to flooding during short duration storms while for large
catchments longer durations will be more critical. For example, considering the relatively small size
of the study area catchments, they are potentially more prone to higher flooding from intense storms
extending over a few hours rather than a couple of days.

6.1 Design Rainfall

Design rainfall parameters are derived from standard procedures defined in AR&R (2001) which are
based on statistical analysis of recorded rainfall data across Australia. The derivation of location
specific design rainfall parameters (e.g. rainfall depth and temporal pattern) for Coogee Bay is
presented below.

6.1.1 Rainfall Depths

Design rainfall depth is based on the generation of intensity-frequency-duration (IFD) design rainfall
curves utilising the procedures outlined in AR&R (2001). These curves provide rainfall depths for
various design magnitudes (up to the 1% AEP) and for durations from 5 minutes to 72 hours.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 45

The Probable Maximum Precipitation (PMP) is used in deriving the Probable Maximum Flood (PMF)
event. The theoretical definition of the PMP is “the greatest depth of precipitation for a given duration
that is physically possible over a given storm area at a particular geographical location at a certain
4 7
time of year” (AR&R, 2001). The ARI of a PMP/PMF event ranges between 10 and 10 years and is
beyond the “credible limit of extrapolation”. That is, it is not possible to use rainfall depths determined
for the more frequent events (100 year ARI and less) to extrapolate the PMP. The PMP has been
estimated using the Generalised Short Duration Method (GSDM) derived by the Bureau of
Meteorology.

A range of storm durations were modelled in order to identify the critical storm duration for design
event flooding in the catchment. Design durations considered included the 0.25-hour, 0.5-hour, 0.75-
hour, 1-hour, 1.5-hour, 2-hour, 3-hour, 4.5-hour, 6-hour and 9-hour durations.

Table 6-2 shows the average design rainfall intensities based on AR&R adopted for the modelled
events.

Table 6-2 Average Design Rainfall Intensities (mm/hr)

Duration Design Event Frequency

(hours) 20% AEP 5% AEP 1% AEP 0.5% AEP 0.2% AEP
0.25 112 145 187 205 229
0.5 82 107 140 153 172
0.75 66 87 114 127 143
1 56 74 97 110 123
1.5 43.9 58 76 84 96
2 36.1 47.8 63 69 79
3 27.4 36.3 47.9 53 60
4.5 20.8 27.3 36.0 39.8 45.0
6 17 22.4 29.5 32.7 36.9
9 12.7 16.7 22 24.7 27.9

Areal Reduction Factor

The areal reduction factor takes into account the unlikelihood that larger catchments will experience
rainfall of the same design intensity (eg 1% AEP) over the entire area. Areal reduction factors
typically apply to catchments significantly larger than those at Coogee Bay and no reduction factor is
2
required for the study area catchment of 2.9km .

6.1.2 Temporal Patterns

The IFD data presented in Table 6-2 provides for the average intensity that occurs over a given storm
duration. Temporal patterns are required to define what percentage of the total rainfall depth occurs
over a given time interval throughout the storm duration. The temporal patterns adopted in the current
study are based on the standard patterns presented in AR&R (2001).

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 46

The same temporal pattern has been applied across the whole catchment. This assumes that the
design rainfall occurs simultaneously across each of the modelled sub-catchments. The direction of a
storm and relative timing of rainfall across the catchment may be determined for historical events if
sufficient data exists, however, from a design perspective the same pattern across the catchment is
generally adopted.

6.1.3 Rainfall Losses

The rainfall losses adopted for the design floods were the same as those used for model calibration
and verification. For the initial and continuing rainfall losses, values of 50mm and 5mm/h were used
for pervious portions of the catchment, with 5mm and 0mm/h being used for impervious areas. These
are higher than those generally recommended for design event losses in AR&R (2001), but are
appropriate for well-draining sandy soils such as those of the Coogee Bay catchment. A sensitivity
test using standard AR&R losses has been carried out and is discussed in Section 6.3.

6.2 Design Flood Results

A range of design event durations were simulated to determine the critical duration for flooding
throughout the study area. In general, the model simulations indicated the peak water levels in most
areas corresponded to the 90-minute or two-hour durations. For local depression storage areas the
critical conditions are flood volume driven for which a longer duration is required. In these areas the
9-hour duration (which was the longest duration considered) is the critical duration. The design flood
results are the maximum condition from the combined 90-minute, 2-hour and 9-hour duration events,
for which the distribution at the 1%AEP event is presented in Figure 6-1. For the PMF event, the
critical durations (and those from which the results have been derived) are the 15-minute, 30-minute,
45-minute and 4.5-hour durations.

The design flood results are presented in a flood mapping series in Appendix A. For the key
simulated design events including the 20% AEP (5-year ARI), 5% AEP (20-year ARI), 1% AEP (100-
year ARI), 0.5% AEP (200-year ARI), 0.2% AEP (500-year ARI), and PMF events, a map of peak
flood depth, velocity and hydraulic hazard is presented covering the modelled area.

