Proceedings of the Eighth International Conference on Probabilistic Safety Assessment & Management (PSAM)
2 Impact of CCF Assumptions on PSA and PSA Application Calculations (PSAM-0172)
Download citation file:
- Ris (Zotero)
- Reference Manager
A research project on the impact of CCF treatment in PSA and PSA application calculations, and especially during calculation of the risk achievement and risk decrease importance measures was performed in 2005 for SKI, the Swedish Nuclear Power Inspectorate. SKI project 14.2-041049/200402008.
CCFs are dominating contributors to risk. It is therefore important with both development and correct use of CCF data. It is also important to treat CCFs in a correct and realistic manner in the calculations of the core damage frequency and when quantifying importance and sensitivity for CCF-related events, both in the basic PSA and in PSA applications.
Risk increase and risk reduction measures for independent basic events (not CCF events) are calculated by setting the basic event probability to 1,0 and 0,0, respectively. This approach can also be used to calculate importance measures for CCF basic events that are created when CCF is considered. However, the interpretation of the importance measures for CCF basic events is more complex. CCF modeling usually means that CCF basic events, which really are combinations of the failures of two or more components, are introduced. In these cases, what should the probability of the different CCF basic events be during calculation of risk achievement and risk reduction measures? This paper presents results regarding risk increase and risk reduction measures for a set of example cases covering two to four trains, and with different assumptions on which CCF basic events should be set to 1.0 and 0.0. The results are discussed and are the basis for the presented recommendations, which are of importance both in calculating the basic risk importance measures, and in risk follow-up and risk monitoring applications when a component really is unavailable due to preventive or corrective maintenance, and there either is or is not knowledge of whether the unavailability is related to CCF or not. A planned activity is a no CCF case while a case without knowledge can be CCF and test of other components gives information and can be a ground for a case with complete dependence.
Sometimes, an importance measures are calculated for a group of basic events, e.g., basic events for the same component type or basic events representing components in a specific physical location. In this case, it is of interest to know what attributes are assigned to the CCF basic events. Shall the original basic event attributes be assigned also to the CCF basic events?
Finally, a special importance measure for CCF groups is proposed.