Monday, October 12, 2015

IAEA International Conference on Operational Safety, including Safety Culture

IAEA Building
Back in June, the International Atomic Energy Agency (IAEA) hosted an International Conference on Operational Safety.*  Conference sessions covered Peer Reviews, Corporate Management, Post-Fukushima Improvements, Operating Experience, Leadership and Safety Culture and Long Term Operation.  Later, the IAEA published a summary of conference highlights, including conclusions in the session areas.**  It reported the following with respect to safety culture (SC):

“No organization works in isolation: the safety culture of the operator is influenced by the safety culture of the regulator and vice versa. Everything the regulator says or does not say has an effect on the operator. The national institutions and other cultural factors affect the regulatory framework. Corporate leadership is integral to achieving and improving safety culture, the challenge here is that regulators are not always allowed to conduct oversight at the corporate management level.”

Whoa!  This is an example of the kind of systemic thinking that we have been preaching for years.  We wondered who said that so we reviewed all the SC presentations looking for clues.  Perhaps not surprisingly, it was a bit like gold-mining: one has to crush a lot of ore to find a nugget.

Most of the ore for the quote was provided by a SC panelist who was not one of the SC speakers but a Swiss nuclear regulator (and the only regulator mentioned in the SC session program).  Her slide bullets included “The regulatory body needs to take different perspectives on SC: SC as an oversight issue, impact of oversight on licensees’ SC, the regulatory body’s own SC, [and] Self-reflection on its own SC.”  Good advice to regulators everywhere.

As far as we can tell, no presenter made the point that regulators seldom have the authority to oversee corporate management; perhaps that arose during the subsequent discussion.

SC Presentations

The SC presentations contained hearty, although standard fare.  A couple were possibly more revealing, which we’ll highlight later.

The German, Japanese and United Kingdom presentations reviewed their respective SC improvement plans.  In general these plans are focused on specific issues identified during methodical diagnostic investigations.  The plan for the German Philippsburg plant focuses on specific management responsibilities, personnel attitudes and conduct at all hierarchy levels, and communications.  The Japanese plan concentrates on continued recovery from the Fukushima disaster.  TEPCO company-wide issues include Safety awareness, Engineering capability and Communication ability.  The slides included a good system dynamics-type model.  At EDF’s Heysham 2 in the UK, the interventions are aimed at improving management (leadership, decision-making), trust (just culture) and organizational learning.  As a French operator of a UK plant, EDF recognizes they must tune interventions to the local organization’s core values and beliefs.

The United Arab Emirates presentation described a model for their new nuclear organization; the values, traits and attributes come right out of established industry SC guidelines.

The Entergy presenter parroted the NRC/INPO party line on SC definition, leadership responsibility, traits, attributes and myriad supporting activities.  It’s interesting to hear such bold talk from an SC-challenged organization.  Maybe INPO or the NRC “encouraged” him to present at the conference.  (The NRC is not shy about getting licensees with SC issues to attend the Regulatory Information Conference and confess their sins.)

The Russian presentation consisted of a laundry list of SC improvement activities focused on leadership, personnel reliability, observation and cross-cultural factors (for Hanhikivi 1 in Finland).  It was all top-down.  There was nothing about empowering or taking advantage of individuals’ knowledge or experience.  You can make your own inferences.

Management Presentations

We also reviewed the Management sessions for further clues.  All the operator presenters were European and they had similar structures, with “independent” safety performance advisory groups at the plant, fleet and corporate levels.  They all appeared to focus on programmatic strengths and weaknesses in the safety performance area.  There was no indication any of the groups opined on management performance.  The INPO presenter noted that SC is included in every plant and corporate evaluation and SC issues are highlighted in the INPO Executive Summary to a CEO.

Our Perspective

The IAEA press release writer did a good job of finding appealing highlights to emphasize.  The actual presentations were more ordinary and about what you’d expect from anything involving IAEA: build the community, try to not offend anyone.  For example, the IAEA SC presentation stressed the value in developing a common international SC language but acknowledged that different industry players and countries can have their own specific needs.

