An energy management system (EMS) is formulated for a hybrid power plant (HPP) participating in bidding stages in the German energy market. The stochastic optimization within the EMS aims to maximize the HPP economic profit by balancing the revenue obtained by market bidding and the cost of battery cyclic damage, while explicitly considering the uncertainties associated with generation and price forecasts. The formulation utilizes data-driven probabilistic models trained using site-specific historical measurements, as well as a novel online cycle counting approach. When compared with standard formulations, the proposed approach shows an accurate estimation and balancing of revenue and costs and a significant reduction in the power deviation penalty, which leads to significantly higher overall profit.
Keywords: Economic dispatch, Battery degradation, Online cycle counting, Power tracking,
Hybrid power plant control, Probabilistic forecasting
Link to paper: https://zenodo.org/records/15005498
