Optimizing the size of a V2G parking deck in a microgrid

Microgrids are usually powered partially and sometimes fully by small scale distributed generators, renewable energy resources and batteries. The batteries of the electric vehicles (EVs) in a parking deck can also be used as a power source of the microgrids. However, unlike fixed batteries, the storage capacity of a parking deck depends on the number of installed V2G (vehicle to grid) stations and the number of available EVs. In this paper, we propose a mathematical model that aims to determine the optimal number of V2G stations in a microgrid parking deck by minimizing the total cost of investment and operation. The EVs in the parking deck are discharged to shift energy purchases from high-price to low-price hours and vice versa. The uncertainty of the demand and the EVs arrivals and departures and the intermittency of the renewable resources are taken into account. We develop a methodology to compute the operation cost of the system using Benders' decomposition and search for the optimal number of V2G stations using the Nelder-Mead heuristic algorithm. Numerical experiments are conducted on a 14-bus microgrid test system for a planning horizon of five years. We also assess the effects of uncertainties, EVs battery size, chargers' rated capacity, EVs arrivals and departure rates, dispatch constraints and interest rate on the determined number of V2G station and the discounted payback period. The experimental results show that investing in the V2G technology can reduce the long-term cost of electricity supply for the microgrid.