Increasing grid stability through accurate infeed forecasts of renewable energies

The rapid growth of renewable energies is leading to increasing imbalances between electricity production and consumption. During high-wind periods with local overproduction of energy, the stability of the distribution grid can only be ensured if a fraction of (wind) farms is throttled or shut down. Precise infeed forecasts of wind and solar power production for individual sites assist in reducing these risks because controlling measures for renewables as well as the regulation of conventional power plants (coal, gas etc.) can be planned with longer lead times. A forecast system is presented which includes the whole chain from weather forecast to infeed forecast. The foundation is made up of weather model forecasts from the UK Met Office and from the high resolution regional model WRF. The conversion of forecast weather elements (e.g. wind profiles) into infeed energy from the turbines/farms is realized by artificial neural networks (ANNs). ANNs are able to recognize and eliminate the systematic errors produced by weather models. The quality of the infeed forecasts is validated with results for transmission nodes across the grid of the company E.ON edis, situated in the Northeast of Germany.