An experimental investigation of supercritical heat transfer in a three-rod bundle equipped with wire-wrap and grid spacers and cooled by carbon dioxide

Abstract

Heat transfer measurements in a three-rod bundle equipped with wire-wrap and grid spacers were obtained at supercritical pressures in the Supercritical University of Ottawa Loop (SCUOL). The tests were performed using carbon dioxide, as a surrogate fluid for water, flowing upwards for wide ranges of conditions, including conditions equivalent to the nominal and near-normal operating conditions of the proposed Canadian Super-Critical Water-Cooled Reactor. The test section contained three heated rods and three unheated rod segments with an outer diameter of 10mm and a pitch-to-diameter ratio of 1.14; the heated length was 1500mm. Detailed surface temperature measurements along and around the three heated rods were collected using internally traversed thermocouples. The following ranges of test conditions were covered, with equivalent water conditions given inside parentheses: pressure from 6.6 to 8.36MPa (19.7–25MPa); inlet temperature from 11 to 30°C (330–371°C); mass flux from 200 to 1175kgm−2s−1 (340–1822kgm−2s−1); and wall heat flux from 1 to 175kWm−2 (11–1847kWm−2). For one set of tests, the heated rods were fitted with a 1.3mm OD wire wrap, having an axial pitch of 200mm along the entire heated length; for a second set, the heated rods were fitted with grid spacers having a 5.3% flow blockage and located at 500mm axial intervals. The effects of spacer configuration on heat transfer at supercritical pressures were documented and analyzed. The observed experimental trends were compared to those obtained in a experiment in a heated tube at similar conditions and with predictions of a supercritical heat transfer correlation. Heat transfer was observed to deteriorate near the start of the heated section of the rod bundle equipped with either type of spacers as well as in a circular tube for mass fluxes between 200 and 700kgm−2s−1. The onset of deterioration occurred at a higher heat flux for the wire-wrapped bundle than for the one with grid spacers, and at the lowest heat flux for the tube. Normal heat transfer in either of the rod bundles behaved similarly to the one in the tube and was also compatible with the predictions of the correlation of Jackson. The grid spacers introduced a strong local enhancement, but their effects disappeared only about 10 rod diameters downstream.