Wednesday, January 7, 2015

Human Performance at a Nuclear Power Plant

2015 is off to a slow start in the safety culture (SC) space but we recently saw two mid-2014 articles worth a few words: “Putting People in the Mix” Parts I and II by Ken Ellis, both originally published in Nuclear Engineering International.*  The basic premise is that an incident investigation finding of human error is only “the tip of the iceberg” in understanding human performance issues.

Part I

Part I describes how people add a probabilistic aspect to nuclear plant performance.  Ellis begins by reviewing the nuclear industry’s defense-in-depth: physical barriers, safety systems and contingency plans.  If an incident occurs, then linear root cause analysis starts with the outcome and works back to identify what happened.  Lessons learned are used to update the defense-in-depth system.

But people don’t always behave according to the laws of physics.  People “can circumvent both equipment and process, either unwittingly or wittingly” because of their personal history, perceptions, stress and other factors.  Adding people makes a complicated system (a nuclear plant) a complex one.  One consequence is that a complete and accurate reconstruction of the events preceding an incident may not be possible.  Incident analysis should include investigating the dynamic context in which any relevant human behavior occurred. 

Part II

Part II describes risk management.  It begins with a list of factors that can increase risks at nuclear plants, including lack of leadership, time pressures, complacency and normalization of deviance.  The organization’s primary goal in risk space “is to narrow the band of what constitutes acceptable risk.” The strategy should be to control human behavior by making the boundaries of the work space “explicit and known, and giving workers opportunities to develop coping skills at boundaries.”

Ellis on goes to list  practices that can help improve safety including communication protocols, conservative decision-making and a questioning attitude.  He concludes with some suggestions for managing human performance risk including explicit discussion of complexity and risk boundaries, seeking divergent opinions and understanding how workers interpret messages from corporate. 

Our Perspective 

There is really nothing wrong with these articles.  Ellis covers the ground fairly well in 2400 words intended for a general nuclear industry audience.  But there is nothing new here.  More importantly, this is a brisk treatment of some important concepts about human behavior, the nature of human and system errors, competing mental models of nuclear operations, and desirable management attributes.  The author’s lack of references means a curious reader is left to his own devices.  One really needs direction to key sources, e.g., Dekker, Hollnagel, Reason, Taleb, Vaughan, Woods and the HRO people to gain a meaningful understanding of such concepts.  If you’ve been following Safetymatters for awhile, you know we’ve covered these folks and their ideas at length.

*  K. Ellis, “Putting People in the Mix: Part I,” Nuclear Engineering International (July 18, 2014) and “Putting People in the Mix:Part II,” Nuclear Engineering International (July 21, 2014).  Mr. Ellis is the Managing Director of the World Association of Nuclear Operators (WANO).  Thanks to Dr. W.R. Corcoran for publicizing Part I in the LinkedIn Nuclear Safety Culture group.


  1. I'm just today catching up on the postings from December and this one regarding the Ellis articles.

    My issue is similar to your's: Is that the best you've got?

    More than any other factor contributing to my unease in each of these three instances is the continued official willingness to focus theory, graphical representation, and even incentive structures at the WTP, on the "flatworld" goal of Doing Safety. All the while the sophisticated, beyond linear systems, Do Work Safely - Throughout the Life Cycle, efforts of those accomplished researchers and systems thinkers you noted go uncataloged, unremarked, and uninvited to this conference. Haage and company have not theory of governance for the nuclear power enterprise; rather they have the Betty Crocker Cookbook.

    In each of these representations we have those in observer functions - often without much in the way of day to day nuclear facilities risk management experience - broadcasting "truths" in the direction of those who are in fact charged with such duties. How do the stack of presentation and related materials improve at all upon the work of the INSAG membership that produced the foundational document on Nuclear Safety Management of Residual Uncertainty (INSAG-4) more than two decades ago. Rickover published a primer on the thinking executives guide to this subject in 1983.

    Ultimately these artifacts, official pronouncements of persons in a position of considerable influence are sending a message intended to be soothing - "Trust us public, see how up-to-date we are; the serious people we are listening to for clues about how to run our businesses better." It is the embodiment of Group Think as that institutional leadership defect was described in the 1990's.

    The WTP is of particular frustration to me; it has been, one the evidence of hazards actually encountered since sometime in the 1980's, a challenge in the active chemistry of non-Newtonian slurries which are encumbered with lots of comparatively inert (to the plant materials) radionuclides. Why then are the particular forms of NRC-licensing rules, design for safety and quality control, which Bechtel long employed in building nuclear power plants, deemed to be the ones suitable to the Tank Farms?

    The WTP challenge is a materials management predicament of such novelty as to surely require a fixed and durable capacity to experiment forward until such time as the filling of waste containers becomes routine. At best Bechtel has backed its way into the needed R&D and even then has persisted in the notion that a simple crosswalk of nuclear power QA practices would suffice.

    DOE in a succession of responsible leaders, on occasion with direct involvement of a willingly hands-on Secretary of Energy, has chosen to preserve the myth that Tank Waste is a socio-techno problem unfolding under the umbrella of a politicized RCRA Environmental Protection Compliance Agreement. That for Tank Waste, unlike the majority of CERCLA recovery at Hanford, this method lacks the affordances for management acknowledging that "The Chemistry is in Charge."

  2. Continuing:

    If Oppenheimer, Groves, the War Department and Congressional oversight had operated this way in 1942 it's hard to imagine that most of the tank waste would have ever been created in the first place. The problem with lack of progress at WTP is a contractor who is ill-suited to the nature of the work, engineering away without an ounce of Strategic Governance over the project.

    Remember this is Hazard Reduction work; the material performance of the systems designed by Bechtel are utterly indifferent to productions schedules, court orders, and the wishes of DOE executives, Congressional budgeteers and State Officials. The project does not lack for funds; it lacks for progress in the reduction of uncertainty about how to stabilize the Tank Waste in a socio-techno-politically acceptable manner. It demands a unique multidisciplinary structure similar to that of the Manhattan Project and with the added demand of independent multidisciplinary oversight. Impossible; not at all - rather untried.

    The situation at the WTP has never been a candidate for serious conversation about developing Nuclear Safety Culture within the DOE System of Regulation by Contract. Do other sector's experience with establishing a Do Work Safely Culture have valuable risk insights for the navigation of the Tank Waste Chemistry Predicament - surely they do. But there is more to learn from NASA and the two Shuttle disasters than from Davis Besse; more from Piper Alpha and Deepwater Horizon than from rehashing Millstone 15 years later. And frankly more to learn from Bechtel's demonstrated lack of corporate agility in its global projects.

    My kingdom for an honest conversation without the endless PPT slide shows!


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