Nuclear Cover-up: Censorship in the USA?

by Lorna Salzman

The question of world energy needs and of the role of nuclear power in filling those needs is complex, but in the United States two basic views on the subject can be identified.

Nuclear advocates say that the world must continue to increase energy consumption and production to maintain economic stability and progress and that we are depleting rapidly our known, usable fossil fuel reserves. They call for widespread development of nuclear reactors as the safest, least polluting and most economical way to provide more energy and to forestall the eventual and complete depletion of fossil fuel supplies.

Nuclear opponents argue that economic growth can continue while the world reduces energy use through easily-tolerated conservation methods, that nuclear power is neither safe, non-polluting, nor economical, and that safer, cheaper alternative energy sources (wind, sun, water, geothermal, etc.) can and should be developed.

A major area of conflict is over safety, but even many advocates acknowledge that certain hazards are inherent in atomic power. Catastrophes are to be feared from various quarters: terrorists could obtain plutonium, a radioactive, poisonous explosive which is a by-product of the atomic energy process and from which atomic bombs can be made. Human or mechanical accidents, earthquakes, fires or floods could take place at refining facilities, reactors, waste disposal sites, or in the transport of nuclear fuel and wastes. Any event like this would release dangerous quantities of radioactive poisons near population centres. Other fears are that men cannot devise a permanent, fail-safe method of storing the waste products of nuclear power (plutonium, for example, has been called 'fiendishly toxic' by one of its discoverers, and it remains lethal for 250,000 years), and that proliferation of nuclear power will cause a concomitant rise in birth defects and cancer due to the increase in low-level 'background radiation' from nuclear plants. Concerned advocates call for more stringent safety measures; opponents, for an end to nuclear power altogether.

Despite the significance of these issues, the American press has paid so little attention to them over the years that a nationwide media research study recently put the nuclear safety story on a list of 'ten best censored stories'. Commenting on the study, columnist and former presidential press secretary Jerry ter Horst cited 'media dereliction, neglect and lack of perception'. But this article will show that the major responsibility for keeping the public poorly informed lies with the nuclear power industry and, more disturbingly, with the agencies of the United States government which have been charged with regulating the nuclear industry.

Albert Einstein said that the question of nuclear power would eventually be decided in the village square. Apparently, industry and government fear the villagers' decision, for while they ask for public acceptance of the nuclear power programme, and while they repeatedly state that the public is in no danger, they have contrived to keep the public ignorant of key safety studies, technical data, unresolved problems, even of internal dissent on these subjects among government-hired experts themselves. Some of this valuable information is now reaching the village square where the decisions must be taken, mostly as a result of Freedom of Information requests and threat of lawsuits by citizen-action groups.

But so far the larger problem of absolute control over nuclear information exerted by Federal agencies under the Atomic Energy Act of 1954 (the law setting out the general regulatory process, as well as control over reactor licensing procedures and radiation exposure standards) has been immune to public challenge. This becomes more and more of a scandal as the government proves itself to have been and still to be, not an impartial watchdog, but an all-out supporter of the nuclear industry.

The following examples are not the only instances of suppression of nuclear information to have been exposed as a result of citizen investigation. And what has been exposed is only the tip of the iceberg. It is becoming clear who censored this 'best censored story' of the past two decades, and we should also ask ourselves why.

Accident Studies - One Denied

In the mid-1950s, at the start of the commercial nuclear power programme, and as a result of Congressional hearings on the thorny problem of nuclear accident liability, Brookhaven National Laboratories were asked to study the theoretical consequences of a nuclear accident at a 500 megawatt reactor within thirty miles of a populated area. Their study, published in 1957, showed that–with a 50 percent release of the radioactive contents of the reactor–several thousand people would die immediately, tens of thousands would be irradiated, and property damages could amount to $7 billion. The report caused serious concern in government and in the insurance industry, which up till then had refused to provide more than $110 million in liability for a single accident. The government, faced with threats from the nuclear industry that they could not expand without sufficient liability insurance, agreed to provide an additional $450 million and simultaneously absolved utilities and reactor manufacturers from any damages or responsibility over the total of $560 million. This limited liability law was passed in the form of the Price-Anderson Act for a period of ten years.

