Practical aspects of dam break analysis

Louis C Hattingh

Hattingh Anderson Associates CC

Dam break analysis

• It is a model

• You need to understand what you model & have an idea of the answers that you expect

• Very little known inputs & lots of assumptions

Baldwin Hills Reservoir

What is important?

• Mode of failure (including mechanism) – Breach width

– Time

• Assumptions – Reservoir water level

– Floods or sunny day

• Level of accuracy

Failure modes

• Internal erosion

• Structural

• Hydrologic

• Hydraulic

• Seismic

• Operational

• Other

Failure modes • Internal erosion • Structural

– Concrete gravity dams failures – Concrete arch dam failures – Concrete buttress dam failures

• Hydrologic – Overtopping

• Hydraulic – Failure due to erosion of rock – Failure due to overtopping of spillway walls and stilling basins – Stagnation Pressure Failure of Spillway Chutes – Cavitation Damage Induced Failure of Spillways

• Seismic – Failure of embankment dams during to seismic loads – Seismic failure of retaining walls

• Operational • Other

– Landslide failures and incidents – Trunnion Friction Radial Gate Failure – Drum Gate Failures

El Guapo Dam, Venezuela

• Built 1975 to 1980 • No proper hydrologic studies - based on

similar basin • Spillway system

– Original uncontrolled ogee with downstream chute

– Tunnel spillway added after chute wall overtopping during construction

• Failure in 1999

El Guapo Dam, Venezuela

Is not a flip bucket but a hydraulic jump basin

Flow outside the spillway chute

Walls Began to overflow at 1:15 am on 12/16/1999

Water level behind dam decreased at 9:00 am on 12/16/1999

12/1/2014 ENTRO Dam Safety Training

Module Louis

Hattingh

10

Water level rose again – erosion had undercut basin, chute and spillway weir at 4:00 pm on 12/16/1999

Approach channel collapsed at 5:00 pm on 12/16/1999

Flood wave reached 1st village at 6:00 pm on 12/16/1999 – reservoir lowered 30 meters in 40 minutes

Zoeknog Dam

• Weathered granites

• No provision for dispersive material during construction

• Incorrect blanket drain position

• No attention to piezometer warning during first filling

• Failed on 25 January 1993

Gleno Dam, Italy

• 50 m high multiple concrete arch dam 213 m long

• Masonry gravity plug built in deep central valley gorge (use lime mortar instead of cement mortar)

• Original concrete gravity • Changed to multiple arch but not

approved

Gleno Dam, Italy

Gleno Dam, Italy

Gleno Dam, Italy

• 1923: – Failure of one of the buttresses leading to

multiple arch failure – 356 fatalities

• Change in design • Iffy concrete quality • Inappropriate material

– Lime mortar for masonry section • Settlement of masonry plug?

Malpasset Dam, France

• Thin arch dam 66.1 m high and 6.7 m thick at base

• Foundation = gneiss

• No foundation grouting or drainage features

• Designed by Andre Coyne

• Completed 1954

Malpasset Dam, France

Malpasset Dam, France

Malpasset Dam

• 1959: – Failure

– 421 fatalities • View of left

abutment and thrust block following the failure

• Thrust block moved about 1 m into abutment and slightly d/s

• FAILURE MODE?

• U/S dipping fault and D/S dipping foliation shear formed lt. abut block

• Arch thrust in direction of foliation decreased permeability

• Tensile stress at u/s face opened foliation shear

• Nearly full uplift developed on foliation

• Block slid out on fault (phi = 30o) and dam went with it

After P. Londe

Malpasset Dam, France

Kariba Dam, Zambia/Zimbabwe

• 128 m high concrete arch • Built between 1956 & 1959 • World’s largest artificial lake • Gated spillway sill = 33 m below crest • Spillway use created 80 m deep eroded

plunge pool over 20 years • Geological feature (discontinuity) in the

river section that was not picked up during planning and design

• Plans are abreast to deal with the issue

Kariba Dam, Zambia/Zimbabwe

Taum Sauk Dam, USA • Concrete-faced

earthfill “ring-dike” structure

• Upper reservoir of pumped-storage project

• Water routinely stored on 3 m high parapet

• NO SPILLWAY!!!

