Mishaps at 3 US Nuclear Plants

By Matthew L. Wald, The New York Times

11 May 11

he Nuclear Regulatory Commission said this week that in hindsight, a problem at the Browns Ferry nuclear plant in Alabama last fall was quite serious. And its records indicate that there have been reports in recent days of engineers' flubbing a basic calculation of reactor operation at two other plants.

The math errors and the Browns Ferry problem are not related to each other but come at a moment of heightened concern about reactor safety after the meltdown at the Fukushima Daiichi reactors in Japan.

On Tuesday, the commission staff announced that a valve that got stuck last October at the Browns Ferry nuclear plant near Athens, Ala., posed a safety threat that fell into the "red" category, the most serious on its four-color scale.

It is only the fifth time since the scale was established in 2001 that the commission has put a problem into that category.

"It took us a while to work through the inspection findings," said Roger Hannah, a spokesman for the agency's Atlanta regional office, explaining the delay. That process involved conferring with the plant's operator, the Tennessee Valley Authority.

And last week, operators of the Oyster Creek nuclear plant near Toms River, NJ, and Nine Mile Point 1, near Syracuse, reported separately that General Electric, which supplies their reactors with fuel, had notified them that it had made some math errors that could have resulted in the reactor fuel's getting hotter than plant operators thought. But General Electric notified the companies in time for them to make changes to their reactors.

Neither the fuel miscalculation nor the valve problem caused any damage or injuries, but both exposed flaws in the system. The valve problem at Browns Ferry could result in a fine, Mr. Hannah said.

The valve, at Browns Ferry Unit 1, is located in the reactor's residual heat removal system, which enables the reactor to cool after it has shut down. The failure was discovered when operators intentionally shut the reactor down for refueling, one of the few times that the heat removal system is used. It would also be used if there were an accident that required the reactor to shut down.

The failure of the residual heat removal system is a cause of the meltdowns at the Fukushima reactors, which are of a similar design to those at Browns Ferry, Oyster Creek and Nine Mile Point 1.

The valve consists of a flat metal disk that sits inside the pipe and is moved by a rod. The rod had become disconnected from the disk, apparently some time weeks or months earlier. The residual heat removal system consists of two separate sets of pumps, valves and piping, and one set would not have worked, Mr. Hannah said.

"The public was never endangered because no actual event occurred," the commission said. "However, the system is counted on for core cooling during certain accident scenarios, and the valve failure left it inoperable, which could have led to core damage."

Such an accident would nonetheless involve "a series of unlikely events," the commission said.

The Tennessee Valley Authority told the commission staff that the valve had a manufacturing defect but that it would still have opened if needed; the commission staff disagreed.

Other "code red" findings by the commission include the notorious case of the Davis-Besse reactor near Oak Harbor, Ohio, where an unnoticed leak allowed acid to flow onto the vessel head. Workers discovered in 2002 that it had corroded away all but a fraction of a thick stainless steel liner. A red finding triggers an extensive inspection; at Browns Ferry that will include equipment, management and procedures to see if anything else is wrong there, Mr. Hannah said.

The fuel problem at the Oyster Creek and Nine Mile Point plants was odder. General Electric notified Oyster Creek on May 4 and Nine Mile Point on May 6 of four separate calculation errors it had made in determining the peak cladding temperature of a new fuel that the plants were using in their reactor cores, operators said.

In some cases the errors caused GE to underestimate the peak temperature, and in others, overestimate it. The calculation is important because it helps determine the maximum power level at which a reactor can safely run.

The mistakes also involved the calculation of how much of the cladding would be damaged in an accident.

Scott Burnell, a spokesman for the Nuclear Regulatory Commission in Washington, said that when the calculations were corrected, each plant was still legal to operate but had less of a safety margin than the operators believed.

The operators can solve the problem by changing the power distribution within the reactor core with no loss of output, he said. Mr. Burnell said this would result in the fuel being used up slightly sooner, meaning that the operators were not getting all the possible energy out of their uranium.

He said it was not yet clear if the same errors had been made in fuel dispatched to other plants.

Mark Sullivan, a spokesman for Constellation Energy, which owns the Nine Mile Point reactors, said, "There is no safety concern with the vendor's calculation changes." Operators had made changes in the plant's software to assure the fuel was not overheated and to preserve their safety margins, he explained.

Nine Mile Point got the notification before it started up after a shutdown for refueling, but Oyster Creek did not; still the operators of the latter said that "this event is not significant with respect to the health and safety of the public."

Both plants are already the focus of petitions brought before the Nuclear Regulatory Commission by a member of the public, Mark Leyse, arguing that the commission has not paid adequate attention to experimental data that shows how severe a reaction could occur if cooling water were lost.

Mr. Leyse says that damage to the fuel cladding and production of hydrogen would occur faster than the commission assumes in its calculations, and that the peak temperature should be lowered. The commission has not ruled.