Localization and control of very low-frequency oscillations in Ireland's power system

Previous research has discovered that many “new” oscillatory behaviours on Ireland’s power system could not be reproduced through traditional model-based analyses. Initial investigations have proven that most VLFOs are sourced by specific generators, either under specific grid conditions or with controlled delays within generator control systems. The work presented in this thesis will apply source location and detection algorithms to PMU measurements from the Irish system as well as studying model enhancements to capture oscillatory behaviour and possible mitigation measures.

To address power system event detection, the second chapter develops a MQTT based framework to detect power system events such as faults, frequency/RoCoF excursions and oscillations on the network. The third chapter primarily focuses on identifying conditions where VLFOs are observed in the power system and then analysing the PMU measurements using the DEF method to identify the units contributing to the oscillations. From over 70 events studied it was found 2 units exhibited behaviour that was consistently exciting oscillations on the system.

The fourth chapter shows how traditional models may not capture oscillations and studies the impact of a transport delay can have on governor response. Furthermore, the impact of synchronizing a fast-starting synchronous generator with a positive damping is also studied. The fifth chapter addresses the impact active power control from wind farms can have for damping VLFOs.

This is studied through simulation and from PMU measurements from the Irish power system. The sixth chapter develops a Sequential Forward Selection (SFS) linear regression method to study the impact geographical concentration wind generation in Ireland can have on voltage phase angle difference between 3 locations on the Irish network. Finally, this thesis addresses a long-standing iii dynamic issue existing on the Irish network and presents possible mitigation schemes and algorithmic enhancements to improve situational awareness. <br><br><i>Thesis is under embargo until 31 July 2029.</i>