Integrated fuel performance and thermal-hydraulic sub-channel models for analysis of sodium fast reactors

Abstract

Sodium Fast Reactors (SFR) show promise as an effective way to produce clean safe nuclear power while properly managing the fuel cycle. Accurate computer modeling is an important step in the design and eventual licensing of SFRs. The objective of this work was to couple a model for metal fuel performance to a sub-channel analysis code to more precisely predict critical phenomena that could lead to pin failure for steady-state and transient scenarios. The fuel code that was used is the recently developed and benchmarked FEAST-METAL code. The sub-channel analysis code that was selected is COBRA-IV-I. This code was updated with current correlations for sodium for pressure drop, mixing, and heat transfer. The new code, COBRA-IV-I-MIT was then validated with experimental data from the Oak Ridge National Laboratory (ORNL) 19-Pin Bundle, the Toshiba 37-Pin Bundle, and the Westinghouse Advanced Reactors Division (WARD) 61-Pin Bundle. Important topics that were addressed for coupling the codes include the following. The importance of azimuthal effects in the fuel pin: FEAST only evaluates the fuel in two-dimensions, assuming azimuthal symmetry; however, coupling to COBRA produces an azimuthal temperature distribution. The acceptability of assuming a two-dimensional fuel rod with an average temperature was examined. Furthermore, how the fuel pin evolves over time affects the assembly geometry. How well a two-dimensional fuel rod allows for an accurate description of the changing assembly geometry was also considered. Related to this was how the evolution of the assembly geometry affects its thermal hydraulic behavior, which determined the exact form of coupling between the codes. Ultimately one-way coupling was selected with azimuthal temperature averaging around the fuel pin. The codes were coupled using a wrapper, the COBRA And FEAST Executer (CAFE), written in the Python programming language. Data from EBR-II was used to confirm and verify CAFE. It was found that the number of axial nodes used in FEAST can have a large effect on the result. Finally FEAST was used to parametrically study three different pin designs: driver fuel, radial blanket, and tight pitch breed and bum fuel. This study provides data for pin expected life in assembly design.