Vehicle-to-Grid Fleet Service Provision Considering Nonlinear Battery Behaviors

The surging adoption of electric vehicles (EVs) calls for accurate and efficient approaches to coordinate with the power grid operation. By being responsive to distribution grid limits and time-varying electricity prices, EV charging stations (EVCSs) can minimize their charging costs while aiding grid operation simultaneously. In this study, we investigate the economic benefit of vehicle-to-grid (V2G) using real-time price (RTP) data from New York State and a real-world charging network dataset. Our V2G approach incorporates nonlinear battery models and price uncertainty to provide a realistic estimation of cost savings from different V2G options. The proposed control method is computationally tractable when scaling up to real-world applications. We show that our proposed algorithm leads to an average of 21% charging cost savings compared to uncontrolled charging (UC) when considering unidirectional charging, and bidirectional V2G enables additional 17% cost savings compared to unidirectional smart charging. Our result also shows the importance of using more accurate nonlinear battery models in V2G controllers and evaluating the cost of price uncertainties over V2G.