Probabilistic Transient Stability Assessment and Corrective Control of Power Systems with Increasing Penetration of Non-synchronous Generation

Abstract

In recent times, the operation of electrical power systems has been changed by two key factors. Firstly, by the liberalisation of the electricity market in many parts around the world, and more recently, by a growing integration of power plants using Renewable Energy Sources (RES) into the network, particularly wind and solar based. The main features of these sources of energy are their inherently highly stochastic behaviour, and that their connection to the network is currently made in a non-synchronous inertialess manner through power inverters. This means that in the years to come, power networks will become much more uncertain while at the same time operating with considerably lower levels of inertia. This thesis investigates the effects of the penetration of highly stochastic, non-synchronous RES generation, on power systems transient stability. The thesis contributes to the area of power systems stability research, particularly surrounding the effects of non-synchronous RES generation on first-swing transient stability. The detailed transient stability assessment of different test networks with increasing non-synchronous RES generation was performed, starting from the study of the basic principles through deterministic analyses of single-machine networks, and extending to probabilistic analyses of realistic multi-machine systems. The results provide a full view of how different parameters and operating conditions of the network are altered by the inclusion of RES generation, and how all these changes actually affect the transient stability of the system, making it easy to generalise the learned concepts to any network so that the actual influence of the high penetration of RES on first-swing transient stability can be readily assessed and understood. A probabilistic approach using Monte Carlo (MC) simulations is formally proposed for the stability assessment of uncertain power networks, including the introduction of new indices for complementing the analysis. Following these investigations, a robust methodology for the identification of the critical rotor oscillation patterns occurring in realistic multi-machine power networks using Hierarchical Clustering (HC) is developed. The method also includes the proposal of new statistic measures for the assessment of the quality of clusters in order to find the optimal clustering parameters, so as to guarantee that the most representative and actually critical oscillation patterns are found. Finally, a practical systematic procedure using the information from the proposed HC approach is developed for the optimal implementation or deployment, either in real time or for planning purposes, of corrective actions for the improvement of transient stability. The main outcomes of this research are the comprehensive assessment of how the increasing penetration of inertialess RES generation actually affects the transient stability of power systems, particularly in the first-swing, and the development of a methodology for the effective deployment of corrective measures for its improvement.

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Jovica Milanovic (Supervisor) & Robin Preece (Supervisor)