Dam Failures

DAM FAILURES – at Durrington Community Centre, 9th March, 1988, by A.C. Twort, Member.

The talk was well illustrated by diagrams and photographs, and covered 14 dams that failed. Only a selection of Failures are reported here. Dams over 45′ high are registered. Of the 500 registered in this country at least half are 100 to 150 years old – illustrating the considerable dam building in the latter half of the 19th century.

Dam foundations are vital. At Bouzey, France in 1884 a section moved out because of poor foundations. It’ was repaired only to fail again 11 years later, killing 150 people. It was a gravity dam, 1705′ long 60′ high.

Its failure contributed to the adoption of the “middle third” rule, which ensures no tension on the up stream side due to the overturning moment of the stored water overcoming the weight of the dam. Where this occurs water can seep through the dam and scour away material, causing slippage.

The worst case of such failure was at Los Angeles – The St. Francis dam . About 200 feet high it failed about 1923 creating a flood wave 125 ft. high running at 18 mp.h. for 40 miles. Over 400 people died.

Sometimes when foundation conditions are difficult with irregular rock a buttress dam is used. While theoretically more economical than gravity ones, construction is difficult. Also, should one buttress fail the rest follow “like a pack of cards”! They are seldom used.

Water seeping into a crack or under the foundations causes uplift of the dam due to the head below the surface of the reserviour. This can seriously modify the stability of the dam derived from the “middle third” rule. Porosity of concrete in dams produces similar effects. It is particularly so in Masonry dams, in which the joints cannot be sealed effectively.

Slippage due to this was dramatically illustrated by the Baldwin Hills, Los Angeles dam- in 1963. It was well built and maintained. Nevertheless a leakage was observed near the top. It developed rapidly, causing the road across the top to fall and within 4 hours the reserviour was empty!

The Dolgarrog, disaster in 1925 was due to poor foundations of the upper of two dams in Snowdonia. They were only 2 feet into the clay base, allowing seepage enough to scour the soil away. The flood caused 16 deaths; fortunately many villagers were at the local cinema, which was away from the route of the flood. The dam watchman failed to warn the village because the flood disrupted the telephone line.

Following this a new Act was passed in 1930, placing full responsibility on the certifying engineer alone for the integrity of the dam. There have since then been no fatalities due to such failures in the U.K.

An arch (concrete) dam depends on the abutments at each end. If one fails the dam disrupts very quickly. The Frejus disaster in France,
1959 is a dramatic example. It was a thin arch 200 ft, high, 22 ft. thick at the base tapering to 5 ft. at the top. Some 10,000 million gallons of water were released. Over 400 people died and the Marseille – Nice railway and road were washed away some 12 miles away by the coast.

The problem of what excessive flood conditions to allow for during construction is illustrated by the Oras Dam, Brazil. A tunnel had been constructed to by-pass 8,000 million gallons a day. But one week 25″ of rain fell, vastly in excess of any expectations, in March 1950 and overwhelmed the dam, flooding some 200 miles down stream.

In the discussion John Fowler mentioned a dam at Fontana, U.S.A., where a crack was “stitched” with cables! And Ken Lambert mentioned the danger of a number of “amateur” dams built by farmers, for instance – which were poorly maintained. The problem of shortage of experienced examiners because few dams are now being built was mentioned.