Identification of Forced Oscillations in Power Systems
Time: Tue 2025-06-10 10.00
Location: 3412 Sten Velander, Teknikringen 33
Video link: https://kth-se.zoom.us/s/65793561998
Language: English
Subject area: Electrical Engineering
Doctoral student: David Bergman , Elkraftteknik
Opponent: Teknologie doktor Jonas Persson, Vattenfall AB
Supervisor: Professor Mehrdad Ghandhari, Elkraftteknik; Professor Robert Eriksson, Uppsala universitet
20250512
Abstract
With the ongoing energy transition come new challenges in the power system. When generation gradually transitions from being produced by large synchronously connected machines to being produced by renewable energy sources it impacts the system dynamics. The power system becomes more sensitive to disturbances, and new stability phenomena arise. Thus, it becomes more important to monitor the power system. With the developmentof Phasor Measurement Units (PMUs) it became possible to monitor the power system at a high sampling rate. A phenomenon that became possibleto detect and monitor with PMUs is forced oscillations. Together with natural oscillations these form two different types of electromechanical oscillations. Whereas natural oscillations are the natural response of the power system due to random load changes or faults, forced oscillations are the response of the system due to an external source that is continuously exciting the system dynamics. If the frequency of the forced oscillation is close to an oscillation mode of the system, resonance can occur between the forced oscillation and the system mode. If specific conditions are fulfilled, the forced oscillation can be greatly amplified in the system and it can spread over large parts of the system. It is therefore important to monitor for forced oscillations and to take corrective actions if they occur. Detailed theory of a few carefully selected methods to detect forced oscillations is presented and their performance is validated in a case study. Once an oscillation is detected the next step is to determine if the oscillation is a forced oscillation or a natural oscillation due to poor damping. To this end a method to distinguish between forced and natural oscillations is presented. It distinguishes between the oscillation types by looking at the damping ratio of the underlying ambient noise and comparing it to a threshold value. The proposed methodology utilizes a parametric model that also includes an undamped sinusoidal component to separate the undamped oscillation from the ambient noise in PMU data. The method is validated using the two-area test system and the Nordic 44 test system.