Human reliability analysis expert judgment

ABSTRACT In Human Reliability Analysis (HRA) the assessment of dependence between human failure events refers to evaluating the influence of the failure on one task on the performance of the subsequent task. In Probabilistic Safety Assessments (PSAs), human action dependencies are commonly evaluated with the THERP method, often extended with Decision Tree (DT) models, to reduce the expert judgment element. This paper compares different DT models used in the HRA practice. The comparison addresses the factors entering the models and the underlying relationships. The comparison shows that, depending on the features of the task under analysis, the results may vary substantially if different DTs are used. Also, often there is limited guidance for the analyst in the assessment of the DT factors this prejudices the repeatability of the assessments because different analysts may very well decide for different assessments. 

The technique for human error-rate prediction (THERP) [ Swain and Guttmann, 1980] is a widely applied human reliability method (Meister, 1984] used to predict human error rates (i.e., probabilities) and the consequences of human errors. The method relies on conducting a task analysis. Estimates of the likelihood of human errors and the likelihood that errors will be undetected are assigned to tasks from available human performance databases and expert judgments. The consequences of uncorrected errors are estimated from models of the system. An event tree is used to track and assign conditional probabilities of error throughout a sequence of activities. 

THERP involves performing a task analysis to provide a description of performance characteristics of human tasks being analyzed. Results are represented graphically in an HRA event tree, which is a formal representation of the required actions sequence. THERP relies on a large human reliability database containing HEPs, which is based upon both plant data and expert judgments. 

---