Application of CFD towards the thermal-hydraulic analysis of Spent Fuel Pool accidents

After a spent fuel assembly is removed from the reactor core, its decay heat production is still too large for pure conductive cooling. It is placed in a Spent Fuel Pool, where the decay heat is removed from the assembly by means of natural convection, while the pool cooling system keeps the water temperature at about 40-50°C. The fuel assemblies are typically arranged in high density racks which consist of borated steel in order to prevent criticality accidents. The rack cells are closed to the sides and force the coolant flow along the axial direction. In the event of a failure of the cooling system followed by boil-off, the water level might decrease below the top of the assemblies. Then the natural circulation path in the water phase is blocked and the dominant cooling mechanisms for the uncovered section of the fuel assemblies are the forced convection due to steam production and the solid heat conduction into the remaining water, as well as towards the upper end. The latter mechanism depends on the boundary conditions at the fuel assembly head, which are determined by the temperature and velocity field in the pool and reactor building atmosphere. Additionally, if the decay heat load of the neighboring fuel assemblies differs significantly, some heat will be exchanged in the radial direction. In the nuclear community, system codes are widely used for safety studies. They deliver fast and reliable results for the flow and heat transfer inside the fuel assemblies. But since three-dimensional convective phenomena can only be taken into account in the form of simplified assumptions, they do not entirely qualify for studies related to Spent Fuel Pools. In this work, CFD is used to study the flow field above and around the exposed storage racks in order to identify large scale convective phenomena. It is expected that the large scale flow field is dictated by the temperature field in the pool, which in turn influences the cooling of the individual fuel assemblies. These interdependencies need to be quantified as a function of the storage rack arrangement and the overall distribution of the fuel assemblies with respect to their decay heat. A porous body approach is employed for the modeling of the fuel assemblies. Best Practice Guidelines are applied as far as possible, since there is no experimental data available that allows a straightforward validation of the simulation results. In this work, the Spent Fuel Pool design of Fukushima's Unit 4 serves as a test case. A loss of coolant by boil-off is postulated, leading to partially uncovered fuel assemblies. Several pool loading strategies for a constant total decay heat load will be presented and conclusions will be drawn as to which configuration is the most favorable from a thermohydraulic standpoint.