Taum Sauk Dam, USA

Taum Sauk Dam, USA • Membrane liner installed in 2004 • Reservoir level instrumentation could not

be reinstalled properly due to liner warranty issues

• Instruments were loose and not reading reservoir level properly

• Resetting of reservoir sensors did not account for settlement of embankment

• Alarms wired so high level and high-high level sensors needed to trigger for alarm

• Over-pumping was not detected and dam overtopped and failed

Taum Sauk Dam, USA

Taum Sauk Dam, USA

Taum Sauk Dam

0.667

0.1000.100

0.0000.000

0.1000.100

0.667

0.831

0.658

0.6900.6200.580

0.845

0.000

0.100

0.200

0.300

0.400

0.500

0.600

0.700

0.800

0.900

0.000 20.000 40.000 60.000 80.000 100.000 120.000

Re

lative

le

ve

l (m

)

Chainage (m)

Spillway

Right bank of embankment

Left bank of embankment

Vaiont Dam, Italy • 265 m high concrete arch dam • Completed in 1960 • Left side reservoir foundation = steep

slopes in bedded limestone with clay interbeds

• 1 month after completion & after heavy rain = first landslide = 700 000 m3 & 2 m wave

• Exploratory adits, piezometers & level of reservoir adjusted to limit slide movement

Vaiont Dam, Italy

Vaiont Dam, Italy

Vaiont Dam, Italy • 1963

– Massive slide of 267 million m3

– 100 m high over dam wall – 2 600 fatalities – Arch survived

• Dam abondoned • Low strength clay layers between

limestone beds • Reservoir geology not fully

understood

Have you considered all relevant failure modes?

Practical example: Gariep Dam will be

destroyed should either Katse or Mohale Dam fail

Katse Dam

Failure mode?

How to check

• Classical dam break/pull the plug

• Maximum discharge for sudden failure – Q = v x A

– v = 2/3 x (g x y)0.5

Mohale Dam

Failure mode?

Uncertainty for both width & time

300

250

200

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach

(minutes)

Mohale dam break @ Mohale

300

250

200

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach

(minutes)

Mohale dam break @ Gariep

The bottom line

• Katse Dam: – Peak flow @ Katse ≈ 500 000 m3/s

– Peak flow @ upper end of Gariep ≈ 28 000 m3/s

– Peak flow routed through Gariep ≈ 5 400 m3/s

• Gariep SEF: – 26 600 m3/s (unrouted)

– 16 000 m3/s (routed)

The bottom line

• Mohale Dam: – Peak flow @ Mohale ≈ 320 000 m3/s

– Peak flow @ upper end of Gariep ≈ 18 000 m3/s

– Peak flow routed through Gariep ≈ 2 000 m3/s

• Gariep SEF: – 26 600 m3/s (unrouted)

– 16 000 m3/s (routed)

300

250

200

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach

(minutes)

Mohale dam break @ Mohale

300

250

200

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach

(minutes)

Mohale dam break @ Gariep

What about 60 minutes?

300

250

200

30%

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach (minutes)

Mohale dam break @ Mohale

300

250

200

30%

40%

50%

60%

70%

80%

90%

100%

Breach width (m)

Time of breach (minutes)

Mohale dam break @ Gariep

Water levels & exceedance probabilities

1 325

1 330

1 335

1 340

1 345

1 350

0.1

1

10

100

1000

10000

71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 00 01 02 03 04 05 06 07 08 09 10 11

Re

lative

wa

ter

leve

l (m

)

Un

rou

ted

in

flo

ws (

m3/s

)

Time

Inflow (m3/s)

Water level

0

10

20

30

40

50

60

70

80

90

100

1 330 1 332 1 334 1 336 1 338 1 340 1 342 1 344 1 346 1 348 1 350

rese

rvo

ir e

xce

ed

an

ce

%

Relative water level (m)

Concluding remarks

• Understand your problem – failure modes

• Consider required accuracy level

• Knowledge of the uncertainties - use sensitivity analysis

• Check the model outputs

Acknowledgements

• Dr Chris Oosthuizen

• Dam Safety Surveillance – present & past

• Gregg Scott – formerly USBR

Good luck & enjoy