Date of Award

8-2018

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Department

Industrial Engineering

Committee Member

Dr. Sandra D. Eksioglu, Committee Co-Chair

Committee Member

Dr. Scott J. Mason, Committee Co-Chair

Committee Member

Dr. Russell Bent

Committee Member

Dr. Burak Eksioglu

Abstract

An electric power system is a network of various components that generates and delivers power to end users. Since 1881, U.S. electric utilities have supplied power to billions of industrial, commercial, public, and residential customers continuously. Given the rapid growth of power utilities, power system optimization has evolved with developments in computing and optimization theory. In this dissertation, we focus on two optimization problems associated with power system planning: the AC optimal power flow (ACOPF) problem and the optimal transmission line switching (OTS) problem under geomagnetic disturbances (GMDs). The former problem is formulated as a nonlinear, non-convex network optimization problem, while the latter is the network design version of the ACOPF problem that allows topology reconfiguration and considers space weather-induced effects on power systems. Overall, the goal of this research includes: (1) developing computationally efficient approaches for the ACOPF problem in order to improve power dispatch efficiency and (2) identifying an optimal topology configuration to help ISO operate power systems reliably and efficiently under geomagnetic disturbances. Chapter 1 introduces the problems we are studying and motivates the proposed research. We present the ACOPF problem and the state-of-the-art solution methods developed in recent years. Next, we introduce geomagnetic disturbances and describe how they can impact electrical power systems. In Chapter 2, we revisit the polar power-voltage formulation of the ACOPF problem and focus on convex relaxation methods to develop lower bounds on the problem objective. Based on these approaches, we propose an adaptive, multivariate partitioning algorithm with bound tightening and heuristic branching strategies that progressively improves these relaxations and, given sufficient time, converges to the globally optimal solution. Computational results show that our methodology provides a computationally tractable approach to obtain tight relaxation bounds for hard ACOPF cases from the literature. In Chapter 3, we focus on the impact that extreme GMD events could potentially have on the ability of a power system to deliver power reliably. We develop a mixed-integer, nonlinear model which captures and mitigates GMD effects through line switching, generator dispatch, and load shedding. In addition, we present a heuristic algorithm that provides high-quality solutions quickly. Our work demonstrates that line switching is an effective way to mitigate GIC impacts. In Chapter 4, we extend the preliminary study presented in Chapter 3 and further consider the uncertain nature of GMD events. We propose a two-stage distributionally robust (DR) optimization model that captures geo-electric fields induced by uncertain GMDs. Additionally, we present a reformulation of a two-stage DRO that creates a decomposition framework for solving our problem. Computational results show that our DRO approach provides solutions that are robust to errors in GMD event predictions. Finally, in Chapter 5, we summarize the research contributions of our work and provide directions for future research.

Share

COinS