WILLOW, Wholistic and integrated digital tools for extended lifetime and profitability of offshore wind farms, aims to achieve an integrated system that will provide an open-source, data-driven health aware curtailment strategy to the offshore wind farm operators. This integrated wind farm control system will look for a trade-off between the power production and the lifetime consumption. Therefore, physical models and data-driven models (AI/ML) will be used to assist decision-making and planning of wind turbine operation and maintenance (O&M) activities. With a 5.8 million euro budget granted within the framework of the Horizon Europe programme, it is expected to contribute to a 50% reduction on the inspection costs, a 5-years lifetime extension of offshore wind farms, a 4% reduction in noise pollution and up to 10% reduction of LCOE (Levelized Cost of Energy), between 3.5 and 4.5 €/MWh.

As wind energy gains ground on the energy market, wind farms will play an increasingly important role in the stability of the electric system. Nowadays, wind farms must deliver commanded output power following the needs of grid operators, as electricity generation has to match demand on real time, which implies producing less power than available. Today this is done either by shutting down a few turbines and letting others produce maximum power, or by down-regulating each turbine by the same amount.

Although these strategies may negatively affect the fatigue life of the turbine, the optimization of these decision-making schemes is extremely complex due to the need to better understand and include many factors such as component degradation, the particular complexity of grid integration, or specific offshore issues like corrosion or the additional loads from waves, tides and currents.

In order to solve all these challenges, WILLOW aims to achieve the following objectives:

1.    Development of a global structural health monitoring (SHM) based on loads, accelerations, images, and thickness losses, considering fatigue progression, pitting corrosion and coating degradation by using physical and virtual sensors combined with Machine Learning techniques.

2.    Development of prognosis tools by combining SCADA and SHM data, using physical models and ML methods to predict the consumed lifetime and the remaining useful life.

3.    Development of a decision-making support tool for smart power dispatch in curtailed conditions and O&M scheduling.