Analysis of the residual two-layer extrapolation method for aggregated wind power estimation

In order to operate power grids with high penetration of renewable energies but low coverage of real time measurements of wind and solar power plants, extrapolation methods to derive the total wind power or PV generation for large regions are indispensable. Primary requirements for extrapolation techniques are robustness in operation against missing or wrong data and flexibility regarding aggregation to different grid areas like control zones or grid nodes. The method used at Fraunhofer IEE, here called standard method, is in particular suitable for extrapolation with a low coverage of power plants with available real time measurements. The gradually increasing number of available measurements in Germany allows new approaches of the extrapolation method. We present a two-layer extrapolation method, called residual extrapolation, which aggregates the measurements or forecasts of available power plants directly within a grid area of known capacity and only extrapolates the residual capacity. In this paper we are investigating the performance of the residual extrapolation method compared to the standard method according aggregation zone, effective radii for extrapolation weights and artificial/simulated changes of available measurement coverage within the aggregation levels.