(May 18, 2007) March 2007 marked the 50th anniversary of the WASH-740 report. Commissioned by the U.S. Atomic Energy Commission (AEC), and also known as "The Brookhaven Report," it was the first official study of the "Theoretical Possibilities and Consequences of Major Accidents in Large Nuclear Power Plants." But "large" at the time meant only 100 to 200 megawatt-electric reactors, small by today's standards.
(656.5802) NIRS - The WASH-740 report looked at the health and economic impacts of a large-scale radioactivity release at a reactor 30 miles upwind of a major city. WASH-740 concluded that up to 3,400 deaths and 43,000 injuries could result; up to 150,000 square miles could be contaminated with radioactivity; and up to $7 billion (in 1957 U.S. dollars) in property could be damaged (over $51 billion -38 billion euro- in 2007, when adjusted for inflation by the Consumer Price Index).
Rather than proceed with caution in light of such catastrophic risks, AEC instead used WASH-740 to lobby for, and rush the passage of, the Price-Anderson Act just six months later. As originally enacted, the private liability for nuclear utilities and their insurance companies was capped at just $560 million (in 1957 dollars; $4 billion in 2007). But that amounted to less than 10% of the property damages estimated by WASH-740. It would take an act of Congress to compensate victims for the additional 90%+ of the damages, to be paid for by U.S. taxpayers. Price-Anderson paved the way for the opening of the first civilian U.S. atomic reactor at Shippingport, Pennsylvania in December, 1957, not to mention the 130 additional commercial reactors in the U.S. in the following decades.
To account for the increasing size of atomic reactors, AEC ordered an update on WASH-740's analysis from its Brookhaven National Lab. This was finished in1965, and found that a major accident at a large reactor could cause up to 45,000 deaths, and property damages many times worse than WASH-740 had predicted. However, AEC covered up these results for 8 years, fearful that their release would increase public opposition to new reactor licensing. The Union of Concerned Scientists (UCS) forced the update's publication through a Freedom of Information Act request in 1973.
While WASH-740, and its long concealed update, had focused on the consequences side of the risk equation, the 1975 AEC report "WASH-1400" focused on the probability of a major reactor accident. Also known as the Reactor Safety Study, NUREG-75/014, or the Rasmussen Report (after Massachusetts Institute of Technology (MIT) professor Norman Rasmussen, the lead author), it applied "Probabilistic Risk Assessment" to just one pressurized water reactor and one boiling water reactor design, then extrapolated the results to the nearly 100 operating reactors in the U.S. Making unrealistic assumptions (such as all reactors are built as designed, with no flaws; components do not degrade with age; reactor vessels cannot fail, as by embrittled vessels failing under pressurized thermal shock), the analysis was significantly flawed. Absurdly, the Rasmussen Report's executive summary claimed that the probability of being hit by a meteorite was higher than being harmed by a nuclear reactor accident.
The Rasmussen Report was intensely criticized on a number of technical fronts. In 1977, Henry Kendall at UCS published "The Risks of Nuclear Power Reactors: A Review of the NRC Reactor Safety Study WASH-1400," which found that Rasmussen likely underestimated accident probability and consequences.
Also in 1977, due to the controversy swirling around the Rasmussen Report, the U.S. Nuclear Regulatory Commission (NRC, formed after the breakup of AEC) appointed a review panel, the Lewis Commission. It reported that Rasmussen significantly underestimated uncertainties.
The Lewis Commission issued strongly worded criticisms of Rasmussen's methodology: "The statistical analysis of WASH-1400 leaves much to be desired. It suffers from a spectrum of problems, ranging from lack of data on which to base input distributions to the invention and use of wrong statistical methods. Even when the analysis is done correctly, it is often presented in so murky a way as to be very hard to decipher.
For a report of this magnitude, confidence in the correctness of the results can only come from a systematic and deep peer review process. The peer review process of WASH-1400 was defective in many ways and the review was inadequate."
It also reported that the executive summary "is a poor description of the contents of the [Rasmussen] report, and should not be portrayed as such…[and] has lent itself to misuse in the discussion of reactor risks."
In response to the Lewis Commission, NRC officially disavowed the Rasmussen Report executive summary, and concluded "…in light of the Review Group conclusions on accident probabilities, the Commission does not regard as reliable the Reactor Safety Study's numerical estimate of the overall risk of reactor accidents." Despite this, the Rasmussen Report is still used today, as in a nuclear engineering textbook at MIT, an epicenter of the "nuclear renaissance."
The next, and last, federal report on risks of major reactor accidents was done over 25 years ago. NRC commissioned Sandia National Lab to prepare the CRAC-2 report (Calculation of Reactor Accident Consequences for U.S. Nuclear Power Plants), which was completed in 1981. However, like the WASH-1400 cover up, CRAC-2 was not initially released to the public. Yet again, UCS, working with U.S. Congressman Ed Markey and the House Committee on Oversight and Investigation, forced NRC to divulge the findings in 1982.
CRAC-2 revealed that a worst case nuclear accident (a maximum "credible" accident combined with worst possible meteorological conditions) could cause up to 100,000 peak early fatalities (a death resulting from radiation exposure that occurs within the first 9 years), and up to 40,000 latent cancer deaths (namely, at the Salem 1 and 2 reactors near Wilmington, Delaware). So many deaths would be comparable to the numbers killed at the Hiroshima or Nagasaki atomic bombings. The costs of a worst case accident (including "estimates of lost wages, relocation expenses, decontamination costs, lost property and the cost of interdiction for property and farmland") could top $657 billion (in year 2000 dollars, or $783.5 billion in 2007 dollars). Despite population growth and increased concentration near many nuclear power plants, as well as recent scientific findings that radioactivity is more harmful to human health than previously thought, a significant update on CRAC-2 has not been performed in over 25 years (NUREG-1150, "Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants," was published in 1991, however, using reactor-specific PRAs).
Until now, that is. On May 7, NRC issued a press release headlined "NRC Announces First Phase of State-of-the-Art reactor Consequences Analyses (SOARCA)." It seems clear that NRC intends to revise downward the estimates of casualties and costs from a major reactor accident. "We're undertaking this research to replace work that's almost 25 years old - studies that were so conservative that their predictions are not useful for characterizing results or guiding public policy. Those predictions have sometimes been misinterpreted and often misused," said Farouk Eltawila, Director of the Division of Risk Assessment and Special Projects in the NRC's Office of Nuclear Regulatory Research. "Today's computer-based analytical tools are much more capable of realistically evaluating potential nuclear power plant accidents, and this project should improve everyone's understanding of the realistic consequences of such potential accidents." The results of SOARCA will be published in 2009.
SOARCA resembles sorcery in more than spelling-the waving of a magic wand that makes atomic reactor risks disappear, on paper at least. Especially given the increased risks of aging reactors suffering breakdowns, and proposed new reactors whose bugs have not yet been worked out, nuclear watchdogs must remain vigilant against WHITEWASH-2007.
Source: Brice Smith's "Insurmountable Risks: The Dangers of Using Nuclear Power to Combat Global Climate Change," IEER Press, Takoma Park, MD, 2006, pages 188-195 and 205-214. This can be ordered at www.eggheadbooks.org/books/insurmount.htm; a summary of the book can be viewed at http://www.ieer.org/sdafiles/14-2.pdf.
Contact: Kevin Kamps, NIRS