6.2.1 Peak Flood Levels, Depths and Velocities

The flood extents for the 1% AEP design event have been presented in Figure 6-2. The figure also
shows the distribution of 15 reporting locations that are referenced in the design flood level summary
Table 6-3. The alignment of a flood long section (presented in Figure 6-3) is also indicated on the
figure. Five distinct sub-catchment areas can be distinguished from the flood model results and the
boundaries of these have also been identified on Figure 6-2. The main sub-catchment is that draining
to Coogee Oval, which is some 200ha in size. The Rainbow Street sub-catchment is located at the
south of the study area. It is around 26ha in size and drains to a depression centred around 303
Rainbow Street. Located between these two sub-catchments is a smaller 19ha sub-catchment
centred around the Havelock Avenue alignment, which drains to Coogee Bay near the Surf Life-
saving Club. The Beach Street sub-catchment covers only 10ha and includes the urban area around
Beach Street and the Goldstein Reserve, draining to Coogee Bay via Coogee Beach. The remaining
sub-catchment is around 16ha in size and covers the headland area to the south of Coogee Beach
that drains directly to Coogee Bay over the cliffs in Grant Reserve and Trenerry Reserve.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 47

Figure 6-1 Critical Durations for the 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 48

Figure 6-2 Distribution of Design Result Reporting Locations

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 49

Table 6-3 Summary of Design Flood Levels

Design Event Flood Level (m AHD)

ID Location
20% AEP 5% AEP 1% AEP 0.5% AEP 0.2% AEP PMF
1 Coogee Oval 5.4 6.1 6.9 7.0 7.1 7.8
2 Brook Street 6.6 6.8 7.0 7.1 7.2 8.1
3 Coogee Tennis Club 7.9 8.1 8.2 8.3 8.4 9.2
4 Coogee Bowling Club 8.9 9.0 9.1 9.1 9.2 10.1
5 Dolphin Street 12.2 12.3 12.4 12.5 12.6 13.5
6 Carrington Road 17.2 18.4 19.0 19.2 19.3 20.3
7 Oswald Street 23.8 23.9 24.1 24.2 24.2 25.2
8 Glen Avenue 28.2 29.1 30.3 30.7 31.4 32.6
9 32 Coogee Street 20.1 20.2 20.7 20.8 20.9 21.6
10 9 Coogee Street 25.3 25.8 26.0 26.1 26.2 26.9
11 Bardon Park 12.5 12.7 13.0 13.1 13.3 13.9
12 Mount Street 13.8 14.4 15.0 15.1 15.2 16.0
13 Pauling Avenue 17.2 17.7 18.0 18.1 18.2 18.9
14 Leeton Avenue 19.4 19.6 19.8 19.8 19.9 20.1
15 Rainbow Street 42.1 42.4 42.8 43.0 43.3 47.6

Figure 6-3 Design Flood Level Long Section

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 50

The main flowpath of the Coogee Oval catchment is aligned with Dolphin Street from Carrington
Road to Arden Street. This corresponds to the areas designated as “Water Reserve” on a historic
map of Randwick by Higinbotham and Robinson, dating from c.1885. This map is shown in Figure
6-4, overlain by the 1% AEP modelled flood depth results. The creek alignment along Dolphin Street
and through the Coogee Bowling Club site is evident and corresponds well with the alignment of the
modelled overland flow path. Flood depths in this area are driven by local obstructions to the flow.
Depressions in Dolphin Street and Brook Street flood to depths of around 0.5m for the 20% AEP
event, rising to around 1m for the 1% AEP event. Flood depths at the bowling club (at green level)
are around 0.3m for the 20% AEP event and 0.5m for the 1% AEP event. The deepest flood depths
are experienced in the Coogee Oval depression, which is separated from the sea by the higher
ground of Arden Street and Goldstein Reserve. Here flood depths are around 0.5m for the 5% AEP
event, rising to over 2m for the 1% AEP flood. Properties that are likely to be impacted by flooding
include those located immediately to the south of the oval, those near the Dolphin Street – Brook
Street intersection and the bowling and tennis clubs.

There are two main tributaries that contribute to the main flowpath, both aligned in a roughly north-
south direction. The first tributary flowpath is generated from the catchment area to the north of
Coogee Bay Road and west of Carrington Road. It includes a remnant creek alignment upstream of
Alison Road. This was marked on the historic map and still exists today. Downstream of Alison Road
it is conveyed through a series of underground culverts. The flood depths upstream of Alison Road
are driven by the volume of runoff and are around 0.3m for the 20% AEP event, rising to around 2m
by the 1% AEP event. The events up to the 0.2% AEP are not sufficient enough to overtop Alison
Road. Downstream of Alison Road the flood depths are locally high in natural depressions in the
topography. These depressions correspond to the creek and flow path alignments identified in 1885
The three most significant of these depressions are:

 Upstream of the Courland Street – Oswald Street intersection. Flood depths here are over 1.5m
for the 20% AEP event and almost 2m for the 1% AEP event;

 Between Clyde Street and Coogee Street. Flood depths here are around 0.5m for the 20% AEP
event and 1.2m for the 1% AEP event;

 Upstream of the Dolphin Street – Carrington Road intersection. Flood depths here are up to 2m
for the 20% AEP event and almost 4m for the 1% AEP event.

