Determination of intermediate-level performance criteria for a nuclear power plant based on reliability and risk models

Abstract

There is a hierarchy of levels at which performance criteria for a nuclear power plant might be set, ranging from overall top-levels (e.g., offsite health effects or radioactive releases) through intermediate-levels (e.g., safety functions/ systems performance) to low-levels (e.g., component or subcomponent specifications). The intermediate-level criteria on safety functions/systems would, if properly determined, constitute more useful and workable risk management framework than the top-level criteria, and more flexible and les prescriptive regulation than the low-level criteria. This study develops a methodology for determining the intermediate-level performance criteria that are consistent with the top-level criteria such as the safety goals. The methodology derives the intermediate-level criteria from the top-level criteria by means of the noninferior solution technique for a multiobjective programming problem where unavailabilities of the safety functions/systems are to be maximized. The methodology uses a probabilistic risk assessment (PRA) model as a basic safety model for the plant. However, it does not require reliability cost functions of the components constituting the plant. The first element of the methodology is to find two-tier noninferior solutions in the objective space and the risk space. Once the two-tier noninferior solutions are obtained, an informal preference assessment can be performed by comparing the risk values associated with the two-tier noninferior solutions with the top-level safety criteria, and obviously unattractive noninferior solutions are discarded resulting in a much narrower band in the two-tier noninferior subspace. The lower limits of the band are then proposed as the intermediate-level performance criteria. As an illustration, the methodology is applied to a nuclear power plant with an existing PRA model and several results are provided.