You wouldn't know it from the bland pronouncements of the Nuclear Regulatory Commission (NRC), but the U.S. nuclear industry just had its closest brush with disaster since the 1979 Three Mile Island accident. The Davis-Besse nuclear power plant, located about 30 miles east of Toledo, Ohio, was operating with a rust hole in the top of its reactor pressure vessel -- a hole wide and deep enough to put your fist into. All that was left to contain the reactor's highly pressurized supply of cooling water around the reactor core was a three-eighths inch liner of stainless steel, and the liner had started to bulge ominously. If the liner had burst, it would have drained cooling water vital for safety and also threatened the reactor's emergency shutdown system.
The plant operator's neglect is bad enough. If this had occurred in Russia, we would be saying it could never happen here. Equally disturbing is the NRC's barely audible response.
The preliminary report of FirstEnergy, the nuclear plant owner, details what happened. During a routine refueling shutdown in February, the company inspected several dozen nozzles to check for cracks, as required by the NRC. The nozzles, located on the head of the reactor vessel, permit control rods to enter the vessel to shut down the reactor, quickly if necessary. A workman discovered the rust hole by luck -- when he happened to bang into one of the control rod tubes coming out of the top of the reactor and it moved. If the reactor had gone back into operation, as it very nearly did, the consequences could have been enormous in terms of public safety as well as the future of the nuclear industry.
It turned out that corrosion had reduced 70 pounds of steel, half a foot thick, to rust. The corrosion was caused by boric acid on the outside of the head. How did the acid get there? The water inside the reactor vessel contains dissolved boric acid, which is used to assist reactor control. Because boric acid corrodes carbon steel, the reactor vessel's interior is lined with stainless steel. The boric acid is not supposed to get to the vessel's exterior, which remains vulnerable to corrosion. But at Davis-Besse the reactor's water leaked through cracks -- it still isn't clear which ones -- and created a boric acid crust on the outside of the reactor head.
This accumulation and damage doesn't happen overnight. The company report explains the hole hadn't been found earlier because, "Boric acid that accumulated on the top of the [Reactor Pressure Vessel] head over a period of years inhibited the station's ability to confirm visually that neither nozzle leakage nor vessel corrosion was occurring." In plain English that means that the company watched the boric acid crust cover an increasing area of the head for years and did nothing about it. That's not all. Some of the reactor vessel rust became airborne and clogged the reactor building's air filters. The filters had previously been changed monthly, but from 1999 on they had to be changed every other day. The company's report says the possibility of corrosion "was not recognized as a safety significant issue by the staff and management of the plant." Obviously the NRC, which had inspectors on site, did not recognize it either.
How important is this? The reactor vessel head resembles a rounded lid that is bolted to the vessel. It's about 15 feet in diameter. The reactor vessel and the vessel head are designed and manufactured with exquisite care from special steel a half-foot thick (with the thin liner of stainless steel). The vessel and head of every reactor have to be monitored throughout their life to make sure that radiation has not caused the metal to become brittle. This is vital because the NRC licensed the plant on the assumption that a break in the reactor vessel is not credible. As a result, the reactor's safety analysis does not deal with breaks in the vessel wall. The reactor's emergency actions operators are trained to cope with breaks in pipes, not the vessel. Some safety systems might work for such a break; then again they might not. The problem was not studied. There would likely be unforeseen complications.
An obvious complication would involve malfunctioning of the control rod system that is supposed to stop the chain reaction in an emergency. There is no backup to the control rods for immediate shutdown. The plant's safety analysis considers the possibility that a limited number of rods, out of several dozen, could fail to drop. The control rod adjacent to the rust hole would have been one of these. But what about the damage that might be caused to other control rod drives above the head if a hole in the vessel unleashed a jet of steam and water coming out of the pressurized vessel? A telling sign that the industry understands the seriousness of the Davis-Besse problem is the silence from the Nuclear Energy Institute, the industry's lobbying arm, which is usually quick to spin a nuclear story. All in all, what happened at Davis-Besse was a narrow escape.
But that isn't the way the NRC has described it in public. The agency's spokesperson told the media that the rust hole didn't pose a safety threat. If the last bit of metal had failed and "allowed steam to escape," the NRC official said, safety systems would have immediately cooled the reactor. Anyway, he said, there would have been no danger to the public. "It's only when you get into the what-ifs that you would have had any leakage from the reactor cooling system." The man was talking through his hat. In reality, the NRC doesn't know what would have happened because the possibility has been considered too unlikely to plan for.
The failure to face up to reality reflects an unhealthy situation. Such spokesmen say what their bosses want them to say, and for several years, the NRC has been knocking itself out to please the industry. The situation worsened in 1998 when the NRC's Senate oversight committee, Environment and Public Works, with strong prompting from the industry association, threatened the NRC with a sharp budget cut. The NRC chairman got the message and revamped the agency's regulatory approach along the lines suggested by the industry. The current commission has by and large continued the same approach, but with a less experienced senior staff. The previous chairman had forced the resignation of the agency's most experienced and competent top officials, who had showed an unwelcome independence of mind.
Just before Davis-Besse's problem surfaced, the NRC gave the plant its quarterly rating under the new rating system. Davis-Besse got the top grade in all 18 categories. From my experience in two terms as an NRC commissioner, during which I visited most of the plants, including this one, I find it inconceivable that everything was fine at Davis-Besse except for one corrosion hole in the reactor vessel. If the plant managers let this problem go, they must have let others go, too. People working in nuclear plants are pretty smart and generally want to do a good job. But they stop asking questions about things that aren't right when they know what answer management is going to give them. At that point, danger lurks.
The NRC has investigated and has now asked other plants to check to make sure they are not suffering from the Davis-Besse problem, but on an unhurried schedule. To a greater extent than ever before we are relying for nuclear safety on the self-regulation of the nuclear operators. Most of them have done a good job, steadily improving their performance. But there are limits to the idea put forward by the industry that post-deregulation financial pressures make for better safety because the operators want to protect their investment. As we know, short-term bottom line orientation also leads some to overreaching, defer necessary modifications or neglect maintenance. Congress and NRC management need to acknowledge that private and public incentives differ.
The late Morris Udall, who as chairman of the House Interior Committee was the principal congressional overseer of the NRC in its eary years, used to say that a forceful and respected NRC was an essential condition of nuclear power. It is still true.