Modelling and analysis of a coal-fired thermal power plant to identify physically observable dominant low-frequency oscillatory modes

This paper presents a systematic approach to study the dominant electromechanical oscillatory modes in a large-scale coal-fired thermal power plant. The proposed methodology first involves analysing the fault recorder data. Next a detailed linearized model is developed, and the oscillatory modes are identified via the eigenvalue analysis. Computations of the participation factors and mode shapes are useful in identifying the contributing sources on the oscillatory modes and their types. The importance of verifying the accuracy of the dynamic data used in the simulation environment for practical power systems and validation of the linear model with nonlinear simulation are highlighted in the paper. The proposed methodology is applied to a 900 MW coal-fired real power plant in Sri Lanka power system. A dominant oscillatory mode of frequency 1.2 Hz where the coal power plant oscillates against the network is identified using the proposed methodology. This systematic approach can be applied to any real power system including renewable energy sources using correct dynamic models.