From the Executive Summary
The Accident Investigation Board (the Board) concluded that a more thorough hazard analysis, coupled with a better filter system could have prevented the unfiltered above ground release. (p. ES-1)
The root cause of the incident was Nuclear Waste Partnership’s (NWP**, the site contractor) and the DOE Carlsbad Field Office’s (CBFO) failure to manage the radiological hazard. “The cumulative effect of inadequacies in ventilation system design and operability compounded by degradation of key safety management programs and safety culture [emphasis added] resulted in the release of radioactive material . . . and the delayed/ineffective recognition and response to the release.” (pp. ES 6-7)
The report presents eight contributing causes, most of which point to NWP deficiencies. SC was included as a site-wide concern, specifically the SC does not fully implement DOE safety management policy, “[t]here is a lack of a questioning attitude, reluctance to bring up and document issues, and an acceptance and normalization of degraded equipment and conditions.” A recent Safety Conscious Work Environment (SCWE) survey suggests a chilled work environment. (p. ES-8)
The report includes 31 conclusions, 4 related to SC. “NWP and CBFO have allowed the safety culture at the WIPP project to deteriorate . . . Questioning attitudes are not welcomed by management . . . DOE has exacerbated the safety culture problem by referring to numbers of [problem] reports . . . as a measure of [contractor] performance . . . . [NWP and CBFO] failed to identify weaknesses in . . . safety culture.” (pp. ES 14-15, 19-20)
The report includes 47 recommendations (called Judgments of Need) with 4 related to SC. They cover leadership (including the CBFO site manager) behavior, organizational learning, questioning attitude, more extensive use of existing processes to raise issues, engaging outside SC expertise and improving contractor SC-related processes. (ibid.)
The body of the report presents the details behind the conclusions and recommendations. Following are some of the more interesting SC items, starting with our hot button issues: decision making (esp. the handling of goal conflict), corrective action, compensation and backlogs.
The introduction to section 5 on SC includes an interesting statement: “In normal human behavior, production behaviors naturally take precedence over prevention behaviors unless there is a strong safety culture - nurtured by strong leadership.” (p. 61)
The report suggests nature has taken its course: WIPP values production first and most. “Eighteen emergency management drills and exercises were cancelled in 2013 due to an impact on operations. . . .Management assessments conducted by the contractor have a primary focus on cost and schedule performance.” (p. 62) “The functional checks on CAMs [continuous air monitors] were often delayed to allow waste-handling activities to continue.” (p. 64) “[D]ue consideration for prioritization of maintenance of equipment is not given unless there is an immediate impact on the waste emplacement processes.” (p. ES-17) These observations evidence an imbalance between the goals of production and prevention (against accidents and incidents) and, following the logic of the introductory statement, a weak SC.
The corrective action program has problems. “The [Jan. 2013] SCWE Self-Assessment . . . identified weaknesses in teamwork and mutual respect . . . Other than completing the [SCWE] National Training Center course, . . . no other effective corrective actions have been implemented. . . . [The Self-Assessment also ]“identified weaknesses in effective resolution of reported problems.” (p. 63) For problems that were reported, “The Board noted several instances of reported deficiencies that were either not issued, or for which corrective action plans were not developed or acted on for months.” (p. 65)
Here is the complete text of Conclusion 14, which was excerpted above: “DOE has exacerbated the safety culture problem by referring to numbers of ORPS [incident and problem] reports and other deficiency reporting documents, rather than the significance of the events, as a measure of performance by Source Evaluation Boards during contract bid evaluations, and poor scoring on award fee determinations. Directly tying performance to the number of occurrence reports drives the contractor to non-disclosure of events in order to avoid the poor score. [emphasis added] This practice is contrary to the Department’s goals of the development and implementation of a strong safety culture across our projects.” (p. ES-15) ‘Nuff said.
Maintenance was deferred if it interfered with production. Equipment and systems were allowed to degrade (pp. ES-7, ES-17, C-7) There is no indication that maintenance backlogs were a problem; the work simply wasn’t done.
Other SC Issues
In addition to our Big Four and the issues cited from the Executive Summary, the report mentions the following concerns. (A listing of all SC deficiencies is presented on p. D-3.)
- Delay in recognizing and responding to events,
- Bias for negative conclusions on Unreviewed Safety Question Determinations, and
- Infrequent presence of NWP management in the underground and surface.
