Applications of probabilistic safety assessment in the Nuclear Power Plant Philippsburg (KKP)

This paper presents the experiences gained from the probabilistic safety assessment (PSA) in the Nuclear Power Plant KKP. Although the primary goal for carrying out the PSA in the framework of periodic safety review is understood to be a complementary tool to the deterministic analysis, there were several improvements in the plant resulting from the PSA. The main technical objective of the paper is to show that the PSA not only provides insights in the balance of design, identifies important contribution to the core melt frequency, it also identifies the weak points in the plant. As a result of this, several hardware modifications as well as modifications in the existing administrative procedures, such as instruction manuals and other safety related procedures were carried out. In some cases new procedures had to be written. Furthermore, the PSA in KKP is being used to evaluate the effectiveness of the planned plant modifications and for getting permission from the licensing authorities to carry our preventive maintenance during power operation. There was also an enormous amount of experience gained from the shutdown PSA. There were several weak points identified, mostly with respect to the administrative procedures. The operation manuals have to be rewritten and lot of work needs to be done with respect to the availability of safety related systems during the shutdown period. It is however important that more attention should be given to crucial issues rather than to matters which are not relevant. The experience shows that in many cases unnecessary attention is given to issues which have very low contribution to the safety of the plant such as postulates of pipe breaks in systems which are not subjected to dynamic loads and where leakages are more probable than breaks. Such postulates, which are not realistic and highly improbable, lead to undesirable contribution to the core melt frequency and also to changes in the written procedures which are contra-productive from the viewpoint of shift personnel. It should also be mentioned that in the PSA, models for Common cause failures be used which are realistic. Another issue is the Human Reliability Analysis (HRA). There has been an attempt to make research-projects out of HRA despite the fact that HRA contains a lot of uncertainties anyway. One should be careful that the PSA reflects the reality in the plant. PSA should be rightly used for the purpose it is meant for, such as optimization of the plant processes and plant design.