Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)


Electrical Engineering

First Advisor

Mads Almassalkhi


The growing utilization of renewable energy resources (RES) within power systems has brought about new challenges due to the inherent uncertainty associated with RES, which makes it challenging to accurately forecast available generation. Further- more, the replacement of synchronous machines with inverter-based RES results in a reduction of power system inertia, complicating the task of maintaining a balance between generation and consumption. In this dissertation, coordinating Distributed Energy Resources (DER) is presented as a viable solution to these challenges.DERs have the potential to offer different ancillary services such as fast frequency response (FFR) when efficiently coordinated. However, the practical implementation of such services demands both effective local sensing and control at the device level and the ability to precisely estimate and predict the availability of synthetic damping from a fleet in real time. Additionally, the inherent trade-off between a fleet being available for fast frequency response while providing other ancillary services needs to be characterized. This dissertation introduces a fully decentralized, packet-based controller for a diverse range of flexible loads. This controller dynamically prioritizes and interrupts DERs to generate synthetic damping suitable for primary frequency control. Moreover, the packet-based control methodology is demonstrated to accu- rately assess the real-time availability of synthetic damping. Furthermore, spectral analysis of historical frequency regulation data is employed to establish a probabilis- tic bound on the expected synthetic damping available for primary frequency control from a fleet and the trade-off of concurrently offering secondary frequency control. It is noteworthy that coordinating a large number of DERs can potentially result in grid constraint violations. To tackle this challenge, this dissertation employs con- vex inner approximations (CIA) of the AC power flow to address the optimization problem of quantifying the capacity of a three-phase distribution feeder to accommo- date DERs. This capacity is often referred to as hosting capacity (HC). However, in this work, we consider separate limits for positive and negative DER injections at each node, ensuring that injections within these nodal limits adhere to feeder voltage and current constraints. The methodology dissects a three-phase feeder into individual phases and applies CIA-based techniques to each phase. Additionally, new approaches are introduced to modify the per-phase optimization problems to mitigate the inherent conservativeness associated with CIA methods and enhance HC. This includes selectively adjusting the per-phase impedances and proposing an iterative relaxation method for per-phase voltage bounds.



Number of Pages

154 p.