Entropy Production Evaluation within a Prototype Pump-Turbine Operated in Pump Mode for a Wide Range of Flow Conditions

Abstract

Inside the pump-turbine, energy is irreversibly lost due to turbulent pulsations in the high Reynolds number zone and actions of viscous forces close to the wall. The conventional differential pressure method cannot obtain specific details of the hydraulic loss within the machine’s flow passages; on the other hand, the entropy production method can provide accurate information on the location of irreversible losses and the spatial distribution of energy dissipation. Therefore, based on the entropy production theory, this study investigates the composition and distribution of hydraulic losses under different flow conditions for a prototype pump-turbine in pump mode. Study results indicated that total hydraulic losses significantly decreased, then slowly increased with an increase in flow rate. The entropy production rate caused by turbulence dissipation (EPTD), direct dissipation (EPDD), and wall shear stress (EPWS) displayed the same variation patterns as that of total hydraulic losses, with EPTD and EPDD being the most dominating. The location of hydraulic loss within the pump-turbine’s flow domain strongly depended on flow conditions. High hydraulic losses primarily occurred in the guide vanes (GV) and draft tube under low flow rates. Under high flow conditions, however, high hydraulic losses were mostly concentrated in the stay vanes (SV), spiral casing, and GV. Hydraulic losses at low flow rates were primarily caused by flow separation within the GV flow channels, vortices in the vaneless region, and inlet flow impacts on the runner blade’s leading edge. On the other hand, large vortices within the GV and SV flow channels, GV wake flow, and unsteady flow at the spiral casing were the main contributors to hydraulic loss under high flow conditions. EPDD was mainly caused by strain rate, so it was closer to the main vortex regions, whereas EPTD was affected by turbulence intensity and had a wider distribution range in the unsteady flow.