Forensic Engineering for Dam Failures

Forensic Investigationsofof

Dam FailuresPaul C. Rizzo

October 7,2010October 7,2010

Case Histories

• A few introductory notes:All j liti ti h j t i l t d– All major litigation on each project is completed

– All information reported here is in the public domain• Dam Failures discussed todayy

– The “Classics” plus– Taum Sauk– Silver Lake

Click to edit Master title style

– Saluda Dam– Swinging Bridge

– Will not talk about Tampa Bay, Lake Dehli, Levee Failures

2

References

• Dr. A.J. “Skip” Hendron

• Brian Green

• Prof J David Rogers• Prof. J. David Rogers

• Euler Cruz & Rafael Cesário


• Dr. David Petley

3

Definition of Failure

• Failure at a dam is “an uncontrolled release of water” • Causes of Failure• Causes of Failure

– Overtopping – the most common cause– Piping thru Embankment – 2nd!!

i i h d i d!!– Piping thru Foundation – 3rd!!– Foundation Degradation

• Karst developmentLi f ti


• Liquefaction• Loss of fines over time

– Earthquakes – ground motion and faultsFailure of gates stoplogs flashboard malfunction– Failure of gates, stoplogs, flashboard malfunction, penstocks, tunnels, and reservoir slopes

4

Some “Classics”Some ClassicsWe learn from out mistakes!!


We learn from out mistakes!!

5

Classic Failures

• Had a major impact on dam or hydro engineering

• Had a major impact on dam safety legislation and dam safety regulations


legislation and dam safety regulations

• Had a major impact on society

6

Folsom DamFolsom, California


July 17, 1995

7

Folsom Dam – “Before”


8

Folsom Dam – “After”


9

Folsom Dam ‐ 1995


10

Folsom Dam

• Increased profession’s awareness of pthe need for maintenance and inspection of auxiliary featuresinspection of auxiliary features associated with a dams

Click to edit Master title style• Changed FERC’s Inspection Guidesg p

11

Shih Kang DamgTaiwan


September 21, 1999

12

Shih Kang Dam – TaiwanImpact of building a dam over a faultImpact of building a dam over a fault


13

Shih Kang Dam ‐ Taiwan


14

Shih Kang Dam ‐ Taiwan


15

Austin (Bayless) Dam( y )Austin, Pennsylvania


September 30, 1911

16

Austin Dam – “Before”


17

Austin Dam – “After”


18

Austin Dam

• Destroyed the myth that “gravity dams can’t fail”

• Illustrates the need for detailed,Illustrates the need for detailed, comprehensive foundation investigations and foundation preparation


and foundation preparation

• Illustrates the need for temperature expansion/contraction joints in gravityexpansion/contraction joints in gravity dams

19

V i DVaiont DamItalyItaly

October 1963


20

Vaiont Dam – Double Arch245m (800) ft high245m (800) ft high


2121

Village of Longarone “Before”Vaiont Dam FailureVaiont Dam Failure


22

The Vaiont Dam Disaster

• Dam completed and filled

• Reservoir cycled up and down several times overthree years

• Left Reservoir Slope Failed catastrophically

• Within 30 – 40 seconds some 270 million m3 of


• Within 30 40 seconds, some 270 million m ofrock crashed into the reservoir.

• Resulting wave overtopped dam by about 100 –Resulting wave overtopped dam by about 100150 m.

23

Vaiont Dam – Reservoir Slope Failure


24

Longarone Village “After”Vaiont Dam FailureVaiont Dam Failure


25

Vaiont Dam at the Top


26

T DTeton DamTeton IdahoTeton, IdahoJune 5, 1976


27

Teton – “Before”


28

Start of Piping – June 5 – 7:00 AM


29

June 5 ‐11:20 AM


30

June 5 – 11:50 AM(Crest still intact)(Crest still intact)


31

June 5 – 11:50 AM(Crest now eroding)(Crest now eroding)


32

June 5 – 12:00 noon(Crest lost – maximum flood)(Crest lost maximum flood)


33

Lessons Learned

• Compaction is extremely important throughout the dam and especially at the abutments

• Grout Curtains are critical to prevent piping

• Keyway design is critical


• Keyway design is critical

• Filters are a key feature to resist piping

(When in doubt, add a filter)

34

Compaction Against Right Abutment


35

Compaction Equipment after Change


36

Sayano‐Shushenskayah kKhakassia, RussiaAugust 17 2009


August 17, 2009

37

Accident at Russia’s LargestHydroelectric Power StationHydroelectric Power Station


38

PenstocksSpillway

Dam

Units

Powerhouse

12345678910

Control Building

39

40

The Accident – 2009 Aug 17


4141

Before the Accident

Crane


Power Units

Air‐Oil Tanks

42Sump Tank Governor

Pumps Generators floor42

After the Accident

Sump Tank

Air‐Oil Tanks

Sump Tank

Floor

Click to edit Master title styleColector Ring

Unit 2

Crosshead –Unit 2

Colector Ring

Unit 1

43

Generator floor

43

General View

Floor


4444

General View


4545

Before the Accident

Generator Rotor – Unit 5


46

Generator Runner

The accident started here

46

Sump Tank (turned)

Air‐Oil Tank

Crosshead

47

After the Accident

Unit 7: destroyed

Click to edit Master title styleUnit 9: destroyed

4848

Unit 9

Unit 7

49

50

51

Unit 2

Units 7Units 7 and 9

52

FloodFlood

53

Lower Van Norman DamLos Angeles, California

bFebruary 9, 1971

Click to edit Master title styleA Classic that was not a failure!

