Vit Plant Glop

Hanford WTP

DOE’s Waste Treatment Plant at Hanford, the “Vit Plant”, is being built to process a complex mixture of radioactive waste products from 1950s nuclear weapons production.  The wastes, currently in liquid form and stored in tanks at the site, was labeled “gorp” by William Mullins in one of his posts on the LinkedIn Nuclear Safety thread.*  Actually we think the better reference is to “glop”.  Glop is defined at merriam-webster.com as “a thick semiliquid substance (as food) that is usually unattractive in appearance”.  Readers should disregard the reference to food.  We would like to call attention to another source of “glop” accumulating at the Vit Plant.  It is the various reports by DOE and Hanford regarding safety culture at the site, most recently in response to the Defense Nuclear Facilities Safety Board’s (DNFSB, Board) findings in June 2011.  These forms of glop correspond more closely to the secondary definition in m-w, that is, “tasteless or worthless material”.

The specific reports are the DOE’s Implementation Plan (IP)** for the DFNSB’s review of safety culture at the WTP and the DOE's Office of Health, Safety and Security (HSS) current assessment of safety culture at the site.  Neither is very satisfying but we’ll focus on the IP in this post.

What may be most interesting in the DOE IP package are the reference documents including the DNFSB review and subsequent exchanges of letters between the Secretary of Energy and the DNFSB Chairman.  It takes several exchanges for the DNFSB to wrestle DOE into accepting the findings of the Board.  Recall in the Board’s original report it concluded:

“Taken as a whole, the investigative record convinces the Board that the safety culture at WTP is in need of prompt, major improvement and that corrective actions will only be successful and enduring if championed by the Secretary of Energy.” (IP, p. 33)***

In DOE’s initial (June 30, 2011) response they stated:

“Even while DOE fully embraces the objectives of the Board’s specific recommendations, it is important to note that DOE does not agree with all of the findings included in the Board’s report.”  (IP, p. 42)****

It goes on to state that “specifically” DOE does not agree with the conclusions regarding the overall quality of the safety culture.  Not surprisingly this brought the following response in the DNFSB’s August 12, 2011 letter, “...the disparity between the [DOE’s] stated acceptance and disagreement with the findings makes it difficult for the Board to assess the response….”  (IP, p. 46)*****  Note that in the body of the IP (p. 4) DOE does not acknowledge this difference of opinion either in the summary of its June 30 response or the Board’s August 12 rejoinder. 

We note that neither the DNFSB report nor the DOE IP is currently included among the references on Bechtel's Vit Plant website.  One can only wonder what the take away is for Vit Plant personnel — isn’t there a direct analogy between how DOE reacts to issues raised by the DNFSB and how Vit Plant management respond to issues raised at the plant?  Here’s an idea: provide a link to the safetymatters blog on the Vit Plant website.  Plant personnel will be able to access the IP, the DNFSB report and all of our informative materials and analysis.

In fact, reading all the references and the IP leave the impression that DOE believes there is no fundamental safety culture issue.  Their cause analysis focuses on inadequate expectation setting, more knowledge and awareness and (closer to the mark) the conflicting goals emerging in the construction phase (IP, pp. 5-8).  While endlessly citing all the initiatives previously taken or underway, never does the DOE reflect on why these initiatives have not been effective to date.  What is DOE’s answer?  More assessments and surveys, more training, more “guidance”, more expectations, etc.

We do find Actions 1-5 and 1-6 interesting (IP, p. 16).  These will revise the BNI contract to achieve “balanced priorities”.  This is important and a good thing.  We have blogged about the prevalence of large financial incentives for nuclear executives in the commercial nuclear industry and assessments of most, if not all, other significant safety events (BP gulf disaster, BP refinery fire, Upper Big Branch coal mine explosion, etc.) highlight the presence of goal conflicts.  How one balances priorities is another thing and a challenge.  We have blogged extensively on this subject - search on “incentives” to identify all relevant posts.  In particular we have noted that where safety goals are included in incentives they tend to be based on industrial safety which is not very helpful to the issues at hand.  Our favorite quote comes from our April 7, 2011 post re the gulf oil rig disaster and is taken from Transocean’s annual report:

“...notwithstanding the tragic loss of life in the Gulf of Mexico, we [Transocean] achieved an exemplary statistical safety record as measured by our total recordable incident rate (‘‘TRIR’’) and total potential severity rate (‘‘TPSR’’).”

