Geometrically-Accurate-Three-Dimensional Simulations of a Used Nuclear Fuel Canister Filled With Helium

Abstract

This paper presents preliminary results of heat transfer simulations performed in geometrically-accurate-three-dimensional model of nuclear fuel canister filled with helium. The numerical model represents a vertical canister, which relies on natural convection as its primary heat transfer mechanism, containing 24 PWR fuel assemblies. The model includes distinct regions for the fuel pellets, cladding and gas regions within each basket opening. Symmetry boundary conditions are employed so that only one-eighth of the package cross-section is included. The canister is assumed to be filled with helium at atmospheric pressure. A constant temperature of 101.7°C is employed on the canister outer surfaces, assuming the canister to be surrounded with water. These conditions of pressure and temperature were considered, in this paper, for comparison purpose with previous work. The effects of buoyancy-induced gas motion and natural convection, along with radiation and conduction through gas regions and solid are considered. Steady state simulations using ANSYS/Fluent were performed for different heat generation rates in the fuel regions. Simulations that include the effect of natural convection and others that do not include this effect are conducted. The peak cladding temperature and its radial and axial locations are reported. The maximum allowable heat generation that brings the cladding temperatures to the radial hydride formation limit (TRH=400°C) is also reported. The results of the three dimensional model simulations were compared to two dimensional model simulations for the same heat generation rate. The results showed that the two-dimensional simulations overestimate the temperature in the canister by almost 70°C.

Heat Transfer Modeling of Spent Nuclear Fuel Using Uncertainty Quantification and Polynomial Chaos Expansion

“…A three-dimensional FLUENT model will be created for future simulations. Previous studies have shown that two-dimensional simulations overestimate the temperature in SNF assembly holding canisters [15]. The uncertainty quantification method could be used to reduce the computational intensity required by large assemblies due to a large number of nodes.…”

“…The second range has a minimum temperature of 420 K and a maximum value of 425 K, and is referred to as the high wall temperature range. These temperature values are typical for SNF assemblies in dry cask storage [15] and correspond, respectively, to storage assemblies placed near the outer boundary and near the center of the cask. Figure 3 shows a picture of a typical cask with multiple vertical assemblies inside it.…”