An adaptive MPEC-based hierarchical framework for Co-optimizing virtual power plant profit and battery energy storage system health

A growing number of Distributed Generators (DG) will be part of Virtual Power Plants (VPPs), thus necessitating the ability to optimize profit on markets as well as equipment longevity; especially for battery energy storage systems (BESS). An economic dispatch process can widen the "Dispatch-Control Gap", and accelerate the degradation of a BESS by forcing it into partial-cycles. This paper presents a new hierarchical two-stage procedure that closes this gap. The upper stage utilizes a mathematical program with equilibrium constraints (MPEC) for optimization of the markets, including new monotone constraints to prevent the partial-cycle degradation of the BESS when planning is performed. The lower stage employs an on-line controller to control the operation of the BESS Fleet while controlling degradation. A feedback loop enables the economic model to learn the true cost of degradation from the experience of the controller. The results of simulation over a 90 day period demonstrate the proposed method of reducing the long term degradation of the BESS by approximately 22% relative to traditional methods. This improved sustainability provides a statistical increase to the overall life cycle profitability of the VPP, demonstrating that a health conscious, hierarchical methodology provides superior economics.