Nodal equivalence theory for hexagonal geometry, thermal reactor analysis

An important aspect of advanced nodal methods is the determination of equivalent few-group parameters for the relatively large homogenized regions used in the nodal flux solution. The theoretical foundation for light water reactor (LWR) assembly homogenization methods has been clearly established, and during the last several years, its successes have secured its position in the stable of dependable LWR analysis methods. Groupwise discontinuity factors that correct for assembly homogenization errors are routinely generated along with the group constants during lattice physics analysis. During the last several years, there has been interest in applying equivalence theory to other reactor types and other geometries. A notable effort has been the work at Argonne National Laboratory to incorporate nodal equivalence theory (NET) for hexagonal lattices into the nodal diffusion option of the DIF3D code. This work was originally intended to improve the neutronics methods used for the analysis of the Experimental Breeder Reactor II (EBR-II), and Ref. 4 discusses the success of that application. More recently, however, attempts were made to apply NET to advanced, thermal reactor designs such as the modular high-temperature gas reactor (MHTGR) and the new production heavy water reactor (NPR/HWR). The same methods that were successful for EBR-II havemore » encountered problems for these reactors. Our preliminary analysis indicates that the sharp global flux gradients in these cores requires large discontinuity factors (greater than 4 or 5) to reproduce the reference solution. This disrupts the convergence of the iterative methods used to solve for the node-wise flux moments and partial currents. Several attempts to remedy the problem have been made over the last few years, including bounding the discontinuity factors and providing improved initial guesses for the flux solution, but nothing has been satisfactory.« less