Workers at Millstone Unit 2 in Connecticut discovered two jets of steam coming from the casing of an isolation valve for the letdown system on May 24, 1993. The letdown system is connected to the primary coolant system of this pressurized water reactor. A small flow rate of water is withdrawn from the primary coolant system by the letdown system, filtered and purified before being returned to the primary coolant system.
The leaking isolation valve casing allowed primary coolant water to leak into the containment. An attempt to stop the leaks failed, but since the leakage was within allowable limits, the plant restarted.
Ten days later, on June 4, workers attempted to stop the leaks by injecting a paste-like substance into the valve casing. Workers drilled holes into the valve casing and injected the sealant. The plan called for the sealant to be forced into the hole by the escaping water. The sealant would then harden to plug the leak. The initial repair worked – for about an hour. Then the leak resumed.
W. C. Fields said, “If at first you don’t succeed, try, try again. Then quit. There’s no use being a damn fool about it.”
The damn fools at Millstone tried 28 more times to stop the stubborn leaks. Eventually, they drilled so many holes in the valve casing that it broke. The reactor was shut down on August 5, because the leakage now exceeded legal limits.
The Millstone gang might still be squirting paste into the valve if it hadn’t broken. Where’s a Dutch boy when you need one?
In this episode, the first attempted repair was a reasonable effort consistent with industry practice. It failed. The second attempt was also deemed reasonable because it was like the first. And the third and succeeding attempts were equally justified as being little different from the prior attempts.
How can one avoid falling into such incremental traps? One way would be to define failure criteria before the first attempt. In other words, if the attempted repair is not successful, pre-determine what the next step would be rather than merely repeat the first one. Another way would be to formally ask what is different during repetitive attempts. In other words, if five prior attempts have failed, what will be different about the next attempt to justify trying it?
Better still, before undertaking another attempted repair, try to finish this answer to the boss’s question: “Well, boss, the reason I thought this was a good idea was …”. If there’s not a solid answer, there’s also not a solid reason to proceed.
“Fission Stories” is a weekly feature by Dave Lochbaum. For more information on nuclear power safety, see the nuclear safety section of UCS’s website and our interactive map, the Nuclear Power Information Tracker.
Support from UCS members make work like this possible. Will you join us? Help UCS advance independent science for a healthy environment and a safer world.