1

Urban Flood Modelling

Adrian J Saul

Pennine Water GroupDepartment of Civil and Structural

EngineeringUniversity of Sheffield

a.j.saul@sheffield.ac.uk

Need for Integrated Urban Drainage within an  integrated catchment

Intra Urban Area

Peri-urban area

urban area

Local area

house and curtilage

THE PUBLIC HOUSE

Pluvial floodingcaused byoverloaded buildingdrainage

curtilagefloodingdue to domesticdrain overloadby rainfall andsaturatedground

THE PUBLIC HOUSE

THE PUBLIC HOUSE

THE PUBLIC HOUSE

THE PUBLIC HOUSE

THE PUBLIC HOUSE

THE PUBLIC HOUSE

local area flooding causedmainly by pluvial drainoverload, overland flows,ponding on roads,watercourse spills andinadequate sewers

Urban area flooding from pluvialupstream effects, includingsurface flood waves andoverloaded sewers plus culvertedand other watercoursesspilling orbacking up

Peri-urban floodingfrom rivers backingup from rural areasand also fromupstream dischargesand overland floodwaves

URBAN FLOODING

Types of flooding

Pluvial flooding Flooding due to asset performanceFluvial floodingCo‐incident floodingGroundwater flooding

Future change considerations• Changes in Rainfall• Urbanisation and urban creep• Impact of asset deterioration and renewal• Emerging and  new technologies• Changes in Groundwater level and Infiltration• Changes in local flood pathways and urban form

Developing an initial understanding  of the problem

• Flood mechanisms and interactions between  different urban drainage systems;

• Scale of the flooding (e.g. localised, town‐wide  or river catchment

wide);

• Frequency of the flooding;• Consequence of the flooding (e.g. degree of 

nuisance, cost).• Basis of modelling is an extremely accurate 

urban surface DEM/DTM

6

• The research challenge is to develop a generalised  tool to deal with the interactions of any above 

ground flows and their interaction with the below  ground drainage system

Types of Data

• OS Mastermap

• LiDAR– Flown by plane or helicopter– Ground drive‐overs

‐

yes

• Topographic/GPS surveys• Historical data – videos, photographs, flood 

levels

Typical DEM and drainage system

LegendStudy Area 2ManholePipeRiver

Elevation110 m

80 m

0 200 400100Meters

­River Aire

Delineated Ponds

Sinks & Exits

Connecting Paths

Approximate Geometry

1D Surface Network (Nodes & Links)

1. Major System

Sewer Network (Manhole & Pipes)

Sub-catchment Delineation

Sewered areas Undrained Areas

Out of Catch

R-R Model parameters

2. Minor System

Minor-major Model (SIPSON) 1D surface path way + 1D sewer network

DTM Enhancement

R-R model parameters

Sewer network

Pond catchment

Reduced pond

Catchment

Interactions

12

Flows ‐

in and out of the sewer system  and subsequent overland flow

• 1D sewer network model  prior to flood

• Sewer model has to be  coupled with overland  flow model when 

flooding occurs.

• 1D/1D or 1D/2D

13

Surface storages and sewer  interactions

PipeManhole

Exit point

Exit direction

Terrain

Lowest point

Exit direction

PipeManhole

Exit point

Exit direction

Terrain

Lowest point

Exit direction

14

Flow directions alternate during an event

• Surface runoff enters drainage system through gulleys

and  manholes.

• Sewer flows surcharge from the manhole.

• Surface overland flow – 1D or 2D

a)

H

b)

H

c)

H

15

Modelling urban overland flow

• Cascade of natural retention ponds

17

Determination of surface pathways

• Connectivity analysis by “rolling ball”,  “bouncing ball”

or “sliding ball”

algorithms 

– Pond to pond– Pond to sewer– Sewer to pond– Sewer to sewer– Out of catchment

manholes

pond

pond

pond

18

Need accurate cross sections of flow paths

19

Need very accurate definition of  streets

ManholeTerrain

Sewer Pipe

ManholeTerrain

Sewer Pipe

Overflow from one  street to another

Lowest pt

20

Buildings

21

bAA

α =

Need building coverage ratio 

bAA

xαΔ

yαΔ

( )1 yα− Δ

Conveyance width  for flux in the y 

direction

Conveyance 

width for flux in 

the x direction

y

x

( )1 xα− Δ

Computational gridBuilding

22

Need building alignment

a b c

d e f

y

x

Computational gridBuilding

23

Need conveyance reduction factors 

( )12

1i , jx yβ−

− Δ ( )12

1i , jx yβ+

− Δ

( )12

1i , jy xβ

+− Δ

12i , jy xβ

+Δ

12i , jx yβ

+Δ1

2i , jx yβ−

Δ

( )12

1i , jy xβ

−− Δ

[ ]i x, j yΔ Δ

xΔ

yΔ

y

x

Computational gridBuilding

Problems with bridges

Complex bridges 

Here is an example map of  Spaghetti Junction

We can highlight road  features

Aggregate road  polygons

Buffer aggregated  polygons

Intersect analysis to  identify possible bridge 

boundaries

DEM/DTM requires very detailed topography of catchment surface – grid resolution 1m x 1m, vertical accuracy 50mm.   Need to distinguish cover types and infiltration capacity –

runoff. Need exact position of gulleys

and manholes, kerb heights, 

drop kerbs, walls, fences, permeable structures, etc

FRMRC

• FRMRC has developed new urban interactive  model

• Models trialled in 3 UKWIR funded case studies

• These have highlighted the need for a joint  1D/1D and 1D/2D approaches

• The detail of the catchment

surface is the critical  factor and enhanced DEM’s

are the way forward

Conceptualisation of integrated urban drainage model  –

includes rivers and coasts 

River

1D model of drainage system

1D model of drainage system

and surface pathways

1D model of drainage system and local 2D models of surface

pathways

1D model of drainage system and local 2D model of floodplain

Simple modelling of local problems within guidelines derived

from integrated urban drainage modelling

What we require Flood risk maps

Location and depth of flooding Damage and vulnerability assessment

Acknowledgement and thanksThe research reported in this presentation was conducted as part

of the Flood Risk Management Research Consortium with support  from the: 

– Engineering and Physical Sciences Research Council 

– Department of Environment, Food and Rural  Affairs/Environment Agency  Joint Research Programme 

– United Kingdom Water Industry Research

– Office of Public Works Dublin

– Northern Ireland Rivers Agency

Data were provided by the EA and the Ordnance Survey.