A bi-level distributed optimization framework to unlock flexibility in grid-connected energy storage systems and electric vehicle fleets

The growing integration of renewable energy sources (RES) into power grids has introduced significant operational variability, amplifying the need for robust flexibility solutions to maintain grid reliability. Demand-side resources, such as flexible loads and electric vehicle (EV) fleets, present cost-effective avenues for balancing supply and demand dynamics. This study proposes a decentralized bi-level optimization framework to enhance the utilization of demand-side flexibility and energy storage systems while ensuring market participant privacy. A Virtual Storage Plant (VSP) model is introduced to coordinate distributed energy storage assets under the supervision of the Transmission System Operator (TSO). The upper-level problem represents the TSO's strategic planning, while the lower-level problem addresses the operation of VSPs, EV parking facilities, and flexible loads. To optimize market interactions and minimize information exchange between the TSO and service providers, an adaptive Alternating Direction Method of Multipliers (ADMM) is employed. The proposed framework is validated using a 30-bus power transmission system, solved through the GUROBI solver within the GAMS environment. The results indicate an 18.7 % reduction in energy balancing costs and a 12 % decrease in transmission losses, alongside a 60 % improvement in convergence speed, demonstrating enhanced coordination, cost efficiency, and privacy preservation. • A bi-level structure to unlock the VSPs, FLs and EVs potential in the balancing market • An adaptive ADMM to improve convergence speed while satisfying market efficiency • Reducing TSO costs and transmission losses, while improving the voltage profile • Reducing thermal units' market share by harnessing demand-side resources' potential