Improved Neutronics Treatment of Burnable Poisons for the Prismatic HTR

In prismatic block High Temperature Reactors (HTR), highly absorbing material such a burnable poison (BP) cause local flux depressions and large gradients in the flux across the blocks which can be a challenge to capture accurately with traditional homogenization methods. The purpose of this paper is to quantify the error associated with spatial homogenization, spectral condensation and discretization and to highlight what is needed for improved neutronics treatments of burnable poisons for the prismatic HTR. A new triangular based mesh is designed to separate the BP regions from the fuel assembly. A set of packages including Serpent (Monte Carlo), Xuthos (1 st order Sn), Pronghorn (diffusion), INSTANT (Pn) and RattleSnake (2 nd order Sn) is used for this study. The results from the deterministic calculations show that the cross sections generated directly in Serpent are not sufficient to accurately reproduce the reference Monte Carlo solution in all cases. The BP treatment produces good results, but this is mainly due to error cancellation. However, the Super Cell (SC) approach yields cross sections that are consistent with cross sections prepared on an "exact" full core calculation. In addition, very good agreement exists between the various deterministic transport and diffusion codes in both eigenvalue and power distributions. Future research will focus on improving the cross sections and quantifying the error cancellation. I. INTRODUCTION It is essential for the analysis of prismatic HTRs to accurately determine the effects of Burnable Poison (BP) pins. Local effects from BP pins are less pronounced in an HTR core than in a Light Water Reactors (LWR). However, with burnable poison pins located in the corner of a block surrounded by reflector blocks on both adjacent sides, the flux depression can be quite significant. Additionally, due to the long neutron migration lengths in HTRs, the area of influence of BPs is spread over neighboring blocks. Error in calculation of the absorption reaction rates of these regions can have a significant impact on the local poison depletion and the isotopic distribution in the surrounding fuel as a function of time (1) . Consequently, these accumulative errors would affect power distribution and shutdown margin predictions through the cycle. In a similar fashion to the method used by CEA in the analysis of the High Temperature Test Reactor (HTTR) (2) , the Idaho National Laboratory (INL) has developed a triangular based mesh to isolate the BP regions with a new meshing routine available in Pronghorn (3) , an application of the Multi-Physics