Answers are grounded in this paper and cite their sources.
BackgroundNuclear data has been identified as the primary source of uncertainty in nuclear reactor physics calculations, which directly impacts the predictive accuracy of reactor parameters. Quantitative research on the propagation of these uncertainties is essential to enhance calculation precision, thereby ensuring the economic efficiency and safety of nuclear reactor systems.PurposeThis study aims to evaluate the impact of nuclear data uncertainties on the effective multiplication factor (keff) using a sampling-based method and to validate its applicability in complex full-core models.MethodsFirstly, the nuclear data sampling tool SANDY was employed to perform multiple stochastic perturbations on nuclide cross-sections within the covariance ranges specified in the ENDF-6 format files, and a high-fidelity physics model was constructed based on the BEAVRS commercial pressurized water reactor benchmark at hot zero-power (HZP) conditions. Then, the Monte Carlo transport code OpenMC was utilized to calculate the keff for each perturbed nuclear database on a high-performance computing cloud platform; to improve computational efficiency, the perturbation module was isolated for parallel execution, and Message passing interface (MPI) technology was implemented to accelerate the OpenMC transport simulations. Finally, the resulting keff distribution was fitted and statistically analyzed to quantify the sensitivity and uncertainty propagation.ResultsResults of the sensitivity analysis indicate that the 235U fission cross-section is the most sensitive parameter under the HZP condition of the BEAVRS benchmark, with a sensitivity coefficient of 0.45. The uncertainty analysis reveals that the 238U inelastic scattering cross-section has the most significant impact on keff, resulting in a standard deviation (σ) of up to 9.1×10-4 whilst the 235U fission and 238U capture cross-sections are highly sensitive, their contribution to the overall keff uncertainty remains moderate due to their relatively small covariance values in the library.ConclusionsThe uncertainty analysis workflow based on SANDY and OpenMC proposed in this study is feasible and highly applicable for complex, large-scale reactor geometries, as demonstrated by the BEAVRS full-core benchmark under HZP conditions. The generated perturbation files support both ACE and HDF5 formats, ensuring broad compatibility across different Monte Carlo simulation platforms including OpenMC and MCNP.
No takes yet. Share an insight, caveat, or question.
Cai et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: