Detecting and Analyzing Frequency Events in Power Systems Using Tunable Parameters-Based Algorithms: Development, Optimization, and Analysis

Abstract

The rapid growth in the integration of renewable energy sources into power grids has driven a transition from conventional thermal-based generation to inverter-based resources. As a result, power system inertia has decreased, and the rate of change of frequency has increased. This presents a challenge for frequency stability in modern power systems. Power systems disturbances, such as significant faults or major disruptions in generation or load, cause imbalances between power supply and demand, which may result in severe frequency fluctuations known as frequency events. Following such events, fast frequency response is needed to provide frequency support and prevent system collapse. Therefore, monitoring and detecting frequency events promptly and accurately is critical to stabilizing power systems. This dissertation addresses the challenge of detecting frequency events in diverse power systems by enhancing existing frequency event detection methods through detection process modifications and developing unique tunable parameters. Since system characteristics differ across regions, frequency event detection algorithms must be customized by domain experts for each balancing area using tunable parameters. By optimizing these parameters for specific power system, the algorithms can accurately detect frequency events and can also be used for further analysis to determine trends in frequency events over time, ensuring system stability. This dissertation focuses on the enhancement and optimization of frequency event detection algorithms. These detection algorithms are compared with other state-of-the-art frequency detection methods. The study examines the impact of signal denoising techniques on detection accuracy, analyzes frequency performance over time, reviews global frequency performance standards, and conducts comprehensive sensitivity analyses. The five primary contributions of this dissertation are: the development of frequency event detection algorithms with tunable parameters for specific balancing areas; optimization of the developed algorithms parameters to enhance results and adaptability, conducting a comprehensive analysis of signal denoising methods and their impact on frequency event detection; the proposal of criteria-based tunable parameters to assess frequency events trends and severity; presentation of an enhanced understanding of global frequency performance standards; and deeper insights into frequency specifications across diverse power systems.