Development and verification of covariance data generation module in NECP-Atlas

Abstract

BackgroundThe quantification of uncertainty has become a common requirement in reactor physics analysis whilst the covariance data of nuclear data serves as the foundational data for conducting uncertainty quantification.PurposeThis study aims to develop a covariance data generation module, named covar_calc, embedded in the nuclear data processing software NECP-Atlas, a nuclear data processing program independently developed by the laboratory of nuclear engineering computational physics (NECP) of Xi'an Jiaotong University, to produce continuous energy covariance data for Monte Carlo programs and multi-group covariance data for deterministic programs.MethodsWithin the framework of NECP-Atlas, covar_calc module was developed to process all covariance data provided in evaluated nuclear data, according to the different storage formats of nuclear data and different computational methods. Covariance data of various parameters such as average fission neutron multiplicities, cross sections, angular distributions of secondary particles, fission spectrum, resonance parameters, and neutron activation cross-sections, were all be processed by covar_calc. Comparative verification was carried out with the covariance data production module in the nuclear data processing software NJOY21. Finally, sensitivity coefficients for different benchmarks were calculated using both the Monte Carlo calculation code NECP-MCX and uncertainty analysis code NECP-UNICORN, and the final uncertainties were computed by incorporating both continuous energy covariance data and multi-group covariance data, and utilizing the "Sandwich formula".ResultsComparison results demonstrate that the accuracy of the multi-group covariance data produced by NECP-Atlas is equivalent to that of NJOY21 and the maximum bias is less than 0.1%. The uncertainties calculated using the multi-group covariance data generated by NECP-Atlas exhibit comparable accuracy to those obtained with NJOY21.ConclusionsThe precision in creating both continuous energy covariance and multi-group covariance presented in this study meets the requirements for usage in Monte Carlo programs and deterministic programs, validating the efficacy of covar_calc module within NECP-Atlas for uncertainty quantification in reactor physics analysis.

Keywords

• covariance data
• uncertainty quantization
• evaluated nuclear data
• nuclear data process
• covariance calculation