Considerations on Current Methodologies for the Assessment of Engineering Correlations for Heat Transfer at Supercritical Pressures

Abstract The design of industrial equipment operating at supercritical pressures, especially in innovative nuclear reactors like Supercritical Water Reactors (SCWRs) at 25 MPa, demands accurate modeling of convective heat transfer. The prediction of fuel surface temperature is vital for safety and regulatory approval of new reactor concepts. Over the years, numerous engineering correlations have been developed based on experimental data for various fluids at supercritical pressures. Some of these correlations claim high accuracy, but their performance often falls short when applied to different datasets, particularly under deteriorated heat transfer (DHT) conditions. DHT is a complex phenomenon sensitive to boundary conditions, challenging traditional formulations like the Dittus-Boelter equation with property ratio corrections. Even more sophisticated formulations have not significantly improved accuracy. Within the EU ECC-SMART Project, a comprehensive review of these correlations was performed, evaluating their performance using two diverse datasets covering a wide range of boundary conditions. Great care was applied in avoiding common pitfalls in assessing correlation accuracy, such as the inappropriate use of experimental wall temperature data, which can bias the results and mislead about the most critical variable to be predicted by a correlation when used in a system code. The paper presents sample results from this assessment, highlighting the potential and limitations of existing formulations. It also addresses cases where multiple solutions for wall temperature arise, even with simple formulations. We consider correlations both in their original form and with optimized coefficients, shedding light on possible enhancements. This study offers insights into the current state-of-the-art in heat transfer modeling for supercritical fluids and suggests directions for future research.