Statement on JCO Criticality Accident

Human-Machine Systems Division, Atomic Energy Society of Japan

JCO Accident Special Work Group Chairman Kazuo Furuta

1. Introduction

It is coming clear that the criticality accident occurred at the nuclear fuel conversion facility of JCO in Tokai-mura on September 30, 1999 was a typical organizational accident, where the organization that had admitted an unauthorized manual, the workers who deviated even from the manual, and the company and the regulatory body that overlooked such misconduct caused the accident. Greatly interested in this nature, the Human-Machine Systems Division of the Atomic Energy Society of Japan has set up a special work group to discuss causes and preventive measures of the accident. Many organizations including this work group are now investigating the accident, but it will take for a while to reveal the truth. We will state, however, some important points at this moment from what is already known, because we believe it is very meaningful to do so.

2. Goal of Investigation

As stated above, this accident was not caused by failures or malfunctions of equipment and facility, but it was caused by workers' unsafe actions and organizational factors of the company and the regulatory body. In order to prevent similar accidents, it is therefore inevitable to reveal causes related to human factors. We believe it is impossible to understand accident mechanism without referring to human factor issues. The accident is now under investigation from the civil, criminal and regulatory viewpoints, and causes and responsibilities are already argued. Though social need for accusation is undeniable, detailed and academic investigation that do not aim at accusation is highly necessary for accident prevention. Since the accident occurred from unbelievable violation of rules in a fuel conversion facility, which is relatively a peripheral sector of nuclear power utilization, there is an argument that the accident is an exception and problems are not general in the nuclear industry of our country. But we believe this argument is wrong, because we do not certainly know there are no other latent pitfalls of nuclear safety. In addition, it is unreasonable to claim that a favorable situation, though it presently exists, will last forever, because defense in depth may be threatened by organizations that cannot adapt themselves to the changing environment. All those who have an interest in the nuclear industry must make every effort to learn from the accident and make use of the lesson.

3. Scope of Investigation

In addition to misconduct of JCO, human factors of the regulatory body that overlooked it caused the accident. Hence problems of not only JCO but also the regulatory body should be focussed on. It is criticized again that inappropriate response to the emergency delayed its termination and enlarged the influence of the accident. It is also due to human factors of the company, the central government, and the municipal governments. Causes of inappropriate response after the accident as well as those of accident itself must be investigated both for the company and the authorities. As for investigation on the company before the accident, short-term and long-term viewpoints are necessary. The short term viewpoint deals with workers' unsafe actions that lead directly to the accident, and events occurred within a short time period before accident initiation are the targets of investigation. In investigation from the long-term viewpoint, formation process of organizational factors behind the workers' unsafe actions is to be revealed by chronologically tracing not only adoption process of the unauthorized manual but also gradual changes of company's business climate, organizational culture, organizational structure, staffing, and so on. Having finished investigation from all of the above viewpoints, it is possible to understand the truth of the accident, and then lessons for accident prevention will be available.

4. Organizational Error

An unsafe human action is generally called a human error or simply an error, but it includes a wide spectrum of actions from a simple error due to inattention to sabotage. The workers' unsafe actions that directly caused the accident are different from the basic human error types, that is, slip, lapse, and mistake, because their actions intentionally deviated from the normative procedure. At the same time, they are clearly different from sabotage, because the workers did never expect bad outcomes but they were motivated to high productivity, though they might have been aware of illegitimacy of the procedure. Moreover, while a human error is apt to be attributed to individuals, organizational factors deeply affected the workers' actions and they can hardly be attributed to individual problems. The organizational error observed in this accident must be distinguished from ordinary human errors as well as sabotage. Human errors that do not include violation have been studied extensively in previous studies of human factors, but lately organizational errors with violation have come to be serious issues. In order to prevent such organizational errors, common patterns where organizational errors are likely to emerge have to be found by analyzing previous cases of accidents, detecting these patterns, and blocking the event sequence along the error emerging process. We should aim also to clarify how local factors of the working environment are affected by global organizational factors, and to establish performance indicators for assessment of organizations and the system for continuous assessment. Investigation of the JCO accident from the viewpoints described in the previous section is the first step to obtain valuable knowledge on organizational errors. 5. Study on Organizational Factors and Safety Culture Organizational factors and safety culture have attracted world wide attention after the Chernobyl accident. Some researchers conducted basic studies of high performance organizations, and many sessions in international conferences have been dedicated to this topic. Still we have insufficient knowledge on their well-established taxonomy, their actual effects on human performance, and practical performance indicators for assessment; arguments tend to float around conceptual levels. Particularly in Japan, people are likely to discuss organizational factors and safety culture as a spiritual issue so that efforts for scientific analysis and systematic practice are lacking compared with western countries. We should therefore start extensive studies of organizational factors and safety culture that can be engineered as a lesson of the accident.