A decade later, when the law was close to expiration, Brookhaven was quietly asked by the Atomic Energy Commission (AEC) to update the study. No public acknowledgment of the new study (called 'the up-dated WASH 740 Report') was made. In fact the government heatedly denied its existence for six years until 1972. The Friends of the Earth, Ralph Nader and a citizens' group called 'Business and Professional People in the Public Interest' threatened a Freedom of Information lawsuit and obtained a look at the internal memoranda and working papers of the study. The update, which dealt with reactors much larger than those originally considered, stated that a similar accident could kill 45,000 people, irradiate another 100,000, cause property damage worth over $17 billion, and contaminate an area of 150,000 square miles with radioactivity, rendering it uninhabitable for centuries.

Several members of the Brookhaven task force had originally urged suppression of the update, but it was turned over to the AEC where it was kept confidential for over six years. Interestingly, the study director, Dr. Clifford Beck, stated in a memo to the AEC commissioners that…'there is no objective, quantitative means of assuring that all possible paths leading to catastrophe have been recognized and safeguarded…here is encountered the most baffling and insoluble enigma existing in our technology: it is in principle easy and straightforward to calculate potential damages that might be realized under such postulated accident conditions; there is not even in principle an objective and quantitative method of calculating probability or improbability of accidents or the likelihood that potential hazards will or will not be realized' (emphasis added). At a committee meeting Dr. Beck later stated about accident probability: 'We feel that we cannot predict if, or when, it might happen'. Beck's conclusions are significant in the light of the findings of other government-sponsored studies.

Hearings on Reactor Safety

Accidents such as those studied by Brookhaven can only occur under certain circumstances. When uranium atoms are split, energy is produced to heat water circulating through the reactor. The resultant steam powers a turbine which then generates electricity. If a pipe carrying water to the fuel breaks, emergency cooling water must reach the fuel within 60 seconds to prevent overheating, melting and release of radiation from the fuel core of the reactor. If the emergency core cooling system (ECCS), designed to prevent such an accident, fails to work effectively, the reactor core would overheat and a major release of radioactive matter in gaseous form could be blown across land and water and to nearby cities.

Obviously the ECCS is an extremely important safety feature, and only if its design and operation are properly carried out can there be any assurance that meltdowns can be prevented or mitigated. Although some citizens have been concerned about the hazards of nuclear power for years, the public had been given little evidence that all was not well with reactors until 1972-3, when the independent Union of Concerned Scientists (UCS) forced the AEC into holding extended technical hearings on the ECCS and other safety features. The hearings revealed not only that the ECCS had never been tested under realistic accident conditions, but that the government-sponsored-and-funded studies relied upon computer codes that AEC engineers themselves said were incomplete and inadequate as a basis for reactor licensing.

During the hearings the government repeatedly denied requests by the Union of Concerned Scientists for access to expert witnesses and internal documents, but some documents were leaked to them anyway. Internal studies brought out as a result of the hearings indicated that nuclear engineers working on the Emergency Core Cooling System and on its computer programme had grave doubts the safety feature would work at all. Small-scale tests of it had failed completely, and full-scale tests had been postponed, with the government farming out paper tests to the manufacturers of the system, rather than to independent evaluators.

Moreover a special internal task force, which had been set up after the small-scale tests had failed, questioned the efficacy of the ECCS as well as the computer simulations upon which claims of its safety were based. One report stated that it was beyond the present capability of engineering science to predict how well the system would perform. While some task force members were optimistic, others strongly disagreed. Dr. Morris Rosen, an AEC official in charge of ECCS analysis wrote, in an internal memo of 1 June 1971, that '…the system performance cannot be defined with sufficient assurance to provide a clear basis for licensing'. Dissenting reports like this, however, were disregarded by the task force when it prepared interim criteria for the ECCS. Another report by the Idaho Reactor Testing Station listed 28 areas where information on the ECCS was 'missing', 'inadequate' or 'unverified'. This report was withheld by the AEC division of Reactor Development and Technology, even from its own AEC colleagues who were responsible for licensing.

