Real-time tracking of electromechanical oscillations in ENTSO-e Continental European Synchronous Area

Abstract

Small signal stability is a crucial aspect of accurately keeping control in modern interconnected power systems, in order to ensure the required security and reliability standards. Such an aspect could represent a serious limiting factor in the search for ever higher power system exploitation levels. Hence, in order to exploit the existing networks as much as possible, ensuring at the same time adequate security levels, the system operators must monitor the electromechanical oscillations continuously, estimating their fundamental parameters (e.g. frequency, damping factor/ratio, amplitude and phase), and determining the actual dynamic stability margins. Nowadays, electromechanical oscillations tracking can be performed in real time thanks to the quick development of wide area measurement/monitoring systems, and the employment of so-called measurement-based methods (a.k.a. mode meters). In this regard, the paper presents a performance comparison among several measurement-based methods currently operating on the Italian Wide Area Monitoring (WAM) platform. The main objective of this investigation is to analyze the behavior of each of them with respect to actual critical/not critical electromechanical oscillations recorded on the European Network Transmission System Operators for Electricity (ENTSO-e) Continental Synchronous European System (CESA). The considered case studies cover a large variety of situations that can be encountered in the ordinary ENTSO-e CESA operation: local and inter-area oscillations; single mode and multi-mode power signals; ambient and ringdown data; and so on. By keeping in mind that usually no best estimator exists, owing to the lack of an optimality definition, the present analysis aims at examining whether some algorithms are generally more suitable than others, with respect to the variety of aforementioned situations. In detail, the measurement-based methods considered in this work are the Tufts–Kumaresan method, Hankel Singular Value Decomposition-VarPRO, Hilbert–Huang Transform, its refined Empirical Mode Decomposition with Fourier-Based Masking Technique (R-EMD), and a recently published Hilbert Transform-based estimation algorithm. Obviously, this set of estimation techniques cannot be considered exhaustive for the research problem under analysis. Nonetheless, it represents a collection of the major, recent and most fashionable methodologies implemented in actual WAM architectures.