Properties that are likely to be impacted by flooding include those at the locations mentioned above,
some additional properties located on Coogee Street and some located between Coogee Bay Road
and Dolphin Street.

The second tributary flowpath is generated from the catchment area to the north of Bream Street and
east of Carrington Road. It contains two main flowpaths which converge on Bardon Park. The
approximate alignment of these is marked on the historic map. One originates from Clovelly Road,
progressing through Marcel Avenue and then along Pauling Avenue. The other originates from
around Division Street, flowing along Leeton Avenue. Flood depths in this area are most significant
upstream of the Abbott Street – Mount Street intersection, where flood depths of around 1m are
modelled for the 20% AEP event, rising to 2m by the 1% AEP event. Flooding is also significant in a
depression at the bottom of Pauling Avenue and on Bardon Park. Flood depths are around 1m for
both location in the 1% AEP event, being 0.2m and 0.5m respectively for the 20% AEP event.
Another flowpath into Bardon Park is also significant. It is aligned approximately along Hill Lane and

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 51

Figure 6-4 Comparison of Design Flood Results with Historic Map

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 52

Smithfield Avenue. Flood depths are typically shallow for the 20% AEP event, rising to over 0.5m for
the 1% AEP event. Properties that are likely to be impacted by flooding include those located
between Alison Road and Abbott Street and those located along Smithfield Avenue or between
Arcadia Street and Hill Lane.

Elsewhere in the Coogee Oval catchment, flooding is generally less substantial. Perhaps the most
significant flooding other than the issues already discussed occurs between Carr Street and Dudley
Street, where local flood depths can exceed 0.5m.

The Rainbow Street catchment consists primarily of two overland flow alignments. One begins at the
Oberon Street – Hendy Avenue intersection and flows east, the other from the Dudley Street - Mount
Street intersection and flowing south. The flood depths on these alignments are typically shallow but
are locally higher than 0.5m in topographic depressions. The flood waters collect in the bottom of the
catchment depression on Rainbow Street, which is relieved to some extent by the capacity of the
stormwater drainage network. The stormwater pipe runs along Rainbow Street and under Blenheim
Park, where it is buried over 30m underground. The flood depths in the Rainbow Street depression
are typically around 0.3m for the 20% AEP event and 1m for the 1% AEP event. The depths are
around 1m greater than this in the deepest part of the depression. Properties likely to be impacted by
flooding are located in the areas described above, predominantly on Oberon Street and Rainbow
Street, and particularly those situated within the Rainbow Street depression.

Elsewhere in the study area the flow paths are confined largely to the roadways, with only some
localised flooding issues.

Flood velocities within the study area vary significantly due to local conditions. They are typically
lower than 0.5m/s in flood storage locations, such as the depressions but are significantly higher in
the roadways. Typical flood velocities in the road alignments are over 1m/s for the 20% AEP event,
rising to over 2m/s for the 1% AEP event.

6.2.2 Hydraulic Categorisation

There are no prescriptive methods for determining what parts of the floodplain constitute floodways,
flood storages and flood fringes. Descriptions of these terms within the Floodplain Development
Manual (NSW Government, 2005) are essentially qualitative in nature. Of particular difficulty is the
fact that a definition of flood behaviour and associated impacts is likely to vary from one floodplain to
another depending on the circumstances and nature of flooding within the catchment.

The hydraulic categories as defined in the Floodplain Development Manual are:

 Floodway - Areas that convey a significant portion of the flow. These are areas that, even if
partially blocked, would cause a significant increase in flood levels or a significant redistribution
of flood flows, which may adversely affect other areas.

 Flood Storage - Areas that are important in the temporary storage of the floodwater during the
passage of the flood. If the area is substantially removed by levees or fill it will result in elevated
water levels and/or elevated discharges. Flood Storage areas, if completely blocked would cause
peak flood levels to increase by 0.1m and/or would cause the peak discharge to increase by
more than 10%.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 53

 Flood Fringe - Remaining area of flood prone land, after Floodway and Flood Storage areas
have been defined. Blockage or filling of this area will not have any significant affect on the flood
pattern or flood levels.

A number of approaches were considered when attempting to define flood impact categories across
Coogee Bay. Approaches to define hydraulic categories that were considered for this assessment
included partitioning the floodplain based on:

 Peak flood velocity;

 Peak flood depth;

 Peak velocity * depth (sometimes referred to as unit discharge);

 Cumulative volume conveyed during the flood event; and

 Combinations of the above.

The definition of flood impact categories that was considered to best fit the application within Coogee
Bay was ultimately provided by Council and was based on a combination of velocity*depth, velocity
and depth parameters. The adopted hydraulic categorisation is defined in Table 6-4.

The hydraulic category map is included in Appendix A. It is also noted that mapping associated with
the flood hydraulic categories may be amended in the future, at a local or property scale, subject to
appropriate analysis that demonstrates no additional impacts (e.g. if it is to change from floodway to
flood storage).

