Detection and Analysis of Oscillation Events with SCADA-Based Statistical Algorithms

SCADA systems, can delay or hinder the accurate and timely identification of oscillation sources. The abundance of SCADA measurements can enhance system observability in this regard, particularly from the generation side. Although SCADA measurements are reported at significantly lower rates than synchrophasors, the inherently asynchronous polling structure of

SCADA can be leveraged to extract useful information about oscillation in the system.

Specifically, this asynchronous nature enables the estimation and comparison of oscillation amplitudes across different system locations, even in the presence of low reporting rates of

SCADA. In this paper, two rigorous algorithms based on inferential statistics are used to analyze the amplitude of oscillations seen in the real and reactive power measurements in the

SCADA data. These algorithms can operate independently of synchrophasor data and detect high-amplitude oscillatory signals associated with forced oscillation events using SCADA data alone. The amplitude analysis offers valuable insights into the potential source of an oscillation event. Forced oscillations can be of any frequency and the statistical methods can analyze that due to the asynchronous nature of SCADA data. The effectiveness and practical applicability of the proposed algorithms are validated by analyzing several real-world oscillation events from the RTE system, where the methods consistently and accurately identify the high-amplitude signals and the sources of oscillations in all these events.