Operations and Planning in Sustainable Power Systems

Abstract

Renewable generation, e.g., wind power and photovoltaic resources, and non-traditional energy providers, e.g, flexible loads and energy storage, have demonstrated their potential to eliminate or, at least, to alleviate dependency on costly and depletable energy resources, as well as to reduce gas emissions, thus fostering a transition toward a sustainable power sector. As result of this transition, future power systems will require new decision-making tools that would rigorously account for unique features of the renewable generation and non-traditional energy providers to adopt these means at socially acceptable costs. However, operational and long-term planning decision-making tools for power systems have not kept pace with this dramatic growth in renewable generation. Hands-on experience in real-life power systems has revealed the inefficiency of these tools and demonstrated the need for an overhaul of the current approaches to power system operation and planning. This dissertation examines existing approaches to account for the stochasticity of renewable generation in short-term planning tools and proposes two new unit commitment models based on stochastic and interval optimization techniques. These models are demonstrated to maintain acceptable levels of reliability and reduce the system-wide operating cost, as well as to increase utilization of available renewable generation. Furthermore, this dissertation presents a new framework that enables participation of renewable generation in providing ancillary services, e.g., active power reserve, that facilitates higher penetration levels of this generation. This dissertation describes a new bilevel model that determines the optimal location and size of merchant storage devices to perform the spatiotemporal energy arbitrage. This method aims to simultaneously reduce the system-wide operating cost and the cost of investments in ES while ensuring that merchant storage devices collect sufficient profits to fully recover their investment cost. This model is used to demonstrate that existing power system with perspective renewable generation portfolios will have sufficient profit opportunities to install merchant storage.