Electricity forecast adapted to ocean conditions: The Mutriku case study

It is essential for any wave power plant at the commercialising stage to accurately model and forecast its production to ensure a good functioning and an efficient trading in the daily wholesale electricity market. Altogether, a reliable forecast is a required step for the establishment of wave energy as a sustainable and viable source of electricity. We propose models whose coefficients adapt to the variation of ocean conditions and apply them to model/estimate and forecast the electricity production of two oscillating wave converters from the Mutriku Wave Power Plant during 2018. After a statistical analysis comparing different approaches, we recommend to model the production using a regression with coefficients depending on the wave height and to do short-term forecast using an autoregressive function with coefficients also depending on the wave height.