Saqib Ehsan, M. Sc. Universität Stuttgart Institut für Wasserbau Lehrstuhl für Wasserbau und Wassermengenwirtschaft Prof. Dr.-Ing. Silke Wieprecht Risk

Saqib Ehsan, M. Sc.

Universität StuttgartInstitut für Wasserbau

Lehrstuhl für Wasserbau undWassermengenwirtschaftProf. Dr.-Ing. Silke Wieprecht

Risk and Planet Earth Conference 2009, Leipzig

Estimation of possible damages due to catastrophic flooding

for long-term disaster mitigation planning

Risk and Planet Earth Conference, Panel 2 for Junior Scientists, 4th March 2009, Leipzig

Contents

- Introduction- 1D-Hydrodynamic modeling with MIKE 11- Development of an improved method for loss

of life (LOL) estimation- Loss of life (LOL) estimation for different

scenarios- Conclusions and Suggestions


Introduction

- Role of climate change in disaster management

- Possible extreme changes in climate as guidelines for the development of new concepts for disaster mitigation

- Drastic weather change - Heavy rainfall- Catastrophic flooding downstream of the dam- Risk to people and property


Introduction cont‘d



- Jhelum river valley downstream of Mangla dam in Pakistan

- One of largest earth and rock-fill dams in world- Main dam height ~125 m high above riverbed

(by Google earth)



Gross storage (original) 7.25 E+9 m3

Net storage (original) 6.59 E+9 m3

Catchment area of reservoir (original)

33,360 km2

Water surface area of reservoir (original)(at maximum conservation level)

253 km2

Power generation 1,000 MW

Crest length of main dam 2,561 m

Design capacity of main spillway 28,583 m3/s

Design capacity of emergency spillway

6,452 m3/s


1D-Hydrodynamic modeling with MIKE 11

Chenab River

Upstream Trimmu Barrrage

Jhelum Bridges

Rasul BarrageMalikwal

BridgeKhushab Bridge

Confluence Point

Suketar Nallah

Bandar KasJabba Kas

Kahan River

Mangla dam

Bunha River

-Project Reach: about 329km

-Different Hydraulic

structures

-Five tributaries between

Mangla and Rasul Barrage;

No gauges are existing there

-1D-modeling for unsteady

flow conditions


1D-Hydrodynamic modeling with MIKE 11cont‘d

Maximum Discharges

0

5000

10000

15000

20000

25000

30000

35000

40000

45000

50000

55000

60000

65000

70000

0 50000 100000 150000 200000 250000 300000 350000

Downstream chainage (m)

Max

. Q (

m3 /s)

40000 m3/s (withbridges)

40000 m3/s (withoutbridges)

50000 m3/s (withbridges)

50000 m3/s (withoutbridges)

MDF (61977 m3/s: withbridges)

MDF (61977 m3/s:without bridges)


1D-Hydrodynamic modeling with MIKE 11cont‘d

Rasul Barrage

High Flooding Scenarios (maximum water level)

150

160

170

180

190

200

210

220

230

240

250

260

270

280

290

0 30000 60000 90000 120000 150000 180000 210000 240000 270000 300000


Max

. wat

er le

vel (

m)

40000 m3/s (with bridges)

50000 m3/s (with bridges)

MDF (61977 m3/s: with bridges)

40000 m3/s (without bridges)

50000 m3/s (without bridges)

MDF (61977 m3/s: without bridges)


1D-Hydrodynamic Modeling with MIKE 11cont’d

Dam break Flood Routing (maximum discharges)

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000

220000

240000

260000

280000

300000

320000

0 30000 60000 90000 120000 150000 180000 210000 240000 270000 300000


Max

. Q (

m3 /s

)

Case1 (with bridges)



Case1 (without bridges)




Rasul Barrage

1D-Hydrodynamic Modeling with MIKE 11cont’d

Dam break Flood Routing (maximum water level)

150

160

170

180

190

200

210

220

230

240

250

260

270

280

290

300

310

0 30000 60000 90000 120000 150000 180000 210000 240000 270000 300000


Max

. wat

er le

vel (

m)








Development an improved LOL estimation method

LOLi = PARi x FATBASE x Fsv x Fage x Fmt x Fst x Fh x Fwar x Fev

LOLi = loss of life at a particular location ´´i`` downstream of the dam

PARi = Population at risk at a particular location ´´i`` downstream of the dam

FATBASE = Base Fatality rate of 0.15 (worst case of medium severity) (Graham, 1999), assuming an average value of 1.0 for all other factors with average conditions.