Table 6-4 Hydraulic categories

Floodway Defined at the 1% AEP Areas and flowpaths where a significant proportion
event using the following of floodwaters are conveyed (including all bank-to-
criteria: bank creek sections).

Velocity * Depth > 0.3 OR

Velocity > 0.5 m/s

Flood Storage Defined at the 1% AEP Areas where floodwaters accumulate before being
event where Depth > 0.15 conveyed downstream. These areas are important
metres for detention and attenuation of flood peaks.

Flood Fringe Defined at the PMF event Areas that are low-velocity backwaters within the
where Depth > 0.15 floodplain. Filling of these areas generally has little
metres consequence to overall flood behaviour.

6.2.3 Provisional Hazard

The NSW Government’s Floodplain Development Manual (2005) defines flood hazard categories as
follows:

 High hazard – possible danger to personal safety; evacuation by trucks is difficult; able-bodied
adults would have difficulty in wading to safety; potential for significant structural damage to
buildings; and

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 54

 Low hazard – should it be necessary, trucks could evacuate people and their possessions; able-
bodied adults would have little difficulty in wading to safety.

The key factors influencing flood hazard or risk are:

 Size of the Flood

 Rate of Rise - Effective Warning Time

 Community Awareness

 Flood Depth and Velocity

 Duration of Inundation

 Obstructions to Flow

 Access and Evacuation

The provisional flood hazard level is often determined on the basis of the predicted flood depth and
velocity. This is conveniently done through the analysis of flood model results. A high flood depth will
cause a hazardous situation while a low depth may only cause an inconvenience. High flood
velocities are dangerous and may cause structural damage while low velocities have no major threat.

Figures L1 and L2 in the Floodplain Development Manual (NSW Government, 2005) are used to
determine provisional hazard categorisations within flood liable land. These figures are reproduced in
Figure 6-5.
Excessive
Velocity

High
2.0
2.0 Hazard
Haz
ard
Leve

1.5
1.5
Vehicles unsta

Da p
Wad

m os

l De
ag s

Velocity (m/s)
e ibl
Velocity (m/s)

ing u

to e

pen
lig fro
ht m
nsaf

den

1.0
st he

1.0
ru re
e fro

ct
ble from here

to
ur
es

n Si
m he

te C
re

0.5
VxD 0.5
ond

=1
Low
ition

Excessive
Depth Hazard
s

0.0 0.0
0.0 0.5 1.0 1.5 2.0 0.0 0.5 1.0 1.5 2.0

Depth (m) Depth (m)

Velocity Depth Relationships Provisional Hazard Categories

c:\aac4\n0003\d101.vsd c:\aac4\n0003\d102.vsd

Figure 6-5 Provisional Flood Hazard Categorisation

The provisional hydraulic hazard is included in the mapping series for each simulated design event
provided in Appendix A.

6.3 Sensitivity Tests

A number of sensitivity tests have been undertaken on the modelled flood behaviour in Coogee Bay.
These tests consider blockage of the stormwater drainage system, reduction in rainfall losses and

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 55

increased sea level. The details of the sensitivity tests and results of the modelled scenarios are
presented below.

6.3.1 Stormwater Drainage Blockage

For the overland flows, the blockage considerations are mainly associated with the underground
stormwater drainage network.

A 100% blockage assumption was applied to all pipes in the modelled subsurface network, thereby
eliminating pipe flow. This results in all of the runoff remaining in the 2D model domain as overland
flow. Blockage scenarios were undertaken using the 1% AEP event for both the 90-minute, 2-hour
and 9-hour storm durations. The results of the blockage scenario simulation are presented in Figure
6-6 and Table 6-5.

Table 6-5 Summary of Blockage Sensitivity Results

Modelled Flood Level (m AHD)

Flood Level
ID Location Baseline 1% AEP with Increase (m)
1% AEP Blockages
1 Coogee Oval 6.9 7.1 0.2
2 Brook Street 7.0 7.2 0.2
3 Coogee Tennis Club 8.2 8.4 0.2
4 Coogee Bowling Club 9.1 9.2 0.1
5 Dolphin Street 12.4 12.6 0.2
6 Carrington Road 19.0 19.3 0.3
7 Oswald Street 24.1 24.5 0.4
8 Glen Avenue 30.3 32.0 1.7
9 32 Coogee Street 20.7 21.0 0.3
10 9 Coogee Street 26.0 26.3 0.3
11 Bardon Park 13.0 13.3 0.3
12 Mount Street 15.0 15.3 0.3
13 Pauling Avenue 18.0 18.2 0.2
14 Leeton Avenue 19.8 19.9 0.1
15 Rainbow Street 42.8 44.6 1.8

The key findings of the stormwater drainage blockage sensitivity test are summarised below:

 Blockage impacts are greatest upstream of significant topographic obstructions, which restrict the
progression of overland flow.

