2D and 3D Thermal Simulations for Storage Systems with Internal Natural Convection for Canistered Spent Fuel

Year
2004
Author(s)
Mike Yaksh, Ph.D., PE - NAC International INC
Christine Wang - NAC International Inc.
File Attachment
5-8_261.pdf522.6 KB
Abstract
actors burn advanced fuel assemblies to higher burnups, the dry storage systems will be required to accommodate higher heat loads. This is due to the increasing capacity of the systems and the need to store higher burnup fuel with reasonable cooling periods (i.e., five to six years). As the storage systems heat rejection design must be passive, natural convection is an efficient means for rejection of heat from the spent fuel to the surface of the canister boundary. The design presented in this paper is a canistered system that employs conduction, radiation and convection to reject heat from the canister, which is stored in a vertical concrete cask. The canister containing the spent fuel in this design is a right circular stainless steel vessel capable of storing 37 PWR fuel assemblies with a total canister heat load of 40 kW. Accompanying any design effort is the use of a numerical methodology that can accurately predict the peak-clad temperatures of the fuel and the structural components of the system. The main challenge to any analysis employing internal natural convection may be perceived as a practical limitation due to the size of the model. Since canisters are typically cylindrical, a two-dimensional model can be used to represent the canister. The fuel basket structure, which maintains the configuration of the spent fuel, is an array of square tubes, and is non-axisymmetric. Flow up through the fuel region in the basket encounters a complex cross section due to the fuel assembly rod array (up to 17×17). The flow region of the heated gas down the outside of the basket in the annulus between the canister shell and the basket assembly (downcomer) is also an irregular shaped area. To confirm that a two-dimensional (2D) modelling methodology is appropriate, a benchmark using results from a thermal test is required. The thermal test focuses on the accuracy of the simulation internal to the canister. The actual design of a canistered system inside a concrete cask requires additional modelling effort, since the flow along the external surface of the canister must be included. The target design, however, employs a nonuniform heating of the PWR fuel assemblies, which permits the heat load of an individual PWR assembly to range from .88 kW to 1.35 kW in a zoned configuration. As nonuniform loading adds more complication to the 2D model, an additional confirmation of the 2D modelling methodology may be obtained by performing a three-dimensional (3D) simulation of a simplified version of the target design.