COMPARISON OF SELECTED FORCED-CONVECTION SUPERCRITICAL-WATER HEAT-TRANSFER CORRELATIONS FOR VERTICAL BARE TUBES

This paper presents an extensive study of heat-transfer correlations applicable to supercritical-water flow in vertical bare tubes. A comprehensive dataset was collected from 33 papers by 27 authors, including more than 125 graphs and wide ranges of parameters. The parameters ranges were as follows: pressures 22.5 - 34.5 MPa, inlet temperatures 85 - 350 degrees C, mass fluxes 250 - 3400 kg/m(2)s, heat fluxes 75 - 5,400 kW/m(2)), tube heated lengths 0.6 - 27.4 m, and tube inside diameters 2 - 36 mm This combined dataset was then investigated and analyzed. Heat Transfer Coefficients (HTCs) and wall temperatures were calculated using various existing correlations and compared to the corresponding experimental results. Three correlations were used in this comparison: Bishop et al., Mokry et al. and modified Swenson et al. The main objective of this study was to select the best supercritical-water bare-tube correlation for HTC calculations in: 1) fuel bundles of SuperCritical Water-cooled Reactors (SCWRs) as a preliminary and conservative approach; 2) heat exchangers in case of indirect-cycle SCW Nuclear Power Plants (NPPs); and 3) heat exchangers in case of hydrogen co-generation at SCW NPPs from SCW side. From the beginning, all these three correlations were compared to the Kirillov et al. vertical bare-tube dataset. However, this dataset has a limited range of operating conditions in terms of a pressure (only one pressure value of 24 MPa) and one inside diameter (only 10 mm). Therefore, these correlations were compared with other datasets, which have a much wider range of operating conditions. The comparison showed that in most cases, the Bishop et al. correlation deviates significantly from the experimental data within the pseudocritical region and actually, underestimates the temperature at most times. On the other hand, the Mokry et al. and modified Swenson et al. correlations showed a relatively better fit within the most operating conditions. In general, the modified Swenson et al. correlation showed slightly better fit with the experimental data than other two correlations.