Monday, October 8, 2012

DOE Nuclear Safety Workshop

The DOE held a Nuclear Safety Workshop on September 19-20, 2012.  Safety culture (SC) was the topic at two of the technical breakout sessions, one with outside (non-DOE) presenters and the other with DOE-related presenters.  Here’s our take on the outsiders’ presentations.

Chemical Safety Board (CSB)

This presentation* introduced the CSB and its mission and methods.  The CSB investigates chemical accidents and makes recommendations to prevent recurrences.  It has no regulatory authority. 

Its investigations focus on improving safety, not assigning blame.  The CSB analyzes systemic factors that may have contributed to an accident and recognizes that “Addressing the immediate cause only prevents that exact accident from occurring again.” (p. 5) 

The agency analyzes how safety systems work in real life and “why conditions or decisions leading to an accident were seen as normal, rational, or acceptable prior to the accident.” (p. 6)  They consider organizational and social causes, including “safety culture, organizational structure, cost pressures, regulatory gaps and ineffective enforcement, and performance agreements or bonus structures.” (ibid.)

The presentation included examples of findings from CSB investigations into the BP Texas City and Deepwater Horizon incidents.  The CSB’s SC model is adapted from the Schein construct.  What’s interesting is their set of artifacts includes many “soft” items such as complacency, normalization of deviance, management commitment to safety, work pressure, and tolerance of inadequate systems.

This is a brief and informative presentation, and well worth a look.  Perhaps because the CSB is unencumbered by regulatory protocol, it seems freer to go where the evidence leads it when investigating incidents.  We are impressed by their approach.

The NRC presentation** reviewed the basics of the Reactor Oversight Process (ROP) and then drilled down into how potential SC issues are identified and addressed.  Within the ROP, “. . . a safety culture aspect is assigned if it is the most significant contributor to an inspection finding.” (p.12)  After such a finding, the NRC may perform an SC assessment (per IP 95002) or request the licensee to perform one, which the NRC then reviews (per IP 95003).

This presentation is bureaucratic but provides a useful road map.  Looking at the overall approach, it is even more disingenuous for the NRC to claim that it doesn’t regulate SC.


There was nothing new here.  This infomercial for IAEA*** covered the basic history of SC and reviewed contents of related IAEA documents, including laundry lists of desired organizational attributes.  The three-factor IAEA SC figure presented is basically the Schein model, with different labels.  The components of a correct culture change initiative are equally recognizable: communication, continuous improvement, trust, respect, etc.

The presentation had one repeatable quote: “Culture can be seen as something we can influence, rather than something we can control” (p. 10)


SC conferences and workshops are often worthless but sometimes one does learn things.  In this case, the CSB presentation was refreshingly complete and the NRC presentation was perhaps more revealing than the presenter intended.

*  M.A. Griffon, U.S. Chemical Safety and Hazards Investigation Board, “CSB Investigations and Safety Culture,” presented at the DOE Nuclear Safety Workshop (Sept. 19, 2012). 

**  U. Shoop, NRC, “Safety Culture in the U.S. Nuclear Regulatory Commission’s Reactor Oversight Process,” presented at the DOE Nuclear Safety Workshop (Sept. 19, 2012).

***  M. Haage, IAEA, “What is Safety Culture & Why is it Important?” presented at the DOE Nuclear Safety Workshop (Sept. 19, 2012).


  1. Another interesting feature of the IAEA presentation is found at Slide 9/56; here a note is made that there are different "views" of culture: 1) The Variable Approach: Culture as one variable amongst others (e.g. strategy, technology, culture, management system, and safety system; 2) The Metaphor Approach: Culture as something inherent in all aspects of the organization.

    The presenter doesn't really make much use of this distinction and that is unfortunate in my perspective. I've not been able to find any other reference to "The Metaphor Approach," nor does it seem intuitively obvious what that label is suggesting.

    From my reading, I might conclude that when Schein talks about the Executive, Engineering, or Operator subcultures that he is referring to sub-components of some whole envisioned as a specific Corporate Culture. If we think of total performance as an algorithmic accumulation of these subculture influences then it might make sense to say the subcultures reflect "variables." But that seems a strained analogy at best - how to account for the interaction among the subcultures seems to lead to pretty serious complication very quickly in this linear deterministic formulation.

    The Venn diagram approach that portrays Culture as the surround of all the other "variables" listed above seems intuitively more sensible, but does it get beyond being a suggestive illustration?

    In Haage's slide I would say that neither representation qualifies as a model of any useful sort. The Variable Approach could reinforce the notion that identifying the state of an Enterprise Culture is a deductive exercise (e.g. given the NRC SC traits, criteria for assessment can be developed)
    The Metaphor Approach might suggest a more Leadership set and directed place for Culture. Or it might not. There must be a better way.

    1. On further thought I recall that some describing culture do use the metaphor that "culture is to the organization and character is to the individual." In that sense the all inclusive sense of personal character might bear some relationship to an all pervasive institutional presence (e.g. the Marine Corps "a few good men" self-image).

      But it would seem from experience that attribution of monolithic character (e.g. to a Washington or Lincoln) belies the historical reality - certainly many organizations have quite fragmented "faces" owing to stove-piped subcultures that persist right alongside others in large enterprises.

      We need real models with testable dynamics.


Thanks for your comment. We read them all. The moderator will publish comments that are related to our content.