Enhancing Flexibility, Operation, and Control in Modern Electricity Grids Through Robust Data-Driven Methods
Loading...
Authors
Olowolaju, Joshua
Issue Date
2025
Type
Dissertation
Language
en_US
Keywords
Alternative Title
Abstract
The modern electric grid faces numerous challenges, including aging infrastructure, increasing demand, managing more frequent extreme weather events due to extreme climate events, and integrating renewable energy sources to promote infrastructure decarbonization. Furthermore, the growing digitalization of power systems has heightened vulnerability to security threats, complicating operational and planning processes. These changes, shaped by economic, technological, environmental, and political factors, have transformed the traditional grid into a smart grid with more complex power flows and system requirements, making reliable operation, monitoring, and control considerably more challenging. Traditional grid management techniques have become increasingly complex and require improvement to handle the growing interdependency between the grid and other infrastructure, such as cyber, and the increasing uncertainties resulting from variable renewable generation. Leveraging data-driven methodologies, such as machine learning and artificial intelligence, presents promising solutions to assist traditional grid monitoring and control platforms by analyzing complex interactions and behaviors among the grid and other systems. Data-driven approaches offer real-time management capabilities and effective forecasting tools, ensuring the smooth operation and control of energy and electricity systems. Furthermore, as grid-edge resources like distributed energy sources and flexible demand continue to grow, the urgency for increased grid flexibility to handle supply and demand fluctuations has intensified. These resources are inherently flexible, as they can be strategically monitored and controlled to shift electricity generation and consumption, providing ancillary services during peak periods. When managed effectively, flexible resources can offer critical grid services, such as day-ahead generation dispatch and peak load management, improving grid reliability and resilience. This dissertation proposes utilizing machine learning and artificial intelligence to enhance energy systems' operation, monitoring, and control. Given the vast amount of data available in today’s grid and the advancements in computing, these technologies present innovative solutions. Intelligent sensors, like smart meters, produce a wealth of data that facilitates advanced grid monitoring. Machine learning can identify hidden patterns and anomalies within extensive datasets, providing computational efficiency and scalability that surpass traditional methods. These techniques enable near-real-time solutions, enhancing grid reliability and resilience. Also, this dissertation proposes quantifying and coordinating the flexibility of grid-edge resources, evaluating the energy flexibility these resources can offer for grid services from a pricing perspective. In this quantification, grid network constraints are considered, as they are essential to the grid's operation and can significantly affect the reliability and efficiency of the overall energy system. These constraints represent the limitations in the grid’s physical and operational characteristics, influencing energy production, transmission, and consumption. Various customer types and their consumption patterns are examined to assess how quickly and to what degree customers can modify their consumption in response to energy savings incentives or signals. In summary, this dissertation explores leveraging data-driven methods to enhance grid operation, monitoring, and control. It also proposes quantifying and optimizing grid-edge resource flexibility for grid services while considering network constraints and customer consumption patterns.