engleski

Some Insights Gained from On-Line Risk Monitoring of the Complete Plant Operating Cycle

This paper is presenting numerical comparison between Probabilistic Risk As-sessment (PRA) results for the so-called average model and for the realistic model. From the one side average PRA model is regularly used model, which is modeling average behavior of the all-modeling parameters during one-year period. On the other side realistic PRA model is considering various plant configurations and other im-portant conditions that are changing during one-year period. Results are obtained for the specific modeling segment related to the plant safety power supply. Diesel gen-erators are regularly maintained during the year, and probability for initiating event loss of offsite power (LOOP) is changing depending on the power grid state and weather conditions. Influences of these two facts are quantified and result presented. Two different PRA models with two different levels of details were used. Both results are showing that core damage probability (CDP) varies in the range of 10% because some activities are rescheduled (i.e., DG is maintained or not during the highest LOOP likelihood). Results are also clearly indicating that extreme instantaneous CD fre-quencies are avoided with better activities schedule. This also results in CDP reduction. Potential for risk reduction is significant when results from on-line risk monitoring of the complete plant operating cycle is considered.

probabilistic safety assessment; risk; nuclear power plant; plant configuration

nije evidentirano