A Methodology for Developing Experimental Benchmark Sets for New Reactor Designs using Sensitivity Analysis

Abstract

A method is examined for generating sets of benchmark experiments that can be used to determine computational biases in new reactor designs. Potential experiments for the ZED-2 test reactor are simulated. The effects on the applicability of a benchmark set resulting from modification of the ZED-2 lattice are studied. The applicability of a potential benchmark set is determined based on a sensitivity analysis using TSUNAMI-3D. Potential experiment sets are developed for a proposed supercritical water reactor design and for thorium fueled CANDU reactor designs. An analogous methodology can be applied to new small modular reactor designs.