Modelling and analysis of turbogenerators in single machine and multi- machine subsynchronous resonance studies.
Subsynchronous Resonance (SSR) is a condition which occurs when turbogenerators are connected to series capacitively compensated transmission systems and it can cause large scale damage to the turbogenerators. The accuracy of predictions of this phenomenon are limited by the accuracy of the mathematical models used for the various system elements. The modal method of modelling a turbogenerator shaft, in which parameters are associated with each natural torsional mode of the shaft, is investigated in detail and the sensitivity of SSR predictions (both small signal and transient) to uncertainties in the mode parameters is evaluated. The modal model is then used to obtain reduced order shaft models and the accuracy of these reduced order modal models in SSR predictions is ascertained. The determination of mode parameters from generator transient response waveforms is investigated. A continuing problem in this field is the separation of damping values obtained from measurements on a synchronized generator, into their mechanical and electrical components. A method is proposed in this thesis which uses eigenvalue scanning techniques together with FFT analysis to achieve this separation. The SSR stability of, and the torsional interaction between two adjacent generators at a power station is studied. The analysis covers identical generators, nominally identical generators with small differences between their mode parameters and different generators with a coincident torsional mode. In addition, the torsional interaction between generators at different power stations which are remote from each other is investigated. This entire analysis is greatly assisted by modelling the turbogenerator shafts in modal form. Finally the damping of SSR oscillations in two non-identical adjacent turbogenerators with a single controlled shunt reactor, which uses the sum of the generator speed signals as an input to the controller, is investigated.