An Offline Approach for Frequency Event Detection in Power Systems: TKEO and Statistical Analysis

The increasing integration of renewable energy sources in power grids and the transition from conventional thermal-based generation to inverter-based resources for power generation have reduced power system inertia and increased the rate of change of frequency, which constitutes a challenge for frequency stability in modern power systems. These changes necessitate robust frequency event detection algorithms that can rapidly and accurately identify events, providing essential support to maintain stability. This paper proposes a frequency event detection algorithm with six tunable parameters for offline post-processing, leveraging the Teager-Kaiser Energy Operator method and statistical analysis. The algorithm is tested on a dataset comprised of Phasor Measurement Unit data from the Western Interconnection of the USA. Detection performance is evaluated against another algorithm with four optimized parameters, using key metrics and binary classification to assess the accuracy and reliability of the featured algorithm. The results demonstrate solid performance of the proposed method in detecting frequency events, showcasing its potential and paving the way for further parameter optimization.