Evaluation of Human Error Probability from Simulator Experiments based on Operator Response Times

Human Reliability Analysis (HRA) is an essential part of Probabilistic Safety Assessment (PSA) that incorporates the influence of human errors towards the overall risk, as Human Error Probability (HEP). One of the major concerns currently in HRA is shortage of human performance data. The widely used HRA methods generally evaluate HEP values based on the aggregation of empirical evidence and expert opinion. As a result, the HEP values generally evaluated with HRA methods have uncertainty in their assessments. In order to reduce the uncertainty in HEP estimates, reliance on expert opinion needs to be reduced and quality HRA data needs to be gathered. Simulators are a good source of generating operator performance data. As a result, to address this gap area in HRA, simulator experiments were done in a reactor simulator to gather operator response time for estimation of plant particular HEP values. An appropriate probability density function (pdf) was fitted to the operator response times and the corresponding distribution parameters were identified based on the goodness of fit test. The deterministic analysis revealed that the time available to the operator for the chosen simulated scenario was nearly 12 minutes. The area under the pdf for the operator response time value greater than 12 minutes was evaluated, which corresponds to HEP value of 2E-2, for the scenario of interest. The implications of this research will be helpful in the evaluation of HEP values to account for human performance in the PSA of advanced reactors.