Economic model predictive control for the energy management problem of a virtual power plant including resources at different voltage levels

In response to the requirement of a sustainable energy system, diversified energy resources and liberalization of electricity markets, the energy sector is experiencing worldwide a huge penetration of Distributed Energy Resources (DER). To maximize the benefits of these assets, DER can be aggregated in a Virtual Power Plant (VPP) and operated as a single system. In this work we consider a VPP given by the aggregation of a cascade of hydropower stations (CHPS) connected at the High-Voltage (HV) grid and integrating a large portfolio of variable Renewable Energy Sources (VRES) connected at Medium-Voltage (MV) grids. Then, we tackle the problem of VPP profit maximization on the joint energy and ancillary services market, under complex technical constraint, safety constraints and unavailability of VPP resources due to faults. First, we propose a generic model of the VPP. Second, we present a two-level sequential VPP energy management strategy composed by long-term bidding optimization and real-time control via Economic Model Predictive Control (EMPC), both receiving forecast as input. Simulations employ realistic models and real forecast provided by the French aggregator Compagnie Nationale du Rhône (CNR). Compared to traditional Reference Tracking MPC (RTMPC), the EMPC increases by 6% the VPP profit and enhances the provision of ancillary services when faults occur.