Our advice for the Vit Plant would be as follows.  In terms of expectations, enforcing rather than setting, might be the better emphasis.  Then monitoring and independently assessing how specific technical and safety issues are reviewed and decided.  Training, expectations setting, reinforcement, policies, etc. are useful in “setting the table” but the test of whether the organization is embracing and implementing a strong safety culture can only be found in its actions.  Note that the Board’s June 2011 report focused on two specific examples of deficient decision processes and outcomes.  (One, the determination of the appropriate deposition velocity for analysis of the transport of radioactivity, the other the conservatism of a criticality analysis.)

There are two aspects of decisions: the process and the result.  The process includes the ability to freely raise safety concerns, the prioritization and time required to evaluate such issues, and the treatment of individuals who raise such concerns.  The result is the strength of the decision reached; i.e., do the decisions reinforce a strong safety culture?  We have posted and provided examples on the blog website of decision assessment using some methods for quantitative scoring.

*  The link to the thread is here.  Search for "gorp" to see Mr. Mullins' comment.

**  U.S. Dept. of Energy, “Implementation Plan for Defense Nuclear Facilities Safety Board Recommendation 2011-1, Safety Culture at the Waste Treatment and Immobilization Plant”  (Dec. 2011).

***  IP Att. 1, DNFSB Recommendation 2011-1, “Safety Culture at the Waste Treatment and Immobilization Plant” (June 9, 2011).

****  IP Att. 2, Letter from S. Chu to P.S. Winokur responding to DNFSB Recommendation 2011-1 (June 30, 2011) p. 4.

*****  IP Att. 4, Letter from P.S. Winokur  to S. Chu responding to Secretary Chu’s June 30, 2011 letter (Aug. 12, 2011) p. 1.

Will Safety Culture Kill Palisades?

To tell the truth, I have no idea.  But the plant has an interesting history and reviewing it may give us some hints with respect to the current situation.

If Palisades were a person, we would think it existed in almost laboratory-like conditions for developing a distinct cultural strain.  It’s elderly, a little “different” and a singleton, with a stillborn sibling and a parent who never really loved it.

Palisades is the 9th oldest of U.S. units that are still operating and was/is Combustion Engineering’s first commercial reactor.  C-E reactors were not as popular as GE or Westinghouse; about 13 percent of the current U.S. fleet uses C-E reactors.  The other old units were owned by companies that developed additional nuclear plants but that didn’t happen for Palisades.  It was supposed to have a big brother, Midland, but the project collapsed, primarily because of construction problems, in 1984 when Midland was about 85% complete, almost bankrupting the owner, Consumers Power (which morphed into CMS Energy and then Consumers Energy.) 

Consumers was looking for someone else to operate or take over the plant as far back as the early 1990s.  Eventually, in 2001, they hired the Nuclear Management Company to operate the plant.  That relationship continued until the plant was sold to Entergy in April 2007.

New managers were able to increase performance in terms of capacity factor (CF).  Under Consumers management, 1996-2000 average CF was 85.2%; under NMC, 2002-2006 CF was 90.0 %; and under Entergy, 2007-2010 CF was 93.0%.  In addition, each of those averages was higher than the average CF of the entire U.S. nuclear fleet for the same period.  (I deliberately omitted 2001; it was a terrible year, with a normal refueling outage followed by a six-month maintenance outage to replace control rod drive assemblies.)

More important from the standpoint of trying to infer something about the safety culture, Palisades kept its nose clean with respect to the NRC.  There were three Severity Level III violations during the Consumers era, and one SL-III and one White violation in 2001.  It looks like three different management regimes were able to maintain an effective safety culture but there has been a recent lapse with three White violations since 2009 and preliminary White and Yellow findings in process.

Conclusion

What does this tell us, if anything?  Has Entergy been squeezing the plant too hard?  Did the CF success under Entergy lead to complacency?  Are there any long-standing material condition problems to sap morale and depress safety culture?  Have there been regular, in-depth independent assessments of organizational issues?  I have no insight into this situation although in our Jan. 12 post, I said it looked like the process of normalization of deviance had occurred.  But there is one thing that should jolt the staff into paying attention to detail, at least for awhile: Some Entergy MBA is carefully watching the numbers.  If the NRC shuts down Palisades, it won’t be long before Entergy folds up its tent and walks away.  No generation means no revenue.  And I can’t believe the PSC or ratepayers in economically depressed Michigan have much interest in bailing out a carpetbagger owner.

