BAW critical experiment benchmark analysis using PRAGMA GPU-based Monte Carlo code

This study validated the GPU-based continuous-energy Monte Carlo (MC) code PRAGMA under light water reactor (LWR) conditions using Babcock & Wilcox (BAW) critical experiment benchmarks. The benchmark calculation results with ENDF/B-VII.0, VII.1, and VIII.0 were compared with the experimental values for both criticality (53 cases) and the radial fission rate distributions (RFDs, 19 cases). The differences among the ENDF/B library versions were marginal and the results were consistent with other MC codes, while showing reduced deviations (C–E) compared with the traditional two-step code reported in the literature. A negative trend in k eff deviation (C–E) was observed as the number of Pyrex rods increased, suggesting a possible systematic bias induced by Pyrex. Furthermore, the uncertainties of criticality and the lattice peaking factor in PRAGMA were quantified using the 95/95 tolerance limit method. These results provide a quantitative basis for validation and accuracy assessment of PRAGMA in LWR analyses.