Imbalance Minimization in Virtual Power Plants using Industrial Demand Response

Abstract

Virtual Power Plants can aggregate and dispatch distributed energy sources (DER) to gain revenue in the Day-Ahead market, however as a Balancing Responsible Party they can risk imbalance cost due to deviations between the Day-Ahead forecast and actual production of their renewable energy portfolio. This study analyses the ability of industrial processes to minimize these imbalances by flexibly adjusting the energy consumption, which is also known as industrial demand response. Firstly, the advantages and disadvantages of major industries to provide demand response in a short-term redispatch scheme were identified by means of a literature survey. Secondly, the capability of the chlor-alkali and hydrogen production industry to minimize imbalances in a Virtual Power Plant was analysed. A simulation setup was developed in Python of a Virtual Power Plant with photo-voltaics (300 MW), onshore wind (300 MW), a chlor-alkali (203.5 MW) and hydrogen plant (193.4 MW) and a controller based on Model Predictive Control. The MPC integrated industrial process dynamics using data-driven Hammerstein-Wiener models which enabled rescheduling of load consumption without violating process constraints. The results show that industrial demand response provided by the chlor-alkali and hydrogen plant can minimize imbalances significantly between -89% to -99%. Furthermore, a sensitivity analysis revealed the effect of important plant and controller parameters on the imbalance minimization which notably indicated that the storage capacity of hydrogen/chlorine and high utilization rate (>95%) of the industrial plants can be a major limiting factor for minimization of imbalances. The findings of this study are expected to contribute to the development of renewable energy based Virtual Power Plants and the wider participation of industrial processes in distributed energy systems.