Problems at Palisades—A Case of Normalization of Deviance?

The Palisades nuclear plant is in trouble with the NRC.  On Jan. 11, 2012 the NRC met with Entergy (the plant’s owner and operator) to discuss two preliminary inspection findings, one white and one yellow.  Following is the NRC summary of the more significant event.

 “The preliminary yellow finding of substantial significance to safety is related to an electrical fault caused by personnel at the site. The electrical fault resulted in a reactor trip and the loss of half of the control room indicators, and activation of safety systems not warranted by actual plant conditions. This made the reactor trip more challenging for the operators and increased the risk of a serious event occurring. The NRC conducted a Special Inspection and preliminarily determined the actions and work preparation for the electrical panel work were not done correctly.”*

At the meeting with NRC, an Entergy official said “Over time, a safety culture developed at the plant where workers thought if they had successfully accomplished a task in the past, they could do it again without strictly following procedure [emphasis added]. . . .

Management also accepted that, and would reward workers for getting the job done. This led to the events that caused the September shutdown when workers did not follow the work plan while performing maintenance.”**

In an earlier post, we defined normalization of deviance as “the gradual acceptance of performance results that are outside normal acceptance criteria.”  In the Palisades case, we don’t know anything more than the published reports but it sure looks to us like an erosion of performance standards, an erosion that was effectively encouraged by management.

Additional Background on Palisades

This is not Palisades’ first trip to the woodshed.  Based on a prior event, the NRC had already demoted Palisades from the Reactor Oversight Process (ROP) Licensee Response Column to the Regulatory Response Column, meaning additional NRC inspections and scrutiny.  And they may be headed for the Degraded Cornerstone Column.***  But it’s not all bad news.  At the end of the third quarter 2011, Palisades had a green board on the ROP.****  Regular readers know our opinion with respect to the usefulness of the ROP performance matrices.


*  NRC news release, “NRC to Hold Two Regulatory Conferences on January 11 to Discuss Preliminary White and Preliminary Yellow Findings at Palisades Nuclear Plant,” nrc.gov (Jan. 5, 2012).

**  F. Klug, “Decline in safety culture at Palisades nuclear power plant to be fixed, company tells regulators,” Kalamazoo Gazette on mlive.com (Jan. 11, 2012).

***  B. Devereaux, “Palisades nuclear plant bumped down in status by NRC; Entergy Nuclear to dispute other findings next week,” mlive.com (Jan. 4, 2012).

****  Palisades 3Q/2011 Performance Summary, nrc.gov (retrieved Jan. 12, 2012).

2011 End of Year Summary

We thought we would take this opportunity to do a little rummaging around in the Google analytics and report on some of the statistics for the safetymatters blog.

The first thing that caught our attention was the big increase in page views (see chart below) for the blog this past year.  We are now averaging more than 1000 per month and we appreciate every one of the readers who visits the blog.  We hope that the increased readership reflects that the content is interesting, thought provoking and perhaps even a bit provocative.  We are pretty sure people who are interested in nuclear safety culture cannot find comparable content elsewhere.

The following table lists the top ten blog posts.  The overwhelming favorite has been the "Normalization of Deviation" post from March 10, 2010.  We have consistently commented positively on this concept introduced by Diane Vaughan in her book The Challenger Launch Decision.  Most recently Red Conner noted in his December 8, 2011 post the potential role of normalization of deviation in contributing to complacency.  This may appear to be a bit of a departure from the general concept of complacency as primarily a passive occurrence.  Red notes that the gradual and sometimes hardly perceptive acceptance of lesser standards or non-conforming results may be more insidious than a failure to challenge the status quo.  We would appreciate hearing from readers on their views of “normalization”, whether they believe it is occurring in their organizations (and if so how is it detected?) and what steps might be taken to minimize its effect.

