Inspection report Ringhals 3 and 4 concerning fire analysis

The Inspectorate has evaluated the fire analysis reported by the Ringhals-3 and -4 PWR-plants. On the whole, SKI finds the report well structured. Methods, conditions and data are well described and follow the established principles for the area. However, SKI have made some observations that are relevant for future analyses. The most important are: the analysis does not consider different possibilities for failure modes in the signal system at a fire event (which is noted in the documentation). The possibility that a system unintentionally starts/stops is a very important parameter, that must be analyzed. Conclusions and interpretations of the analysis need to be improved. It is not clear how the result relates to other PSA-studies, which leaves an overall picture missing. Ideas for improvement of future analyses are lacking, as is an overall sensitivity analysis. Spreading of fire is not thoroughly analyzed. The study does not cover experiences (or lack of experiences) at the plant. The documentation should be improved regarding the description of different chains of events. In the present report, it is not possible to follow an initiating event to the moments when different functions are activated. This is partly due to the conservative assumption that the reactor always is shutdown in case of fire, which need not be the case