NRC's Process Of Calculating Risks, Seriously Flawed

by Brian Hansen

ROCKVILLE, Maryland, October 5, 2000 (ENS) - The U.S. Nuclear Regulatory Commission's new process to calculate and limit the environmental and public health risks posed by America's nuclear power plants is "seriously flawed," an environmental watchdog group charged today.

David Lochbaum, a nuclear safety engineer at the Union of Concerned Scientists (UCS), outlined the argument this morning at a meeting of the Nuclear Regulatory Commission's Advisory Committee on Reactor Safeguards.

According to Lochbaum, the Nuclear Regulatory Commission's power plant risk assessments cannot be trusted because they are based, in part, on "biased" information compiled by plant owners and operators. Moreover, the Nuclear Regulatory Commission (NRC) is analyzing the industry supplied information using a number of unrealistic assumptions, Lochbaum said.

"The NRC is guessing when it makes safety decisions using the results from incomplete and inaccurate [information]," Lochbaum said.

The nuclear power industry catalogs the information it supplies to the NRC in documents known as probabilistic risk assessments, or PRAs. These assessments are designed to give the Nuclear Regulatory Commission an analytical means with which to evaluate the severity of an accident that might occur at a nuclear power plant.

However, the risk assessments that the Commission draws up from those PRAs are necessarily inaccurate, Lochbaum said, because the NRC relies on a number of unfounded assumptions in the course of analyzing them.

Those assumptions, which are spelled out in official Nuclear Regulatory Commission documents, were outlined in an UCS report released earlier this summer. According to the report, the Nuclear Regulatory Commission drafts risk assessments for nuclear power plants assuming that:

* Nuclear power plant reactor pressure vessels never fail.

* Nuclear power plant workers make few serious mistakes.

* Nuclear power plant design and construction are completely adequate.

* Nuclear power plants operate within technical specifications and other regulatory requirements.

* Risk at nuclear power plants is limited to reactor core damage.

* Equipment at nuclear power plants fails at a constant rate.

"History shows there is a greater probability of a flipped coin landing on its edge than of these assumptions being realistic," Lochbaum concluded in the UCS report. "In computer programming parlance, 'garbage in, garbage out.'"

But beyond that list of unfounded assumptions, one of the most serious problems with the NRC's nuclear power plant oversight program is that there are no objective standards that plant owners must follow in compiling their probabilistic risk assessments, Lochbaum said.

Lochbaum cited the case of the Wolf Creek nuclear reactor in Kansas and the Callaway nuclear reactor in Missouri to make his point. Although the two facilities were built using the exact same blueprints and the exact same materials, their PRA results - and their resultant NRC risk assessments - are "as different as Dr. Jekyll and Mr. Hide," Lochbaum said.

"The numbers make it look like Wolf Creek is the good twin and Callaway is the bad twin," Lochbaum wrote in the UCS report. "In reality, these risk assessments cannot be used to decide this sibling rivalry, [because] they were developed using different methods and different assumptions.

"It is therefore no surprise that their results differ so radically," Lochbaum said. "The data do not allow the safety levels of these identical plants to be evaluated, even on a relative basis."

Lochbaum also cited cases where nuclear power plants built to significantly different design and operational specifications were declared by the NRC to have the same exact risk factors.

For Lochbaum, those types of findings indicate that the Nuclear Regulatory Commission is not capable - or not interested - in determining which plants are safe, and which are not.

"Apparently, the NRC views being the same as OK, and being different as OK, too," he said. "That's curious."

George Apostolakis, the vice chairman of the Nuclear Regulatory Commission committee, challenged Lochbaum on that point. Apostolakis acknowledged that nuclear power plant owners have been known to submit inaccurate or inconsistent information to the NRC, but he denied that those reports would result in the drafting of misguided risk assessments.

"It's a private industry, and if they want to use optimistic numbers, it's up to them," Apostolakis said. "But when they try to use them here, that's a different story."

According to Apostolakis, nuclear power plant risk assessments are based on far more criteria than simply the "optimistic" PRAs provided by the nuclear power industry.

"You give the impression that our decisions rely on PRAs," said Apostolakis, who also serves as a professor of nuclear engineering at the Massachusetts Institute of Technology. "You use the word 'rely' too much."

Lochbaum, asked if he thought he put too much emphasis on the industry's contribution to the nuclear regulatory process, said, "I think I underestimated it."

Apostolakis acknowledged Lochbaum's point, but the NRC committee member would not admit that nuclear power plant risk assessments have been compromised as a result.

"I grant you that there may have been situations where people have misused the process," Apostolakis said. "It would be interesting to find out why these things are happening - whether it's just judgment, or accounting."

But Apostolakis would not concede Lochbaum's point that "all possible" measures should be taken to prevent and mitigate every risk posed by the construction and operation of nuclear power plants.

"Are you seriously proposing that all possible measures be taken?" said Apostolakis, whose research interests include mathematical methods for risk and reliability assessment of complex technological systems. "Is that an exaggeration to make a point, or are you asking us to take you literally?"

Apostolakis noted that no activity is risk free, and that many endeavors - such as automobile and airplane travel - have been regulated to pose certain levels of risk.

As a matter of policy, the NRC attempts to limit risks associated with nuclear power plant operations to less than one percent of the risks that the public faces from other types of accidents.

However, current NRC risk assessments consider only the probability of an accident happening -not the associated consequences. When consequences are very high - as they are from nuclear plant accidents - prudent risk management strategies dictate that probabilities be kept very low, the Union of Concerned Scientists argues.

"An accident at a U.S. nuclear power plant could kill more people than were killed by the atomic bomb dropped on Nagasaki (Japan)," the UCS report declares. "

The Union of Concerned Scientists has called on the Commission to abandon its "unrealistic" risk assessment program in favor of one with objective and verifiable standards.

The group has also appealed to Congress to provide the NRC with the budget it needs to "restore the safety margins" at America's nuclear power plants.

The Reactor Oversight Process and Plant Assessment Results are available on the Nuclear Regulatory Committee website at:

The Union of Concerned Scientists website is:

Back to