Bottom line: Read the summary and go to the conference materials if something piques your interest—but keep your expectations modest.


*  International Atomic Energy Agency, International Conference on Operational Safety, June 23-26, 2015, Vienna.

**  IAEA press release, “Nuclear Safety is a Continuum, not a Final Destination” (July 3, 2015).

Friday, October 2, 2015

Training Materials for Teaching NRC Personnel about Safety Culture

This is a companion piece to our Aug. 24, 2015 post on how the NRC effectively regulates licensee safety culture (SC) in the absence of any formal SC regulations.  This post summarizes a set of NRC slides* for training inspectors on SC basics and how to integrate SC information and observations into inspection reports.

The slides begin with an overview of SC, material you’ve seen countless times.  It includes the Chernobyl and Davis-Besse events, the Schein tri-level model and a timeline of SC-related activities at the NRC.

The bulk of the presentation shows how SC is related to and incorporated in the Reactor Oversight Process (ROP).  The starting point is the NRC SC Policy Statement, followed by the Common Language Initiative** which defined 10 SC traits.  The traits are connected to the ROP using 23 SC aspects.  Aspects are “the important characteristics of safety culture which are observable to the NRC staff during inspection and assessment of licensee performance” (p. 13)  Each SC aspect is associated with one of the ROP’s 3 cross-cutting areas: Human Performance (14 aspects), Problem Identification and Resolution (6 aspects) and Safety Conscious Work Environment (3 aspects).  During supplemental and reactive inspections there are an additional 12 SC aspects to be considered.  Each aspect has associated artifacts that indicate the aspect’s presence or absence.  SC aspects can contribute to a cross-cutting theme or, in more serious cases, a substantive cross-cutting issue (SCCI).***

The integration of SC findings into inspection reports is covered in NRC Inspection Manual Chapter 0612 and NINE different NRC Inspection Procedures (IPs). (p. 30)  In practice, the logic chain between a SC aspect and an inspection report is the reverse of the description in the preceding paragraph.  The creation of an inspection report starts with a finding followed by a search for a related SC cross-cutting aspect.  Each finding has one most significant cause and the inspectors should “find the aspect that describes licensee performance that would have prevented or precluded the performance deficiency represented by that cause.” (p. 33)

Our Perspective

This is important stuff.  When NRC inspectors are huddled in their bunker evaluating their data and observations after reviewing your documentation, crawling around your plant and talking with your people, the information in these slides provides the road map for their determination of how one or more alleged SC deficiencies contributed to a performance problem which resulted in an inspection finding.

Think of the SC aspects as pegs on which the inspectors can hang their observations to beef up their theory of why a problem occurred. Under routine conditions, there are 23 pegs; under more stringent inspections, there are 35 pegs.  That’s a lot of pegs and none of them is trivial which means your organization’s response may consume sizable resources.

We’ll finish with a more cheery thought:  If you get to the point where the NRC is going to conduct an independent assessment of your SC, their team will follow the guidance in IP 95003.  But don’t worry about their competence, “IP 95003 inspection teams will receive "just-in-time" training before performing the inspection.” (p. 43)

Bottom line: If it looks like controlling oversight behavior and quacks like a bureaucrat, then it probably is de facto regulation.


*  NRC Training Slides, “Safety Culture Reactor Oversight Process Training” (July 10, 2015).  ADAMS ML15191A253.  The slides include other material, e.g., a summary of the conditions under which the NRC can “request” a licensee to perform a SC assessment, a set of case studies and sample test questions for trainees.

**  The Common Language Initiative led to NUREG-2165, “Safety Culture Common Language” which was published in early 2014 and we reviewed on April 6, 2014.

***  There are some complicated decision rules for determining when a problem is a substantive cross-cutting issue and these are worth reviewing on pp. 27-28.