Optimal Dispatch Scheduling of a Wind-battery-System in German Power Market

Due to the principle of supply and demand, the real-time electricity price in the German spot market varies throughout a single day. Since wind power is non-controllable and partially unpredictable, it is difficult to schedule its power output. Suppliers lose profits by not taking full advantage of the price variation. Moreover, the forecast errors can represent a financial risk, in case of the provision of reserve energy. In this paper, an approach is presented where a battery energy storage system (BESS) is used to make wind power plants (WPPs) scheduled. First, the BESS is used to adjust the dispatch plan of wind power output, in order to exploit price variations beneficially in day-ahead and intraday markets. Second, the BESS is applied to address forecast errors during the real-time operation, in order to balance particularly the expensive forecast errors, which can endanger the stability of the power system. Deviations between forecast and real power output are characterized as expensive forecast errors, if the payment for the deviations is more than 50 €/MWh. In order to realize a multiuse of the battery, a genetic algorithm is employed to optimize the portion of power and energy capacity for the BESS, which participates in above-mentioned different energy market auctions. In contrast, for the daily operation strategy of the BESS, an hourly-discretized linear optimization algorithm is employed. The wind power forecasts for a pool of wind farms with an overall nominal power of 238 MW are generated with WEPROG's (WEPROG GmbH, Wetter & Energie PROGnosen, Böblingen, Germany, http://www.weprog.com/.) multi-scheme ensemble prediction system (MSEPS). In addition, the measured data of the wind farm pool were also available. Time-series data of electricity prices for the German/Austrian area are taken from EEX European Energy Exchange AG, Leipzig, Germany, https://www.eex.com/de/. and used partially for the optimization and verification. The results show that, by applying the proposed method, financial benefits are achieved for the wind farm and the cost caused by forecast errors can be decreased.