Federal Probe of Maine Nuclear Plant

An anonymous employee alleges engineers manipulated computer simulations to win federal approval to boost its power output

AN anonymous letter has triggered an intensive safety investigation of a Maine nuclear power plant in Wiscasset.

A team of inspectors from the US Nuclear Regulatory Commission began a probe yesterday at Yankee Atomic Electric Company in Bolton, Mass., which provides engineering and technical services for the Maine plant.

The letter, sent by someone claiming to be an employee, alleges that engineers manipulated computer simulations to hide potentially serious deficiencies in the reactor's emergency-cooling system.

The allegations, true or not, highlight two areas of growing concern among nuclear-safety advocates: a lack of NRC vigilance in checking documents that utilities file to change operating procedures and the willingness of some nuclear utilities to exploit this laxity to save money.

''We're taking these allegations very seriously,'' says Kelly Smith, spokeswoman for Yankee Atomic. Since the allegations first surfaced last week, she says the company has formed two review teams to check their validity.

The federal response

The NRC pulled together its own inspection ''after we learned of the letter,'' says Victor Dricks, a spokesman for the NRC's Region I office in King of Prussia, Pa.

The allegations are contained in a three-page letter to Robert Pollard, senior nuclear safety engineer at the Union of Concerned Scientists in Washington, D.C., a nuclear energy watchdog group.

The anonymous letter, Mr. Pollard says, was accompanied by 128 pages of documents. The person, he says, knows the subject ''and has access to documents that are not publicly available.''

The writer claims to have worked at Yankee Atomic for several years. He alleges that the plant operators knowingly used faulty data to create a computer simulation showing the plant could safely operate at higher power outputs.

The letter writer is asking that Maine Yankee's power rating be cut by roughly 300 megawatts to its original 2,400 MW and that both the plant and Yankee Atomic be fined.

The letter asserts that company officials openly spoke of timing their request for a higher power rating so as to ensure that one NRC official, ''considered to be a particularly lenient person,'' was still in a position to review the request and that the request would arrive between Thanksgiving and Christmas, ''when the NRC staff is least vigilant.''

In addition, the letter asserts that the NRC granted two prior increases in Maine Yankee's reactor output based on assessments improperly altered to show that the containment building would withstand the increase in pressure if a large coolant leak occurred at the higher power ratings.

According to the letter, after the 1979 accident at the Three Mile Island nuclear-power plant outside of Harrisburg, Pa., the NRC required all nuclear plants to improve computer models that simulate how their safety systems would hold up under various accident scenarios. One involves a so-called small-break leak in the core cooling system, which, if unchecked, could lead to a core meltdown.

From 1980 to 1983, Yankee Atomic developed a new model, which according to the letter predicted the results of a number of experiments ''reasonably well.'' Yet under scenarios involving a small-break, loss-of-coolant incident, the plant's emergency cooling system failed to prevent the metal tubing (known as fuel cladding) that surrounds the fuel pellets from exceeding 2,200 degrees F. Fuel damage begins to occur at around 1,800 degrees F, and the exterior cladding begins to melt if the temperature exceeds 2,200 degrees for a prolonged period.

According to the letter, Maine Yankee officials refused to even discuss the possibility of upgrading the emergency core-cooling system. Over the next four years, Yankee Atomic tried a variety of changes to the program. ''But with any reasonable code modification and input parameters ... the fuel-rod cladding temperature was calculated to exceed 2,200 degrees F,'' according to the letter.

Under pressure from the NRC to close out the code-revision project, Yankee Atomic submitted the code to the agency for review in 1987. According to the letter, company officials decided to apply for another increase in reactor output before the NRC could finish reviewing the new program. Instead of using its new code, the writer alleges, the company used its pre-Three Mile Island simulation program and asked that the plant's reactor be allowed to run at 2,700 MW, rather than 2,630 MW. It also told the NRC that it was working on a new analysis that would meet the post-TMI requirements.

Still too hot

In 1989, the NRC granted Maine Yankee's request for the power upgrade. In 1990, under pressure from the NRC to supply the more up-to-date analysis, the company used its new code. But the results were still too hot. The letter alleges that officials then replaced two key parameters in the model with estimates from sources that either lacked NRC approval or differed from those used to license the plant. Down came the temperatures, and this estimate became the final basis for granting the power increase. The NRC acknowledged receiving this estimate but never checked it, according to the letter.

Maine Yankee has been shut down since January for repairs to its steam generators. The repair technique being used - inserting metal sleeves into 17,000 corroded steam-generator tubes - retards the flow of coolant through the reactor. This reduces the cooling system's ability to carry away heat. Large-scale sleeving typically forces operators to run reactors at lower power ratings to compensate for the reduced coolant flow. Plant officials, however, have assured Maine state regulators that the plant will be able to run at its full 2,700 mw, despite the large number of sleeves.