Modeling of Thermohydraulics in a Lead-Cooled Reactor System - A CFD Study Using OpenFOAM

Sammanfattning

This study investigates the thermal-hydraulics of a simplified model of Blykalla's Swedish Advanced Lead-cooled Reactor (SEALER) using computational fluid dynamics (CFD) in OpenFOAM. The objective is to develop and verify a system-level model for forced and natural convection. The reactor cooling circuit is represented using porous media and volumetric source terms. Simulations are conducted using Reynolds-averaged Navier-Stokes (RANS), employing both eddy-viscosity and Reynolds stress turbulence models. The model is initially validated for single-phase steady-state forced convection by comparison with reference data. Strong agreement is obtained with respect to mass flow rate, core temperature difference, and pressure drop, indicating a consistent implementation of the system components. A noticeable deviation of \SI{18.6}{\%} is observed in the steam generator pressure drop, which suggests sensitivity to the porous media implementation. The turbulence model has a negligible impact on system-level performance during forced convection. For natural convection, the results show a strong dependence on reactor geometry, where increased system height enhances circulation and thermal distribution. The choice of turbulence model has a more pronounced effect under natural convection conditions, where the Reynolds stress model predicts a lower mass flow rate and exhibits reduced numerical stability compared to the eddy-viscosity models. The influence of thermophysical modeling is assessed by comparing a temperature-dependent property model with the Boussinesq approximation. The Boussinesq model predicts a higher mass flow rate and a lower temperature rise across the core, consistent with its simplified treatment of density variations. Multiphase simulations using a volume of fluid (VOF) approach are conducted to assess the capability to predict buoyancy-driven flow in the presence of large density differences. The results indicate that the presence of a gas phase has a negligible impact on system performance and that the OpenFOAM treatment of buoyancy is sufficient for system-level analysis.