Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
1728960 | Annals of Nuclear Energy | 2012 | 9 Pages |
In this paper we investigate the influence of various up-to-date nuclear data libraries, such as ENDF/B-VI.6, ENDF/B-VII.0 and JEFF 3.1, on the multiplication factor of the TRIGA benchmark with fuel made of enriched uranium and zirconium hydride and SB light-water reactor benchmarks with fuel made of fissile material in zirconium matrix. The calculations are performed with the Monte Carlo computer code MCNP. Differences of ∼600 pcm in keff are observed for the benchmark model of the TRIGA reactor, while there are practically no differences in the kinf of the fuel. Therefore, an investigation is performed also for hypothetical homogeneous and heterogeneous systems with different leakage. The uncertainty analysis shows that the most important contributors to the difference in keff are the Zr isotopes (especially 90Zr and 91Zr) and thermal scattering data for H and Zr in ZrH. As the differences in keff due to the use of different cross section libraries are relatively large, there is certainly a need for a review of the evaluated cross section data of the zirconium isotopes.
► Calculations with ENDF/B-VII.0 nuclear data overpredict keff of Zr benchmarks. ► TRIGA criticality benchmark sensitive to Zr data. ► Zr scattering cross section responsible for differences in keff. ► Need for new experimental data on Zr cross sections.