Abstract
CASMO-4 is a multigroup two-dimensional transport code for LWR lattice physics calculations. MCNPX and TRITON/T6-Depl are two general-purpose transport codes with depletion capability for various fuel designs. MCNPX can use continuous-energy cross sections while TRITON currently only supports multigroup depletion calculations. This study presented a systematic comparison of these three codes for depletion calculations of a typical BWR fuel assembly. Key parameters for sensitivity studies were neutron cross-section libraries, burnup steps, modeling of poison rods, inclusion of additional nuclides for depletion, thermal expansion, pin-by-pin depletion, and Dancoff factors. The CASMO-4 results were arbitrarily taken as a reference base on which the differences of MCNPX or TRITON calculations were evaluated. Useful observations from the comparisons were as follows: The ENDF/B-VII cross-section library gave the most consistent result with CASMO-4. At least five radially subdivided zoning of a Gd-bearing rod was necessary for depletion calculations. MCNPX calculations were more sensitive to choices of burnup steps and numbers of nuclides being traced in fuel inventory than TRITON did. Applying the same thermal expansion corrections in TRITON reduced its differences with CASMO-4 in the middle of cycle. Pin-by-pin depletion is necessary but only slightly changed k∞ profiles in this case compared with average depletion. Using more accurate Dancoff factors in TRITON resulted in an excellent agreement of k∞ values with CASMO-4 at the early stage of burnup, but they still gradually deviated at later burnups. Overall, both MCNPX and TRITON predicted k∞ profiles in this problem were within 500 pcm agreement with CASMO-4 in the entire burnup period.