BEAVRS BENCHMARK EVALUATION WITH CASMO5 AND SIMULATE5

Abstract

The MIT BEAVRS benchmark problem, which was primarily setup for the verification and validation of high-fidelity tools that have coupled neutron transport, thermal-hydraulics, and fuel isotopic depletion models, has also found extensive usage in the reactor physics community for validating core analysis tools. The primary purpose of this paper is to provide an accurate, comprehensive evaluation of the BEAVRS benchmark with CASMO5 and SIMULATE5 codes. The CMS5 calculated results for low-power physics tests (hot zero power critical boron, control rod worth and isothermal temperature coefficients) and full power operation (boron let-down and flux map reaction rate distributions) are compared to plant measured data provided in the benchmark specification. The CMS5 model, using ENDF/BVII.1 nuclear data library, predicts HZP critical boron concentration for all-rods-out conditions within 10 ppm for Cycle-1, and 25 ppm in Cycle-2; the control rod worth is predicted with a difference of 0.7% ± 3.8%, where the maximum difference is less than 10%. For the core follow calculations at the hot full power condition, the average difference in predicting the critical boron concentration is less than 20 ppm. In addition, the radial and nodal reaction rate distributions are predicted with a mean difference of about 1.6% and 3.8%, respectively. The CMS5 calculations are repeated using the most recent ENDF/B-VIII.0 library. No significant difference is observed in predicting measured plant parameters with different nuclear data libraries. Additionally, the impact of various modeling options, which are typically employed with nodal diffusion codes, on the predictions of important core parameters are presented as part of the benchmark evaluation.