You are using an outdated browser.
Please upgrade your browser
and improve your visit to our site.
Skip Navigation

Lessons of Chernobyl

For pro-nukes, for anti-nukes.


After the accident at Three Mile Island in 1979, some critics of nuclear power claimed that we had come within a hairsbreadth of radiation releases that would have left many thousands dead. Some advocates of nuclear power claimed that the worst conceivable accident had occurred, proving that reactors are not agents of catastrophe, except perhaps to the reactors themselves. It was impossible to know which of those claims had more truth in it. Perversely, the ambiguity arose because no one was hurt. Most of us were left to wonder whether a much worse accident lay ahead.

Now, with the accident at Chernobyl, we are no longer wondering. About 30 people are known to be dead, and at least 80 more are gravely ill. The reactor’s confinement was completely breached, and American scientists have estimated that more than 50 percent of the motile radioactive material was released into the atmosphere (a total estimated release of over 40 million curies, compared to 15 curies at TMI). No one here yet understands precisely what happened in the Chernobyl reactor, and it’s unlikely that the Soviets do either. Nevertheless, there are already important lessons to be drawn from the accident, for both the critics and the advocates of nuclear power. The lessons don’t depend on the causes of the accident, only on the fact of it and the extent of its damage.

Lessons for Critics. The immediate observation is that the accident has produced about 100 prompt deaths and may eventually be judged responsible for as many as a few hundred more. Evacuation was necessary to keep the casualties this low. Richard Wilson, a physicist at Harvard University and the chairman of last year’s American Physical Society panel on severe nuclear accidents, points out: “They evacuated probably everyone within 10 miles downwind, and if you look at our emergency planning regulations, they say: evacuate people 12 miles downwind. The Russians . . . probably got them out in time, so the number of prompt deaths in that area would not be particularly big.” This is not to say that Chernobyl was the worst accident that could possibly happen. A little applied stupidity can make almost any situation worse. But it is an excellent sample of a true nuclear disaster: events were for a time completely beyond control; a gas explosion occurred, and the reactor core melted down; the release of radioactivity was very large. The accident occurred in a populous area, close to a major city. In spite of all that, the fatalities are in the hundreds, not the thousands, certainly not the tens of thousands. This is indeed a disaster, and the victims deserve all the sympathy and assistance we can manage (most of the press coverage focused on Soviet secrecy and American tourists). But it is not the cataclysm some have predicted.

One can argue that an accident of precisely this kind would be much less likely in the United States. Reactor safety procedures here are more rigorous than in the Soviet Union, and we have only one reactor like Chernobyl No. 4. Such graphite-moderated reactors are inherently more dangerous than the light water reactors (LWRs) that make up the bulk of the commercial nuclear industry in this country. Our only graphite model is the N-Reactor at Hanford, Washington, operated by the defense materials complex of the Department of Energy to produce materials for nuclear weapons. The containment features of the Chernobyl reactor are rudimentary compared with U.S. power reactor requirements and were not part of the reactor’s basic design. Perhaps most important, the Chernobyl plant has what is known as a “positive reactivity coefficient” (the nuclear reaction increases as the production of power increases), which seriously aggravates control of internal problems; no U.S. reactors have positive coefficients.

But the dominant fact of Chernobyl is that we now have empirical evidence on the scope of a real reactor disaster, and its dominant characteristic is hundreds of fatalities. Such an accident is comparable in scale to other disasters whose risks society has judged acceptable. It would be irrational to compare the casualties to those caused, say, by the Mexican earthquake, whose occurrence is an unfortunate condition of life there. But it is reasonable to compare them to the results of a major airplane crash, whose occasional occurrence is a condition of our society. A crash that kills 350 people does not produce serious exhortations from public leaders to stop commercial airplane flights. Rather, it produces a demand that the cause of the accident be discovered, and that, if possible, no one be placed in further risk of such a crash.

This is a responsible reaction. Society is not willing to forgo the benefits of air travel, but society should not bear unnecessary risks. Nevertheless, air travel inherently involves some risk. By any scientific account, the risks an average person runs by living in a country with nuclear power are far smaller than the risks that same person runs riding in airplanes. Still, it could be possible that, despite the minimal risk to any individual, the dire prospects of any single accident are too great for society to bear. This is the prospect the Chernobyl accident, the only evidence we have, seems to confront.

A nuclear power accident is different from an airline crash in two important ways. First, the radiation and its damage are invisible. The air crash takes its lives immediately; the reactor disaster may cause casualties and fatalities for many years as cancers develop. To most people, radiation is simply a mysterious killer associated with nuclear reactors. Thus although rationally the cases are the same, emotionally they are very different. But if we need nuclear power, the choice must be made on a rational basis.

Second, we have no practical alternative to airplanes. With the first great “energy crisis” behind us, we can easily become complacent about our future energy supply. The facile answer of some critics of nuclear power has been that risks of any kind are unnecessary: we will simply get our electricity elsewhere. Still, the oil will run out. The principal “inexhaustibles” (solar electric and fusion power) remain high-tech dreams of the distant and uncertain future. We will all be well off if the benign inexhaustibles prove practical and economic, but we would be improvident to wager ht e future on them. There remain only two sure sources of energy: coal and nuclear fission. But large-scale coal-burning cumulatively poisons both the atmosphere and the water supply. In the long term its effects could be devastating worldwide.

