Seismic risk assessment for nuclear power plants: implications of a stop

Abstract

Nuclear power plants (NPPs) are designed and assessed for severe earthquake shaking, which can pose significant demands on the structural system and internal equipment. Base isolation is a viable means to substantially reduce seismic risk. In the United States, isolation systems will comprise seismic bearings designed and tested to maintain the axial load carrying capacity with a 90+% confidence at the 90th percentile displacement for beyond design basis (BDB) shaking, where BDB shaking is linked to a uniform hazard response spectrum with a return period of 100,000 years. The earthquake risk is conservatively quantified herein in terms of annual frequency of unacceptable performance of an individual isolator, which provides a metric to compare the three risk-reduction strategies, namely, 1) testing prototype isolators to achieve higher confidence than 90%, 2) testing isolators to a greater displacement and corresponding axial load, and 3) providing a stop at the 90th percentile displacement for BDB shaking. Testing the isolators to 90% confidence at the 90th percentile BDB displacement results in a risk significantly smaller than 1×10-6 if the isolation system is complemented by a stop. Increasing either the confidence level and/or the displacement for prototype testing may not reduce risk to less than 1×10-6.