 The largest flood level increase was around 1.8m above base case conditions. This increase
occurs in the depression of the Rainbow Street catchment. Here the only flow outlet is via the
pipe network and so there is a significant increase in flood level. The peak flood level represents
the total volume of runoff from the local catchment.

 A flood level increase of 1.7m was simulated in the remnant creek line upstream of Alison Road.
Here the road forms around a 3m high obstruction, serviced by a 2.5m x 2.0m cross-drainage

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 56

Figure 6-6 Impact of Stormwater Pipe Blockage on 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 57

capacity. The blockage of the stormwater drainage network provides no outlet from the upstream
depression until a flood level of over 31.7m AHD is reached. Overtopping of Alison Road occurs
during the blockage scenario but not under baseline conditions.

 Flood level increases of around 0.2m to 0.3m were typically simulated in areas located along the
major flowpath alignments, i.e. the previous creek/gully locations. Larger increases were not
experienced for any of the topographic depressions other than those specifically mentioned
above. This is because the baseline 1% AEP event is substantial enough to fill the remaining
depressions.

 Elsewhere, blockage impacts were 10cm or less. Areas where larger impacts were simulated
have been marked on Figure 6-6.

6.3.2 Sea Level

The model developed for this study adopted a fixed water level boundary of 1m AHD to represent the
sea level in Coogee Bay. This is similar to a conservative spring tide level, but is unlikely to have any
impact of the model results in the study area as the lowest-lying areas are situated a few metres
above this at around 4.5m AHD. Higher sea-level conditions could coincide with a flood event if for
example local catchment flooding occurred during a significant coastal flood. Future climate change
predictions also suggest a 0.9m increase in sea-levels by 2100. To test the influence of higher sea-
level conditions on flood levels within the study area an extreme water level of 3.5m AHD was
adopted. This would be similar to a 0.5% coastal flood event with a 0.9m increase for climate change.
The results of the increased sea-level scenario simulation are presented in Figure 6-7 and Table 6-6,

Table 6-6 Summary of Sea-level Sensitivity Results

Modelled Flood Level (m AHD)

Flood Level
ID Location Baseline 1% AEP with 3.5m Increase (m)
1% AEP AHD Sea-level
1 Coogee Oval 6.9 7.0 0.1
2 Brook Street 7.0 7.0 0.0
3 Coogee Tennis Club 8.2 8.2 0.0
4 Coogee Bowling Club 9.1 9.1 0.0
5 Dolphin Street 12.4 12.4 0.0
6 Carrington Road 19.0 19.0 0.0
7 Oswald Street 24.1 24.1 0.0
8 Glen Avenue 30.3 30.3 0.0
9 32 Coogee Street 20.7 20.7 0.0
10 9 Coogee Street 26.0 26.0 0.0
11 Bardon Park 13.0 13.0 0.0
12 Mount Street 15.0 15.0 0.0
13 Pauling Avenue 18.0 18.0 0.0
14 Leeton Avenue 19.8 19.8 0.0
15 Rainbow Street 42.8 42.8 0.0

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 58

Figure 6-7 Impact of Increased Sea-level on 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 59

The key findings of the sea-level rise sensitivity test are summarised below:

 The sea-level conditions have a minimal impact of upstream flooding conditions;

 Coogee Oval and Goldstein Reserve is the only area (other than the beach itself) that is
impacted by the sea-level rise scenario. Here a flood level increase of around 0.1m is modelled,
due to the slight reduction in drainage outlet capacity;

 Elsewhere the impacts are negligible.

6.3.3 Rainfall Losses

The rainfall losses that were determined through the calibration process found an initial loss of 50mm
to be appropriate for the study area. This is outside of the normal range recommended by AR&R, but
is reasonable for well-draining sandy soils. The sensitivity of the 1% AEP design results was tested
by adopting the standard recommended initial loss for eastern NSW of 15mm. The results of the
decreased initial rainfall loss scenario simulation are presented in Figure 6-8 and Table 6-7.

Table 6-7 Summary of Initial Loss Sensitivity Results

Modelled Flood Level (m AHD)

Flood Level
ID Location Baseline 1% AEP with Increase (m)
1% AEP 15mm Loss
1 Coogee Oval 6.9 7.0 0.1
2 Brook Street 7.0 7.2 0.2
3 Coogee Tennis Club 8.2 8.4 0.2
4 Coogee Bowling Club 9.1 9.2 0.1
5 Dolphin Street 12.4 12.6 0.2
6 Carrington Road 19.0 19.2 0.2
7 Oswald Street 24.1 24.2 0.1
8 Glen Avenue 30.3 31.2 0.9
9 32 Coogee Street 20.7 20.9 0.2
10 9 Coogee Street 26.0 26.2 0.2
11 Bardon Park 13.0 13.3 0.3
12 Mount Street 15.0 15.2 0.2
13 Pauling Avenue 18.0 18.2 0.2
14 Leeton Avenue 19.8 19.8 0.0
15 Rainbow Street 42.8 43.2 0.4

The key findings of the reduced initial rainfall loss sensitivity test are summarised below:

 As for the blockage scenario, initial loss impacts are greatest upstream of significant topographic
obstructions, which restrict the progression of overland flow. This is because the peak flood levels
in such locations are driven by volumes rather than peak flows;

 The largest flood level increase was around 0.9m above base case conditions. This increase
occurs in the remnant creek line upstream of Alison Road. Here the road forms around a 3m high

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 60

Figure 6-8 Impact of Decreased Initial Rainfall Loss on 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 61

obstruction, serviced by a 2.5m x 2.0m cross-drainage capacity. The reduction in initial loss of
35mm significantly increases the flood level here, but overtopping of Alison Road does not occur.

