For over 20 years, a corps of ex-Navy engineers, lawyers, and inspectors has worked quietly from two towers here in Maryland to ensure that the most dangerous technology ever devised operates safely in civilian hands.
It's likely that most Americans seldom give this federal agency, the Nuclear Regulatory Commission, a second thought. That's because it has worked relatively well over the years, if the safety record of the 110 nuclear power plants in the US is any measure. There's been only one major mishap on the NRC's watch: the 1979 Three Mile Island partial core meltdown.
But the NRC, like the facilities it oversees, today faces an uncertain future. Whistleblowers say the agency is lax in its oversight and too cozy with the industry it regulates. The result, they say, is a growing danger of another serious nuclear accident.
Such criticism isn't necessarily new - but it's increasing in volume at a time when fundamental change in the nuclear business has the potential to undercut the industry's own safety efforts.
A review of NRC documents and transcripts, as well as conversations with inspectors and industry insiders across the country, indicates that the agency may, indeed, have problems. Specifically, the Monitor has found that:
*NRC inspectors may too often be willing to take a utility's word that it is addressing troubling problems, instead of doing inspections at the plant or verifying calculations to ensure the problems are fixed.
*Higher-level NRC managers sometimes downgrade the severity of safety problems identified by on-site inspectors without giving reasons for the change.
*The agency is too slow to act when confronting potentially dangerous problems that could affect plants using similar reactor designs.
*NRC inspectors who persist in pressing safety issues have been subjected to harassment and intimidation by their supervisors.
Whether such flaws, in themselves, could lead to another Three Mile Island is debatable. But they arguably reflect on the thoroughness with which the federal government monitors the awesome power generated by splitting atoms.
"What I'm afraid of is the next nuclear accident," says one veteran NRC inspector. "We need nuclear power, but the watchdogs are not watchdogs."
What critics find particularly troubling is that the NRC's apparent problems come at a time when demands on the agency are growing, the nuclear utilities they regulate face an uncertain economic future, and key parts of many plants are wearing out faster than anticipated.
Drawing a parallel to some of the pressures on the National Aeronautics and Space Administration before the Challenger explosion in 1986, David Lochbaum, a longtime industry consultant, says: "It is eerie how applicable those concepts are to the current condition of the nuclear industry."
Deregulation is a major cause of the pressure on the industry. In many states, consumers will soon be able to pick the cheapest supplier of electricity they can find, similar to the way they now opt for the best deal in long-distance phone service. This puts nuclear power at a disadvantage because it's a relatively expensive way to produce kilowatt-hours.
For instance, seven of the eight nuclear power reactors in New England may soon be too high-cost to stay in business, according to a recent study conducted for the Massachusetts State Attorney General's Office by the Resource Institute, a private research firm.
Aging plants are another source of fiscal problems. As nuclear facilities get older, they face expensive technical troubles, such as weakening and cracking in the steel shell that surrounds nuclear reactors. In some cases, utilities have elected to pay hundreds of millions of dollars to repair plants with age-related problems. At least two plants - Yankee Rowe in Massachusetts and Trojan in Oregon - have shut down permanently because age-related repairs were too expensive.
Utility executives acknowledge the challenges ahead. But they dismiss concerns that plants are being pushed beyond safe limits. They point out that, by many measures, US plants are today the safest, best-run nuclear facilities in the world.
All the industry's indicators to measure plant problems are trending down and, by the end of last year, bettered five-year goals set in 1990, says Joe Colvin, president of the Nuclear Energy Institute (NEI) in Washington. Take automatic "scrams" - unplanned, automatic reactor shutdowns caused by equipment problems. The number of such malfunctions has fallen markedly, Mr. Colvin notes. In 1990, the industry adopted a goal of an average of one automatic scram for every 7,000 hours (roughly one year) of plant operation. Last year, it bettered the goal by achieving 0.9 scrams per 7,000 hours.
This decline has contributed to an overall improvement in the efficiency of US nuclear power, according to industry figures. Last year, US nuclear plants ran at an average of 82.6 percent of capacity, up from 75.1 percent in 1994.
"Several years ago people finally realized that if you put more effort into improving safety and reliability, you would reduce unplanned outages," Colvin says.
He says increased efficiency is helping the industry become more competitive with other suppliers of electricity. NEI calculations show that the inflation-adjusted average price for a kilowatt-hour of electricity produced by nuclear power fell 12 percent in 1994, the last year for which full data are available.
But the economic viability of a nuclear utility rises and falls on its individual record, not industry averages. Entergy, a New Orleans-based holding company that operates five nuclear plants in the South, recently examined its economic prospects. According to spokesman Terry Young, Entergy's facilities will have to run at an average of 87 percent capacity within three years if the electricity it produces is to be even marginally competitive with other sources in the region.
