General assessment of convection heat transfer correlations for multiple geometries and fluids at supercritical pressure

The objective of this paper is to assess different correlations independently against a diversified databank - the Canadian Nuclear Laboratories multi-fluid and multi-geometry supercritical heat transfer databank. This databank was recently expanded by adding compiled and original experimental data, obtained through collaboration with the Nuclear Power Institute of China. The databank was subjected to screening for outliers, duplicates, and unreliable data. In addition, inappropriate data, not satisfying certain conditions, were removed. Nevertheless, the used databank comprised more than 41,000 measurements of heat transfer to different fluids flowing vertically upward in different geometries. Following a literature review and a compilation of correlations, an assessment of the tabulated correlations was performed against the databank. In total, 24 correlations were considered and applied to the entire database for different fluids including water, and different flow geometries including tube, annulus and rod bundle. Graphical comparison of best-estimate correlations and representative experimental data is presented in this paper. In addition, statistical error analysis was performed and leading correlations were identified. Although the leading correlation showed a standard deviation of less than 6%, variation of predicted wall temperature and heat transfer coefficient with fluid temperature followed the scatter of the experimental data. (author)