As the share of renewable energy sources increases, electricity prices and thus consumption will become more weather-dependent. This may result in significant congestion during cold spells when Combined Heat and Power (CHP) units and heat pumps run near maximum capacity, but also during sunny, windy times when electricity is cheap. Current markets and regulations assume the network to be like a copper plate, obliging network operators to facilitate any market transaction at any cost.
We propose a novel approach to congestion management, with significant lower social costs than current solutions. This approach relies on the development of algorithms that can efficiently schedule flexible loads of multiple customers within the capacity constraints of the local energy networks, exploiting stochastic information on generation and consumption. Additionally, we design rules for a game that customers can play to stimulate the communication of their planned usage and flexibility in shifting loads, in such a way that they cannot gain anything by manipulating this information. For this purpose we develop new planning approaches, combining plans for different horizons and temporal resolutions, and exploiting hierarchy, all under such incentive constraints.
In addition to flexible loads, we include decentralised CHP production in our analysis, extending recent work on multiple energy-carrier optimisation from individual customers to the local distribution grid. These methods are evaluated in simulations based on the distribution network and data from a real case. The results are obtained by a strong inter-disciplinary team of researchers in power engineering, computer science, economy, policy, and management.