54

Lower Van Norman February 9 1971February 9, 1971

L V N D b ilt b th Cit f L t A l


• Lower Van Norman Dam was built by the City of Lost Angelesas part of the Los Angeles Aqueduct in 1916 – 1918, using thehydraulic fill and puddled fill techniques. A rolled fill additionwas placed in 1924.

• The embankment failed during the February 9, 1971 M. 6.7 SanFernando Earthquake but no water was released

55

Fernando Earthquake, but no water was released.

Lower Van Norman


56

Lower Van NormanForerunner to Modern Liquefaction AnalysisForerunner to Modern Liquefaction Analysis

Susceptibility to Liquefaction

Click to edit Master title style•Careful forensic evaluations by the geotechnical engineering group at U.C.Berkeley unraveled the dam’s failure by liquefaction of a zone of low densitysandy hydraulic fill, shown in blue in the above sections.Th St t b tl l t d 30 th h d li fill d f t fitti

57

•The State subsequently slated 30 other hydraulic fill dams for retrofittingbetween 1973 – 75.

Swinging Bridge Damg g gSullivan County, New York


May 5, 2005

58

Swinging Bridge Dam Cross SectionSwinging Bridge Dam Cross Section


59

Swinging Bridge Dam


60

Swinging Bridge Dam


61

Swinging Bridge Dam

• Over time leaks developed

• Fines transported to penstock during dewatering

• Sinkholes developed on crest

• Reservoir drained

• Penstock & Unit abandoned


• Dam repaired and refilled (3 yr shutdown)

62

Silver Dam Lakel hUpper Peninsula, Michigan

May 2003Click to edit Master title style

May 2003

63

Silver Lake DamDead River Hydroelectric ProjectDead River Hydroelectric Project

M t i T St L k• Mountain Top Storage Lake– Main Concrete Dam with Spillway & Outlet Pipe/Gate Control– Auxiliary Dams and Dikes

Emergency Spillway with Erodible Fuse plug– Emergency Spillway with Erodible Fuse plug

• May, 2003 – snowmelt, Gate Control not Activated• Fuse Plug Eroded


• Fuse Plug Eroded• Main Dam Overtopped – minimal damage• Dam No. 2 overtopped & eroded away entirely

• Failure drained approximately 25,000 acre feet of water from Silver Lake Reservoir

• $100 million damage – no fatalities

64

• $100 million damage no fatalities

Silver Lake – Overtopping of Main Dam


65

Saluda Dam

Columbia, South Carolina

Click to edit Master title styleRehabilitation

66

Saluda DamA Failure that was Prevented!A Failure that was Prevented!

WATEREERIVER

SALUDA RIVER

COLUMBIA

SANTEERIVER

COLUMBIA

LAKEMURRAY


67

Aerial View of Saluda Dam

Intake TowersNORTH

Spillway

McMeekin Station

Click to edit Master title styleAsh Ponds &

Saluda Hydro

BushAsh Ponds & Landfill

Bush River Rd.

68

Saluda Dam

HydroelectricPlant

Existing Dam

Plant

Lake Murray

Lake Area 78 square miles


Lake Area 78 square milesLake Capacity 1,600,000 acre feet Dam Length 7,800 feetMax Dam Height 200 feetPowerhouse Capacity 260 MWOriginal Construction Hydraulic Fill

69

Original Construction Hydraulic FillOriginal Completion 1930

Original Dam ConstructionOriginal Dam Construction

•Dam was constructed 1927 ‐ 1930

•11,000,000 cy‐‐semi‐hydraulic method


7085

As‐Built Dam Configuration

• Semi‐Hydraulic Fill

• Most soils “Dumped”

• No compaction

Rolled CoreWashed Fill


Rip RapUnwashed Fill

Washed Fill

Washed Fill

Sluiced Core Residual SoilUnwashed Fill

71

Dynamic Stability Analysis

• Liquefaction Evaluation of Saluda Dam

– Earthquake record selection

Liquefaction methodology– Liquefaction methodology

• Post‐Seismic Stability Analysis


72

Liquefaction Analysis

Factor of Safety against Liquefaction (FSL)


-0.02 0.32 0.66 1.001.00 1.33 1.67 2.002.00 3.78 5.57 7.35

Extensive Liquefaction, Post‐Seismic FS < 1

73

Section AA’ (typ.)

