Optimal demand side management in networks of multi-energy prosumers

Recently, energy systems have evolved from fully centralized to decentralized systems, in which end-users become prosumers, namely customers equipped with multi-energy systems, that are, at the same time, energy producers and consumers. This emerging paradigm, originally intended for the electrical sector, can bring benefits also if extended to district heating networks, as it increases the penetration of local renewable sources and waste heat, thus improving energy efficiency. In addition, prosumers can operate Demand Side Management (DSM), i.e. that heat can be stored in the building heat capacity or taken back when considered more efficient. However, distributed prosumers highly interact with the energy network, and the optimal operation of an individual sub-system may not correspond to the most profitable operation of the global system. Hence, the optimal management of these systems constitute a large-scale problem with significant computational challenges. This work presents a novel strategy for concurrent management of networks of multi-energy prosumers, achieved through a distributed optimization architecture. The strategy is enhanced with a heuristic iterative procedure which proposes various DSM actions and verifies if they lead to a more efficient solution for the global network. The first evaluation of the method on a University Campus leads to 3.6% reduction in global cost and 30% reduction of the thermal peak, which allows to avoid the use of peak boilers.