Report Of The President's Commission On
The Accident At Three Mile Island           > TMI-2 > Kemeny

Quick Links:

Report Index

The Commission

Senior Staff







The Accident




We had a broad mandate from the President to investigate the Nuclear Regulatory Commission. When NRC was split off from the old Atomic Energy Commission, the purpose of the split was to separate the regulators from those who were promoting the peaceful uses of atomic energy. We recognize that the NRC has an assignment that would be difficult under any circumstances. But, we have seen evidence that some of the old promotional philosophy still influences the regulatory practices of the NRC. While some compromises between the needs of safety and the needs of an industry are inevitable, the evidence suggests that the NRC has sometimes erred on the side of the industry's convenience rather than carrying out its primary mission of assuring safety.

Two of the roost important activities of NRC are its licensing function and its inspection and enforcement (I&E) activities. We found serious inadequacies in both.

In the licensing process, applications are only required to analyze "single-failure" accidents. They are not required to analyze what happens when two systems fail independently of each other, such as the event that took place at TMI. There is a sharp delineation between those components in systems that are "safety-related" and those that are not. Strict reviews and requirements apply to the former; the latter are exempt from most requirements "- even though they can have an effect on the safety of the plant. We feel that this sharp either/or definition is inappropriate. Instead, there should be a system of priorities as to how significant various components and systems are for the overall safety of the plant. There seems to be a persistent assumption that plants can be made sufficiently safe to be "people-proof." Thus, not enough attention is paid to the training of operating personnel and operator procedures in the licensing process. And, finally, plants can receive an operating license with several safety issues still unresolved. This places such a plant into a regulatory "limbo" with jurisdiction divided between two different offices within NRC. TMI-2 was in this status at the time of the accident, 13 months after it received its operating license.

NRC's primary focus is on licensing and insufficient attention has been paid to the ongoing process of assuring nuclear safety. An important example of this is the case of "generic problems," that is, problems that apply to a number of different nuclear power plants. Once an issue is labeled "generic," the individual plant being licensed is not responsible for resolving the issue prior to licensing. That, in itself, would be acceptable, if there were a strict procedure within NRC to assure the timely resolution of generic problems, either by its own research staff, or by the utility and its suppliers. However, the evidence indicates that labeling of a problem as "generic" may provide a convenient way of postponing decision on a difficult question.

The old AEC attitude is also evident in reluctance to apply new safety standards to previously licensed plants. While we would accept a need for reasonable timetables for "backfitting," we did not find evidence that the need for improvement of older plants was systematically considered prior to Three Mile Island.

The existence of a vast body of regulations by NRC tends to focus industry attention narrowly on the meeting of regulations rather than on a systematic concern for safety. Furthermore, the nature of some of the regulations, in combination with the way rate bases are established for utilities, may in some instances have served as a deterrent for utilities or their suppliers to take the initiative in proposing measures for improved safety.

Previous studies of I&E have criticized this branch severely. Inspectors frequently fail to make independent evaluations or inspections. The manual according to which inspectors are supposed to operate is so voluminous that many inspectors do not understand precisely what they are supposed to do. There have been a number of incidents in which inspectors have had difficulty in getting their superiors to concentrate on serious safety issues. The analysis of reported incidents by licensees has tended to concentrate on equipment malfunction, and serious operator errors have not been focused on. Finally, while the statutory authority to impose fines is fairly limited, a previous study shows that I&E has made minimal use of even this authority.

Since in many cases NRC does not have the first-hand information necessary to enforce its regulations, it must rely heavily on the industry's own records for its inspection and enforcement activities. NRC accumulates vast amounts of information on the operating experience of plants. However, prior to the accident there was no systematic method of evaluating these experiences, and no systematic attempt to look for patterns that could serve as a warning of a basic problem.

NRC is vulnerable to the charge that it is heavily equipment-oriented, rather than people-oriented. Evidence for this exists in the weak and understaffed branch of NRC that monitors operator training, in the fact that inspectors who investigate accidents concentrate on what went wrong with the equipment and not on what operators may have done incorrectly, in the lack of attention to the quality of procedures provided for operators, and in an almost total lack of attention to the interaction between human beings and machines.

In addition to all the other problems with the NRC, we are extremely critical of the role the organization played in the response to the accident. There was a serious lack of communication among the commissioners, those who were attempting to make the decisions about the accident in Bethesda, the field offices, and those actually on site. This lack of communication contributed to the confusion of the accident. We are also skeptical whether the collegial mode of the five commissioners makes them a suitable body for the management of an emergency, and of the agency itself.

We found serious managerial problems within the organization. These problems start at the very top. It is not clear to us what the precise role of the five NRC commissioners is, and we have evidence that ~ they themselves are not clear on what their role should be. The huge bureaucracy under the commissioners is highly compartmentalized with insufficient communication among the major offices. We do not see evidence of effective managerial guidance from the top, and we do see evidence of some of the old AEC promotional philosophy in key officers below the top. The management problems have been made much harder by adoption of strict rules that prohibit the commissioners from talking with some of their key staff on issues involved in the licensing process; we believe that these rules have been applied in an unnecessarily severe form within this particular agency. The geographic spread, which places top management in Washington and most of the staff in Bethesda and Silver Spring, Maryland (and in other parts of the country), also inhibits the easy exchange of ideas.

We therefore conclude that there is no well-thought-out, integrated system for the assurance of nuclear safety within the current NRC.

We have found evidence of repeated in-depth studies and criticisms both from within the agency and from without, but we found very little evidence that these studies have resulted in significant improvement. This fact gives us particular concern for the future of the present NRC.

For all these reasons we recommend a total restructuring of the NRC. We recommend that it be an independent agency within the executive branch, headed by a single administrator, who is in every sense chief executive officer, to be chosen from outside NRC. The new administrator must be provided with the freedom to reorganize and to bring new blood into the restructured NRC's staff. This new blood could result in the change of attitudes that is vital for the solution of the problems of the nuclear industry.

We have also recommended a number of other organizational and procedural changes designed to make the new agency truly effective in assuring the safety of nuclear power plants. Included in these are an oversight committee to monitor the performance of the restructured NRC and mandatory review by HEW of radiation-related health issues.