Such statistics, say nuclear critics, show how hard the industry must push its equipment to stay in business. They predict that financial pressures will force many plant operators to push even harder, imperiling safety.
One example of the types of problems that can arise - cited by the NRC itself - is the Wolf Creek nuclear plant in Kansas. Two years ago, in an effort to shorten shutdown time for repairs, workers tried to do two tasks at once. They began testing repairs to one cooling system while a twin heat-removal system was still in operation.
The unexpected result: In 66 seconds, 9,200 gallons of water that should have been cooling the reactor core instead drained into a holding tank. The NRC says that if another 3 to 5 minutes had elapsed, the reactor would have been 30 minutes from a meltdown.
While the nuclear industry faces growing financial pressures, so does the agency overseeing it. The NRC's budget reached $540 million in 1994 but has since slipped to $473 million for the current fiscal year.
Ironically, the NRC's budget has no direct effect on the federal balance sheet. The agency's money comes from fees imposed on nuclear utilities, not from Uncle Sam's pocket. In principle, NRC officials could just charge higher fees if they felt they needed more cash. In practice, however, NRC budgets must still clear the tight-fisted congressional appropriations process.
"We are the only check on the resources committed," says a staff member of the House appropriations subcommittee that has jurisdiction over the NRC's budget. "We're the only voice for the utilities and the ratepayers."
Budget problems have already caused the NRC to consider cutting some of its oversight responsibilities. For example, the agency has been experimenting with a process that would give utilities a greater role in testing the fitness of their own reactor operators.
To some insiders, this experiment represents a dangerous precedent. In late 1994, an NRC official turned down a request from the Virginia Electric Company to do pretty much the same thing - develop and administer its own reactor-operator license tests. Such tests could lead to wide variations in evaluations, open the door to biased exams, and create a conflict of interest for utilities, which have a major stake in seeing their operators pass, said an NRC official at the time.
But last year the NRC went ahead and approved a modified version of this self-test approach. It gave utilities at 21 plants the authority to draft and partially administer their own exams, which the NRC would grade. Among the benefits of the change: The NRC could save $3 million to $4 million a year.
Another uncertainty now facing the nuclear agency is the future of the very industry it is supposed to watch. Nuclear power is a stagnant industry in the United States - and some NRC officials worry about regulating it out of existence.
"We've had no new license applications for years, and maybe we'll get none for the rest of the NRC's life," says one insider at NRC headquarters. "Whether it's said or not, an awful lot of managers in the NRC have in the back of their mind the viability of the nuclear industry."
It's just this kind of problem that may lie at the heart of some inherent conflicts at the NRC. Many critics claim, for instance, that higher-ranking NRC managers are often less strict than on-site inspectors when it comes to problems at plants. On-site overseers have a more by-the-book approach to their work. They say: "This is the regulation, and you've passed or not passed," says the NRC insider. "For resident inspectors, it's a black-and-white issue. If there's a serious accident, they know the first question will be: Why didn't the resident pick up on that?"
"[NRC] managers interact with utility vice presidents," he continues. "They're more likely to say: This is a technical violation, not a threat to public health and safety. There's a concern that if you apply regulations written in the abstract, you'll get into a situation where plants shut down prematurely for minimal safety issues."
The gist of the problem, surmises Tom Devine, legal director for the Government Accountability Project in Washington, is that "if the NRC enforces safety laws, nuclear power plants that already are lemons will become financial disasters." The NRC, he adds, "has never transcended the conflict of interests inherent in the old Atomic Energy Commission," the NRC's predecessor, which had the contradictory role of promoting the use of nuclear energy while regulating its use.
The old AEC, whose top officials often came from the nuclear Navy, may have left another legacy in today's NRC: the "Navy way."
"There are cells scattered around the agency" of managers schooled in the Navy way, says a senior plant inspector who joined the NRC in 1976. "They tend to collect people who'll do what they're told to do, no matter what. I've had managers who when you say, 'You know that's wrong,' will say, 'Yeah, but that's what my boss wants me to do.' "
In theory, the NRC isn't a top-down autocracy. Opportunities for dissent exist, under the rubrics of "peer review" and "differing professional opinion." These processes are "superficially collegial, but intolerant and repressive when it counts," Mr. Devine says. "Typically, the NRC will cordially begin an open-ended process of debating an issue. Things don't get ugly until an employee wants corrective action taken that is costly to industry."
While conceding some problems, NRC officials insist their agency has been effective. Moreover, the NRC is tightening its processes under its new chairwoman, Shirley Jackson. Individual incidents notwithstanding, she says in an interview, "the staff is a very technically competent, very able group of people who are very focused on safety. I think one should not lose sight of that."
* Next: The Ballad of Watts Bar.