Uncertainty analysis of the OECD/NEA LWR UAM benchmark Phase I using CASMO5 with JENDL-4.0 and ENDF/B-VII.1

Abstract

ABSTRACT An uncertainty analysis of the OECD/NEA LWR UAM benchmark Phase I was performed by CASMO5 using JENDL-4.0 and ENDF/B-VII.1, and the individual influence of nuclide/reaction covariance data was estimated to preliminary discuss the important covariance data on the uncertainties of lattice physics parameters used in core calculations. In the boiling water reactor and pressurized water reactor UO2 fuel assembly problems, the random sampling was used to estimate the uncertainties of lattice physics parameters. The difference in the covariance data between JENDL-4.0 and ENDF/B-VII.1 was discussed by comparing the influence of nuclide/reaction covariance data. In terms of the k-infinity uncertainty, it was confirmed that the large discrepancy between the two nuclear data libraries was caused by the covariance data of 235U [-] (i.e. the magnitude of the 235U nubar (an average number of neutrons per fission) uncertainty), which was one of the dominant covariance data. In addition, even though the covariance data of 238U [elastic-(n,n’)], 90Zr [elastic-elastic], and 157Gd [(n,γ)-(n,γ)] had the large influence on the assembly-homogenized 2-group macroscopic cross section and pin power distribution, those data were lack in JENDL-4.0. Thus, it is important to take into account the discrepancy and lack of such covariance data in future uncertainty analyses.