Optimal Aggregator Bidding Strategies for Vehicle-to-Grid

Electric vehicles offer many benefits over traditional vehicles such as lower operating costs and the potential to run on locally produced renewable energy. However, EVs cost more to purchase and mass unregulated charging of EVs can cause energy shortages in the power grid and distribution system overloads. Vehicle-to-Grid (V2G) has been proposed as a way to increase the adoption speed of Electric Vehicles (EVs). V2G is defined as the provision of energy and ancillary services from an EV to the electricity grid. Power flow in V2G can be either unidirectional or bidirectional.;In this dissertation optimal bidding strategies for V2G are developed for use by aggregators. Since unidirectional V2G is the logical first step in V2G implementation, algorithms for it are developed first. These formulations account for unplanned EV departures during scheduling and the necessary corrections. Simulations show the benefits of these optimal algorithms for aggregators, utilities, and customers over using simpler smart charging algorithms. The effect of real world driving efficiencies on V2G scheduling is explored. The necessity of aggregated bidding in the near term is also addressed.;Some of the negative impacts of EV charging on the distribution system are increased losses, poor voltage profiles, and feeder overloads. One way to mitigate these is through coordinated charging with the objective to minimize losses. Feeder load factor and load variance objectives are formulated and shown to be approximately equivalent to minimizing losses. The load factor objective is formulated as a constraint for a profit maximization problem. Integrating this constraint reduces feeder losses and eliminates line overloads and unacceptable voltage sags when compared to unconstrained V2G scheduling.;Bidirectional V2G algorithms are formulated since bidirectional V2G has the added utility of distributed energy storage. Ancillary services dispatch algorithms are developed and an optimal V2G scheduling algorithm is formulated. Simulations show that the aggregator receives significant profits and the system peak load is reduced when scheduling in this manner. Customer costs of charging are also extremely low.