Numerical investigation on the unstable flow and its interaction with the hump characteristic in a pump turbine at pump mode

Pumped storage power is the most mature and cleanest energy in the world, which is universally used in energy recovery fields. As the pumped storage units operate at part-load conditions, typical unstable flow phenomena are universally observed. Guaranteeing the operation stability of the unit is being increasing significant. Since much attention has been paid to the unstable flow characteristic in the hump region, the interaction between the unstable flow and hump characteristic is seldom involved. In present research, comprehensive analysis is systematically conducted to explore unstable flow and its interaction with the hump characteristic in a pump-turbine operated at pump mode. Three typical models, corresponding to the test model, small lean angle model (lean angle θ = 2°) and medium lean angle model (lean angle θ = 10°) representing different flow conditions, i.e., hump peak, hump valley and stable condition respectively, are selected for analysis. First, characteristic curves predicted by a modified shear stress transport k-ω partially averaged Navier-Stokes model show good agreements with the experiment data. Then, comprehensive discussions are performed in the blade passage of runner and guide vane. In the runner blade passage, the scale of flow separation near the suction side of the blade increases with increment of the lean angle. In the guide vane, although rotating stall is observed in three models, its scale and intensity are largest under the hump peak condition, and alleviated under the hump valley and stable condition. After frequency and its dynamics analysis, it is found that one dominant frequency of 0.088fn is induced by the rotating stall evolution. Moreover, as the rotating stall is alleviated under the stable condition, the dominant frequency captured by the proper orthogonal decomposition (POD) method is transferred to 3fn, and the hump is also alleviated. © 2024 Elsevier Ltd