A common denominator among a number of the popular posts is safety culture assessment, whether in the form of surveys, performance indicators, or other means to gauge the current state of an organization.  Our sense is there is a widespread appetite for approaches to measuring safety culture in some meaningful way; such interest perhaps also indicates that current methods, heavily dependent on surveys, are not meeting needs.  What is even more clear in our research is the lack of initiative by the industry and regulators to promote or fund research into this critical area.   

A final observation:  The Google stats on frequency of page views indicate two of the top three pages were the “Score Decision” pages for the two decision examples we put forward.  They each had a 100 or more views.  Unfortunately only a small percentage of the page views translated into scoring inputs for the decisions.  We’re not sure why the lack of inputs since they are anonymous and purely a matter of the reader’s judgment.  Having a larger data set from which to evaluate the decision scoring process would be very useful and we would encourage anyone who did visit but not score to reconsider.  And of course, anyone who hasn’t yet visited these examples, please do and see how you rate these actual decisions from operating nuclear plants.

From SCWE to Safety Culture—Time for the Soapbox

Is a satisfactory Safety Conscious Work Environment (SCWE) the same as an effective safety culture (SC)?  Absolutely not.  However, some of the reports and commentary we’ve seen on troubled facilities appear to mash the terms together.  I can’t prove it, but I suspect facilities that rely heavily on lawyers to rationalize their operations are encouraged to try to pass off SCWE as SC.  In any case, following is a review of the basic components of SC:

Safety Conscious Work Environment

An acceptable SCWE* is one where employees are encouraged and feel free to raise safety-related issues without fear of retaliation by their employer.  Note that it does not necessarily address individual employees’ knowledge of or interest in such issues.

Problem Identification and Resolution (PI&R)

PI&R is usually manifested in a facility’s corrective action program (CAP).  An acceptable CAP has a robust, transparent process for evaluating, prioritizing and resolving specific issues.  The prioritization step includes an appropriate weight for an issue’s safety-related elements.  CAP backlogs are managed to levels that employees and regulators associate with timely resolution of issues.

However, the CAP often only deals with identified issues.  Effective organizations must also anticipate problems and develop plans for addressing them.  Again, safety must have an appropriate priority.

Organizational Decision Making

The best way to evaluate an organization’s culture, including safety culture, is through an in-depth analysis of a representative sample of key decisions.  How did the decision-making process handle competing goals, set priorities, treat devil’s advocates who raised concerns about possible unfavorable outcomes, and assign resources?  Were the most qualified people involved in the decisions, regardless of their position or rank?  Note that this evaluation should not be limited to situations where the decisions led to unfavorable consequences; after all, most decisions lead to acceptable outcomes.  The question here is “How were safety concerns handled in the decision making process, independent of the outcome?”

Management Behavior

What is management’s role in all this?  Facility and corporate managers must “walk the talk” as role models demonstrating the importance of safety in all aspects of organizational life.  They must provide personal leadership that reinforces safety.  They must establish a recognition and reward system that reinforces safety.  Most importantly, they must establish and maintain the explicit and implicit weighting factors that go into all decisions.  All of these actions reinforce the desired underlying assumptions with respect to safety throughout the organization. 

Conclusion

Establishing a sound safety culture is not rocket science but it does require focus and understanding (a “mental model”) of how things work.  SCWE, PI&R, Decision Making and Management Behavior are all necessary components of safety culture.  Not to put too fine a point on it, but safety culture is a lot more than quoting a survey result that says “workers feel free to ask safety-related questions.”


*  SCWE questions have also been raised on the LinkedIn Nuclear Safety and Nuclear Safety Culture discussion forums.  Some of the commentary is simple bloviating but there are enough nuggets of fact or insight to make these forums worth following.

Nuclear Industry Complacency: Root Causes

NRC Chairman Jaczko, addressing the recent INPO CEO conference, warned about possible increasing complacency in the nuclear industry.*  To support his point, he noted the two plants in column four of the ROP Action Matrix and two plants in column three, the increased number of special inspections in the past year, and the three units in extended shutdowns.  The Chairman then moved on to discuss other industry issues. 

The speech spurred us to ask: Why does the risk of complacency increase over time?  Given our interest in analyzing organizational processes, it should come as no surprise that we believe complacency is more complicated than the lack of safety-related incidents leading to reduced attention to safety.

An increase in complacency means that an organization’s safety culture has somehow changed.  Causes of such change include shifts in the organization’s underlying assumptions and decay.

