Thursday, February 28, 2013

The Safety Culture Common Language Path Forward (Update)

The intent of the NRC's Safety Culture Common Language Path Forward initiative is to describe safety culture (SC) attributes at a more detailed level than the NRC’s SC policy statement.  On January 29-30, 2013 the NRC held a public workshop to finalize the draft SC common language.*  The document they issued after the workshop** contains attribute definitions and examples of behavior and artifacts that support or embody each attribute.  This document will be used by the NRC to develop a NUREG containing the final common language.

Last March we posted on a draft produced by previous workshops, focusing on areas we consider critical for a strong SC: decision making, corrective action, management incentives and work backlogs.  In that post, our opinion was that decision making and corrective action were addressed in a satisfactory manner, the treatment of incentives was minimally acceptable and backlogs were all but ignored.

So, how does the “final” language treat the same subject areas?  Is it better than the draft comments we reviewed last March?  The arrows indicate whether the final version is better
, the same → or worse ↓.

Decision making – Good.  Decision making incorporates “. . . a consistent, systematic approach to make decisions” (p. 51) and a conservative bias, i.e., “. . . decision-making practices that emphasize prudent choices over those that are simply allowable. A proposed action is determined to be safe in order to proceed, rather than unsafe in order to stop.” (p. 52)  In addition, communicating, explaining and justifying individual decisions is mentioned throughout the document. 

Goal conflict is addressed under leader behavior “. . . when resolving apparent conflicts between nuclear safety and production” (p. 12) and leaders “avoid unintended or conflicting messages that may be conveyed by operational decisions” (p. 37); work process “activities are coordinated to address conflicting or changing priorities.” (p. 23) 

Corrective action – Satisfactory
.  The section on problem identification and resolution (pp. 13-17) is suffused with desirable characteristics of corrective actions and the CAP.  A good CAP has a low threshold for identifying issues and problems are thoroughly evaluated.  Corrective actions are timely, effective and prevent recurrence of problems.  Periodic analysis of CAP and other data is used to identify any programmatic or common cause issues.

Management incentives – Unsatisfactory.  The section on incentives appears to focus on workers, not managers: “Leaders ensure incentives, sanctions, and rewards are aligned with nuclear safety policies and reinforce behaviors and outcomes which reflect safety as the overriding priority.” (p. 7)  This is even less complete than the single sentence that appeared in last year's draft: “Senior management incentive program [sic] reflect a bias toward long-term plant performance and safety.”*** The failure to mention the senior management incentive program is a serious shortcoming.

Backlogs – Minimally Acceptable.  Backlogs are specifically mentioned in maintenance and engineering (p. 24) and document changes (p. 25).  In addition, problem evaluation, corrective actions, CAP trending analyses, operating experience lessons and many administrative activities are supposed to be addressed in a “timely” manner.  I hope that implies that backlogs in these areas should not be too large.     

But attention to backlogs is still important.  Repeating what we said last year, “Excessive backlogs are demoralizing; they tell the workforce that accomplishing work to keep the plant, its procedures and its support processes in good repair or up-to-date is not important.  Every “problem plant” we worked on in the late 1990s had backlog issues.”


Overall, this latest document is an improvement over the March 2012 version but still short of what we'd like to see.

*  M.J. Keefe (NRC) to U.S. Shoop (NRC), “Summary of the January 29-30, Workshop to Develop Common Language for Safety Culture” (Feb. 7, 2013)  ADAMS ML13038A059.

**  Nuclear Safety Culture Common Language 4th Public Workshop January 29-31, 2013  ADAMS ML13031A343.

***  U.S. Shoop (NRC) to J. Giitter (NRC), “Safety Culture Common Language Path Forward” (Mar. 19, 2012) p. 12.  ADAMS ML12072A415.

1 comment:

  1. From the NSC Policy statement: “Nuclear Safety Culture is defined as the core values and behaviors resulting from a collective commitment by leaders and individuals to emphasize safety over competing goals to ensure protection of people and the environment.”

    In the elaborated discussion on Leadership Safety Values and Actions, items LA.3, LA. 4, and LA 5 contain the phrase “…nuclear safety remains the overriding priority.”

    If one reads these two expressions and has no difficulty connecting them together then what’s all the fuss about – surely everyone knows what nuclear safety is and how important it is; so as long as it “remains the overriding priority” then all is good. Remind me, the definition of “nuclear safety” is where again?

    But wait, what if a responsible manager or an executive with a publicly owned firm comes up to a situation in which goal conflicts have emerged from different staff analyses starting with different technical and financial assumptions (those do occur in large capital intensive organizations)? Who’s to say if the goal conflict constitutes an instance of competing priorities, or just a misalignment of methods and sources?

    Likewise when it comes to assessment and action on feedback of all kinds from operations or from independent assessments, or internal quality reviews, or external events, presumably the thresholds of significance and the criteria of legitimate and authentic (they’re not the same attribute) evidence for responsiveness to the NRC’s “expectations” are transparently clear – its only me that’s not seeing it!

    In the idyllic world imagined by the searchers for “safety culture common language” such ambiguities probably do not arise. Reading these descriptions it is not hard to imagine that an “expert system” computer program is being crafted that future staff (from corporate HR and office tower security, to the most recently hired auxiliary operator – and of course all those pesky contractors) can carry on their mobile computers to be consulted whenever doubt might arise concerning “overriding priorities.”

    There is an inescapable flaw in this utopian and more than a little puritanical exercise – it has no halting rule – the potential for regression is nearly infinite on: “How much is enough?” Does no one wonder if the benefit of all this hair-splitting is really to promote better sense-making in complex circumstances?

    If one wonders what such an arc to permanent dissatisfaction with Nuclear Safety Culture looks like review the extended oversight for Safety Culture inadequacies accorded the Palisades plant described elsewhere in Safety Matters.

    And does anyone notice that at some point this elaboration begins to duplicate the content of the best practices performance objectives and measurement criteria contained in an entire generation of guidance developed by INPO and another developed by IAEA. This is a mass search for the “golden rivet;” doubtlessly it will continue apace for what is to stop a crew with the license of an “overriding priority?”

    From any social science or neuro-cognitive perspective we look the amateurish at best and psychologically dysfunctional at worst. And how, in any way at all would this “common language” have overcome the collected anxiety (it nurtured by decades of experience with hostile stakeholders) of TEPCO, the Nuclear Industry Owners group, and the responsible Ministry (METI) that the public was not to be trusted with information indicating an increase in risk from tsunami events?


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