Seismic Fragility Analysis for Prestressed Concrete Containments for Nuclear Power Plants Based on Stripe-Based Approach

Abstract

The Separation of Variables (SOV) approach has been widely used for the seismic fragility analysis of nuclear power plant structures. This method begins with linear elastic analysis, and inelastic behavior (e.g., changes in stiffness and damping ratio) is indirectly considered employing the inelastic energy absorption factor. However, recent studies suggest enhancing the accuracy of fragility curves through nonlinear analysis. This study aimed to evaluate the efficacy of a nonlinear analysis-driven approach for assessing the seismic fragility of concrete nuclear containments, and propose suitable damage variables and failure criteria. For the stripe-based nonlinear approach, we addressed the uncertainty of seismic demand by utilizing 30 sets of ground motion time histories, and the uncertainty of seismic capacity by using 30 sets of three material properties, determined by Latin hypercube sampling. Pushover and time-history analysis results suggest that the failure criteria based on lateral drift are more appropriate for consistently capturing failure compared to those based on base shear or strain levels. Additionally, the median capacities obtained from the stripe-based approach with the force and drift variables exceed those from the SOV approach by 11.3% and 1.52%, respectively. These findings indicate that the stripe-based approach can lead to a lower failure probability estimate compared to the SOV approach, by more comprehensively considering nonlinear behavior. However, caution is still required to avoid overestimation of capacity.

Keywords: concrete nuclear containment, seismic fragility analysis, separation of variables, stripe-based approach, damage variable