Underlying Assumptions

We know from the Schein model that underlying assumptions are the bedrock for culture.  One can take those underlying assumptions and construct an (incomplete) mental model of the organization—what it values, how it operates and how it makes decisions.  Over time, as the organization builds an apparently successful safety record, the mental weights that people assign to decision factors can undergo a subtle but persistent shift to favor the visible production and cost goals over the inherently invisible safety factor.  At the same time, opportunities exist for corrosive issues, e.g., normalization of deviance, to attach themselves to the underlying assumptions.  Normalization of deviance can manifest anywhere, from slipping maintenance standards to a greater tolerance for increasing work backlogs.

Decay

An organization’s safety culture will inevitably decay over time absent effective maintenance.  In part this is caused by the shift in underlying assumptions.  In addition, decay results from saturation effects.  Saturation occurs because beating people over the head with either the same thing, e.g., espoused values, or too many different things, e.g., one safety program or similar intervention after another, has lower and lower marginal effectiveness over time.  That’s one reason new leaders are brought in to “problem” plants: to boost the safety culture by using a new messenger with a different version of the message, reset the decision making factor weights and clear the backlogs.

None of this is new to regular readers of this blog.  But we wanted to gather our ideas about complacency in one post.  Complacency is not some free-floating “thing,” it is an organizational trait that emerges because of multiple dynamics operating below the level of clear visibility or measurement.  

     
*  G.B. Jaczko, Prepared Remarks at the Institute of Nuclear Power Operations CEO Conference, Atlanta, GA (Nov. 10, 2011), p. 2, ADAMS Accession Number ML11318A134.

Regulatory Assessment of Safety Culture—Not Made in U.S.A.

Last February, the International Atomic Energy (IAEA) hosted a four-day meeting of regulators and licensees on safety culture.*  “The general objective of the meeting [was] to establish a common opinion on how regulatory oversight of safety culture can be developed to foster safety culture.”  In fewer words, how can the regulator oversee and assess safety culture?

While no groundbreaking new methods for evaluating a nuclear organization’s safety culture were presented, the mere fact there is a perception that oversight methods need to be developed is encouraging.  In addition, outside the U.S., it appears more likely that regulators are expected to engage in safety culture oversight if not formal regulation.

Representatives from several countries made presentations.  The NRC presentation discussed the then-current status of the effort that led to the NRC safety culture policy statement announced in June.  The presentations covering Belgium, Bulgaria, Indonesia, Romania, Switzerland and Ukraine described different efforts to include safety culture assessment into licensee evaluations.

Perhaps the most interesting material was a report on an attendee survey** administered at the start of the meeting.  The survey covered “national regulatory approaches used in the oversight of safety culture.” (p.3) 18 member states completed the survey.  Following are a few key findings:

The states were split about 50-50 between having and not having regulatory requirements related to safety culture. (p. 7)  The IAEA is encouraging regulators to get more involved in evaluating safety culture and some countries are responding to that push.

To minimize subjectivity in safety culture oversight, regulators try to use oversight practices that are transparent,  understandable, objective, predictable, and both risk-informed and performance-based. (p. 13)  This is not news but it is a good thing; it means regulators are trying to use the same standards for evaluating safety culture as they use for other licensee activities.

Licensee decision-making processes are assessed using observations of work groups, probabilistic risk analysis, and during the technical inspection. (p. 15)  This seems incomplete or even weak to us.  In-depth analysis of critical decisions is necessary to reveal the underlying assumptions (the hidden, true culture) that shape decision-making.

Challenges include the difficulty in giving an appropriate priority to safety in certain real-time decision making situations and the work pressure in achieving production targets/ keeping to the schedule of outages. (p. 16)  We have been pounding the drum about goal conflict for a long time and this survey finding simply confirms that the issue still exists.

Bottom Line

The meeting was generally consistent with our views.  Regulators and licensees need to focus on cultural artifacts, especially decisions and decision making, in the short run while trying to influence the underlying assumptions in the long run to reduce or eliminate the potential for unexpected negative outcomes.

*  IAEA Technical Meeting on Safety Culture Oversight and Assessment, Vienna, Feb. 15-18, 2011.

**  A. Kerhoas, "Synthesis of Questionnaire Survey."