On the Influence of Model Uncertainty in Thermal Stratification Simulations of Pool-Type Sodium-Cooled Fast Reactors using Computational Fluid Dynamics

Abstract

Thermal stratification in the upper plenum of pool-type sodium-cooled fast reactors (SFR) can occur during transient and accident conditions, potentially impacting the natural circulation performance and adversely affecting heat removal performance. Having reliable simulation tools and a good understanding of simulation uncertainty is crucial for designing a robust reactor. In this research, we demonstrate an approach to quantify the effect of model uncertainty on thermal stratification simulations using computational fluid dynamics (CFD). A 3D Reynolds Averaged Navier-Stoke (RANS) model is established based on the SUPERCAVNA facility, in which the flow scenario is an approximate representation of the complex flow in the upper plenum of SFR. To assess the influence of model uncertainty, the turbulent viscosity is treated as a Gaussian random field, in which the hyper-parameters are inferred in phenomena-based separate effects tests and then applied in a forward context to the application of interest. The uncertain mixed-convection phenomenon is investigated by quantifying the variation of temperature distribution; the uncertainty mixing flow rate at different elevation levels is also accessed. The preliminary result shows that the size and shape of the flow recirculation zone are significantly affected by a small perturbation of the eddy viscosity field. As a result, large variations in temperature profile and mixing flow rates are observed.