CALCULATING ALPHA-FACTORS WITH THE PROCESS-ORIENTED SIMULATION MODEL

摘要

Introductory information is provided regarding a recent update on regulatory requirements for PSA for nuclear power plants in Germany. As a complement to established approaches to quantification of common cause failures (CCF) a process-oriented simulation model (POS-model) for CCF has been developed. It is based on the following sequence of stochastic variables which is supposed to describe the CCF process adequately: time of CCF impact, number of components of the component group affected by the impact, times of failure of the impacted components, and time of detection of the CCF process by inspection or functional testing. Based on simulation of this sequence, the associated unavailabilities can be calculated. Recently, the procedure of parameter estimation for the POS-model has been optimized and assessed as an essential step to rendering the model applicable to practical problems. The estimation procedure has been tested in different applications including simulated CCF data and benchmark exercises carried out earlier. The results obtained were encouraging. As a further step a field of application for which a greater amount of published CCF data is available has been identified. In NUREG/CR-5497 common cause failure parameter estimations have been provided for some 40 different component types, various failure modes and common cause component group sizes from two up to six. One of the models for which parameter distributions have been derived is the Alpha Factor Model. From the point of view of demonstrating the usefulness of the POS model, this large amount of systematically derived information was seen as a challenge. An approach has been selected to derive alpha factors for component group sizes greater than four from the data for lower degree of redundancy using the POS model. This program has been carried out for six different combinations of components and failure modes that were selected primarily based on large numbers of dependent failures to make sure that the comparison has a solid statistical basis and on having a good mix of technically different components. Details of the approach including a simplified parameter estimation and the results obtained are described in the paper. The conclusion is that the POS model has passed this exercise based on a considerable amount of failure data in a satisfactory manner.