Executive Summary

The Project on Government Oversight's findings include:

• All 110 operating nuclear reactors in the U.S. are operating despite documented significant safety problems that have not been fixed or field-verified by the government regulating agency, the Nuclear Regulatory Commission (NRC):
  
  
  • The NRC has yet to verify 389 high priority safety improvements that the operators claim to have fixed, or implemented, at every nuclear power plant in the United States.
    
    
  • Currently, there are 76 high priority safety improvements that remain unimplemented at a minimum of 62 different nuclear power plants. Some of these issues were "resolved" as far back as 1978 by the NRC, yet these resolutions are known not to have been applied at the nuclear plants.
    
    
• The longer safety issues remain unresolved, unimplemented and unverified by the NRC, the closer the United States gets to another Three Mile Island or Chernobyl:
  
  
  • One safety issue that has the potential to melt the nuclear core and/or break the containment of radioactivity, emergency sump performance, has reappeared four times since the issue was "resolved" eleven years ago -- most recently in September 1995. This critical process involves recirculating water through the reactor system to maintain core cooling.
    
    
  • Degraded steam generator tubes, a problem "resolved" by the NRC in 1988, have ruptured at six nuclear power plants since its resolution. Steam generator tubes are used to convert water into steam which eventually drives the reactor's turbine-generator. According to a former NRC Commissioner, this problem is " . . . a loaded gun, an accident waiting to happen." This problem caused one plant to shut down in 1994, and caused an alert to be sent out to all nuclear plant-operating licensees in April 1995.
    
    
  • In 1990, as a result of diesel generator failures, Plant Vogtle in Waynesboro, Georgia was within hours of a meltdown. These generators are needed to restore power in case of an emergency and keep the reactor from melting down, yet they do not always start when they are needed. The NRC "resolved" this problem in 1993 with no new requirements. However, 89% of the offsite power outages occurring from 1965 to 1989 occurred when the major electrical power source was either already out of service or it failed during the outage.
    
    
  • The Thermal-Lag 330-1 Fire Barrier System, which is required to protect safe shutdown equipment, has yet to be replaced at 80 reactors even though the NRC and the industry know that it is actually flammable. Instead, the industry developed "fire watch," a resolution that will only identify a fire but not retard it as required by NRC regulations.
    
    
  • In 1993, TVA "... administered an examination to determine the Chemical Technicians current level of basic chemistry knowledge". The mean scores from TVA's three operating facilities ranged from 26% to 42%.
    
    
  • Water Hammer, a problem originally identified by the NRC in 1978 and resolved without any corrective requirements for operating plants in 1984, causes vibrations that rupture pipes and can spray radioactive water. Five water hammer events have been documented at Watts Bar (TN) with one recently listed by the plant on March 20, 1996.
    
    
  • In 1992, operators of TVA's Sequoy facility (TN) discovered they had been operating the facility despite unreliable radiation monitors for fifteen years -- these monitors are vital to protecting the outside environment. The monitors were not likely to be triggered until a significant amount of radiation had already escaped into the atmosphere.
    
    
• A trend of resolving issues with "No Requirement" exists in which the NRC dismisses high priority safety issues without actually fixing them. From 1984 to the present, more than one-half of 62 high priority Generic Safety Issues (33 high priority GSIs) have been dismissed with "No Requirement". Essentially, the NRC does not require operators to make any safety improvements.
  
  
• In 1985, an NRC Commissioner asserted, " I can think of no other instance in which a regulatory agency has been so eager to stymie its own ability to carry out its responsibilities. . . . the most compelling evidence to date of the Commission majority's open hostility to the regulatory mission of this agency. . . . The Commission has said there is about a 50-50 chance of another severe accident in the next twenty years." The NRC's cavalier handling of high priority safety issues will not reduce this accident risk.
  
  
• The NRC's acquiescence to the nuclear industry is blatant:
  
  
  • In 1987, an NRC Chief stated, "Who the hell is regulating who?(sic) . . . In my opinion, the NRC did not have to buckle to industry pressure. . . . As a result, the NRC has essentially left it to the nuclear industry to regulate itself."
    
    
  • In 1984, an NRC spokesperson stated, "Generic Safety Issues were developed deliberately as an accommodation to the nuclear power industry" -- in order to get plants licensed without resolving significant safety questions.
    
    
  • The NRC allows licensees to decide if they will follow the NRC's requested actions or if they would like to propose their own alternatives. In one case, the NRC management accepted all 25,000 opinions made for new technical specifications by the nuclear industry, despite NRC staff objections.
    
