Comparison of MCNP and WIMS-AECL / RFSP Calculations With High Temperature Substitution Experiments in ZED-2 Using CANFLEX-LVRF

Abstract

This paper summarizes comparisons of calculation results from MCNP5 and WIMS-AECL / RFSP with experimental results obtained from the Zero Energy Deuterium (ZED-2) critical facility, examining CANFLEX Low Void Reactivity Fuel (CANFLEX-LVRF) in heated channels, substituted into a reference lattice and cooled under ACR like coolant conditions, with H₂O, air, or CO₂ as an air substitute. CANFLEX LVRF shares features in common with the ACR 1000 fuel, notably an increase in enrichment (over natural uranium) in the outer elements of the fuel bundle, and presence of a neutron absorber in the central element. The reference and substituted fuel channels were arranged in a 24.5 cm hexagonal lattice in order to provide neutron similarity to the 24 cm square lattice pitch of the ACR 1000. These results therefore provide useful data for validation of the reactor physics toolset for use in ACR 1000 applications. For the mixed lattices, results for both MCNP5 and WIMS-AECL / RFSP show small biases in keff, ranging from -7 mk to -5 mk, small biases in coolant void reactivity, ranging from -1 mk to +0.5 mk, and good agreement for copper activation rate distributions (based on calculated neutron flux). Bare core MCNP and WIMS-AECL stand-alone results, based on substitution analysis, also show small biases in keff, ranging from 6 mk to 0.4 mk, and small biases in coolant void reactivity, ranging from -0.3 mk to +3.7 mk. This validation exercise thus gives good agreement between measurement and calculation and provides confidence in the accuracy of the physics toolset.