Analysis of Simulator Data for HRA Purposes (PSAM-0037)


The paper addresses the use of nuclear power plant simulators for Human Reliability Assessment (HRA) purposes as an extension of the normal training of control room personnel. The paper opens with a brief history of the use of simulators for this purpose, starting with the US Nuclear Regulatory Commission (USNRC) program in the early 80s and continuing with projects undertaken by the Electric Power Research Institute (EPRI), Electricité de France and later by other countries. The USNRC program lead to the use of Time Reliability Curves (TRC) in the HRA field and to the development of the Human Cognitive Reliability method and hence gave rise to various programs to collect and analyze simulator data, including the Operator Reliability Experiments funded by the Electric Power Research Institute and others.

The data collection processes improved over the years to go from the early USNRC approach, which collected only time data and spread to include observer data, time data and video recordings. The increased emphasis on observer data mirrored to some extent the development of the HRA field. The HRA field was challenged by Ed. Dougherty to produce a second generation of HRA methods. Erik Hollnagel responded by identifying relationship of context and human error probability (HEP). Others were seeing through the experience of simulator observations, the limitations of the current HRA methods, including THERP and various TRCs.

The later simulator studies showed that control-room crews could make observable errors, some of which were recovered. The paper depicts a typical TRC which displays not only the non-response characteristics used in the earlier USNRC TRCs, but also the presence of other data points which do not fit within the standard probability distribution. Close review of these data point indicates the effects of context upon the responses of crews.

The TRCs are useful to see the effects of context upon the TRCs, for example the paper contains TRCs from the same plant modified by human factor changes to control boards. Also, changes in the designed response of crews to a given accident can lead to vastly different TRCs. So TRCs can be used as a tool to examine the effect of improvements or otherwise on crew performance. The whole context affects both the shape of the TRCs and deviations from the random distribution curves, so it is possible to redesign the context to change the TRC shape without necessarily enhancing the reliability of the process.

By collecting simulator data one can see the presence of both slips and mistakes. It is known that slips are easier to recover from and that it is more difficult to recover from mistakes. The paper refers to some experiments involving an emergency operating procedures tool, which help to supply the cognitive feedback to crew and reduce the effect of cognitive lockup. The tool supplies the feedback necessary for the crew to recover from mistakes and this bourn out from the simulator results.

Since context is an important effect the simulator data is used to determine the influence of various context elements. Careful observations of crew performance can yield insights into the effects of context. One HRA tool developed based upon context is the Holistic Decision Tree method and it is discussed in the paper. The use of such models can incorporate context from either observations or use of expert judgment. In accident scenarios where crews' error rates are high the HEPs can be used to check relationships between HEP and context. Of course, it would be better to carryout experiments in which context were changed in a controllable manner. The test of the EOP support system was one such case where it was possible to see the impact of improvement in operator performance and reliability.

  • Abstract
  • Introduction
  • Brief History of Simulator Data Collection Studies
  • Data Collection Systems and Processes
  • Insights from Time and Observational Data
  • Use of Simulator Data for HRA Purposes
  • Discussions and Conclusions
  • Acknowledgments
  • References

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related eBook Content
Topic Collections

Sorry! You do not have access to this content. For assistance or to subscribe, please contact us:

  • TELEPHONE: 1-800-843-2763 (Toll-free in the USA)
  • EMAIL: asmedigitalcollection@asme.org
Sign In