Abstract
•Impact of heterogeneous modeling on burnt fuel vector is evaluated.•The effects of fully heterogeneous model based burnt fuel composition on the core integral parameters is analyzed.•Using the respective assembly type burnt vector, instead of the generic burnt vector, shows a significant difference in assembly power fractions.
Computational benchmarks are an important source for verifying the codes and modeling methodology in nuclear reactor analysis. The current study aims at utilizing the newly developed burnup capability of open source code OpenMC to perform analyses of the IAEA 10-MW MTR benchmark reactor. The whole core model developed for the benchmark is verified against the Serpent results from an earlier study. The effect of the volume-weighted homogenized (VWH) unit-cell model against a fully heterogeneous fuel assembly model to obtain burnup dependent number densities is determined. It was observed that the comparison of four factors (resonance escape, fast fission, thermal utilization, and reproduction factor) for two modeling options shows relative differences which increase with burnup. These differences translate to about a 350 pcm difference in the HEU case while the LEU core shows a relatively larger difference of about 500 pcm in integral core parameters. Root Mean Square (RMS) and maximum relative error in assembly power as large as 0.83 and 2 % respectively are seen for HEU and LEU cases. Although standard fuel element (SFE) and control fuel element (CFE) are geometrically significantly different, some studies have used VWH unit-cell based on SFE to generate and use number densities for CFE as well. The effect of this modeling choice is also analyzed. This led to a relative difference of more than 20 % in the concentration of Pu-239 for both HEU and LEU studies, with SFE based number density being overpredicted. Xe-135 concentration also shows about a 20 % difference for the HEU case. The difference in the whole core neutron multiplication factor, SFE vs SFE + CFE individual burnt fuel vector, is negligible. On the other hand, power RMS error and maximum relative difference for power fraction exhibit a maximum value of 0.84 and 2.24 respectively. This study can serve as a reference for future verification efforts using the IAEA 10-MW MTR benchmark reactor.