CFD Modeling of Fuel Rod Heat Transfer to Aid in the Risk Assessment of Fuel Rod Failure Due to Localized Crud

This paper presents a computational fluid dynamics (CFD) modeling methodology that has been developed to provide predictions of very local heat transfer variation in fuel rod assemblies. Results from the CFD analysis are used in HIDUTYDRV and other advanced codes that have been developed and are used internally by Westinghouse to predict very local crud deposition and dryout. This methodology is used in making the EPRI Level IV crud and corrosion guideline assessments, which were developed in response to the INPO 0 by 2010 initiatives. This methodology has been in production use for risk assessment of CE-design 14×14 and 16×16 fuel reloads. The methodology is in the process of being extended to other Westinghouse fuel design reloads. Local crud deposition and dryout are strongly dependent on very local boiling or steaming on small areas of the fuel rod, often referred to as local hot spots. These local hot spots can not be predicted utilizing standard subchannel modeling methodology because subchannel models do not provide sufficient azimuthal detail of individual rods. Local hot spots are also very dependent on the particular grid features, which are not explicitly modeled in subchannel analysis. The commercial code Star-CD by CD-ADAPCO is utilized to develop a detailed CFD model of a single fuel assembly grid span. Detailed azimuthal and axial predictions of the heat transfer coefficient are made for each rod in the model. These predictions are then normalized to a Dittus-Boelter based heat transfer coefficient so that the predictions can be translated to other spans and other fuel assemblies. Details of this translation as well as the use of normalized heat transfer coefficients in the advanced codes used to predict local crud and dryout are provided in a separate follow-on paper ICONE17-75715 also being presented at ICONE17. This paper presents details on the CFD methodology that has been developed to predict local normalized heat transfer coefficients for a fuel rod assembly. Results for a particular application are provided to illustrate the methodology. The application is for a fuel design that contains mixing grids and spans with and without intermediate flow mixers.