High fidelity simulation of conventional and innovative LWR with the coupled Monte-Carlo thermal-hydraulic system MCNP-SUBCHANFLOW

Abstract

In order to increase the accuracy and the degree of spatial and energy resolution of core design studies, coupled 3D neutronic (multi-group deterministic and continuous energy Monte-Carlo) and 3D thermal-hydraulic (CFD and sub-channel) codes are being developed worldwide. At KIT, both deterministic and Monte-Carlo codes were coupled with sub-channel codes and applied to predict the safety-related design parameters such as critical power ratio, maximal cladding, fuel temperature and DNB. These coupling approaches were revised and improved based on the experience gained. One particular example is replacing COBRA-TF with SUBCHANFLOW, an in-house developed sub-channel code, in the COBRA-TF/MCNP coupling, accompanied with new way of radial mapping between the neutronic and thermal-hydraulic domains. The new coupled system MCNP5/SUBCHANFLOW makes it possible to investigate a variety of fuel assembly types. Key issues in such a coupled system are the implementation of the thermal-hydraulic/neutronic feedback mechanisms, the precision of the Monte-Carlo solutions, and the supervision of convergence during the iterative solution process. Another key issue considered is the optimal application of parallel computing. Using multi-processor computer architectures, it is possible to reduce the Monte-Carlo running time and obtain converged results within reasonable time limit. In particular, it is shown that by exploiting the capabilities of multi-processor calculation, large fuel assemblies on a pin-by-pin basis with a resolution at sub-channel level can be analyzed. One of the most important issues addressed in the current work is the temperature effects on nuclear data. For the particular studies pseudo material mixing approach was used to account for the temperature dependence of the nuclear data.