Embed Size (px)
Citation preview
An Examination of Dam Failuresvs.
the Age of Dams
Patrick Regan
FERC
One Simple Question
Does 50 years of, apparently, successful operation assure continued safe operation?
Swift 2 Dam
Swift 2 Dam
Taum Sauk
Taum Sauk
Database Facts and Figures
• Total entries 4283
• # where age can be determined 1158
• # of countries 84
• Earliest construction date 1550
• Oldest dam incident/failure 286
Distribution of Incidents by Age# of Incidents vs. Age All Dams
0
50
100
150
200
250
300
350
400
Age Range
# of
Inci
dent
s
Camara Dam
Camara Dam2 Years Old
IVEX Dam
IVEX Dam152 Years Old
• To answer the question as to does 50 years of “successful” operation means we don’t need to worry about a dam any more, the data suggests that >50% of incidents of dams that survived their first 5 years occur after the dam is 50 years old.
% of Incidents that Occur at an Age Greater Than All Dams that Survive their 1st Five Years
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%>
5
>1
0
>1
5
>2
0
>2
5
>3
0
>3
5
>4
0
>4
5
>5
0
>5
5
>6
0
>6
5
>7
0
>7
5
>8
0
>8
5
>9
0
>9
5
Age
How are Incidents for Different Types of Dams Distributed?
% of Incidents That Occur At An Age Greater Than By Type of Dam
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
Age
Earthfill
Rockfill
Gravity
Arch
All
How are Incidents for Different Failure Modes Distributed after
1st 5 years
% of Incidents that Occur at an Age Greater Than by Failure Mode
After 1st 5 Years of Operation
0%10%20%30%40%50%60%70%80%90%
100%
>5 >15
>25
>35
>45
>55
>65
>75
>85
>95
Age
Flooding
Seepage
Structural
How are Incidents Distributed by the Era of Dam Construction?
# of Incidents by Type of Dam vs. Decade Built Dams that Survived 1st 5 Years of Operation
0
10
20
30
40
50
60
Time Period
Buttress
Crib
Rockfill
Multiple Arch
Gravity Concrete
Gravity Masonry
Earthfill
Arch
Other
How are Incidents Distributed by the Era of Dam Construction?
Earth Dam Seepage Incidents % of Incidents Beyond a Given Year by Era of Construction
(Data Beyond 45 Years of Age Removed)
0%
10%
20%
30%
40%
50%
60%
>5 >10 >15 >20 >25 >30 >35 >40 >45
Age
<1900
1900-1949
1950-1975
Some Data on Older Dams
• 46 Incidents in Database with Age >100• Failure Modes
– % due to flooding (not PMF) 49%– % due to structural issues 31%– % due to seepage/piping 20%
• Oldest Piping/Seepage Incident– 152 years old (10th oldest)
• Oldest Flood Related Incident– 286 years (oldest event)
What Contributes to Older Incidents?
• Exposure– Floods– Earthquakes
• Aging– Piping– ASR– Settlement
• Changed Conditions– Changed Operating Basis– Loss of Knowledge– Modifications that have unforeseen consequences– Changed conditions in vicinity of dam
Merwin Gate Nuts
What’s Causing Flood Related Failures?
• Gates not working– Stuck in place– Operators fail– Power fails– Debris plugging– SCADA errors
• Spillways failing– Overtopping of chute walls– Cavitation, backcutting, stagnation pressure, etc.
• Human Factors– Can you get there? Will it work?
Noppikoski Dam
• Noppikoski Dam was a central core rockfill dam.
• The dam was remotely operated from a location 30 miles away.
• The two spillway openings were controlled by steel stoplogs that were raised by using a hoist located on a moveable trolley.
Noppikoski Dam
• In September, 1985 a major storm occurred in central Sweden.
• During the initial phase of the flooding the stoplogs had been removed from the spillway bays. However, due to a lessening of the storm the left spillway stoplogs had been replaced and one stoplog was placed in the right bay and left hooked to the hoist.
Noppikoski Dam
• On September 6th, a Friday, the storm increased in intensity. An operator was dispatched to the site from his home.
• When the operator was about half way to the dam, he is informed that the road had been cutoff due to overflowing creeks.
• His remaining trip had just changed from 30 miles to 55 miles on lesser roads.
Noppikoski Dam
• When the operator arrived at the dam, it was raining very hard and, he could not raise the remaining stoplog in the right bay because it had become stuck. This effectively took the hoist out of service.
• A similar problem had occurred previously but was believed to have been fixed.
• A crane company was contacted in case the hoist was the problem.
Noppikoski Dam
• The crane company was delayed in dispatching the crane because the call came in late on a Friday evening and they had trouble finding an operator.
• Additional plant staff were dispatched to Noppikoski Dam to help try and raise the stoplogs.
Noppikoski Dam
• In case there weren’t enough problems;– The telephone lines to the dam went down– An upstream dam was in danger of overtopping
and the spill into Noppikoski was increased– The power station was in danger of being
inundated and was shut down thereby losing the plant’s discharge capacity
– The mobile crane could not reach the dam, stopping a couple of hundred meters short due to the road being cutoff by a stream
Noppikoski Dam
• Operators tried to raise the embankment sections along the sides of the spillway with no success and the dam overtopped and failed.
• Luckily, the downstream residents had been warned and there were no fatalities.
Noppikoski Dam
Can you get there?Will it work?
Noppikoski DamAfterthoughts
After the failure the operator presented a paper describing the factors that contributed to the failure. In conclusion he stated:
“In my opinion, the important question of how to adapt the plants – with the exception of augmented discharge functions – to practical operation, in view of the complications of the kind previously listed, has not been considered or documented to the same extent.”
What Can We Do?
• Learn from the past– Case History Workshops– Database of Dam Safety Incidents
• Utilize Risk Assessment– Can help assess issues beyond the scope of
traditional analyses– Consider non-traditional failure modes
• Evolve our Guidelines into Best Practices that continuously capture the state-of-practice
“We are made wise not by the recollection of our past, but by the responsibility for our future.”George Bernard Shaw
