Abstract:

Fast running codes and highly accurate predictions of local reactor safety parameters are needed by manufacturers, utilities and regulators in order to optimize the core designs and to assess the safety features. Nowadays efforts are underway to move from nodal diffusion-based core simulations to more detailed higher-order transport solvers in order for pin-by-pin calculations to become feasible. However, it will take time for such new developments to become standards in industrial applications due to challenging problems such as acceleration techniques, efficient algorithms for parallelization and memory requirements. In parallel, Monte Carlo methods for core simulations have experienced a tremendous increase of usage in the nuclear community due to their general flexibility in treating any kind of geometry without any angular, energy and spatial approximations and also because they are inherently adapted to effectively utilize High-Performance-Computing (HPC) architectures. In line with this trend, during the last decade R&D activities have taken place at INR aiming to develop Monte Carlo-based high-fidelity calculation schemes capable to solve large-scale pin-by-pin depletion and transient problems. These schemes are intended to contribute to the prediction of relevant local safety parameters (at pin/sub-channel level) with less conservatism with respect to the current state-of-the-art methods and they are also appropriate to provide reference solutions to lower order deterministic codes for cases where detailed (pin-wise) experimental data is not available. The developed calcula-tion schemes are based on the coupling between neutron transport, sub-channel thermal-hydraulics and fuel thermo-mechanics in order to properly describe the fundamental reactivity feedback mecha-nisms such as the Doppler and moderator effects. Two different couplings between the Serpent2 Monte Carlo code, the SCF sub-channel code and the TRANSURANUS fuel performance have been developed; the first one is based on an object oriented programming approach with mesh-based fields exchange and the second one on the canonical internal (master-slave) concept. The optimization of the codes to tackle full core pin-by-pin problems and the increased accuracy obtained when adding also the more sophisticated fuel rod thermo-mechanics model to the coupling will be discussed. The verification and validation (V&V) of the couplings against PWR and VVER plant data and the associ-ated use of HPC resources will be presented. Finally, the potential of the developed tools to tackle problems related to spent nuclear fuel characterization in view of the R&D on "Predisposal" will be highlighted.