(A) new two-step method with improved nodal equivalence for PWR analysis and design

Abstract

Despite a lot of progress in high-fidelity transport-based analysis of nuclear reactors, the conventional two-step method is expected to continue to play a key role in commercial pressurized water reactor (PWR) design due to its merits of having acceptable accuracy and cost. However, it is well known that the popularly-used neighbor- and spectrum-irrelevant leakage correction in the conventional method causes some errors in prediction of the core reactivity and power distribution. First, the intrinsic limitations of the conventional two-step methods are discussed by analyzing a two-dimensional small modular reactor (SMR) core with the DeCART2D/MASTER code system. This thesis suggests a new two-step method that could improve solution accuracy by overcoming the inherent drawbacks of the conventional 2-step method. The new method adopts the albedo-corrected parameterized equivalence constants (APEC) method to correct the lattice-based raw cross-sections (XSs) and discontinuity factors (DFs) by accounting for the actual neutron leakage. For a full-scope development and implementation of the APEC correction, the MASTER nodal code was modified so that the group constants can be corrected in the middle of a microscopic core depletion. The color-set models are defined to determine the APEC functions for burnup-dependent XS and DF corrections. Then the new two-step method was applied to depletion analyses of the SMR without thermal feedback, and its validity was evaluated in view of the prediction capability of the reactor eigenvalue and nodal power profile. In addition, four variants of the original SMR core and a large PWR core were analyzed for a further evaluation of the APEC-assisted depletion. In this study, several combinations of the burnup-dependent and -independent XS and DF corrections were also considered. The numerical results show that this advanced method could enhance the nodal equivalence significantly with inexpensive additional costs. Therefore, it is expected that affordable sufficient-fidelity diffusion-based analysis may be possible for PWRs.