 A flood level increase of 0.4m was simulated in the depression of the Rainbow Street catchment.
This is due to the increased runoff volume from the local catchment which drives the peak flood
level in the depression.

 Flood level increases of around 0.2m were typically simulated in areas located along the major
flowpath alignments, i.e. the previous creek/gully locations. Larger increases were not
experienced for any of the topographic depressions other than those specifically mentioned
above. This is because the baseline 1% AEP event is substantial enough to fill the remaining
depressions.

 Elsewhere, initial loss impacts were 10cm or less. Areas where larger impacts were simulated
have been marked on Figure 6-8.

6.3.4 Climate Change

Current practice in floodplain management generally requires consideration of the impact of potential
climate change scenarios on design flood conditions. For the Coogee Bay catchments this requires
investigation of increases in design rainfall intensities. Typically climate change sensitivity tests in this
regard consider increases in design rainfall intensity of 10%, 20% or 30% in accordance with
DECCW Practical Consideration of Climate Change Guideline for Floodplain Risk Management
(2007).

Specific climate change simulations were not undertaken as part of this study but the 0.5% AEP and
0.2% AEP design results can be used, when compared to the 1% AEP results, to give an indication
as to the potential magnitude of climate change impacts. The 0.5% AEP design rainfall depth is
approximately 10% greater than that of the 1% AEP and the 0.2% AEP design rainfall depth is
approximately 25% greater than that of the 1% AEP. Comparing results of the 0.5% AEP and 0.2%
AEP events to the 1% AEP event is comparable to considering a 10% or 25% increase in design
rainfall depths to the 1% AEP event respectively. As discussed previously, these are similar
increases typically considered for climate change assessments.

The assessment of the 10% increase in rainfall intensity scenario is presented in Figure 6-9 and
Table 6-8. The assessment of the 25% increase in rainfall intensity scenario is presented in Figure
6-10 and Table 6-9. The key findings of the potential climate change impacts are summarised below:

 As for the blockage and initial loss scenarios, increased rainfall impacts are greatest upstream of
significant topographic obstructions, which restrict the progression of overland flow. This is
because the peak flood levels in such locations are driven by volumes rather than peak flows;

 The largest flood level increase was around 0.4m above base case conditions for the 10% rainfall
increase scenario and 1.1m for the 25% rainfall increase scenario. This increase occurs in the
remnant creek line upstream of Alison Road. Here the road forms around a 3m high obstruction,
serviced by a 2.5m x 2.0m cross-drainage capacity. The increased rainfall considerations of
+10% and +25% significantly increases the flood level here, but overtopping of Alison Road does
not occur.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 62

Figure 6-9 Impact of 10% Rainfall Increase on 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 63

Figure 6-10 Impact of 25% Rainfall Increase on 1% AEP Event

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 64

Table 6-8 Summary of 10% Increased Rainfall Assessment

Modelled Flood Level (m AHD) Flood Level

ID Location
1% AEP 0.5% AEP Increase (m)

1 Coogee Oval 6.9 7.0 0.1

2 Brook Street 7.0 7.1 0.1
3 Coogee Tennis Club 8.2 8.3 0.1
4 Coogee Bowling Club 9.1 9.1 0
5 Dolphin Street 12.4 12.5 0.1
6 Carrington Road 19.0 19.2 0.2
7 Oswald Street 24.1 24.2 0.1
8 Glen Avenue 30.3 30.7 0.4
9 32 Coogee Street 20.7 20.8 0.1
10 9 Coogee Street 26.0 26.1 0.1
11 Bardon Park 13.0 13.1 0.1
12 Mount Street 15.0 15.1 0.1
13 Pauling Avenue 18.0 18.1 0.1
14 Leeton Avenue 19.8 19.8 0
15 Rainbow Street 42.8 43.0 0.2

Table 6-9 Summary of 25% Increased Rainfall Assessment

Modelled Flood Level (m AHD) Flood Level

ID Location
1% AEP 0.2% AEP Increase (m)

1 Coogee Oval 6.9 7.1 0.2

2 Brook Street 7.0 7.2 0.2
3 Coogee Tennis Club 8.2 8.4 0.2
4 Coogee Bowling Club 9.1 9.2 0.1
5 Dolphin Street 12.4 12.6 0.2
6 Carrington Road 19.0 19.3 0.3
7 Oswald Street 24.1 24.2 0.1
8 Glen Avenue 30.3 31.4 1.1
9 32 Coogee Street 20.7 20.9 0.2
10 9 Coogee Street 26.0 26.2 0.2
11 Bardon Park 13.0 13.3 0.3
12 Mount Street 15.0 15.2 0.2
13 Pauling Avenue 18.0 18.2 0.2
14 Leeton Avenue 19.8 19.9 0.1
15 Rainbow Street 42.8 43.3 0.5

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD CONDITIONS 65

 A flood level increase of 0.2m occurs in the depression of the Rainbow Street catchment for the
+10% rainfall scenario, with a 0.5m increase for the +25% rainfall scenario. This is due to the
increased runoff volume from the local catchment which drives the peak flood level in the
depression.

