Improving wind power forecasts in the Belgian North Sea

With the completion of the first Belgian offshore wind energy zone in 2020, for an installed capacity of 2.26 GW, a significant amount of wind energy is now available in the Belgian part of the North Sea. Due to the relative lack of space in this area, all wind farms lie close together in a narrow band, and each wind farm has a high density, in terms of number of turbines, and/or installed power per area. There are thus considerable wake losses in the Belgian offshore zone. Moreover, in case of a major storm, many wind farms might experience a so called cut-out event, with automatic shut-down of the turbines due to high mean wind speed, at practically the same time. Since this can lead to large imbalance risks on the electricity grid, the Royal Meteorological Institute of Belgium (RMI) has developed a dedicated storm forecast tool for Elia, the Belgian transmission system operator for high-voltage electricity. This storm forecast tool, which has been operational since November 2018, consists of 15 minute wind and power forecasts per wind farm, together with cut-out probabilities and uncertainty quantification, by combining our high-resolution (4km) ALARO model with the ENS ensemble forecasts of the European Centre for Medium Range Weather Forecasting (ECMWF). We report on several approaches to improve the offshore wind power forecasts, as part of the BeFORECAST project (Nov 2022 - Oct 2025), funded by the Energy Transition Fund of the Belgian federal government. In particular, to take into account wake losses, the Fitch et al. wind farm parameterization (WFP) was implemented in our ALARO model, based on an earlier implementation by KNMI into HARMONIE-AROME. Both these models are being developed in the ACCORD consortium, and use the same dynamical core to some extent, with IFS/ARPEGE global codes as basis, but differ greatly in the different physics parameterizations used, and the physics-dynamics coupling (tendencies vs fluxes). For instance, unlike AROME and HARMONIE-AROME, the ALARO model uses an explicit deep convection scheme (3MT) and turbulence is based on the TOUCANS framework. Verification of the improved wind and power forecasts is based on several lidars at different locations, and power data per wind farm from Elia, possibly supplemented with SCADA data from wind farms where available. Other approaches we study are multivariate statistical postprocessing based on historical wind speed observations to generate corrected wind speed scenarios, and postprocessing of forecasts using wake models (since we cannot implement a WFP in the ENS ensemble). Finally, an alternative power forecasting method, using an artificial neural network trained on power observations and NWP forecasts is also looked at. Special consideration is given to wind storms and fast ramping events.