A decentralised energy trading architecture for future smart grid load balancing

Current state-of-the-art electric vehicle charging is found to have a profoundly disruptive effect on decentralised grids, increasing prevailing peak demand and causing network congestion. However, when charging behaviour is aligned with the needs of the grid, the batteries of electric vehicles can be used as a distributed resource to provide ancillary services. This paper proposes an decentralised algorithm that is capable of exposing the benefits of an electric vehicle fleet to grid system operators, taking the user preferences of the individual owners into account and keeping the application lightweight through a decentralised architecture. The algorithm is implemented in an agent-based model based on real Dutch smart metering data. The architecture is shown to decrease local imbalances, offer financial incentives to electric vehicle owners and maintain a minimum state-of-charge at departure for individual system users.