    
  • According to a 1993 NRC Inspector General report, NRC Inspectors stated that they were dissuaded from finding violations at nuclear power plants. Inspectors who made this allegation said they were personally penalized if they found violations -- they would not receive rewards or promotions.
    
    
• NRC Chair Shirley Ann Jackson stated in April 1996, " . . . there is the danger that increased competitiveness in the electricity industry may create pressures to minimize expenditures to the point that safety is compromised." Jackson has repeated this theme in other recent speeches.
  
  
• Currently, the NRC lists 11 nuclear reactors on their "Watch List", of plants needing additional oversight or with a declining trend in performance. They are Indian Point 3 (NY), Millstone 1, 2 and 3 (CT), Browns Ferry 1 and 3 (AL), Dresden 2 and 3 (IL), Hope Creek (NJ), Salem 1 and 2 (NJ).
  
  
• When a plant is listed on the "Watch List", time does not force safety to the forefront -- Dresden's reactors have been on the list for nearly ten years.
  
  
• Nuclear power plants do not get any safer as they age. Aging power plants are a significant safety concern and the NRC's proposal to renew operating licenses, up to an additional twenty years past their current forty-year licenses, will only add to this dilemma.

Background

In 1975, the Atomic Energy Commission (AEC) was dismantled due to its inability to distance itself from the industry it was supposed to be regulating. The result was the creation of the Department of Energy, which was charged with promoting nuclear power, and the creation of the Nuclear Regulatory Commission, which was charged with regulating the U.S. nuclear industry. The Nuclear Regulatory Commission is responsible, according to the Atomic Energy Act of 1954, as amended, for licensing and regulating nuclear facilities and materials and conducting related research.

Currently, there are 116 nuclear power plants in 32 states in the United States: 110 operating plants and 6 plants under construction. These nuclear reactors provide nearly 100% of the Nuclear Regulatory Commission's funding through user fees charged to applicants and licensees.

In 1977, the NRC began categorizing safety problems occurring at many nuclear power plants. Public scrutiny forced the NRC to develop a program, referred to as "Unresolved Safety Issues", to investigate and to disclose annually to Congress issues affecting nuclear power plants. Unresolved Safety Issues are the highest priority safety issues because they have the potential to cause the greatest public harm.

In 1982, "Generic Safety Issues" were conceived, and joined "Unresolved Safety Issues" as issues that had the potential to risk public safety. Generic Safety Issues characterize safety concerns that potentially affect all, or many reactors, and facilities. In 1984, the NRC added GSIs to their Annual Report to Congress.

Each year the NRC promises Congress that safety issues will be resolved quickly, usually within the next year. However, it can take nearly twenty years before the Nuclear Regulatory Commission identifies, resolves, implements and verifies significant safety problems and solutions at nuclear power plants. No one notices. No one is held accountable for these delays. Ironically, licensees are not afforded the same time-frame when issues arise. By law, 10 CFR Part 21, licensees are allotted only 60 days to decide if a potential problem is valid, to submit an action plan for correcting the problem or to report the problem to the NRC. Why then does it take the NRC longer than 60 days to study issues that potentially affect all operating plants?

This delay is due to the Nuclear Regulatory Commission's cozy relationship with the nuclear industry. Recent examples prove that the Nuclear Regulatory Commission's behavior parallels that of its predecessor, the Atomic Energy Commission -- an agency that was unable to regulate the nuclear industry.

Due to the NRC's lack of oversight and enforcement of regulations, it is critical to examine industry documents and identify deficiencies that have existed for many years. Shirley Ann Jackson, Chair of the NRC, stated, "The responsibility for safety rests with the industry. . . . Like any other regulatory body, NRC is essentially an auditing agency." Yet, the mission of the NRC is to regulate the nuclear industry and to enforce safety requirements. There are numerous examples of the nuclear industry making public safety a low priority, and the NRC, which has the power to withdraw a plant's license or shut down the plant, instead passively allows it to proceed.

The Project on Government Oversight (POGO) is a watchdog that investigates and exposes abuse of power, mismanagement, and acquiescence to corporate interests by the federal government. The Nuclear Regulatory Commission, a federal government regulating agency, is guilty of all three of these offenses. POGO has relied on the NRC's own documents, memorandums and regulations to examine an agency that cannot be separated from the industry it is intended to regulate.

