Solving the C5G7 3-D extension benchmark with the SN code TORT

The C5G7 3-D Extension MOX Fuel Assembly Benchmark is designed to test the capabilities of modern deterministic transport methods to calculate whole core flux distributions sufficiently well without relying upon spatial homogenization techniques. The configuration considered is a small three-dimensional core with 90 degrees symmetry containing two UO2 and two MOX fuel assemblies in one quarter of the core. For all material compositions, a seven-group set of isotropic cross sections is provided to be used for the transport calculation, including the Monte Carlo reference solution. In the present paper, we report on our results for this benchmark using the 3-D SN code TORT. Our calculation models for the three control rod configurations are described in detail, in particular the meshing of the circular pin cells within the rectangular geometry. The calculations are repeated for different mesh sizes and quadrature orders in order to obtain estimates for the sensitivity of the results on these modelling features. It is found that with respect to keff, our TORT solutions agree to within 0.3 % with the corresponding Monte-Carlo reference values. Concerning the pin-wise fission rate distributions, the maximum errors do not exceed 1.2 %. With increasing quadrature order and meshes per pin cell, the maximum pin-wise fission rate errors decrease, with the pin cell meshing exhibiting stronger influence than the SN order. Investigations of the axial fluxes and fission rates reveal the influence of the control rods, causing flux depression depending on the insertion depth. (authors)