Results and Insights from the US HRA Empirical Study

摘要

As an effort to improve the robustness of human reliability analysis (HRA), the US Nuclear Regulatory Commission (NRC) participated and supported the International HRA Empirical Study (referred to as the International Study in the paper), in which HRA predictions of different analysts and methods were compared to observed crew performance data at Halden Reactor Project's HAMMLAB (HAlden huMan-Machine LABoratory) simulator facilities. The study produced important findings; however, since there was only one case in the International Study where the same HRA method was applied by different teams, it was difficult to clearly separate method specific effects from variability created by the analysts' application of a given method. Therefore, the US HRA Empirical Study (referred to as the US Study in the paper), which was performed on a US nuclear power plant (NPP) simulator, was conducted to improve the insights developed from the International Study. One major objective of the study was to test the consistency and accuracy of HRA predictions among different analyst teams using the same methods.