The Limit Exceedance Factor Importance Measure: An Application to the Prism Reactor Design

Abstract

Probabilistic Risk Assessment has several different standard importance measures for structures, systems, and components (SSCs). These importance measures are Risk Achievement Worth (RAW), Risk Reduction Worth (RRW), and Fussell-Vesely (FV). In particular, RAW and FV have been used to identify SSCs of low risk significance. However, when designing a conceptual reactor with a safety goal in mind such as those proposed in various technology neutral frameworks, these importance measures are quite crude. RAW is defined such that a component is always in the failed state, and a designer may not be looking to remove a system but to simplify it to improve economics. The Limit Exceedance Factor has been developed as a more informative importance measure when there is a goal in mind. It is defined as the factor by which the failure probability of a component may be multiplied by such that the end state (e.g. core damage) frequency exceeds a limit. With a living PRA available throughout the design process, it could allow a designer to know how much room there is for possible simplification in redundant systems. Alternatively, in the case where a system does not meet the frequency limit it can show which systems might be ideal targets for improvement to reach the limit. For a sodium-cooled fast reactor design, using the goals and limits outlined in NUREG-1860 (Technology Neutral Framework), one end state that must stay below a threshold is a very large release due to an energetic scenario. Several SSCs can be identified as risk insignificant with RAW values below two. The other SSCs tend to have very high RAW values on the order of one thousand. When the LEF is applied to the SSCs that seem to be very important using traditional importance measures, we see that some may actually have a fairly wide margin from their cited failure probabilities and still maintain the frequency goal.