Energy management system for optimal operation of microgrid consisting of PV, fuel cell and battery / Shivashankar Sukumar

Operating a microgrid with intermittent source such as photovoltaic (PV) introduces uncertainty in its operation. Therefore, it is necessary to manage the energy from this type of distributed energy resources (DERs) to optimize it usage in optimal manner. In this work, an energy management system (EMS) is proposed for a grid connected microgrid where the DERs is able to supply the local load based on directives provided by an EMS and minimize the power oscillations caused by PV plant. The proposed EMS is for grid connected microgrid consisting of solar PV plant, battery energy storage system (BESS) and fuel cell. In this work, a novel ramp-rate control strategy is proposed where the energy storage is used to control the ramp-rate of PV output power within the desirable level. In addition, a novel ‘mix-mode’ operating strategy is proposed to reduce the microgrid’s daily operating cost. The objective functions in the proposed strategies are solved using linear programming and mixed integer linear programming. Since battery size influences operating cost of microgrid, a sizing method to determine optimal energy capacity of BESS in kWh is also proposed. The BESS sizing problem is solved using grey wolf optimizer (GWO), particle swarm optimization (PSO), artificial bee colony (ABC), gravitational search algorithm (GSA), and genetic algorithm (GA). It was found that GWO presents the optimal solution. The proposed EMS module integrates the proposed ramp-rate control strategy and mix-mode operating strategy and solved using receding horizon economic dispatch (RHED) approach. Comparison on solving the energy management problem using RHED approach with metaheuristic methods like PSO and evolutionary programming (EP) was also conducted and found that RHED approach provides the optimal solution with less computational run time. The results indicate that the proposed EMS can reduce the operating cost by 10.2% when compared with metaheuristic methods