Conference Agenda

Dispersed flow film boiling (DFFB) is a key flow regime that affects fuel rod integrity during emergency core cooling system (ECCS) injection phase in large-break loss of coolant accident in light water reactors. However, the existing experimental data in the DFFB regime have limitations in the important parameters that were measured, such as the void fraction and vapor superheat. These limitations in the DFFB data lead to remarkable uncertainties in the models/correlations developed for the wall convective heat transfer. To improve our understanding of and obtain experimental data for the wall heat transfer characteristics in the DFFB regime, a series of quasi-steady-state DFFB experiments was performed in the Post-CHF Heat Transfer (PCHT) test facility, over flow conditions of mass flux from 60 to 150 kg/m²-s, and pressure from 1.38 to 4.14 bar. The obtained wall convective heat transfer coefficient shows a transition in its trend with the thermodynamic equilibrium quality due to the transition of the dominant heat transfer mechanism from interfacial heat transfer to the vapor convection. The transition mechanism was confirmed by analyzing the computed vapor Reynolds number and the measured void fraction by an X‑ray radiography system. Based on the collected data, a new wall convective heat transfer correlation was developed by applying the Reynolds analogy to consider the variation of the interfacial area concentration with the droplet size. The newly proposed correlation successfully demonstrated its improved predictive capability compared to the existing models/correlations for a wide range of DFFB wall heat transfer data.

Best Estimate Plus Uncertainty (BEPU) methodologies are gradually establishing themselves as the favored way to perform Deterministic Safety Assessments (DSA) of Nuclear Power Plants (NPPs), because they account for uncertainties in plant states and physical behaviors while employing accurate-to-reality (best estimate) simulation codes. This work is performed in the context of the ATRIUM project, which seeks to establish best practices and standardize the BEPU process to make the results consistent, minimizing user effect. The approach taken in this project is to focus on specific phenomenology that is crucial to the progression of the desired transient (a Loss of Coolant Accident), and work with Separate Effect Tests (SET) experiments to find adequate uncertain parameters and their PDFs to propagate.

The target phenomenology for this study is that of post-Critical Heat Flux (CHF) heat transfer, concretely film boiling. Simulations of the Becker film boiling experiments were performed using RELAP5 to validate its suitability and identify influential parameters. Initial tests showed that RELAP5 overestimated burnout quality, resulting in inaccurate axial temperature profiles. Also, a sensitivity analysis of several RELAP parameters and initial conditions showed they didn’t influence the results.

To address this, RELAP5’s source code was modified to introduce and externalize new parameters for better control over film boiling modeling. Consequently, a sensitivity analysis was carried out on this new group of parameters, helping measure their effect and aiding in fine-tuning them to improve RELAP5’s predictions. This enhanced flexibility demonstrates a promising approach for more accurate RELAP5 analyses of complex post-CHF phenomena.

Prediction of the quench temperature is one of the most important parameters for the accurate estimation of the accident progress in light water reactors (LWR). In this study, a single rod flow quench experiments were conducted in the subcooling range of 0 to 40 K and 600-1100 °C initial cladding temperature conditions. Measurements of quench temperature were made at four levels of elevation from the bottom of the furnace. Based on these comprehensive tests, unique quench temperature trends were identified comprising two distinct regimes: (i) increase of quench temperature with increasing subcooling and initial cladding temperature with less effect of elevation in the high subcooling and low initial cladding temperature regime and (ii) constant quench temperature independent of subcooling and initial cladding temperature with a large effect of elevation in the low subcooling and high initial cladding temperature regime. The criteria delineating the two regimes was determined as a function of both subcooling and initial cladding temperature. A thorough analysis of cooling rate during the film boiling regime and quench front velocity was performed to develop a quench temperature correlation to better understand this new finding on quench temperature behavior. The study covers a critical gap in literature where high initial cladding temperatures (over 800 °C) under wide range of subcooling for such experiments have not been typically explored.

Quenching of hot vertical wall with a falling liquid film is an important thermal-hydraulic process to ensure the safety of nuclear reactors even during emergency situations. In this work, to develop a reliable model to predict the propagation velocity of the quench front, the temperature distribution of heat transfer surface during quenching was measured using a high-speed infrared camera. A silicon wafer that is transparent to the infrared ray was used as the hot wall, and the initial wall temperature, the wall thickness, the cooling liquid temperature, and the liquid flow rate were changed parametrically. The main heat transfer mechanism from the wall to the liquid film near the quench front was found to be the nucleate boiling. The heat transfer coefficient profile derived from the measured temperature distribution was therefore correlated using widely accepted heat transfer correlations such as Zuber's correlation for pool boiling CHF (Critical Heat Flux) and Rohsenow’s correlation for the pool boiling HTC (Heat Transfer Coefficient). The calculated quench velocity was in good agreement with the experimental data not only for the silicon wafer but also for the copper and zirconium walls of different thermal properties.