At the ECCS hearing the AEC tried to prevent Rosen and another official, Robert Colmar, from testifying; and the Commission further distributed to all government witnesses a memo that warned: 'Never disagree with established policy.' Soon after the hearings, Rosen was removed from his position as head of the Systems Performance Branch of the AEC Division of Reactor Standards and given a purely advisory job. He later left the AEC and Colmar asked for a transfer.

At the conclusion of the hearings the AEC adopted its old position on the adequacy of the ECCS, with some minor changes that Dr. Henry Kendall of the UCS termed 'purely cosmetic'. The government continues to license reactors with the same untested emergency system, and even now no actual testing of the system is planned.

The Rasmussen Reactor Safety Study (RSS)

For years the government and industry had been defending themselves against charges that nuclear power is rife with inherent dangers as well as external risks from errors and sabotage. The response had been to claim that 'redundant safety features', 'safety-in-depth', and 'quality assurance-quality control' essentially preclude serious accidents, and that in any case the consequences of an accident would not be unacceptably severe. But the uncovering and publication of the updated WASH-740 report–indicating large numbers of deaths, radiation injuries and widespread radioactive contamination–belied the latter claim. The facts uncovered during the ECCS hearings proved that the major features (and perhaps others) on which the former claim rested had never been tested. The only alternative left to the government was to demonstrate that catastrophic accidents were extremely unlikely to occur.

To this end they commissioned WASH-1400, the Rasmussen (after the study's director) Reactor Safety Study or RSS. At a cost of $4 million and after three years of research, a report that allegedly attempted to assess the probability and consequences of serious nuclear accidents was released in October 1975. Although many technical consultants were employed on the study, two features stand out. First, the computer data on reactor components' behavior and failure rates (and analysis of them) were provided by the nuclear industry itself without any independent evaluation; the accident consequences appendix was actually prepared by a Westinghouse employee. (Westinghouse is a major manufacturer of nuclear reactors and related equipment.) Second, the study was conducted by in-house AEC staff at their headquarters in Bethesda, Maryland.

USA: A Terminal Case

Research indicating low levels of radioactivity in the environment may cause cancer has been terminated early by its Federal government sponsors. Dr. Thomas Mancuso (of the Public Health School, University of Pittsburgh, USA) was working on a study with Dr. Alice Stewart (a cancer epidemiologist at Birmingham University, England) and George Kneale (a British statistician). It was funded by the US government Department of Energy and was meant to focus on the long-term health of workers at a Federal nuclear facility in Hanford, Washington State.

In 1974, when another study by Dr. Samuel Milham for the Washington State Department of Health indicated a positive correlation between radiation exposure and cancer deaths, the Atomic Energy Commission tried to persuade Dr. Mancuso to release his preliminary findings in order to rebut the other research. Dr. Mancuso, who had been working on his study since 1964 using a 30-year database, refused to do so, stating it would be premature. In fact his interim findings (published later) also showed a positive correlation between low levels of radiation exposure–far below present permissible limits–and excess cancers.

Last year Dr. Mancuso was informed that funding for his study would be terminated as of July 1978. The reason given was that, being aged 65, he should retire, although the actual retirement age at the University of Pittsburgh is 70. Simultaneously the government asked him to turn over his data to Oak Ridge-Associated Universities, a government laboratory in a major centre of nuclear research, development and manufacture at Oak Ridge, Tennessee.

A final report was submitted in July 1977 and later published. During this period the DOE had prepared and was circulating internal critiques of the study, but refused to make them available to Drs. Mancuso and Stewart. They obtained copies only after filing a Freedom of Information request.