POGO would like to thank the many individuals who assisted in making this report possible: Jane Fleming, Ann Harris, William Jocher, James Riccio (Staff Attorney, Public Citizen's Critical Mass Energy Project), Bob Pollard, Paul Gunter (Nuclear Information & Resource Service), Peter Stockton, The Safe Energy Communication Council (SECC), the National Nuclear Safety Network (NNSN), NRC staff members who provided us with updated statistics, the entire staff at the NRC Public Document Room, and the many whistleblowers who cannot be mentioned by name. We appreciate the time that everyone has taken in assisting POGO's investigation.

Partners In Crime: The NRC And The Nuclear Industry Endangering The Public

Decades ago, the NRC developed two programs to list and prioritize safety problems that are occurring at many, if not all, nuclear reactors: Unresolved Safety Issues (USI) and Generic Safety Issues (GSI). Unfortunately, these programs have not made nuclear reactors safer. Issues languish on safety lists for more than ten years, and many issues are considered "resolved" without ever making changes. For example, one issue that was listed as a high priority safety issue in 1984 remains unresolved today. Another issue listed in 1984 was finally resolved in 1994 though no safety improvements were required. Still, another safety issue was resolved in 1985, yet four incidents occurred following its "resolution". Former NRC Chairman Ivan Selin believes the nuclear industry is at fault for problems in handling safety issues:

"As I suggested earlier, a related area in which the industry must do better is in anticipating generic problems and in solving them early. This need will become more acute as the universe of regulated reactors gets older and new generic aging issues emerge. Motor-operated valve issues were initially poorly handled by the industry and stand as an example of how we should not deal with emerging issues. When confronted with the problem, the industry's response was to deny its existence without investigation, forcing the NRC to spend much time and resources to prove the problem's existence. We were often limited in our ability to scope the problem accurately without industry cooperation. Later, when the NRC was able to prove that its concern was valid, both of us found ourselves in a position where a safety issue had been known for several years, but corrective action had not been taken. A similar pattern has sometimes been seen in the way industry has handled steam generator tube issues and the [Boiling Water Reactor] level instrument problems. When generic problems such as these are not promptly and fully addressed, both the NRC and the industry find themselves under justifiable criticism. Additionally, unnecessary financial and organizational resources are often required to deal effectively with such long-festering problems."1 (emphasis added) (APPENDIX A)

It is astonishing that the NRC, the federal agency that regulates one of the most hazardous industries in the world, relies on that industry for even the most basic information. A recent GAO report stated that the NRC's reliance on the nuclear industry will not subside in the future:

"Although one purpose of the NRC's inspection program is to prevent significant events at plants, in practice NRC rarely detects such events before its licensees do. All 16 significant events that the NRC reported for 1993, including the event in South Texas, were initially identified by the licensees rather than by NRC. This situation is unlikely to change because, according to NRC, it has initiatives under way to rely more heavily on licensees to identify and correct problems at nuclear plants."2 (emphasis added)

While Selin blames the licensees for the problematic safety program, the program has also struggled due to a lack of NRC oversight. According to Inside NRC, a trade publication, " . . . the agency has not tracked and processed generic issues as well as it could in the past . . . . "3 GAO found that the NRC's role in overseeing the nuclear industry is restricted:

"Specifically, NRC said that the licensees are responsible for the safe operation of their plants and implied that the intent of its inspection program is limited to ensuring that the licensees identify and resolve potential safety issues before they result in significant problems. . . . Because the licensees are ultimately responsible for the safe operation of their facilities, NRC relies heavily on them to identify and report problems at their facilities. However, NRC inspects a small portion of each licensee's activities to provide independent assurance that the licensees are operating their facilities safely."4 (emphasis added)

The NRC's reliance on the nuclear industry has developed over the years into a blind trust of the nuclear industry that is neither justified nor warranted. Watts Bar Nuclear Plant is a perfect example of why nuclear power plant operators cannot be trusted. In 1988, Ann Harris, a Tennessee Valley Authority (TVA) employee, reported 24 allegations to the NRC -- of which 23 were confirmed by the NRC. These allegations included falsification of documentation, tampering with documentation and covering-up improprieties. The NRC responded to one of the allegations with the following, "The completeness, accuracy, and adequacy of QA documentation is a long standing, documented problem at Watts Bar."5 If the NRC knew this to be a " . . . long standing, documented problem", then why didn't they impose severe fines against those who were involved and send out a message that the NRC will not tolerate improper behavior of any kind? Instead, the NRC has defaulted on its responsibility by giving the nuclear industry more power to regulate itself and, in some areas, is " . . . reducing the regulatory burden on licensees."6

Why has the NRC developed this position? Industry pressure. In a recent speech Commissioner Kenneth C. Rogers stated:

