Reducing Wind Energy Forecast Error With a Hybrid Ensemble Prediction Method

When wind energy is traded in ex-ante time frames for Day-Ahead electricity markets, accurate wind forecasts are needed to secure positions up to 36 h after the market gate closes to bidders. Wind energy is traded in discrete quantities; however, it is generated from an intermittent and variable resource. Deterministic forecasts are preferred for energy trading as they are compatible with providing a defined forecast quantity. However, deterministic forecasts cannot capture the stochastic nature of the underlying power source and are therefore suboptimal. Ensemble-based forecasts have the potential to reduce forecast error by accounting for uncertainties not captured in deterministic models. However, ensemble forecasts are not always available at wind turbine hub heights. Therefore, a method is needed to apply ensemble information at turbine hub heights for energy forecasting purposes. This paper presents a novel machine learning-based method that translates the perturbations from a localised Numerical Weather Prediction model's 10-m wind speed output to an ensemble energy forecast at 100 m. The extrapolated ensemble-based forecast has improved the forecast accuracy by up to 9% when compared to the deterministic output, in addition to providing useful information on the forecast uncertainty. The findings have important implications for future energy trading, transmission system operation and meteorological forecasting.