The director, Dr. Norman Rasmussen of the Massachusetts Institute of Technology (MIT), had no training or expertise in reactor safety, statistical methods, risk assessment or any field relevant to accident probability. Furthermore he had, and still has, firm ties with the atomic energy industry, as a consultant to the Nuclear Energy Property Liability Association, to two prominent nuclear engineering companies and to Reddy Communications, a nuclear public relations firm, and as a founding board member of American for Energy Independence–a lobbying group for nuclear power funded by Westinghouse.

Clearly the RSS had built-in inadequacies. Furthermore, it dealt only with intrinsic failures of components and systems, not with external causes of accidents by human error or sabotage. As noted above, it was based on industry computer models, data and analyses not verified independently by the RSS task force. It concerned itself with reactor accidents only, not accidents in other parts of the nuclear fuel cycle. It analyzed only two representative reactors of 800 megawatts in size or less, although much bigger 1100 megawatt reactors were already being constructed.

But the defects of the final report were far greater. As a result of the Freedom of Information request filed by UCS, 50,000 pages of internal studies, memos and comments pertaining to or used in the RSS were released by the government. These proved what many had suspected all along: the findings for safety were pre-determined by the study group, and the report was envisioned from the start as one that would be 'of significant benefit for the nuclear industry'. Data and technical examples had been selected so as to prove a priori assumptions that accidents were extremely unlikely and that if they did occur the consequences would be essentially negligible.

The AEC was keenly aware of the public scrutiny that the report would receive. Memoranda of the period note that 'the sensitive nature of these studies will require careful control of all official information releases', and 'the report to be useful must have reasonable acceptance by people in the industry'. The internal documents show a sharp disparity between the claims of the agency that the study was a 'full, objective and scientific analysis of the risks', and the truth about its finding and also its methodology, as the following example shows.

I have already referred to 'quality assurance quality control' (QA-QC). These are the standards by which components and procedures are designed to operate to guarantee safety , and they were especially relevant to the RSS because findings on the unlikelihood of accidents rests in part on how well QA-QC is implemented. The working papers for the RSS, however, betray a concern that, if a truly comprehensive QA-QC review were conducted at the two plants being studied (Surrey in Virginia and Peach Bottom in Pennsylvania), numerous deficiencies might become known which would cast doubt on a basic element of nuclear safety. Two approaches were suggested: the first would select information to support a 'pre-determined finding' of safety and reliability; the second would not pre-determine finding but, as compensation, would discuss only those major deficiencies already found and resolved in the field by the AEC, thus inspiring public confidence in QA-QC procedures. The first approach, of course, left the door open to public awareness of possibly poor QA-QC programmes that the AEC feared could 'undermine public confidence in the reliability of plant safety systems' (E. Gilbert memo, 10/23/73). The second could be recognized as incomplete and could cast a negative cloud over the whole programme, leading to fears that 'the whole story (may be) much worse…' After much deliberation, the argument was settled–by eliminating the entire subject of QA-QC from the study.

Internal Dissent Stifled

Before the publication of a first draft of the report in August 1974, internal comments were made on a working draft. Twelve persons, mostly AEC officials and some outside consultants, made comments, but these were not compared or correlated. However, looking at the now-released internal documents, one can see recurring criticism in two areas: common mode failures, and the problem of identifying all possible causes of component or system failure. A common mode failure involves two or more failures stemming from a common cause. Such a failure has wide implications for nuclear safety since it could undermine the concept of 'redundant safety features'. An expert in common mode failure had alerted the RSS study group to the problem as early as 1973, and task force comments on the first draft were critical along the same lines, calling some of the examples cited, 'a disaster' that might 'invalidate' much of the study results. The RSS study group, according to the experts, had overlooked some obvious types of common mode failures as well as near-misses, all of which had been omitted from the draft. For example, one AEC staff member pointed out that the RSS paid practically no attention to earthquakes as an accident-initiating factor, even though this was considered a prime cause of potential common mode failures.

