Measurement-based techniques for optimization and control of power distribution systems using distributed energy resources

Widespread integration of distributed energy resources (DERs), such as solar photovoltaic, wind, and storage systems helps to alleviate environmental concerns of conventional fossil fuel-based generation. Moreover, DERs can provide valuable benefits to the reliable and efficient operation of the integrated power grid when dispatched accurately and timely. Furthermore, emerging advanced sensors provide vast amounts of data, which enables new opportunities in the operation of power grids. These trends have motivated recent research into developing techniques to dispatch DERs for optimization and control of power distribution systems. Existing research involves repeated solutions of optimization problems constrained by the nonlinear network power balance, DER capacity limits, and other operational limits. Such methods may be computationally burdensome in practical implementations, especially with rapidly varying operating points. Moreover, many require an accurate and up-to-date network model that may not be available in real time, resulting in DER setpoints that lead to unexpected or undesired system behaviour. In this thesis, the overarching goal is to develop new techniques to optimize and control modern power distribution systems with high penetrations of DERs by leveraging real-time measurements provided by advanced sensors such as distribution-level phasor measurement units. Specifically, the research focuses on (i) centrally dispatching optimal DER active- and reactive-power outputs by recursively estimating sensitivity models relating bus voltages and power injections, (ii) leveraging distributed estimation and optimization methods to alleviate the scalability concerns of the centralized measurement-based DER dispatch method, (iii) capitalizing on system-wide optimization and local measurements to regulate system voltages in cases with highly intermittent renewable generation, and (iv) minimizing network power losses and cost of DER reactive-power outputs based on a measurement-based gradient-descent method. Overall, the proposed research advances the efficient and reliable operation of modern power distribution systems and facilitates the widespread integration of DERs by using measurement-based techniques. As a consequence, this work contributes to transitioning toward a more environmentally-friendly electric power supply by enabling higher penetrations of renewables.Applied Science, Faculty ofEngineering, School of (Okanagan)Graduat