A new method for day-ahead sizing of control reserve in Germany under a 100% renewable energy sources scenario

In Germany, the installed capacity of renewable energy sources, especially that of wind and photovoltaic energy, has increased over the past few years and will continue to increase in the future. Due to errors in forecasting wind and photovoltaic energy, the control reserve needed to balance the electricity system will correspondingly increase if control reserves will be sized statically for several months or one year as it is done in most countries today [1-3]. That is because sizing control reserves this way does not consider the fact that there will be hours with a high penetration of wind and photovoltaic which cause a different demand for control reserves than hours with a lower penetration. Therefore, in this work, we present a new probabilistic dynamic method that sizes control reserves for the single hours of the following day making use of forecasts of the power feed-in of wind and photovoltaic. In contrast to similar approaches [2,3] forecast errors of wind and photovoltaic power are not modeled as normal distributions, which does not reflect reality [4-6], but by kernel density estimation to get more realistic distributions. Under a 100% renewable energy scenario for Germany, the control reserve that would be allocated by the dynamic method is compared with the control reserve that would be allocated by a static method. The static method is similar to the probabilistic Graf-Haubrich method, which is applied in Germany today, but can, in contrast to this method, be applied to future scenarios. It is shown that the dynamic method halves the average required control reserve.