Smart management of integrated energy systems through co-optimization with long and short horizons

Abstract

The integration of all sectors of energy production, distribution and consumption in multi-source energy networks has lately gained attention as an attractive strategy to deal with the challenges raised by decarbonization roadmaps. For such a network to become a smart energy system, however, it needs to be managed and controlled in a smart way. While existing techniques mainly focus either on short-term unit commitment or on yearly scheduling separately, this work presents an original combined optimization algorithm which merges the two methods, in order to enhance system real-time control with long-term evaluations (e.g. incentives and yearly constraints). The control architecture comprises three coordinated optimization levels, each periodically updated through the receding time horizon strategy. A long-term supervisory module performs whole-year optimal scheduling accounting for long-term factors and determines the constraints for a short-term supervisory module which, in turn, optimizes the control action for the energy production system in real-time. In parallel, energy distribution modules minimize energy supply to the different portions of the distribution network downstream. Simulation results on a hospital case study demonstrate a 9.7% reduction in total operating cost over the whole year, as well as an increase in revenues deriving from incentives for high efficiency cogeneration.