Behavioral Modeling of Electric Vehicles Using Price Elasticities

In the future, electric vehicles (EVs) are assumed to assist the day-ahead electricity market (DEM) operators in the form of large-scale controllable and flexible loads. For example, EVs can charge or discharge in response to the DEM prices. Therefore, an important problem to address is to mathematically represent the flexibility (i.e., extent of the hourly demand shifting and/or curtailment) offered by the EVs to the DEM operator over the next day. In this paper, an electric vehicle aggregator (EVA) uses a three step approach to model the flexibility offered by the EV consumers at different load buses in a transmission network using convex programming. First, the 24-hour reference response values (demand/supply) are determined using the data of the EV consumers and some chosen reference price. Second, the optimal 24-hour aggregate response of the EVA (i.e., new hourly demand/supply values) corresponding to each randomly perturbed scenario of the day-ahead prices is determined. Third, using the reference values and the different scenario specific response values, the price elasticities are determined. These price elasticities act as indicators of the flexibility offered by the EV consumers. The results demonstrate the effectiveness of the proposed approach besides providing insights into the degree of demand shifting that can occur across various DEM price scenarios.