Abstract
The fuel element of the High Temperature Engineering Test Reactor (HTTR) presents a doubly heterogeneous geometry, where tiny TRISO fuel particles dispersed in a graphite matrix form the fuel region of a cylindrical fuel rod, and a number of fuel rods together with moderator or reflector then constitute the lattice design of the core. In this study, a series of full-core HTTR criticality calculations were performed with the SCALE6 code system using various geometric and unit-cell options in order to systematically investigate their effects on neutronic analysis. Two geometric descriptions (ARRAY or HOLE) in SCALE6 can be used to construct a complicated and repeated model. The result shows that eliminating the use of HOLE in the HTTR geometric model can save the computation time by a factor of 4. Four unit-cell treatments for resonance self-shielding corrections in SCALE6 were tested to create problem-specific multigroup cross sections for the HTTR core model. Based on the same ENDF/B-VII cross-section library, their results were evaluated by comparing with continuous-energy calculations. The comparison indicates that the INFHOMMEDIUM result overestimates the system multiplication factor (k eff ) by 55 mk, whereas the LATTICECELL and MULTIREGION treatments predict the k eff values with similar biases of approximately 10 mk overestimation. The DOUBLEHET result shows a more satisfactory agreement, about 4.2 mk underestimation in the k eff value. In addition, using cell-weighted cross sections instead of an explicit modeling of TRISO particles in fuel region can further reduce the computation time by a factor of 5 without sacrificing accuracy.