WASHINGTON — When does a nuclear plant become too old?
The nuclear industry is wrestling with that question as it tries to determine whether problems at reactors, all designed in the 1960s and 1970s, are middle-aged aches and pains or end-of-life crises.
This year, utilities have announced the retirement of four reactors, bringing the number remaining in the United States to 100. Three had expensive mechanical problems but one, Kewaunee in Wisconsin, was running well, and its owner, Dominion, had secured permission to run it an additional 20 years. It was losing money, though, because of the low wholesale price of electricity.
“That’s the one that’s probably most ominous,” said Peter Bradford, a former member of the Nuclear Regulatory Commission and a former head of the Public Service Commission in New York. “It’s as much a function of the cost of the alternatives as it is the reactor itself.”
While the other three, San Onofre 2 and 3 near San Diego and Crystal River 3 in Florida, faced expensive repair bills because of botched maintenance projects, “Kewaunee not only didn’t have a major screw-up in repair work, it didn’t even seem to be confronting a major capital investment,” he said.
This is a turnaround because until recently the life expectancy of reactors was growing. When the Nuclear Regulatory Commission began routinely authorizing reactors to run 20 years beyond their initial 40-year licenses, people in the electricity business began thinking that 60 was the new 40. After the past few weeks, however, 40 is looking old again, at least in reactor years, with implications for the power plants still running, and for several new ones being built.
“They were intended to last for as long as they were commercially feasible,” said Robert Curry Jr., who was a member of the New York Public Service Commission from 2006 to 2012. But with low gas prices, additional costs imposed after the Fukushima Daiichi accident of March 2011, and “the general mistrust of nuclear by anyone who saw ‘The China Syndrome,’” commercial feasibility now is shorter, he said.