Analysis of nodalization uncertainty for nuclear system analysis code with Lax-Wendroff numerical scheme

Abstract

In modeling the complex systems such as a nuclear power plant (NPP), the nodalization becomes more important to satisfy Courant number limitation and reduce the numerical diffusion due to the 1st order numerical scheme. Since the higher-order numerical scheme can reduce the numerical diffusion problem, the accuracy can be improved and the nodalization uncertainty can be reduced by improving the current numerical scheme. However, another sources for the user effect uncertainty originates from a nodalization process and this can be affected by the selected numerical scheme. Sufficient sensitivity analyses should be performed on the nodalization to ensure that the calculated results are free from these uncertainties. Therefore, in this study, the nodalization uncertainties of 1st order upwind scheme and Lax-Wendroff scheme are evaluated with the revised MARS-KS code. The analysis of the nodalization uncertainty is performed for the separate effect tests (SETs) such as subcooled boiling and reflood experiments. The nodalization uncertainty is analyzed with the error between the experimental data and the calculation results of 1st order upwind scheme or Lax-Wendroff scheme. Thus, when using Lax-Wendroff scheme, the more accurate results with decreased nodalization uncertainty can be obtained. For the reflood calculations, accurate and consistent prediction of maximum peak cladding temperature (PCT) and accurate prediction of quenching time were observed with Lax-Wendroff scheme. However, the quenching time may not be consistent depending on the user. In terms of the computational efficiency, the central processing unit (CPU) time is generally increased by 40% with Lax-Wendroff scheme.