Predicting Critical Heat Flux in Rectangular Channels: Which Correlation to Trust?

Abstract

The PALLAS-reactor is an advanced nuclear reactor dedicated to the production of medical isotopes and (medical) nuclear technological research currently under construction in The Netherlands RELAP5 is the primary tool used for modeling the accident scenarios relevant for licensing the PALLAS-reactor. The Groeneveld Look-Up Tables (LUT) are the main Critical Heat Flux (CHF) model of RELAP5. These have been extensively validated against experiments with round tubes. However, literature suggests that the LUT over-predict CHF inside the narrow, rectangular fuel channels of research reactors such as the PALLAS-reactor. As literature abounds with CHF correlations of limited applicability and low degrees of validation and confidence, correlations that can accurately predict CHF in rectangular channels are needed. The goal of this work is twofold. First, it aims at selecting a number of promising CHF correlations and to assess these against experiments representative of the PALLAS-reactor design. Secondly, this work ascertains whether the LUT are indeed optimistic with respect to CHF in rectangular channels. To this purpose, a literature review has short-listed the Sudo-Kaminaga (S-K), the Kim-modified S-K, and the Hall–Mudawar (H-M) correlations. These correlations are implemented in the SPECTRA code developed at NRG and assessed against two experiments in rectangular channels. The Kim-modified S-K correlation fits the experiments better than the original S-K scheme for high mass fluxes, while predicting higher CHF values at low flow rates. Interestingly, the H-M correlation predicts the experiments accurately, although the correlation was developed from round tubes data. At high mass fluxes, the LUT predictions converge towards the experiments, thereby suggesting that the effect of the channel geometry becomes less important at high flow rates. At low mass fluxes, on the other hand, the LUT markedly fail to match the experiments. This work contributes to accurately estimating CHF in narrow rectangular channels, which is crucial for licensing research reactors such as the PALLAS-reactor, thus assisting in the betterment of many people’s lives.