TRACE code core reflood thermal-hydraulics phenomena benchmarking against the NRC–PSU Rod Bundle Heat Transfer (RBHT) test facility

Abstract

This paper evaluates the performance of the U.S. Nuclear Regulatory Commission’s (NRC’s) thermal hydraulic code TRAC/RELAP Advanced Computational Engine (TRACE) against experimental reflood data from the NRC/Pennsylvania State University (NRC/PSU) Rod Bundle Heat Transfer (RBHT) test facility, as an integral step in verification of code accuracy. This paper is an extension of the NURETH-20 conference paper by the first author (Garrett et al., 2023) that has been recommended for consideration and submission to Nuclear Engineering and Design. An international study on reflood thermal-hydraulics, sponsored by the Nuclear Energy Agency (NEA) Working Group on Accident Management and Analysis (WGAMA), was conducted with data collected in the NRC/PSU RBHT test facility, located at the Pennsylvania State University. A series of 16 benchmark tests were conducted, with conditions covering a carefully selected range of oscillatory, variable stepped and constant rate reflood injection velocities. These unique conditions are useful for code validation and model improvement. These 16 tests were segmented into 11 open tests, followed by five blind tests. This paper covers the five blind tests as the 11 open tests were covered by Garrett et al. at the NURETH-19 conference (Garrett et al., 2021).

For TRACE code benchmarking, a numerical model with the same dimensions as the RBHT facility was used. The initial and boundary conditions for this model were taken from experimental measurements. Many of the test conditions were chosen to examine sensitivities to important parameters, such as reflood liquid subcooling, reflood rate, and system pressure. The wide range of test conditions served to test the code and provide insight to its strengths and potential areas of improvement. These novel experiments were vital in this effort.

Simulations were made for five reflood tests and comparisons between predicted and measured results were made for the transient cladding temperatures, vapor temperature, bundle liquid mass fraction, carryover fraction, and steam exhaust fraction. The comparison presented in this paper has provided useful insight into code improvements. Studies to more accurately model reflood phenomena are currently underway as a result of the work presented in this paper.