Wave Energy Converters (WECs) need to operate in a harsh environment, where fatigue in materials and salt accumulation regularly lead to premature failures of components. In the course of seven years of continuous operation of the Mutriku Wave Power Plant, several events of this kind have occurred and are described in detail in this article. Preventive maintenance is conducted regularly in the plant. However, the novel machine learning approach could help to optimize maintenance works, as failure can be predicted without unnecessary replacement of components.
