Modeling and results of the dry cask simulator (DCS) with STAR-CCM+

Abstract

To understand the behavior of commercial nuclear fuel during its storage phase in spent fuel dry storage casks, simulations are necessary and to validate the simulation models, full-scale experiments are required. Therefore, an experimental facility called the dry cask simulator (DCS) was built in Sandia National Laboratories with the purpose of producing validation-quality data that can be used to test the accuracy of the modeling used to determine cladding temperatures in modern vertical dry casks. The DCS has been created under the framework of the Extended Storage Collaboration Programme (ESCP), coordinated by EPRI.The ultimate goal of the DCS was to study the thermal–hydraulic response of a single boiling water reactor (BWR) fuel assembly under variety of heat loads, internal vessel pressures, and external configurations. In this article, the simulations conducted by ENUSA Industrias Avanzadas S.A., S.M.E and UPM (Universidad Politécnica de Madrid, Spain) using STAR-CCM+ are validated against the experiments conducted at Sandia National Laboratories.In order to validate the simulation models several sensitivity analyses have been performed: a grid convergence index to estimate the discretization error in the simulations, a turbulence model sensitivity for the inner part of the cask, and three different approaches to model the fuel assembly: Case 1 (model with no spacers and no bottom tie plate), Case 2 (model with spacers and no bottom tie plate and Case 3 (model with spacers and bottom tie plate).Additionally, a thermal–hydraulic analysis is also presented, not only to explain the difference between the three ways to model the fuel assembly, but also to obtain a better understanding of the behavior of the dry cask according to pressure and heat output changes.Among the three cases, the simplest model (Case 1) has proven to obtain the most accurate results, being able to predict the temperature distribution in the fuel assembly and the air mass flow compared to the experiment measurements for all four cases of study. Meanwhile, the two other cases have shown that for steady state solutions the spacers and the bottom tie plate are not necessary and can introduce unwanted errors. Nonetheless, all cases have provided insight into the behavior of a dry storage cask system under different conditions showing the different heat transfer mechanisms interactions.