Technology's Disaster Clock

By Edward Tenner

Remember the Ixtoc I well blowout of 1979, that released about 3.3 million barrels of oil into the Gulf of Mexico over more than ten months? Not many North Americans do -- because they were less environmentally conscious, because it occurred in Mexican rather than U.S. waters, because Iran's Islamic revolution and the Soviet invasion of Afghanistan filled the airwaves and the headlines, or even because many of today's adults were too young to notice, or even unborn.

And that's one of the big problems behind the BP oil spill. In 1977 the University College London civil engineers Paul Sibley and Alistair Walker published a paper suggesting that major bridge collapses occurred at approximately 30-year intervals as new designs succeeded old as a result of the failure's lessons, new generations of designers became increasingly confident in the safety record of their innovations, until they finally pushed them over a tipping point, beginning a new cycle. The civil engineering professor and historian of technology Henry Petroski has developed this idea, which last came to the fore in the Minneapolis bridge collapse of 2007, as discussed here and here. My graduate teacher William H. McNeill coined a mordant phrase for such recurrence of disasters partially as a result of confidence in reforms, the Law of the Conservation of Catastrophe.

Do cycles of disaster apply to oil rigs as well as to bridges? Sibly and Walker thought so. In the February 12, 1976 issue of New Scientist they had the North Sea in mind when they wrote "When Will an Oil Platform Fail?" but their conclusion was prophetic for the Gulf as well:

Our studies have shown that it is a mistake to rely on the success of previous structures as an assurance of safety and that whenever vigilance is relaxed the price must be paid. With the present scale of structures the price will undoubtedly be much higher than the cost of any testing or research that could be done now.

This article available online at:

http://www.theatlantic.com/technology/archive/2010/06/technologys-disaster-clock/58367/