Very short-term probabilistic forecasting of wind power based on dual-Doppler radar measurements in the North Sea

Abstract

Probabilistic very short-term forecasts of wind farm power can provide valuable information for electricity market participants, especially in power systems with high penetration of wind energy. Recently, the first dual-Doppler radar measurements of an offshore wind farm have become available. A probabilistic very short-term forecasting model of wind power is proposed using observed wind speeds and directions from a dual-Doppler radar system. Predictive wind speed distributions are derived from three-dimensional dual-Doppler observations covering the wind farm and its vicinity. To estimate the power generated by the wind turbines, a power curve constructed with dual-Doppler observations is used. A forecasting horizon of five minutes is evaluated. Results show an improved performance of a deterministic radar-based forecast of power versus the benchmarks persistence and climatology, and also a high potential for a wind power probabilistic forecast.