Fsv = Flood Severity factor

High Severity very likely 1.0Medium Severity unlikely 0.3Low Severity very unlikely 0.1

Fage = Age risk factor

A (<10yrs+ (>=65yrs)),B (10-15)yrs and C (15-64)yrs

Fage = 1.25 *A% +1.1* B%+ 0.8* C% (general form) Fmt = Material risk factor

Fmt = 1 * X % + 1.5 * Y % (general form)

Where, X= % of other type of houses, Y= % very low strength houses



Fst = Storey risk factor Fst = 1 (for high severity and all house types)

Fst = 1- S % (for medium and low severity)

Where, S= % of more storey houses

Fh = Health risk factor; 3% disabled people Fh = 1 *H % + 1.25*D % (general form)

Where, H= % of PAR with avg. health, D= % of disabled PAR



Fwar = Warning factor (Graham,1999)

Warning Flood Severity understanding Fwar

No No 1 Some (15-60min) Vague/unclear 0.7 Adequate (>60min) Precise/clear 0.3

Fev = Ease of evacuation factor

Warning Ease of evacuation Fev

No No 1 Some (15-60min) Some 0.7 Adequate(>60min) Good 0.3



Loss of Life estimation

PAR downstream of Mangla dam (98-Census data)

0

20000

40000

60000

80000

100000

120000

140000

160000

180000

200000


PAR

(No.

of

Peo

ple

at r

isk)

PAR

Total PAR : 1178038

Urban PAR : 37%

Rural PAR : 63%

Estimated PAR is related to the highest flood event in the past



Estimated Total Loss of Life downstream of Mangla dam (98-Census data)

0 5000 10000 15000 20000 25000 30000

1

2

3

4

5

Sele

cted

Sce

nari

os

Total Loss of Life

LOL (MDF 61977 m3/s:without bridges)

LOL (MDF 61977 m3/s:with bridges)

LOL (50000 m3/s:without bridges)

LOL (50000 m3/s: withbridges)

1- Warning Initiation 30min after Failure 2- Warning Initiation 15min after Failure

3- Warning Initiation at Failure

4- Warning Initiation 1hr before Failure

5- Warning Initiation 2hrs before Failure


% Total Loss of Life for Different Failure Cases

1

1.5

2

2.5

3

3.5

4

4.5

0 50000 100000 150000 200000 250000 300000 350000

Max. Discharge (m3/s)

% T

otal

LOL

(%

dea

d pe

ople

)

%LOL (with bridges)

%LOL (without bridges)

Worst Case for Warning Initiation:

30 minutes after Failure



Cumulative Loss of Life due to Dam Failure

0

10000

20000

30000

40000

50000

60000

70000

80000

90000

100000

0 25000 50000 75000 100000 125000 150000 175000 200000 225000 250000 275000 300000


Cum

ulat

ive LOL

Failure Case1

Failure Case2

Failure Case3

% Cum. LOL up to 50Km: about 80% of Total LOL


Total LOLWorst Case for Warning

Initiation: 30 minutes after Failure




Conclusions and Suggestions

- Severe climate change can cause extreme flooding downstream of a

dam

- Estimation of possible damages is an important part of any dam

safety study

- Loss of life increases with the delay in warning initiation with respect

to dam failure

- For all dam failure cases, maximum LOL (~80%) occurs in first

50 km downstream of Mangla dam

- % total LOL for the worst case of Mangla dam failure is close to 4%

which seems to be very high


Conclusions and Suggestions

- LOL results clearly show the need of improvement in existing risk Reduction measures in order to reduce possible LOL due to Mangla dam failure

- More research is required to estimate

- ease of evacuation - risks posed by age groups - very low strength houses and more storey houses - Realistic estimation of possible LOL due to natural hazards like floods helps in long-term disaster mitigation planning


THANKS FOR YOUR ATTENTION

QUESTIONS??

[email protected]

www.iws.uni-stuttgart.de

Lehrstuhl für Wasserbau und Wassermengenwirtschaft

Institut für Wasserbau, Universität Stuttgart

Documents

Saqib Ehsan, M. Sc. Universität Stuttgart Institut für Wasserbau Lehrstuhl für Wasserbau und Wassermengenwirtschaft Prof. Dr.-Ing. Silke Wieprecht Risk