Treatment of human and organizational factors for multi-unit HRA: Application of SPAR-H method

Abstract

Human reliability analysis (HRA) refers to a method for evaluating human errors and providing human error probabilities for application in probabilistic safety assessments (PSAs). Since the Fukushima accident, HRA has been considered an important issue in multi-unit PSAs. However, existing HRA methods generally focus on the analysis of human errors in single-unit PSAs, while HRA methods for the multi-unit PSA have not been well established. If an accident occurs that influences more than one unit, the human and organizational factors due to the formation of accident management organizations, the use of shared or mobile equipment, and the influence of a severe accident on another unit may have a crucial impact on the plant safety and the result of the PSA. Therefore, issues have been raised regarding the analysis and evaluation of the human and organizational factors in a multi-unit HRA. This study aims to develop an approach to treat the human and organizational factors for a multi-unit HRA on the basis of Standardized Plant Analysis Risk HRA (SPAR-H) method. This approach is applicable to level 1, 2, and 3 PSA. Firstly, six multi-unit task types were considered based on a previous study that identified and categorized human and organizational factors for multi-unit HRA. Secondly, a task analysis considering HRA event tree analysis, suggested by Technique for Human Error Rate Prediction (THERP), and a timeline analysis were performed for six multi-unit task types. Thirdly, to identify the challenges of the existing SPAR-H method for multi-unit task types, an evaluation of the applicability of the existing SPAR-H method was conducted by qualitative and quantitative analyses. In addition, the dependence assessment approach suggested in the existing SPAR-H was also evaluated in terms of multi-unit situations. Lastly, by treating the multi-unit challenges identified in the previous section, methods to 1) analyze six multi-unit task types and 2) evaluate the dependence between the operator actions were suggested with some examples. The study is expected to contribute to the estimation and calculation of human error probabilities for applications in multi-unit PSAs.