Data and Documents

REALISM -

In Engineering and Analysis

of Radioactivity Sources and Consequences

in Using Nuclear Technologies

Statement on Realism - Introduction
 

GETTING REAL
THE IMPORTANCE OF BASING ANALYSIS AND REGULATION ON PHYSICAL REALITY
 

The benefits of nuclear technology are clear and simple. They are continually manifested in the real world, here and now. The absence of pollution, the abundance of fuel, the reliability of output, the inherent absence of a significant waste problem, are demonstrable, measurable, and undeniable. The arguments against nuclear technology, on the other hand, are hypothetical. The predicted casualties have not happened, even after 40 years with several hundred nuclear power plants. Peer-reviewed analysis shows that physical science precludes creation of a serious public health hazard after any realistic series of mishaps. Why then is there opposition?

We blame it on the media, the politicians, and our inability to communicate, but in fact our message has been heard and understood. We have told people that radiation poses an unprecedented hazard to humanity, that we can control it only by superhuman skill and discipline. We tell them we must create a new priesthood to preserve throughout millennia of changing and warring societies, the dreadful secret of where our waste is buried. We invent the guillotine break, the China syndrome, the mass evacuation plan, the nuclear mutant, and the land forever abandoned to contamination. All this is the product of the proponents! And then we wonder why people are frightened.

In 1982, the US Nuclear Regulatory Commission published “worst case” accident studies for each of the 130 nuclear power plants then operating, a-building or planned. If everything went wrong, how bad could it be? Sandia Lab, as NRC’s contractor, dutifully tabulated the answers: up to 610,000 “early injuries,” 100,000 “early fatalities,” 25-mile “fatal radius,” 70-mile “injury radius” and hundreds of billions of 1980 dollars lost in a single plant casualty. Many other such reports have been produced before and since. A 2002 news report cited “a special report prepared by experts within the NRC” that claimed “fatal, radiation-induced cancers as far as 500 miles from the site,” adding that “millions of people within such a 500-mile zone might have to be evacuated for periods ranging from 30 days to one year, and that people within 10 miles of a nuclear power plant such as Indian Point might never be able to return to their homes,” noting that 93 million people live within 500 miles of the plant.

The report was characterized as “an official NRC planning regulation…[that] provides the basis for any future NRC regulation on evacuation needs, safety requirements and insurance.” A named NRC expert who worked on the study reportedly said, “how many millions of people might have to be evacuated following a real spent fuel pool fire would depend on the cost of evacuation and what is perceived to be an acceptable death rate.”

We can hardly blame the media for giving such reports prominent coverage. And we shouldn’t be surprised that opponents of nuclear power still quote these figures today—they’ve never been effectively repudiated. The fact that these consequences were said to be highly improbable is little comfort. And new claims of this sort are published repeatedly. If there is any validity to such estimates, we should have long ago stopped operating nuclear power plants. If the claims are not valid, the public deserves to be shown why, and we should stop using such unrealistic premises in policy, engineering and operations.

The purpose of this paper is to show that such consequences are not realistic—i.e. not relevant to the real world. They are not merely improbable; they are in fact realistically impossible. The difference is important. No valid prediction can result from calculations based on events that are physically impossible.

For this paper we examine a realistic but highly improbable situation: simultaneous destruction of systems outside containment, rupturing containment penetrations, melting the reactor core, and producing ground-level unfiltered releases within about one hour. This is the worst realistic casualty, and if we found the consequences were intolerable, we would have to examine all the steps that might lead there and show that we had made them sufficiently improbable. But we have found that even this extraordinary situation could yield few if any fatalities, whereas casualties actually experienced with non-nuclear means of generating electricity-- natural gas systems, oil, coal, dams, etc.—have repeatedly caused hundreds, and even thousands, of deaths.

We know a great deal about what is physically possible with nuclear reactors. We have demonstrated that nuclear fuel retains most of its fission products, even when molten. We have measured the limited distribution of the fission products that do escape, especially in a water environment. And we have seen how most of what escapes clumps and drops out and does not stay aloft to be carried far away. We know this with the same physical and chemical certainty that we know a power reactor cannot explode like an atomic bomb.

And we know that trivial amounts of radiation do not cause cancer or other harm, and that computer models that multiply trivial individual radiation doses by millions of people to get thousands of cancer deaths are not just improbable. They are not conservative. They are simply wrong. Hundreds of millions of medical patients have received low to moderate doses that refute this premise, along with hundreds of millions of people that have lived in high natural background areas. We’ve been told that it is merely prudence and good practice to presume that things are worse than they are. “Better safe than sorry.” But good policy does not result from a distorted picture of reality. And we will describe briefly what a high price we pay for assuming otherwise, without any compensating benefit.

There is another type of flaw in the computer models that ‘predict’ massive deaths. Some of the premises that are not impossible in an unrealistic hypothetical sense are in fact impossible in a realistic sense. For example, the DOE requirement that calculations of fatalities from a spent-fuel shipping cask casualty assume the event takes place in a high-density city environment and that no one moves until one year after the explosion and that no steps are taken during that year to reduce any contaminant. Neither of these premises is physically impossible, but most people would agree they are not realistic—i.e. not relevant to the real world. We identify some such premises and replace them with technically justifiable conservative but realistic premises. We also show that the leading individuals and organizations that created and maintain the present radiation protection policy have affirmed clearly and publicly the judgments we make, but nevertheless recommend the regulations be based otherwise, “to be prudent.”

Finally, the most basic reason to change current radiation protection policy, and the regulations that prescribe it and derive from it, is that it is self-contradictory. It defines as dangerous, radiation levels that it elsewhere claims are harmless. It requires tearing down a building if it becomes contaminated, while ignoring the higher radiation level of the National Capitol or New York’s Grand Central Station or the base of the Statue of Liberty, all of which are built of naturally radioactive granite. It has regulated man-made radiation levels down below those encountered in natural backgrounds, and has then tried to obliterate this contradiction by calling Nature “the silent killer.” A responsible professional should not long abide such unscientific policy. This discrepancy between policy and science is acknowledged in the reports of the advisory and regulatory bodies (see numerous quotes in Annexes 2 and 3 below). Thus, changing the policy to match the science cited in these reports would not require challenging any current scientific theory or data. It would merely remove the gratuitous unscientific presumption of damage where none has been observed or predicted scientifically. We do not become safer by presuming the world is different than it is.

TO SUMMARIZE:  We consider the consequences of some very improbable casualties. But we do not presume conditions which are refuted by tests and experiments, or prohibited by the laws of nature. And having made such analyses, we find that nuclear power plants and their fuel are not capable of producing the extreme consequences “predicted” by many simplified, unrealistic, reports and are therefore are not a significant public health hazard under any realistic conditions of operation or hostile attack. We then suggest some actions to bring radiation protection policy, and engineering and operations, into line with these realities. The first step should be for sponsoring government agencies to show and assert clearly that past reports, with “predictions” of thousands of deaths from nuclear casualties, are based on unrealistic premises and therefore these deaths would not occur in the real world.
 

I am convinced nuclear regulation now needs to be anchored in realistic conservatism…if we are to avoid the twin pitfalls of under-regulation and over-regulation…I use “conservatism” in the sense of preserving adequate safety margins, and I use “realistic” in the sense of being anchored in the real world of physics, technology and experience…When engineering margins are applied to input parameters, they can distort our understanding of what is truly important. Safety margins are better discerned when they are applied at the decision-making stage, rather than at the analysis stage.

Nils J. Diaz, Chairman, US Nuclear Regulatory Commission
To Japanese Atomic Industrial Forum Annual Conference, April 21, 2004, Tokyo, Japan