The other criticism amounted to no more than commonsense: the inability of any analysis to ensure that it has included all potential sources of accidents, to which Dr. Beck referred in the updated WASH-740 study. Clearly accident probability calculation, especially for complex nuclear power plants, cannot be estimated without identifying all possible causes of failures. Darrell Eisenhut, a member of the AEC regulatory staff and its internal review group, commented on this and added that the omission of sabotage was proof that the study did not include all significant effects. Richard De Young, another member of the AEC regulatory staff, stated that 'The absolute assurance given in the report that "all" accidents have been considered renders the conclusions vulnerable if it can be shown that even one sequence of significance has been overlooked…' He added: 'A risk assessment that does not address the sabotage issue cannot be considered other than incomplete', and he urged more work, saying that release of the report without such revision would be 'inadvisable and a disservice to the study group'. De Young goes even further: 'The report contains deficiencies and inconsistencies to such an extent that to correct them would likely be a major task requiring many more months of effort.'

Despite these internal comments the draft was released on 20 August 1974 without mention of the substantive internal criticisms of the study or of the fact that the review group's recommendations had been ignored. At this time major independent reviews were also conducted by the UCS together with the Sierra Club, by Intermountain Technologies Inc. on behalf of the US Environmental Protection Agency, and by the American Physical Society Study Group on Light Water Reactor Safety. Again no substantive criticisms were allowed by the AEC to be aired publicly, nor did these groups receive any response from the RSS study group.

The AEC simply set up another review body to consider all comments within sixty days. This second group picked out many safety issues they claimed had not been fully addressed, including earthquakes and also fires. Seven months later, in March 1975, a major fire began in the Tennessee Valley Authority's nuclear plant at Brown's Ferry, Alabama, which had been in operation for only seven months. Starting in the cable spreading room, the fire destroyed all the 'redundant' electrically wired safety systems and brought the plant perilously close to a meltdown. A cable tray fire like this had not been considered by the RSS as a possible initiating event for an accident.

As the regulatory agency, now called the Nuclear Regulatory Commission (NRC) prepared to issue the final RSS draft, the chief of the government's Accident Analysis Branch stated: 'The final RSS cannot therefore be accepted uncritically and without further review.' He too was ignored and the report was published on 30 October 1975.

Congress and the RSS

The Congressional Joint Committee on Atomic Energy (now abolished) was considering another extension of the Price-Anderson Act at the time. It held oversight hearings on the final RSS, but was given the impression that expert critics had already commented on the final version and had taken no strong exception. In fact, internal criticism had been brushed aside and no independent groups had an opportunity to comment on the final study at all. In answer to a query from Congressman Mike McCormack on 26 November 1975 on the existence of any substantial dissent, Dr. Rasmussen stated that there was 'None that I'm aware of'. This of course was deceptive, if only because at the time of the hearings, as of 20 November, interested parties like the UCS had not even received copies of the final version and could not have been expected to compare it with the earlier draft.

The study continues to be used by government and industry–despite expert criticism from impartial sources–as proof that nuclear power plant accidents are extremely unlikely. That many of the RSS staff themselves disagree is a fact that remains largely unknown.

One can only conclude that the government and the nuclear industry cannot tolerate dissent, that is to say, any information which casts a negative light on nuclear power or calls into question the judgement of those 'experts'. For a supposedly open, democratic society, the campaign of suppression and deception that has accompanied the nuclear power debate is unprecedented. Nonetheless, against powerful odds and well-funded adversaries, citizens and public interest groups have managed to ferret out information and use it in such a way that the opposition to nuclear power development continues to grow. The question remains: if nuclear power is as benign and beneficent as its proponents say, why do they go to such lengths to stifle dissent? And, more to the point, if nuclear power sparks such deception and requires such manipulation, is it not inherently incompatible with democracy?

Source: Index on Censorship. Volume 7, Number 5, September-October 1978, pgs. 37-42.

© 2002 Lorna Salzman. All rights reserved. Material may be quoted with permission.