For starters, the Board appears to have a limited view of what SC is. They see it as a cause for many of WIPP's problems but it can be fixed if it is “nurtured by strong leadership” and the report's recommendations are implemented. The recommendations are familiar and can be summed up as “Row harder!”*** In reality, SC is both cause (it creates the context for decision making) and consequence (it is influenced by the observed actions of all organization members, not just senior management). SC is an organizational property that cannot be managed directly.
The report is a textbook example of linear, deterministic thinking, especially Appendix E (46 pgs.) on events and causal factors related to the incident. The report is strong on what happened but weak on why things happened. Going through Appendix E, SC is a top-level blanket cause of nuclear safety program and radiological event shortcomings (and, to a lesser degree, ventilation, CAMs and ground control problems) but there is no insight into how SC interacts with other organizational variables or with WIPP’s external (political, regulatory, DOE policy) environment.
Here’s an example of what we’re talking about, viz., how one might gain some greater insight into a problem by casting a wider net and applying a bit of systems thinking. The report faults DOE HQ for ineffective oversight, providing inadequate resources and not holding CBFO accountable for performance. The recommended fix is for DOE HQ “to better define and execute their roles and responsibilities” for oversight and other functions. (p. ES-21) That’s all what and no why. Is there some basic flaw in the control loop involving DOE HQ, CBFO and NWP? DOE HQ probably believes it transmits unambiguous orders and expectations through its official documents—why weren’t they being implemented in the field and why didn’t DOE know it? Is the information flow from DOE to CBFO to NWP clear and adequate (policies, goals); how about the flow in the opposite direction (performance feedback, problems)? Is something being lost in the translation from one entity to another? Does this control problem exist between DOE HQ and other sites, i.e., is it a systemic problem? Who knows.****
Are there other unexamined factors that make WIPP's problems more likely? For example, has WIPP escaped the scrutiny and centralized controls that DOE applies to other entities? As a consequence, has WIPP had too much autonomy to adjust its behavior to match its perception of the task environment? Are DOE’s and WIPP’s mental models of the task environment similar or even adequate? Perhaps WIPP (and possibly DOE) see the task environment as simpler than it actually is, and therefore the strategies for handling the environment lack requisite variety. Was there an assumption that NWP would continue the apparently satisfactory performance of the previous contractor? It's obvious these questions do not specifically address SC but they seek to ascertain how the organizations involved are actually functioning, and SC is an important variable in the overall system.
Contrast with Other DOE SC Investigations
This report presents a sharp contrast to the foot-dragging that takes place elsewhere in DOE. Why can’t DOE bring a similar sense of urgency to the SC investigations it is supposed to be conducting at its other facilities? Was the WIPP incident that big a deal (because it involved a radioactive release) or is it merely something that DOE can wrap its head around? (After all, WIPP is basically an underground warehouse.) In any event, something rang DOE’s bell because they quickly assembled a 5 member board with 16 advisor/consultants and produced a 300 page report in less than two months.*****
Bottom line: You don't need to pore over this report but it provides some perspective on how DOE views SC and demonstrates that a giant agency can get moving if it's motivated to do so.
* DOE Office of Environmental Management, “Accident Investigation Report: Radiological Release Event at the Waste Isolation Pilot Plant on February 14, 2014, Phase 1” (April 2014). Retrieved April 30, 2014. Our thanks to Mark Lyons who posted this report on the LinkedIn Nuclear Safety group discussion board.
** NWP LLC was formed by URS Energy and Construction, Inc. and Babcock & Wilcox Technical Services Group, Inc. Their major subcontractor is AREVA Federal Services, LLC. All three firms perform work at other, i.e., non-WIPP, DOE facilities. NWP assumed management of WIPP on Oct. 1, 2012. From NWP website. Retrieved May 2, 2014.
*** To the Board's credit, they did not go looking for individual scapegoats to blame for WIPP's difficulties.
**** In fairness, the report has at least one example of a feedback loop in the CBFO-NWP sub-system: CBFO's use of the condition reports as an input to NWP’s compensation review and NWP's predictable reaction of creating fewer condition reports.
***** The Accident Investigation Board was appointed on Feb. 27, 2014 and completed its Phase 1 investigation on March 28, 2014. The Phase 1 report was released to the public on April 22, 2014.