Social Scientists in an Adversarial Environment: The Nuclear Regulatory Commission and Organizational Factors Research

Abstract

This paper examines the Nuclear Regulatory Commission’s (NRC’s) pursuit of social science research that could inform the oversight of nuclear power plant management. Perhaps no nuclear regulator has been as supportive of research on the intersection of organizational factors and reactor safety or as cautious in applying those findings to its regulations.

This dissonance was rooted in the NRC’s long-held conviction that it should regulate power plants not people, which conflicted with its regulatory experience after the 1979 Three Mile Island accident (TMI). Intrusive oversight of a licensee’s “business,” it was believed, would destroy its sense of ownership for safety. TMI challenged that understanding of the NRC’s role, and a series of mishaps at other plants compelled the agency to cross the line between regulation and management. The NRC’s relationship with industry became highly adversarial, and the agency turned to social scientists to help establish an objective basis to judge a licensee’s organizational culture. Behavioral experts joined plant oversight review teams and received generous funding to quantify the contribution of organizational factors to accident risk. Scores of scholars at national laboratories and a dozen universities contributed, but the NRC abandoned the research in the mid-1990s in the face of inconclusive research and industry resistance.

In need of a less controversial oversight program, the NRC abandoned direct assessment of plant management for a more quantitative approach that relied on plant performance indicators. When the 2002 Davis-Besse vessel head erosion event came perilously close to a significant loss-of-coolant accident, it raised questions about the appropriate role for the NRC in assessing a licensee’s safety culture. The NRC revised its oversight program to incorporate qualitative insights from its earlier research while still acknowledging the line between regulation and management. The NRC learned that while there were substantial cultural and technical obstacles to integrating safety culture insights with established management and regulatory practices, it was necessary to overcome them. The agency found stability in its contentious oversight program only when it made appropriate room for safety culture expertise.

Keywords:

• History of nuclear power
• organizational factors
• safety culture
• probabilistic risk assessment, U.S. Nuclear Regulatory Commission

Notes

^a A colloquialism for the useless act of teaching someone to do what they already know.

^b Battelle did not define “organizational factors,” but current definitions capture their 1980s essence as “the organizational structures, processes, and behaviors that influence the actions of individuals at work.” See S. Peters and others, “Organizational Factors in PRA: Twisting Knobs and Beyond,” in the Proceedings of the 2019 International Topical Meeting on Probabilistic Safety Assessment and Analysis (PSA 2019), April 28–May 3, 2019, Charleston, South Carolina, NRC ADAMS ML19057A474.

^c “Safety culture” was coined by the International Nuclear Safety Advisory Group of the IAEA. Initially, safety culture had a narrow definition in the research literature as one factor within a broader organizational culture. In turn, organizational culture was one of many organizational factors, such as resource allocation, training, and organizational knowledge. Unfortunately, this logical hierarchy broke down as safety culture grew into an omnibus concept that subsumed many elements once considered organizational factors. For example, the NRC’s safety culture policy statement and related documents include several traits of a healthy safety culture previously seen as organizational factors. See “Identification and Assessment of Organizational Factors Related to the Safety of NPPs, State of the Art Report,” NEA/CSNI/R(99)21, Vol. 1, pp. 11–21, Nuclear Energy Agency, Committee on the Safety of Nuclear Installations (1999) and S. Peters and others, “Organizational Factors in PRA: Twisting Knobs and Beyond,” in Proceedings of the 2019 International Topical Meeting on Probabilistic Safety Assessment and Analysis (PSA 2019), April 28–May 3, 2019, Charleston, South Carolina, NRC ADAMS ML19057A474.