Lagrangian Simulation of Sediment Erosion in Francis Turbines Using a Computational Tool in Python Coupled with OpenFOAM

Hydraulic erosion from suspended sediment is a major degradation mechanism in Francis turbines of sediment-laden rivers, especially in Andean hydropower plants. This study presents a Python3.9-based computational tool integrating the empirical Oka erosion model within a Lagrangian particle tracking framework, coupled to single-phase CFD in OpenFOAM 10. The novelty lies in a reduced-domain approach that omits the spiral casing and replicates its particle-induced swirl via a custom algorithm, lowering meshing complexity and computational cost while preserving erosion prediction accuracy. The method was applied to a full-scale Francis turbine at the San Francisco hydropower plant in Ecuador (nominal discharge 62.4 m3/s, rated output 115 MW, rotational speed 34.27 rad/s), operating under volcanic and erosive sediment loads. Maximum erosion rates reached ~1.2 × 10−4 mm3/kg, concentrated on runner blade trailing edges and guide vane pressure sides. Impact kinematics showed most collisions at near-normal angles (85°–98°, peak at 92°) and 6–9 m/s velocities, with rare 40 m/s impacts causing over 50× more loss than average. The workflow identifies critical wear zones, supports redesign and coating strategies, and offers a transferable, open-source framework for erosion assessment in turbines under diverse sediment-laden conditions.