Enhancement in heat transfer from heat generating core debris using a passive approach in fast reactors: 3-D CFD analysis

Abstract

The article presents the results of three-dimensional numerical analysis to assess the natural convection heat transfer in pool-type sodium-cooled fast reactors during post-accident heat removal. To improve the coolability of the destroyed core, suitable modifications in core catcher design are suggested. The equations for conservation of mass, momentum and energy are solved in their transient form using finite volume based discretization in (r,θ,z) system. Turbulence is modeled using the k−ω SST model based on sensitivity analysis of different turbulence models and validation exercises. A comparative analysis of 3-D and 2-D simulations is carried out. Various feasible geometrical arrangements for core collection trays are analyzed in detail for better coolability of the core debris by natural convection. Multiple debris collection trays with and without cooling pipes are studied towards achieving coolability of core debris in case of core disruptive accident. Results show that better safety margins are achieved by using 3-D as compared to 2-D simulation. The incorporation of cooling pipes in distributed heat sources on multiple trays leads to enhanced coolability of the debris within the vessel. Maximum temperatures on the core catcher/collection trays and in the core debris (with the selected core catcher design) are 860 K and 932 K respectively. These are substantially below the acceptable thermal design limits. Furthermore, detailed insight of the flow field confirms the turbulence in the lower plenum, partly due to impingement of passive jets from multiple cooling pipes on core debris and partly due to flow bypass through the slots at upper tray. It is established that the use of passive pipes and multi-tray concept for core catcher can successfully accommodate the whole core debris arising from core disruptive accidents within the main vessel without exceeding the safe thermal design limits.