MODELING STUDIES AND CODE-TO-CODE COMPARISONS FOR PRESSURE TUBE HEAVY WATER REACTOR CORES

Preliminary, conceptual studies have been performed previously using deterministic lattice physics (WIMS-AECL) and core physics codes (RFSP) to estimate performance and safety characteristics of various thorium-based fuels and uranium-based fuels augmented by small amounts of thorium for use in pressure tube heavy-water reactors (PT-HWRs). To confirm the validity of the results, the WIMS-AECL/RFSP results are compared against predictions made with the stochastic neutron transport code MCNP. This paper describes the development of a method for setting up an MCNP core model of at PT-HWR for comparison with WIMS-AECL/ RFSP results, using a core with 37-element natural uranium fuel bundles as a test case for sensitivity studies. These studies included evaluating the sensitivity of the bias of the effective neutron multiplication factor (k(eff) ) , a source convergence study, uncertainties correction with multiple independent simulations, the impact of irradiation map binning methods, and the impact of reflector models. A Python-based software scripting tool was developed to automate the creation, execution, and post-processing of reactor physics data from the MCNP models. The software tool and algorithm for creating an MCNP core model using data from the WIMS-AECL and RFSP models are described in this paper. Based on the preliminary evaluations of the simulation parameters with the base model, reactor physics analyses were performed for PT-HWR cores with thorium-based fuels in a 35-element bundle type. Code-to-code results demonstrate good agreement between MCNP and RFSP, giving confidence in the method developed and its applicability to other fuels and core types.