92nd American Meteorological Society Annual Meeting (January 22-26, 2012)

Wednesday, 25 January 2012: 5:15 PM
Wind and Solar Power Forecasting: Using Multi-Model and Refined Ensemble Prediction Output Approaches
Room 345 (New Orleans Convention Center )
Robert Mureau, MeteoGroup, Wageningen, Netherlands; and D. Malda and H. van den Brink

Probability forecasts are of high importance for the wind energy market. The wind power forecasts of MeteoGroup are provided with uncertainty ranges to be able to show the reliability of every forecasted value. Shutdowns of production with very high wind speed as well as rapidly increasing or decreasing wind speeds (ramp events) are of particular interest.

The wind (power) forecasts with its uncertainty ranges are developed by optimal usage of the ECMWF ensemble forecast in combination with a multi-model approach to obtain the optimal forecast. Forecasts are calculated for speed, wind direction and for power production. If available, real-time and historical observational data are used to optimize the forecast and improve the quality. Not only the optimal forecast, but also the uncertainty ranges can be improved via statistical means when observations are available. The method can be applied to individual turbines, wind parks or an entire portfolio.

Besides wind (power) forecasts, solar power forecasts are also developed at MeteoGroup. Radiation forecasts and other parameters from several NWP models are used to create a power forecast of high quality. NWP model radiation output is adapted and optimized for every single solar farm with distinction in direct and diffuse radiation, taken into account the way solar panels directed to the sun. When power observations are available, statistical MOS techniques are used to improve the power forecast. Satellite images are used to adapt the forecast of the first few hours, but also to create a set of historical radiation data.

In the presentation, an overview will be given of the latest developments in wind and solar power forecasting at Meteogroup.

Supplementary URL: