Analysis of sensitivity and uncertainty quantification for transient simulation with the emphasis on changes in the transient structure

The sensitivity and uncertainty quantification (UQ) methods were studied in the context of transient simulation results. The influence of simulation parameters on changes in transient structure and occurrence of important phenomena was the focus of the study. Various statistical and data analysis tools were used to investigate the changes in the transient behavior. The NUPEC BWR Full-size Fine-mesh Bundle Test (BFBT) void fraction transients were analyzed. The sensitivity and UQ were studied for the model created using TRAC/RELAP Advanced Computational Engine (TRACE). The pump and turbine trips were simulated using TRACE system code and the code prediction was validated based on void fraction measurement at three axial elevations. The analysis contains a sensitivity study for various simulation parameters. Simulation parameters include boundary conditions, geometry of the facility and physical model parameters. The sensitivity of important simulation parameters was studied using: Pearson’s product momentum correlation coefficient, Sobol’s variance decomposition, Morris screening method, Delta Moment-Independent Analysis, Fourier Amplitude Sensitivity Test (FAST) and Derivative-based Global Sensitivity Measure (DGSM). The methodology of transient sensitivity analysis is presented. The sensitivity and uncertainty study was performed using the TAPE code (Transient Analysis PackagE), a new uncertainty quantification and sensitivity tool developed at the University of Illinois.