| Summary: | Unstable electromechanical dynamics, or electromechanical modes, may cause large amplitude oscillations that may lead to system break-ups or partial blackouts. Thus, the monitoring and study of these modes are of prominent importance and can be of great help for system operators and planning engineers. In this thesis ambient data analysis has been applied on both simulated and synchronized phasor data from Phasor Measurement Units (PMUs) installed at the distribution network of the Nordic power system at Tampere, Lund and Luleå to estimate frequencies and damping ratios of electromechanical modes in the systems. Different spectral estimators (Yule-Walker, Multitaper and Welch) have been used and their performance has been evaluated. Damping estimates were obtained using an autoregressive Yule-Walker model and the half-power point method. Emphasis on general handling and preprocessing of PMU data is made throughout the thesis. The performed analysis indicates that within the measurement locations available, two main inter-area modes in the Nordic power system can be estimated at approximately 0.4 Hz and 0.5 Hz. In addition relevant system dynamics in the frequency range of 0.6-1.0 Hz can also be observed. Ambient data analysis techniques have great potential for monitoring electromechanical oscillations in power systems. However, there are issues related to data quality that need to be systematically addressed, especially when it comes to calculating accurate damping estimates, in the presence of undamped low amplitude sinusoids and forced oscillations.
|
|---|