 Flood level increases of around 0.1m and 0.2m (for the +10% and +25% scenarios respectively)
are typical in areas located along the major flowpath alignments, i.e. the previous creek/gully
locations. Larger increases were not experienced for any of the topographic depressions other
than those specifically mentioned above. This is because the baseline 1% AEP event is
substantial enough to fill the remaining depressions.

 Elsewhere, increased rainfall impacts are 10cm or less. Areas where larger impacts were
simulated have been marked on Figure 6-9 and Figure 6-10.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
CONCLUSIONS 66

7 CONCLUSIONS
The objective of the study was to undertake a detailed flood study of the local overland flow
catchments of Coogee Bay and establish models as necessary for design flood level prediction.

In completing the flood study, the following activities were undertaken:

 Collation of historical flood information for the study area;

 Consultation with the community to acquire additional historical flood information;

 Development of a 2D/1D hydrodynamic model (using TUFLOW software) to simulate hydrology

and flood behaviour in the catchment;

 Calibration of the developed model using the available flood data, primarily relating to the May
2009 event;

 Prediction of design flood conditions in the catchments and production of design flood mapping
series.

In simulating the design flood conditions for the local catchments in the study area, the following
locations were identified as potential problem areas in relation to flood inundation extent and property
affected:

 Alfreda Street – the properties located immediately to the south of Coogee Oval are liable to be
flooded during significant flood events. This will be from flood waters ponding behind Arden
Street and Goldstein Reserve. The car parking areas situated around the Oval will also be
affected (as well as the oval itself);

 Brook Street – properties situated along Brook Street near the Dolphin Street intersection are in
a local depression which is separated from the oval to the east by an area of higher ground;

 Coogee Bowling Club and Tennis Club – these clubs are situated on the traditional creek
alignment of the Coogee Oval catchment. As such they will experience frequent flooding and
significant flooding during major flood events. Some of the property is situated below ground
level, which will exacerbate the problem;

 Coogee Street and Dolphin Street – the properties situated between these two roads and to the
west of Carrington Road are located on an old creek alignment. Locally deep flooding will occur
on some properties, particularly adjacent to Carrington Road, which is a substantial obstruction
to overland flows. Some properties to the south of Dolphin Street will also be impacted;

 Clyde Street – properties located here between Oswald Street and Coogee Street are impacted
by flood levels building in local depressions, particularly behind Coogee Street, which forms a
significant obstruction to overland flow and has limited cross-drainage capacity;

 Oswald Street – properties located between here and Alison Road are on a traditional creek
alignment and as such are at risk of flooding when the available stormwater drainage capacity is
exceeded, particularly those located in the depression adjacent to Oswald Street;

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
CONCLUSIONS 67

 Abbott Street – properties situated between Abbott Street and Alison Road are located within an
historic creek alignment and are liable to flooding, particularly those in a depression that sits
behind Mount Street;

 Bardon Park – properties situated on around Bardon Park are within overland flow alignments,
including those located between the park and Leeton Avenue and those to the east along
Smithfield Avenue;

 Carr Street – properties situated between here and Dudley Street are located on an overland
flowpath and may experience locally deep flooding in depressions;

 Oberon Street – properties situated south of Oberon Street, between Hendy Avenue and Mount
Street are located on an overland flowpath and may experience locally deep flooding in
depressions. This is also the case for properties situated to the north of Oberon, around Cox
Street and Bay Street;

 Rainbow Street – a large depression in the topography (some 10m deep) is situated between
Rainbow Street and Marian Street, and to the east of Brook Street. The location is liable to
significant flood depths when the available capacity of the stormwater drainage is exceeded.

The flooding issues within the Coogee Bay study area are largely restricted to locations which were
naturally creek/gully lines, but are now occupied by urban development. Along these alignments
natural depressions in the topography and those created by manmade obstructions, such as roads
and other land-raising activities, fill to significant depths during major design flood events. Stormwater
drainage networks are typically designed to around a 20% AEP standard. Once the available
drainage capacity is exceeded the depressions will quickly fill with excess runoff, acting as local flood
storages. For large flood events such as the 1% AEP these storages are filled to capacity and water
flooding progresses via the lowest adjoining point in the topography. This type of flood behaviour is
widespread throughout the study area.

