Multi-objective optimization of a two-stage liquefied natural gas cryogenic submerged pump-turbine in pump mode to reduce flow loss and cavitation

Abstract

To simultaneously minimize the energy loss and optimize the suction performance in a two-stage LNG (liquefied natural gas) cryogenic submerged pump-turbine in pump mode (PTP), a multi-objective optimization work containing numerical models of cryogenic cavitation flow, novel suction performance prediction, entropy production diagnosis and multi-criteria decision-making method is conducted to obtain the optimal geometric variables, as well as the corresponding performances under 0.7 Q0 with typical growth trend in energy loss after cavitation. Results show that the inlet angle β1 at the shroud and the arc section radius of the shroud r1 of the first-stage impeller and the wrap angle Φ at the shroud of the second-stage impeller are the significant parameters affecting the energy loss and suction performance of the PTP. Then, the ideal compromise solution, β1 = 14.34°， r1 = 26.28 mm and Φ = 127.44°, is obtained by multi-objective optimization and decision-making. The active suppression of the impinging flow, flow separation and rotating stall in the two-stage impellers after optimization yields a 6.18 % drop in the total entropy production rate with sacrificing a 1.25 % drop in pump head. In addition, the PTP is more likely to maintain a better suction performance for the optimal design over a wide range of flow rates. This research contributes to the search for the tradeoff between the conflicting hydraulic performance of the multistage pump-fed unit.