"During a meeting last December, the opinion of those representatives [Federal Energy Regulatory Commission and the National Association of Regulatory Utility Commissioners] was that the Commission should be careful to not pursue regulatory action too quickly because of the general uncertainty about restructuring and deregulation and because a premature regulatory action may have a potentially harmful impact on the industry."7 (emphasis added)

However, the NRC already understands the nuclear industry's motives. In a recent speech, James M. Taylor, NRC Executive Director for Operations, explained the reasons why the industry often operates outside requirements:

"These factors include a very narrowly focused approach to technical issues and their resolution, little evidence of a questioning attitude, limited tracking and trending tools to assess performance and program effectiveness, lack of respect for the [Final Safety Analysis Report], inadequate resolution of identified problems and ineffective quality assurance programs that failed to identify or bring this pattern to management's attention.8" (emphasis added)

"Newspeak" At The NRC

In his novel 1984, George Orwell coined the word "newspeak" to mean the deliberate use of ambiguous and deceptive talk, as by government officials, in seeking to mold public opinion. NRC has taken this definition to new levels. According to the NRC, a safety issue can be "resolved" without requiring any safety improvements. An NRC document revealed:

" . . . about half of the HIGH-priority issues resulted in decisions to take regulatory action, i.e., in retrospect, it appears that resources had been devoted to resolving a large number of issues with no resulting safety improvement."9 (emphasis added)

From 1984 to the present, more than one-half of 62 high priority Generic Safety Issues (33 high priority GSIs) have been dismissed with "No Requirement" or "no resulting safety improvement". The trend of resolving issues with "No Requirement" has allowed the NRC to dismiss high priority safety issues without fixing them. Simply put, listing an issue as a GSI is a convenient way for the NRC to stall and later dismiss safety problems within the nuclear industry, even those categorized as high priority.

Remarkably, a substantial number of high priority issues, which raise important safety concerns or point to elevated risks, are "resolved" without any improvements or changes. What does the NRC mean when it claims a safety issue is "resolved"?

"An issue is considered resolved . . . when its resolution has resulted in either: (a) the establishment of regulatory requirements or guidance (by Rule, SRP change, or equivalent); or (b) a documented authoritative decision that no change in requirements is warranted."10

A General Accounting Office investigation concluded more than a decade ago that:

"NRC considers a generic issue 'resolved' when it identifies and approves a solution to the issue. However, this does not mean that the deficiency has been corrected. Neither NRC nor utilities need to make required changes for an issue to be considered resolved. . . . a resolved issue (i.e., one for which a solution has been identified and approved but not carried out)." 11

In 1979, Max Carbon, Chairman of the NRC's Advisory Committee on Reactor Safeguards (ACRS), stated:

"In some cases an item has been resolved in an administrative sense, recognizing that technical evaluation and satisfactory implementation are yet to be completed. . . . In other instances, the resolution has been accomplished in a narrow or specific sense, recognizing that further steps are desirable, as practical, or that different aspects of the problem require further investigation."12

This process remains true today -- merely reaching the so-called "resolution" of a safety issue can take ten or more years. (APPENDIX B -- Table 1 and Table 2) After an issue is considered "resolved", the industry can still delay for years before actually making changes, if any, to the plant.

Congress receives an NRC Annual Report, disclosing all safety issues and the dates they are scheduled for "resolution". However, Congress does not hold the NRC accountable if a safety issue remains unresolved past the expected "resolution" date. Congress accepts the NRC's standard promise that safety issues, especially those considered the highest priority, will be resolved rapidly -- often, by the following year. As seen in Table 1 and Table 2, these dates routinely are extended each year, yet no one holds the NRC accountable.

The industry's use of the word "implement" does not always mean a problem has been corrected. A resolution is considered implemented once the licensee notifies the NRC that they have made the corrective change. Since many corrective actions are not required to be verified by the NRC, and considering industry's poor record on safety issues, one can hardly be assured that "implemented" actually means fixed.

Not only does the NRC use misleading definitions for "resolved" and "implemented" but also, when an accident does occur, the NRC appears to camouflage willingly its occurrence. The NRC limits the use of "accident", which carries a very negative connotation, and, instead, uses terms such as "abnormal occurrence", "event", "failure" and "incident" to describe accidents at nuclear power plants. This clever technique allows the NRC to calm the fears that many people hold concerning this industry. In this manner, the American public and Congress have been duped into believing that the nuclear industry no longer poses a great danger and that there have been no accidents since TMI.