Definition and on-field validation of a microgrid energy management system to manage load and renewables uncertainties and system operator requirements

The present paper proposes an Energy Management System (EMS) to be used in grid connected microgrids. To do this, first a suitable model (for optimization purposes) of all the components that typically appear in a microgrid is presented, then, four possible electric network models are detailed and finally the overall architecture of the optimization problem of the EMS is set up. Moreover, as the optimization of the energy production/consumption of a microgrid relies on the thermal and electric load and on the renewables forecasting, an online empirical correction of forecasted data is proposed, highlighting its positive impact on the overall operational cost of the microgrid. Another aspect which is accounted regards the possibility of allowing the EMS to act as a power plant controller in compliance with the Distribution System Operator (DSO) requirements in terms of reactive power management and voltage control (as requested by the majority of grid codes and national regulations). So, the proposed algorithm structure splits the optimization problem into two sub problems: the first one basically dictates the active power production of the dispatchable units minimizing an economic objective function, while the second accounts for the satisfaction of the DSO requirements. The experimental validation of the proposed EMS is performed on the University of Genoa Smart Polygeneration Microgrid (SPM), where the proposed EMS is currently running. The considered test cases highlight the effectiveness of the proposed EMS in economically operating the microgrid (compared to simpler EMS previously installed in the SPM) and in satisfying the reactive power or voltage regulation requirements provided by the Italian technical requirements.