Hydropower plants across Europe are facing growing pressure from two directions at once. On one side, climate change is intensifying sediment loads in rivers, accelerating wear and erosion on turbines and other components. On the other, the shift to a more flexible, renewables-heavy electricity grid is pushing plants to operate across a wider range of conditions. This increases the risk of cavitation, vibration, and fatigue damage, which ultimately result in unplanned downtimes for maintenance, during which the hydropower plant cannot operate.
We have now taken a big step towards a solution. Building on earlier concept validation work, the ReHydro team has now moved to real-world implementation across three demonstration sites, focusing on gathering reliable data under actual operating conditions.
At the Valeira hydropower plant in Portugal, an integrated monitoring system has been deployed to track both long-term efficiency and cavitation phenomena in Kaplan turbines. The system draws on acoustic sensors, a dedicated data acquisition setup, and both local and cloud-based infrastructure. Operational data from the plant’s SCADA (Supervisory Control and Data Acquisition) system is used to evaluate turbine performance, while several multivariate regression models have been developed to estimate expected efficiency, allowing the comparison between measured and predicted performance. The results confirm that it is possible to build a robust methodology for long-term efficiency monitoring using existing plant data, while also detecting cavitation events and their spatial non-uniformities across different operating regimes.
In Switzerland, at the Vissoie and Bitsch plants, the focus has been on understanding how sediment affects Pelton turbines. A comprehensive monitoring architecture has been installed, combining thermodynamic efficiency measurements, turbidity and density sensors, high-frequency accelerometers, and a dedicated imaging system for tracking bucket erosion. These field measurements are complemented by computational fluid dynamics simulations and laboratory analyses, providing both real-time estimates of sediment concentration and detailed insights into erosion mechanisms. Especially the imaging system makes a case for future automated erosion analysis using artificial intelligence.
In Norway, the Røldal–Suldal system has served as a testing ground for digital integration. Parallel digital twins of the hydropower system have been developed to simulate plant behaviour across different operating conditions, and a transient digital twin captures dynamic processes in real time. By feeding these models with live operational and environmental data, the team has been able to validate simulations, test optimisation strategies, and demonstrate that digital twin frameworks can meaningfully support production planning and operational decision-making.
Across all three sites, the results point in the same direction: advanced monitoring technologies, combining on-site measurements, data-driven analysis, and numerical modelling, are technically feasible under real operating conditions and can deliver consistent, high-quality data with genuine value for predictive maintenance and refurbishment planning. The work reinforces the broader ReHydro ambition to support the digitalisation of hydropower, and with it, the long-term reliability and sustainability of a technology that remains central to Europe’s low-carbon energy future.
A full account of the methods, results, and findings will be made public in the future.