Research On Human Reliability Analysis And Its Application To SGTR Accident Of AP1000 Nuclear Power Plants

The historical experience shows that the human factors is one of the extremely important elements which influence the complex industrial systems. Correctly allocating human-system functions, reasonably designing operational procedures and human-system interfaces have become the key to reducing or mitigating the accidents caused by human errors. This paper researches the human reliability based on steam generator tube rupture(SGTR) accident from the perspective of human errors. It plays an important role in achieving a complete probabilistic risk analysis(PRA), qualitatively and quantitatively assessing how the human errors impact on system safety, and providing a human factors engineering design that will minimize human errors and provide recovery capability.This paper analyzes the comparison of six kinds of HRA methods which are the most widely used at present, and addresses their respective advantages and disadvantages. The operator cognitive model is established and mapped to the digital human system interface. The Rasmussen three-level behavior model is analyzed, whose relationship to the cause of human errors and the operator cognitive model is provided.The THERP-HCR combined model is established based on the comparison of HRA methods and the category of human error events. This paper addresses the assumptions and standardized process of HRA.This paper analyzes a post-accident human error event about SGTR in AP1000 and calculates the human error probability in isolating the faulted steam generator. The case shows that the THERP-HCR combined model is more feasible and reliable than other HRA methods and it can be spread to other nuclear power plants.The focus in the future is addressed finally. It is essential to develop the database handbook applicable to digital nuclear power plants and Chinese operators, and obtain the operation data related to time through stimulator experiments.