When we are dealing with technologies that are in some measure risky, responsible criticism is essential. But campaigning and litigating to prove that nuclear power is inherently unacceptable is unreasonable: no one knows. And playing on public fears and ignorance to press one’s case is irresponsible. The prudent course is to improve LWR safety and economics while we do research and development on the inexhaustibles. Critics should apply their talents, both political and technical, to producing a development program that can find acceptable solutions.

Lessons for Advocates. Reactor accidents must now be acknowledged to be serious possibilities threatening the lives of large numbers of people. Even those advocates who cannot conceive of an American Chernobyl must recognize that the American people cannot be persuaded to join that confidence. The object lesson is too large and too real. Whatever you think about the safety of the current LWRs, they are going to have to be improved before the nuclear power industry in the United States can possibly be revitalized.

Of course, the industry has other serious problems, principally economic ones. Even when oil was cheap, nuclear reactors used to appear cheaper to utilities in the long run, and in the late 1960s they were cheaper. But then safety concerns (manifested in licensing delays and legal challenges from critics) stretched construction times from under five years to ten years or more; high interest rates paid over these periods raised the costs of plants by 40 to 50 percent; and the uncertainties of licensing time made the economics of the investment difficult to calculate. U.S. reactors and their safety processes have not been standardized. Partly because of this, they deliver on average to 55 percent of their operating capacity factor. This is close to the worst record in the world (Swiss, Finnish, and Belgian reactors deliver nearly 90 percent of capacity). Lack of standardization makes licensing more difficult and results in expensive engineering and technical services (now two-thirds of the construction costs) at each construction site.

By contrast, the naval reactors that power our nuclear submarines are built to strict design standards and have an exemplary record of safety (and of cost and schedule; the cost of the rest of the submarines has escalated). The U.S. naval reactors provide an empirical proof that nuclear reactors can be made to operate safely, just as Chernobyl demonstrates that commercial reactors can be dangerous. Naval reactors operate at very close quarters with their populations, yet no significant accidents—with or without casualties—have ever occurred in our Navy (there is some evidence of a Soviet naval reactor accident, unconfirmed of course). On the other hand, naval reactors don’t have to be economically efficient and don’t have to deliver power levels anywhere near the ones required of commercial reactors. LWRs grew out of the naval reactor program in the ’50s, but there never seemed a strong enough reason to submit the commercial reactors to the rigorous controls applied to the design, construction, and operation of the naval ones. Now there is.

Some advocates and friendly critics of nuclear power maintain that the way out of ht e current economic-safety-public relations dilemma is to develop a new kind of commercial reactor that is inherently safe, that even in principle can’t have an accident that leads to a meltdown or a release of radioactivity. Even if such a reactor were developed, there would be a terrific problem convincing the American public it met those criteria. That kind of absolute confidence in technical infallibility is getting very difficult to inspire. But even aside from public suspicions, the inherently safe reactor has two problems: it means starting over in development, commercialization, and licensing, all long processes; and it probably means dealing with reactors whose power production is limited.

A more realistic approach is to undertake a serious reevaluation, improvement, and standardization of the existing LWR. The government could conduct a major program in focused safety and design research to produce a single new standard design for a significantly safer LWR. This is, in fact, the general direction in which the Department of Energy research programs have been headed for the past few years, but the effort is far short of a national commitment that would address both the technical and public acceptability problems. The model that emerged would make construction times shorter and construction processes cheaper; it would simplify licensing so that procedures could be rapid, efficient, and relatively free of political maneuvering; and it would help solve the capacity-factor problem in operating costs. Most important, it would be credibly safer.

The utilities should be made part of this process, not only to establish their commitment but also to make use of their operating experience and their requirements for size, operability, and cost. Critics of the industry should participate as well, to ensure that the full range of potential problems has been recognized and dealt with.

It may be naive to suggest today that this is one problem that only the federal government can deal with. But the situation is clear: the public is deeply concerned, and with good reason; the necessity for a safe nuclear power option in the future (20 or 30 years from now) is manifest in virtually all current analyses; the utilities and nuclear construction industry are financially and politically incapable of changing the situation; and the ideologically motivated fraction of the critics (to whom nuclear power and nuclear weapons are indistinguishable) is politically adept and intractable. A government-sponsored development program to cut through this fearful knot could be established immediately and run for five years for less than the cost of replacing ht e shuttle Challenger. Surely it is as justifiably a government responsibility as a new shuttle.

The great difficulty with such a solution is not that it involves government action (the government has always sponsored some reactor R&D) but that it isn’t exciting. No big new technology for the advocates, no political crisis for the critics, no quick fix for the utilities, no appearances on ABC’s “Nightline” for anybody. But after a decade and a half of taking the energy problem seriously, it’s time we got around to mature attitudes toward solutions.