07/06/02

Staff engineers at the Nuclear Regulatory Commission last fall wanted their bosses to order the shutdown of the Davis-Besse nuclear power station by year's end.

They feared there might be dangerous cracks in nozzles in the reactor's lid. They were concerned about the lack of inspection information they were getting from the plant, according to NRC documents. And they realized FirstEnergy Corp. intended to operate the reactor without a new inspection beyond the time the NRC considered prudent.

But top NRC managers ultimately decided not to issue the order.

They reasoned it was "very unlikely" that cracks would cause the lid to rupture. They also judged that even if such an accident happened, the public's risk was acceptably low. They went along with Davis-Besse's request to operate six weeks beyond the Dec. 31 date the agency set for inspecting other nuclear plants.

What neither the NRC nor FirstEnergy, Davis-Besse's operator, knew was that the reactor not only already had cracks, but also that acid in the reactor's coolant had leaked through them and bored a 5- by 7-inch hole all the way through the heavy steel lid.

Only the lid's thin stainless steel liner, bulging into the hole, kept the radioactive coolant from spewing into the reactor building in an unprecedented accident.

NRC Chairman Richard Meserve revealed the internal debate in a letter this week responding to questions about the close call at Davis-Besse.

The questions came from U.S. Reps. Marcy Kaptur of Toledo and Edward Markey of Boston.

"Had we been aware of the degradation [rust hole], the agency would have taken the appropriate regulatory actions to shut down the reactor for the required inspection," Meserve wrote.

The NRC's inspector general is examining how the decision to delay the Davis-Besse inspection was reached, and whether it mirrored what the inspector general found to be a flawed agency decision in a similar case two years ago involving a reactor near New York City.

"When Arthur Andersen gets it wrong, the shareholders pay the price," said Steve Fought, legislative director for Kaptur. "If the NRC gets it wrong, everybody could end up paying the price." Fought said Kaptur's congressional staff is reviewing Meserve's letter and its hundreds of supporting documents. "It's not just a question about the company," he said. "It's a question about how the NRC acted too."

To justify a shutdown order, the NRC would have had to prove the action was necessary to protect the public, Meserve wrote.

Relying on FirstEnergy's own risk assessment calculations, the NRC staff determined that operating the plant past Dec. 31 could increase the risk of a reactor core-damaging accident beyond what the agency normally would consider acceptable. But the chance of such an accident posing a danger to the public was low enough to be within the agency's allowable limits.

Based on the information the agency had at the time, NRC managers decided they didn't have enough reason to halt Davis-Besse's operation early.

"After considerable deliberation and increased [FirstEnergy] management attention, it is the staff's judgment that sufficient information is available to justify operation of the Davis-Besse facility until Feb. 16," according to a confidential NRC staff report issued Nov. 30, 2001. The date was a compromise, since FirstEnergy originally had sought to operate until March 31.

During the shutdown, workers inspecting for cracks in the nozzles that allow control rods to pass through the reactor lid accidentally found the rust hole, covered by a thick layer of "lava-like" boric acid. The company is estimating it will spend $200 million buying replacement power and repairing the crippled reactor.

The issue that prompted the debate within the NRC about what to do with Davis-Besse began last year. In early 2001, operators of a nuclear plant in South Carolina similar to Davis-Besse reported finding cracks in some of the stainless-steel tubes, or nozzles, that allow the reactor's control rods to pass through the lid and into the nuclear core.

The cracks were large and threatened to encircle the nozzles, increasing the risk they would fracture and shoot out of the lid, propelled by the reactor's high operating pressure.

The NRC last August alerted utilities across the nation operating all 69 pressurized water reactors like Davis-Besse. They asked 13 plants, including Davis-Besse, that were at the greatest risk for having the kinds of cracks found at the South Carolina facility to explain why they believed their reactors could safely operate beyond Dec. 31.

FirstEnergy said it had reviewed inspection records from 1998 and 2000 and had concluded that there was no sign of leakage from nozzle cracking. It made this argument even though its own records showed that much of the reactor lid was caked with brown/red deposits of boric acid more than an inch thick and that workers had to use crowbars to remove the rock-hard deposits.

The company wanted to delay inspection until its long-scheduled April 1, 2002, shutdown for refueling. That way, costs would be lower and inspectors' exposure to radiation would be limited.

When the skeptical NRC staff asked for more information, FirstEnergy promised that it would lower the temperature of the reactor to limit growth of any cracks. It also pledged to defer inspection and maintenance of safety systems to keep them available in case of an accident.

Finally, the company told the NRC it would give reactor operators special training on how to handle the kind of lid rupture the agency believed could happen at Davis-Besse.

In an August internal memo, an NRC analyst worried that the operators' current level of training might be inadequate.

"There is concern that the event would evolve in a different manner than that expected by plant operators and cause confusion," a staffer in the Division of Systems Safety and Analysis wrote.

Union of Concerned Scientists nuclear engineer David Lochbaum said using risk assessment analysis to justify a six-week extension is flawed, charging that the NRC was merely looking for an excuse to allow the company to continue operating.

"If you think nozzles are broken and you are not allowed to operate with broken nozzles, then I think your answer is that that is your answer," Lochbaum said.

"When you don't like that answer . . . you play with the numbers again. That's what they did."