Validation of Nuclear Data Libraries for Reactor Safety and Design Calculations

Abstract

Benchmark calculations are performed for a large number of critical experiments such as homogeneous uranium and plutonium nitrate solutions, UOX and MOX fuel lattices, and MTR fuel. For these calculations, various modern nuclear point data libraries were used, namely JEF-2.2, ENDFIB-VI, and JENDL-3.2, and the not yet released JEFF-3 data; the differences between the corresponding results are discussed. The influence of the unresolved resonance treatment possible within the latest MCNP version is investigated. For selected systems, deterministic and Monte Carlo calculations were performed with a new multigroup library with 292 energy groups, based on JEF-2.2, and the results are compared with MCNP calculations with point data.

First Conclusions of the WPEC/Subgroup-22 Nuclear Data for Improved LEU-LWR Reactivity Predictions

“…Before the creation of the WPEC working group, several Continuous Energy Monte-Carlo (CEMC) studies had raised the problem [1], [2], [3], [4]. The reactivity bias was observed for low enriched thermal benchmarks after extensive integral tests of the ORNL U235 evaluation of the resolved resonances by Leal, Derrien, Wright and Larson [5] adopted in ENDF/B-VI.8, JEFF3.0 and JENDL3.3.…”

Neutron scattering kernel for solid deuterium

“…Within such framework, a complete description of the rotational motion in a gas of H 2 (and D 2 ) was produced by Young and Koppel [3], where the neutron cross-sections for the ortho and para species were explicitly given. Since then, a large body of models and calculational techniques were developed to describe the interaction of slow neutrons with liquid H 2 and D 2 , differing in their treatment of the translational motion and coherence effects [4][5][6][7][8][9][10]. The only prescription to calculate the cross-section of solid hydrogen and deuterium was introduced by Bernnat et al [11], but spin correlation effects were not fully accounted for in their work.…”

Monte Carlo calculations with nuclear point data based on JEF-2.2 and JEFF-3.1 for the VENUS-7 critical benchmarks

“…Over the past few years, we have undertaken extensive benchmark studies for criticality calculations on a broad spectrum of fissile systems with the Monte Carlo code MCNP [1] using different nuclear data libraries [2,3]. With the JEF-2.2 based data we use routinely for water moderated systems, it turned out that the multiplication factors of LEU LWR type lattices tend to be slightly underestimated.…”