Optimization Study for Uncertainty Reduction of In-Pile CHF Experiments in TREAT
Abstract Reactivity-initiated accidents (RIAs) are one of the postulated incidents that can threaten the operational safety of a nuclear reactor. During a RIA, a rapid increase of energy deposition in the fuel can lead to a departure from nucleate boiling (DNB) occurrence which refers to the point where a drastic decrease in heat transfer capabilities occurs and the surface heat flux exceeds the critical heat flux (CHF). Aiming to understand the fundamentals beneath CHF and to predict it, the Transient Reactor Test (TREAT) facility at the Idaho National Laboratory (INL) is a unique facility that will be used to experimentally investigate the transient CHF under in-pile pool boiling condition. As part of a comprehensive effort to utilize TREAT for this project, this study analyzed the expected uncertainties in the experimental data by identifying the key inputs for the uncertainty in the temperature measurements and quantifying their priorities. The sensitivities of key inputs from neutronics modeling, the clad-to-coolant heat transfer, thermophysical properties of the tube, and coolant conditions were quantified using Sobol sensitivity analysis methods, and the significant effect of the occurrence of the CHF on the sensitivity of input was found.