A Microgrid Model With EV Demand Uncertainty and Detailed Operation of Storage Systems

New entities and ways of operating electrical power systems have emerged. An example of this is the microgrid (MG), which can be described as a set of power generation units, energy storage systems (ESSs) and demands that can operate as a single set, in order to optimize their energy resources. Moreover, the recent growth in the use of electric vehicles (EVs) in transportation systems has created new challenges to electrical power systems. This article proposes a two-stage stochastic energy scheduling model for an MG to set up the day-ahead optimal decision in the first stage. Real-time operations, depending on wind and photovoltaic solar power, baseload demand and EV demand variability, are taken into account in the second stage. The proposed model considers conventional generators and alternating current (AC) linearized power flow constraints. Besides, a detailed ESS operation is considered, including degradation costs, and nonlinear charging/discharging as well as efficiency variation dependent of state of charge (SOC), for lead-acid and lithium-ion technologies. EV demand behavior is generated by probability-based algorithms that use the Monte Carlo method to perform their simulations, which are based on data of initial SOC, plugin/out times, type of EVs, and chargers. Finally, the proposed model is successfully validated with an IEEE-37 bus system. The results support the superiority of the proposed model over existing works, showing significant variations in battery bank (BB) operation. The computational time is similar to previous works, despite adding new constraints that increase the complexity.

Automated summary

This article proposes a two-stage stochastic energy scheduling model for optimizing the energy resources of microgrids. The model considers real-time operations based on wind and photovoltaic solar power, baseload demand, and electric vehicle demand, and is validated on an IEEE-37 bus system.