Uncertainty and sensitivity analysis for hot-leg LOCA in two-loop PWR

Overcooling and/or pressurization events have the potential to result in pressurized thermal shock (PTS) in reactor pressure vessels of pressurized water reactors (PWRs). PTS can occur during various overcooling scenarios, including loss-of-coolant accidents (LOCAs). Thermal-hydraulic calculations of overcooling scenarios provide an input to structural analysis. The purpose of this study is to perform advanced thermal-hydraulic calculations of a hot-leg LOCA with uncertainty and sensitivity analyses. A novel approach, utilizing a fast Fourier transform–based method (FFTBM) with signal mirroring (SM), is proposed for a sensitivity study to evaluate the influence of input uncertain parameters with low computational cost before performing an uncertainty analysis. For the analysis, a break size of 45.6 cm2 (equivalent to a 76.2-mm- or 3-in.-equivalent diameter) was selected in a two-loop PWR. A verified and validated RELAP5 input deck was used. In sensitivity study 30, one-at-a-time sensitivity calculations plus reference calculation were performed. These were followed by uncertainty and sensitivity analyses using the GRS (Gesellschaft für Anlagen- und Reaktorsicherheit) Software for Uncertainty and Sensitivity Analyses (SUSA) version 4.2.6 tool, varying 15 uncertain input parameters across 208 samples. The input uncertain parameters were derived from the European Union–funded Advanced PTS Analysis for LTO (APAL) project. Three figures of merit (FOMs), reactor pressure, liquid temperature, and reactor vessel wall temperature below the cold leg connection, were analyzed. The results demonstrated that FFTBM-SM effectively identified the four most influential parameters, aligning closely with the results from the global sensitivity analysis from SUSA, which utilized 208 samples from the uncertainty analysis. Although slight differences in parameter ranking were observed, the findings validate FFTBM-SM as a valuable prescreening tool for sensitivity studies. The limitation of the sensitivity study is that it adopts a local approach. Finally, the results of the uncertainty analysis provide tolerance regions for the selected FOMs that are comparable to those obtained in the original APAL study for a four-loop PWR.