Power System Bifurcation Detection Using Statistical Early Warning Signals With Synchrophasor Data

The increasing deployment of Synchrophasor or Pha-sor Measurement Units (PMUs) across North America helped utilities and system operators enhance monitoring and better power system observability. PMUs can provide synchronized, high-resolution data on voltage and current phasors, enabling real-time situational awareness and control of the power system. This widespread deployment of PMUs has helped system monitoring, offering utilities unprecedented insight into grid dynamics like system phase changes or bifurcations. One of the critical challenges posed by this proliferation of PMUs is the vast volume of data generated, necessitating advanced data analytics techniques for effective decision-making. However, data analytics needs to be done faster and computationally efficiently to implement the developed algorithm in a utility grid control room. In this research, the authors proposed a data-driven method to analyze power system bifurcations with PMU data using statistical metrics known as Early Warning Signals. Generally, bifurcations can lead to voltage instability, frequency fluctuations, and even blackouts. Hence, timely detection of bifurcations and types, both pre-event if possible or post-event for detailed forensics with the proposed data-driven method, helps utilities identify critical patterns and correlations in PMU data, allowing for proactive grid stability and reliability management.