Most of the study area drains to two large depressions – Coogee Oval and Rainbow Street. The
surrounding higher land prevents progression of overland flow and flood waters rise as the available
storage volume is filled. Both are serviced by Council’s stormwater network but when the drainage
capacity is exceeded the flood levels rise. At Coogee Oval the higher ground of Arden Street and
Goldstein Reserve is situated some 2m above the bottom of the Oval. During major flood events the
available storage of the Oval will be exceeded and flood waters will spill through here and on to the
beach. The Rainbow Street depression is some 10m deep and as such the storage capacity will
never be exceeded. In extreme flood conditions such as the PMF event or under a blocked
stormwater drainage scenario, a significant flood risk to this area is posed, with possible flood depths
of several metres.

The potential impacts of future climate change are relatively insignificant in the study area for the 1%
AEP event, with negligible impacts from sea-level rise and only locally significant flood depth
increases under increased rainfall intensity conditions.

The flood study will form the basis for the subsequent floodplain risk management activities, being the
next stage of the floodplain management process.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
REFERENCES 68

8 REFERENCES
Department of Environment and Climate Change (DECC) (2007) Floodplain Risk Management
Guideline – Practical Consideration of Climate Change.

Gary Blumberg and Associates (1999) Assessment of Impacts from January 1999 Flooding.

NSW Department of Infrastructure, Planning and Natural Resources (DIPNR) (2005) Floodplain
Development Manual.

Patterson Britton and Partners (1999) Coogee Oval and Bowling Club – Flood Assessment for 24
January 1999.

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
DESIGN FLOOD MAPPING A-1

APPENDIX A: DESIGN FLOOD MAPPING

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
COMMUNITY QUESTIONNAIRE B-1

APPENDIX B: COMMUNITY QUESTIONNAIRE

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
 COOGEE BAY FLOOD STUDY
Randwick City Council is undertaking a detailed flood study of the Coogee Bay
catchment to help identify flooding problem areas. We are seeking the community’s
help by collecting information on any flooding or drainage problems that you may
have experienced in the past. Please take a minute or two to read through these
questions and provide responses wherever you can. Please return this form to
Randwick City Council in the enclosed envelope (no stamp required).

Coogee Oval 24 January 1999

Darren Lyons (BMT WBM Consultants) Terry Kefalianos (Randwick City Council)
Ph: 02 4940 8882 Ph: 02 9399 0525
[email protected] [email protected]
CALIBRATION DATA C-1

APPENDIX C: CALIBRATION DATA

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
HISTORIC NEWSPAPER ARTICLES D-1

APPENDIX D: HISTORIC NEWSPAPER ARTICLES

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
SUMMARY OF PUBLIC EXHIBITION SUBMISSIONS E-1

APPENDIX E: SUMMARY OF PUBLIC EXHIBITION SUBMISSIONS

K:\N1924_COOGEE_BAY_FLOOD_STUDY\DOCS\R.N1924.001.03.DOCX
Issue Action
Resident Experience of flooding in 1984 A review of information from the Coogee
and 1988 on Rainbow Street in the Bay Flood Study indicates that the
vicinity of Marian Street. resident’s experience is consistent with
Resident experience with flooding in the flooding simulated by the hydraulic
Marian Street. modelling.
Resident claims to have not experienced Flooding at this location is a minor
flooding at their block of units in Mount overland flow path at the rear of the
Street. property in the 1% AEP flood. The
frequency of such storms and the nature
of the “flooding” are likely reasons for
the resident not witnessing such events
previously.
Request that stormwater drains in Marian Council has an established drainage
Street are cleaned on a regular basis maintenance program that includes the
cleaning of stormwater pits. The
stormwater network will be inspected and
cleaned as required.
Concern that Marian Street is omitted as Marian Street is part of a broader trapped
a potential problem area for flood low area located on Rainbow Street
inundation in page iv of the Executive which is mentioned in this section
Summary
Comment that drains on Malabar Road Potential measures to manage the impacts
and Marian Street do not have capacity to of extreme flood events will be
cope with extreme flood events. considered as part of the Coogee Bay
Floodplain Risk Management Study.
Question regarding whether there are any Potential measures to manage the impacts
recommendations for property blocks, of extreme flood events will be
particularly those in flood prone areas. considered as part of the Coogee Bay
Floodplain Risk Management Study.
Council’s Flooding Advice and Flood
Related Development Controls Policy
provides advice on how to deal with the
interim period until the study is complete.
Complaint regarding drainage at the Council will investigate options to
bottom of the stairs in Cairo Street modify the stairway to improve
accessibility and use of the stairway. A
by product of this will be the
improvement of drainage at the bottom of
the stairs
Complain regarding local discharges Drainage of the development was
from a development on Denning Street reviewed and found to comply with the
appropriate standards.
Concern over the potential impact of the Flooding insurance is progressively
study on insurance premiums. becoming available in Australia.
Insurance companies undertake their own
studies and use their own methodologies
independent of Council to determine
premiums. The flood study is part of the
process undertaken in good faith and
aims to determine methods of minimising
private and public losses. The
identification of areas at risk of flooding
is necessary to achieve this and does not
change any property owners actual risk
but does enable them to become informed
of the risk.