Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering

First Advisor

Mads Almassalkhi


The research objective of the proposed dissertation is to make best use of available distributed energy resources to meet dynamic market opportunities while accounting for AC physics of unbalanced distribution networks and the uncertainty of distributed solar photovoltaics (PV). With ever increasing levels of renewable generation, distribution system operations must shift from a mindset of static unidirectional power flows to dynamic, unpredictable bidirectional flows. To manage this variability, distributed energy resources (DERs; e.g.,solar PV inverters, inverter-based batteries, electric vehicles, water heaters, A/Cs) need to be coordinated for reliable and resilient operation. This introduces the challenge of coordinating such resources at scale and within confines of the existing distribution system. It also becomes important to develop efficient and accurate models of the distribution system to achieve desired operating objectives such as tracking a market reference, reduction in operation cost or voltage regulation. This work surveys, discusses the challenges and proposes solutions to the modeling and optimization of realistic distribution systems with significant penetration of renewables and controllable DERs, including energy storage. To contain this increase in system complexity as result of the large number of controllable DERs available, the distribution system has to be adapted from a passive Volt-Var focused operator to a more active manager of resources. To approach this challenge, in this work, we propose two main approaches. The first is a utility centric approach, where the utility controls the dispatch of flexible resources based on solving an optimization problem. This approach would require the utility to have all the network and resource data and also the control over customer devices. Another approach is a more aggregator centric approach, where an aggregator is an entity that represents an aggregation of many diverse DERs or a Virtual Battery (VB). In this approach, it is the role of the aggregator to dispatch DERs, whereas the utility provides certain bounds and limits (calculated offline), which the aggregator (which dispatches resources in real-time) must operate under. The benefits of such an approach lie in improved data-privacy and real-time dispatch. We present simulation results validating the proposed methods on various standard IEEE and realistic distribution feeders.



Number of Pages

261 p.