Insight into the Influences of Blade Configuration on Internal Flow and Pressure Pulsation in High-Head Pump-Turbine: 9-Blade Versus 6+6 Long-Short-Blade Runners

High-head reversible pump-turbines frequently suffer from hump-instability under pump-mode operation, and inherent performance trade-offs exist between pump-mode and turbine-mode characteristics for bidirectional operating requirements. The 6+6 long-short-blade runner is regarded as a promising candidate for mitigating pump-mode hump-instability. Taking a high-head pumped-storage pump-turbine as the research object, this study follows the above-mentioned engineering principle. A full-passage CFD model including volute, stay-ring, guide vanes, runner and draft-tube is established with the SST k-ω turbulence model. Numerical methods are validated against model-test external-characteristic data. Steady and unsteady simulations are carried out for the prototype 9-blade runner and the 6+6 long-short-blade runner. The results show that, under pump-mode operation, the 6+6 configuration shifts the hump-instability region toward lower mass-flow rates away from the unit’s primary operating range. Short blades divide flow passages, suppress large-scale channel vortices within the hump zone, and alleviate flow asymmetry in the vaneless space. Nevertheless, local impingement and secondary-flow losses occur at long-short-blade junctions, and short-blade leading edges bear potential cavitation risk. This configuration only shifts the instability range rather than completely eliminating hump-instability. A typical 50% partial-load turbine-mode condition is selected for constraint-based verification. The 6+6 runner reduces pressure-pulsation amplitudes at draft-tube monitoring points by 25.52% and 19.83%, which suppresses occasional large-amplitude pressure fluctuations and improves signal periodic consistency. The novelty of this work lies in the quantitative revelation of such trade-off characteristics under the pump-priority-design philosophy for a 655 m high-head unit. It demonstrates that the 6+6 long-short-blade scheme is not a universal optimization measure.