Lower Bound

B t C

-0.03 0.31 0.66 1.001.00 1.33 1.67 2.002.00 3.51 5.02 6.53


Best Case

-0.06 0.30 0.65 1.001.00 1.33 1.67 2.002.00 3.96 5.92 7.88

74

Remediation Concept

Competent Rock

Click to edit Master title styleRockfill Damoc a• Length 5,100 feet

• Max. Height 200 feet

75

g

Remediation Concept

Click to edit Master title styleRCC Dam• Length 2,200 feet

• Max. Height 210 feet

a

76

g

Taum Sauk Upper ReservoirLesterville, MissouriDecember 15, 2005


ecember 5, 005

77

Taum Sauk on December 14, 2005


78

Taum Sauk on December 15, 2005Release of 1 4 Billion Gallons of WaterRelease of 1.4 Billion Gallons of Water


79

Original Rockfill Dam X‐Section


80

Original Rockfill Dam X‐Section


81

Panel 72 Erosion


82

Taum Sauk Forensic Investigation

• Field InvestigationsVi l Ob i– Visual Observations

– Geologic Mapping• Breach AreaBreach Area• Breach Channel

– Record Search


– Test Borings & Sampling– Laboratory Testing– Instrumentation InvestigationInstrumentation Investigation– Interviews (operators & mgmt)

83

Forensic Investigation

• Engineering AnalysesH d li A l i ( fl l it– Hydraulic Analysis (overflow zones, velocity, impingement zone)

– Rockfill Dike (seepage, piping & stability)– Concrete Facing (effect on µ)– Liner (effect on µ)– Parapet Wall (stability, structural)


Parapet Wall (stability, structural)– Asphalt Liner (effect on µ)– Grout Curtain (effect on µ)Fil ( ff )– Filters (effect on µ)

84

Rockfill Dike Stability ‐ Sliding


Water PressureResisting

Weight of

Estimated P

Normal F

W

Sliding Weight of Water

85

Pore Pressure at Overtopping

Force

Taum Sauk Re‐build Data

• Same Footprint

• No change in capacity or generation

• 110 to 140 ft high• 110 to 140 ft high

• ~ 6800 ft long ‐ 9 monoliths @ ~800 ft


• RCC with processed “old rockfill” dam

• RCC – 2.8 million CY + 0.5 million CVC

• Symmetrical w Concrete Face

• Overflow Release Structure86

• Overflow Release Structure

• Highly redundant Instrumentation

Rock Foundation Preparation


87

Foundation Preparation


88

Upstream View of the Fracture ZoneCleaning and Mapping in ProgressCleaning and Mapping in Progress


89

Overflow Release Structure


90

Taum Sauk on April 1, 2010


91

Keys to a ForensicKeys to a Forensic Investigation


92

Keys to a Forensic Investigation

• Investigation Work PlanPl f Liti ti f th b i i– Plan for Litigation from the beginning• Civil or Civil & Criminal• Expect Opposing Experts

Lit t S h– Literature Search– Interviews– Review of Existing Calculations & Design Basis


– Field Investigations– Laboratory Testing– “New” AnalysisNew Analysis– Report Preparation

93


• Have a well defined engagement letter• Avoid emails!!• Avoid derogatory remarks & comments• Keep consistent & accurate field logs• Keep confidentiality


• Use NDA’s with Subcontractors• Treat drafts & preliminary documents consistent with standard corporate policy

• Implement a robust QA/QC Program

94

• Be open minded!!


• Public Domain Literature Search – Professional Literature

– News Accounts & Photos/Videos/

• Private File Search– STID’s & PFMA’s


– STID s & PFMA s

– Permit Applications

Inspection Reports– Inspection Reports

– Instrumentation Reports

C t ti R t & Ph t95

– Construction Reports & Photos


• Field Investigations (drilling, geophysics, etc)– Implement a Field Work Plan– Implement a HSEP immediatelyI l t QA/QC P– Implement a QA/QC Program

– Use Calibrated Equipment with traceable records• SPT Energy


SPT Energy• Shelby Tube & Split Barrel Dimensions• Pressure gages, flow meters, DRM Systems

U S l M if P l– Use Sample Manifest Protocols– Watch the field records carefully!!!!– Use NDA’s with subcontractors

96

– Use NDA s with subcontractors


• Laboratory Testing– Implement a Laboratory Testing Plan

– Implement a QA/QC Programp Q /Q g

– Use Calibrated Equipment with traceable records


– Use defendable Standards (ASTM, C of E, etc)

– Use Sample Manifest Protocolsp

– Watch the lab records carefully!!!!

– Should Field Logs be adjusted based on lab

97

Should Field Logs be adjusted based on lab tests?

Thank you for your time!


98

Forensic Engineering for Dam Failures

Documents