Analysis and mitigation of sub-synchronous oscillations induced by inverter-based resources
Author(s)
Javaid, Muhammad Sharjeel
Type
Thesis
Abstract
High shares of Inverter-Based Resources (IBRs) in power systems are triggering complex dynamic interactions and new stability challenges. One common problem is poorly damped Sub-Synchronous Oscillation (SSO) caused by adverse interaction among IBRs through the network. These oscillations are difficult to foresee, pose a serious threat to system security and often force grid operators to limit the instantaneous share of IBRs. This thesis focuses on the detection, analysis and mitigation of these SSO. First, it establishes the modelling adequacy to accurately capture SSO in time- and frequency-domain studies. The role of network dynamics, which is generally ignored in planning tools, is emphasised. Consequently, EMT-dq modelling is established as a middle-ground offering greater accuracy than RMS models with significantly less computational burden than EMT-abc (point-on-wave) to allow more scenario checks. Next, a set of diagnostic indicators is developed to identify which IBR control loop is primarily responsible for a specific SSO mode using time-domain signal sensing. Building on these realisations, novel and practical control strategies for both grid-following (GFL) and grid-forming (GFM) inverters are developed where the known network model is made part of the plant for control design. Such a design ensures higher bandwidths and enhanced stability margins. Finally, to circumvent the opaqueness of proprietary ‘black-box’ IBR models, a data-driven technique is presented to estimate each IBR’s linearised model (as impedance spectra) around an operating point. This technique eliminates the need for dynamic frequency scans and enables system-wide stability assessment and root cause analysis (e.g., participation of each IBR in SSO) using only one-shot input-output measurements for each IBR. Overall, this thesis presents scalable modelling framework to capture and analyse IBR-driven SSO, proposes advanced control techniques to mitigate them, and introduces data-driven IBR estimation method to obtain operating point specific impedance spectra of ‘black-box’ IBR models for analysing SSO.
Version
Open Access
Date Issued
2025-05-24
Date Awarded
2025-11-01
Copyright Statement
Attribution-NonCommercial 4.0 International Licence (CC BY-NC)
License URL
Advisor
Chaudhuri, Balarko
Teng, Fei
Akhtar, Zohaib
Sponsor
The Punjab Educational Endowment Fund, Pakistan
Publisher Department
Department of Electrical and Electronic Engineering
Publisher Institution
Imperial College London
Qualification Level
Doctoral
Qualification Name
Doctor of Philosophy (PhD)