ROCKVILLE, Md. — The
Nuclear Regulatory Commission is approaching completion of an ambitious study that concludes that a meltdown at a typical American reactor would lead to far fewer deaths than previously assumed.
The conclusion, to be published in April after six years of work, is based largely on a radical revision of projections of how much and how quickly
cesium 137, a radioactive material that is created when uranium is split, could escape from a nuclear plant after a core meltdown. In past studies, researchers estimated that 60 percent of a reactor core’s cesium inventory could escape; the new estimate is only 1 to 2 percent.
A draft version of the report was provided to The New York Times by the
Union of Concerned Scientists, a nuclear watchdog group that has long been critical of the commission’s risk assessments and obtained it through a Freedom of Information Act request. Since the recent
triple meltdown at the Fukushima Daiichi nuclear plant in Japan, such groups
have been arguing that the commission urgently needs to tighten safeguards for new and aging plants in the United States.
The report is a synthesis of 20 years of computer studies and engineering analyses, stated in complex mathematical terms. In essence, it states that if a prolonged loss of electric power caused a typical American reactor core to melt down, the great bulk of the radioactive material released would remain inside the building even when the reactor’s containment shell was breached.
Big releases of radioactive material would not be immediate, and people within a 10-mile radius would have enough time to evacuate, the study found. The chance of a death from acute radiation exposure within 10 miles is therefore near zero, the study projects, although some people would receive doses high enough to cause fatal cancers in decades to come.
One person in every 4,348 living within 10 miles would be expected to develop a “latent
cancer” as a result of radiation exposure, compared with one in 167 in previous estimates.
“Accidents progress more slowly, in some cases much more slowly, than previously assumed,” Charles G. Tinkler, a senior adviser for research on severe accidents and one of the study’s authors, said in an interview at a commission office building here. “Releases are smaller, and in some cases much smaller, of certain key radioactive materials.”
The N.R.C. did not intend to release the report until next spring and said its conclusions were still being adjusted after a peer review.
The health effects of a catastrophic meltdown were hypothetical until the
1979 accident at Three Mile Island. That
destroyed a billion-dollar reactor but caused no apparent physical harm to nearby residents, immediately or over time. Debate has persisted over whether the United States skirted a disaster or whether that accident was about as bad as it could get.
Edwin Lyman, a nuclear physicist with the Union of Concerned Scientists, contends that the nuclear commission has consistently painted an overly rosy picture and that its latest study does as well. He noted that the study assumed a successful evacuation of 99.5 percent of the people within 10 miles, for example. The report also assumes “average” weather conditions, he noted.
But if a rainstorm were under way during a release of radioactive materials, he said, it could wash contaminants out of the air into a small area, producing a high dose there.
Jennifer L. Uhle, the deputy director of the commission’s office of nuclear regulatory research, said the report was intended to present the “best estimate” and not the worst case.
Dr. Lyman said the earlier estimate was of a different accident, a major pipe break. The new study considered that accident too unlikely to analyze.
Dr. Lyman suggested that in projections of fatal cancer cases, the focus should be on people who live within 50 miles. The average population within 10 miles of an American nuclear plant is 62,000; within 50 miles, it is about five million.
The commission’s old projection of eventual cancer deaths was one for every 2,128 people exposed within 50 miles; the new study projects one cancer death for every 6,250 people exposed, which still comes to hundreds of cancer deaths within the 50-mile circle, in addition to the hundreds of thousands who would be expected to die of cancer from other causes.
Dr. Lyman countered that when dealing with estimates based on so many variables — including more than 100 reactors of different designs and vintage, in areas with disparate population densities — a difference of a factor of three is not important. In his view, the study reconfirms that reactors pose serious risks.
The commission’s shift in thinking about how much radioactive cesium 137 would escape after a core meltdown is based on a conclusion that most of it would either dissolve in water that stays put or adhere to surfaces within the plant. The authors said previous analyses had made “conservative assumptions” that most of the cesium and other materials would escape. But laboratory studies and computer modeling have not borne out that hypothesis, they said.
Commission experts have said that a total blackout would be extremely rare at an American plant and that backup generators and other machinery would fill the breach until grid power was restored. Nonetheless, the study focused on what would happen in the event of a nuclear station blackout, meaning a complete loss of power from the grid and from backup diesel generators, and then an exhaustion of batteries that supply power, leading to a meltdown. That is what happened at Fukushima.
The study focused on two common reactor types in this country: boiling-water reactors at the
Peach Bottom Atomic Power Station in Pennsylvania, similar to those at Fukushima, and pressurized-water reactors at the
Surry Power Station in Virginia.
The study gives a highly detailed prediction of which equipment would stop operating; what temperatures, steam pressures and flows of water and steam would result; and where and when leaks would begin after a meltdown.
It concluded that Peach Bottom would not release enough radioactive material to kill anyone immediately, although it could increase the rate of cancer deaths over future decades. At Surry, the probability was so low and the number of people living within 10 miles so small that the death toll would be a fraction of a person.
The report was prepared by staff members of the Nuclear Regulatory Commission and
Sandia National Laboratories, a Department of Energy lab. Beyond the revisions to be made as a result of the peer review, the report could undergo further changes after public comments are received next year.
Once completed, it might be used by the commission when it analyzes proposed safety improvements in terms of costs and benefits, or decides where reactors should be located.
“Once we think we know what the best estimate is, we think we can start thinking about applications,” said Jason H. Schaperow, a senior reactor systems engineer and one of the authors.
