Developing Predictive Maintenance Framework for Wind Turbine Blade Erosion: State of the Art and Concept Analysis

The global goal of achieving 2000 gigawatts of offshore wind power by 2050 has driven the development of the wind energy sector. This ambitious goal is facing a significant challenge in maintaining the efficiency and health of the wind turbines. Wind turbine Blade erosion is among the main critical failure modes that lead to high production losses and maintenance expenses. At present, the industry is utilizing manual or drone-based inspection, however, they are targeting more cost-effective and more informative like to predictive maintenance for erosion and severity by using data science techniques based on SCADA and sensors data. This paper reviews existing erosion blade analysis methods in wind turbines, highlighting their strengths and weaknesses. The method is based on the literature review to determine the state of the art in terms of monitoring, classification, and prediction. Moreover study will evaluate potential predictive maintenance concepts based on key performance matrix extracted from key stakeholders. It can be concluded that a combined concept of vibration, aerodynamic, acoustic, and production loss techniques supported with XGBoost, FFT, and LSTM are the most effective methods.