Reactors Prone to Long Closings, Study Finds

New York Times

By Matthew L. Wald

September 18, 2006

WASHINGTON - An analysis of nuclear reactors by a safety group has found that they are prone to costly, lengthy shutdowns for safety problems regardless of their age or the experience of their managers. The finding could have implications for companies considering building new reactors.

The analysis, by David Lochbaum, a nuclear engineer at the Union of Concerned Scientists, counted 51 times that a reactor had been closed for a year or more. Thirty-six of those shutdowns were to restore an adequate level of safety by fixing flaws in equipment, procedures or training; 11 were to replace major components required for operations and safety; and 4 were for damage recovery. In all, of the 130 power reactors ever licensed, 41, were closed for at least a year. Ten were closed twice.

Mr. Lochbaum said the most common reason for a shutdown was for an "attitude adjustment" for workers and managers, so they would be more attuned to safety. He said he was surprised by some of his findings, which are scheduled to be released Monday. "I expected that the first plant off an assembly line would have been challenged, or troubled, but that there was a learning curve, and the fourth or fifth or sixth plant for a company would have avoided these problems," he said. "But it wasn't the case."

But a vice president of the industry's trade association, Marvin Fertell of the Nuclear Energy Institute, said that the industry had, in fact, learned from its errors, and that only experienced operators would build new plants. And at the Nuclear Regulatory Commission, Stuart A. Richards, deputy director of the division of inspection, said his agency had improved its inspections, to focus on "risk-significant areas," and was now able to find problems more promptly.

Extended shutdowns would be a bigger problem for future plants because, in the past, electricity customers of regulated utilities paid for them. But some of the reactor construction projects now being considered would be built as "merchant" plants, with no guaranteed income, only revenue from power sales.

The heart of the problem, Mr. Lochbaum said, is that the Nuclear Regulatory Commission is not good at assessing the ability of a reactor staff to keep the plant in good physical condition and maintain training and other requirements. As a result, he said, plants operate until serious problems accumulate and force a shutdown.

"This is the wrong way to do business, from a safety standpoint and an economic standpoint," he said in a telephone interview.

Mr. Fertell, of the industry trade group, agreed.

The only reactor currently in an extended shutdown is the Tennessee Valley Authority's Browns Ferry Unit 1, in Alabama. It last ran in 1985. The shutdown of more than a year that ended most recently was at Davis-Besse, near Toledo, Ohio, where workers found that an acid used in the plant, boron, had corroded a 70-pound chunk of steel in the reactor's vessel head, leaving only a half-inch stainless steel liner.

Early in the era of commercial nuclear power, analysts theorized that shutdowns were what was known in the industry as "teething problems" and that with experience, reactors would run more smoothly. But most of the shutdowns came after the reactors were 10 years old. The Davis-Besse plant was more than 23 years old when it was closed in 2002. It was closed for more than two years. Besides the hole in the reactor head, engineers later found that crucial pumps that used water for lubrication were prone to break down because of debris in the water. Discovery of decades-old design problems is common during lengthy shutdowns.

While Mr. Lochbaum, a longtime adversary of the nuclear industry, is often critical of the companies that operate reactors, he said regulators were the problem in this area. The rules require reactors to have Corrective Action Programs to keep track of physical and procedural problems, and each lengthy shutdown is an indication that the program itself is flawed, he said. Regulators monitor the physical condition of reactors, he said, but are not good at observing the quality of the corrective programs. For example, the commission gave high marks to the program at Davis-Besse less than a year before inspectors found that operators had let acid eat through six inches of steel, bringing the plant close to a catastrophic rupture.

Mr. Richards said he had not seen the report but acknowledged errors by the commission in handling the Davis-Besse case. But he said N.R.C. inspections had been improved using a new process, of which the Corrective Action Program itself was a major component. And, he said, the commission had previously penalized reactors for accumulations of minor violations, adding them up to count for a major problem; now it focuses only on major problems.

Mr. Lochbaum said that after a reactor was shut down for one reason, other problems were often discovered. In an extended shutdown at the Crystal River plant, in Florida, workers found design defects even though the plant had been running for nearly 20 years. He said the problems included that, in an emergency, the pumps would not have worked as intended and piping would have exposed workers and the public to radiation.

"Did the plant's owner bring in busloads of smarter workers after the N.R.C. put the reactor on notice?" Mr. Lochbaum asked in the report. The problem, he said, was that perception by the inspectors that plant management was competent was blinding them to problems at the reactors.

But Mr. Fertell, of the Nuclear Energy Institute, said an extended reactor shutdown often became "a monster can of worms." "You were basically under a magnifying glass," he said, with inspectors finding issues faster than management could resolve them.