Control and Optimization of Future Electric Grid Integrating Plug-In Electric Vehicles and Wind Power

This dissertation studies the integration and control problems that will arise when large numbers of plug-in electric vehicles (PEVs) and wind power are introduced to the electric grid. Various control and optimization techniques are developed in this dissertation to harnesses the synergy between PEVs and wind power to facilitate the grid operations.;First, a PEV charging control algorithm is developed to utilize the idle generating capacity in evening hours to charge of the newly introduced PEVs on the future Michigan grid. The control algorithm adopts a partially-decentralized structure, so that its implementation does not require excessive computation and communication. At the global level, a SOC threshold command is calculated and broadcasted to all PEVs as the basis of charging level. At each charger, two attributes of individual PEVs, the battery state of charge and plug-off time, are considered to calculate the final charging power. The proposed algorithm allows most PEVs to be fully charged. In the meantime, the grid-level objective "valley filling" is achieved. The algorithm also includes a feedback mechanism to regulate grid frequency to explore the potential of manipulating PEV charging to replace conventional reserves in the valley hours.;Secondly, this dissertation investigates means to mitigate wind power intermittency. Model predictive control (MPC) is used to control the charging and discharging of battery energy storage system (BESS) to provide reserves. Unlike existing MPC studies that focused on state tracking or output regulation, realistic objective functions that capture the reserve costs to cover wind surplus or deficit are used. The effect of BESS capacity sizing is also investigated.;Thirdly, to accommodate both PEVs and wind power on the grid, a hierarchical control algorithm is proposed. The control algorithm has three levels. The top-level controller solves a scheduling optimization problem to minimize the grid-wide cost of electricity generation. The middle- and bottom-level controllers are based on the control algorithms previously developed for PEV charging and wind power scheduling. The hierarchical structure allows the features in the different control algorithms to be preserved.;Next, a carbon disincentive policy is proposed to promote the use of low-carbon power plants for electricity generation to reduce grid CO 2 emissions. The proposed policy can be used to adjust the carbon content in the generation mix, and the tradeoff between the generation costs and grid CO2 emissions is investigated. Analyses show that introducing wind generation can significantly reduce the electricity generation costs, but not grid CO2 emissions if no PEVs are available to mitigate wind intermittency. To address both the generation costs and CO2 emissions, manipulations in both the supply and demand on the grid are needed.;Lastly, the generation planning problem is studied. A systematic methodology is proposed to evaluate the cost of constructing different types of generating capacities. The methodology considers the evolutions in both the supply and demand of the electric grid, including annual increases in the grid load and changes in the merit order when new power plants are commissioned. Furthermore, the renewable intermittency and reserve-related costs are also considered, which are new features not seen in the literature. Based on the used assumptions, the cost evaluation identifies the construction cost as the bottleneck that prevents wind power from entering the market, although the wind intermittency can be addressed by BESS or PEVs on the operation stage.;The modeling and optimization framework developed in this dissertation makes it possible to study the synergy between PEVs and wind power on the electric grid. Simulation results show that PEVs and wind power are complementary to each other, and a proper integration is needed to realize their full potential.