ANALYTICAL INVESTIGATION OF THE EFFECT OF GENERATOR MODELLING ON ELECTROMECHANICAL MODE DAMPING Keren Kaberere 1 , Alexander Petroianu 2 , Komla Folly 3 University of Cape Town, Dept. of Electrical Engineering, Cape Town, South Africa 1 kkanuthu@ebe.uct.ac.za, 2 apetroianu@ebe.uct.ac.za, 3 kfolly@ebe.uct.ac.za Abstract: Power system analytical tools differ in their components modelling. The differences affect electromechanical modes damping. This paper investigates the effect of including rotor speed deviation in stator voltage calculation –with the stator transients neglected- and the modelling of turbine output, on electromechanical mode damping of a single machine infinite bus system. We use a sixth order generator model with different excitation control configurations. We analyse results obtained with EUROSTAG and compare these with results obtained with three other industrial-grade tools. Our results show that: (i) if rotor speed deviation is included in the stator voltage calculation, the results are more conservative than those obtained if speed deviation is neglected. (ii) if the turbine model output is torque, the results are more conservative than those obtained if the output model is power. Copyright © 2006 IFAC Keywords: Eigenvalue analysis, electromechanical mode, speed deviation, turbine model output. 1 INTRODUCTION The advancement of computer technology has facilitated the development of several power system analytical tools. Hence, decision makers increasingly rely on digital simulations for planning and operation. One important function of these tools is eigenvalue analysis. System stability is deduced from the eigenvalue results. The analytical tools differ in their components modelling and numerical methodology; therefore, for the same benchmark network, different tools give different results. From experience (Kaberere, et al., 2005a; 2005b) the differences in results, obtained using different tools, are mainly due to differences in components modelling. The results obtained with different tools differ in damping but agree on frequency of oscillation (Kaberere, et al., 2005a; Slootweg, et al., 2002). Kyriakides and Farmer (2004) acknowledge the need for carrying out studies to determine the modelling aspects that result in “damping errors”. Kaberere, et al. (2005a) highlighted the following generator modelling aspects as causes of variations in eigenvalue results obtained using different tools: i) Stator voltage calculation; include (rotor angular velocity, ω r ≠ 1) or neglect (ω r = 1) speed deviation. ii) Representation of turbine model output; mechanical torque, T m or mechanical power, P m . We investigate the effect of the two generator modelling aspects listed above on electromechanical modes. We use a single machine infinite bus (SMIB) system to demonstrate the effect of the two modelling aspects. We use a sixth order generator model with different excitation control configurations. With the tools at our disposal (PSS/E, PowerFactory, EUROSTAG, and SSAT), it was not possible to investigate the modelling aspects using only one tool. Only EUROSTAG and PowerFactory include rotor speed deviation in stator voltage calculation, but none of the two tools allows the user to neglect the speed deviation. The A matrix is therefore important for investigating the effect of speed deviation. The A matrix in PowerFactory is not accessible. Therefore we use EUROSTAG and MATLAB to investigate the effect of including rotor speed deviation on the electromechanical mode. We compare the EUROSTAG/MATLAB results with results obtained using PSS/E, PowerFactory and SSAT. Our investigation reveals that: (i) if the stator transients are neglected, results obtained with rotor speed deviation included in the stator voltage calculation are more conservative than those obtained with speed deviation neglected, (ii) results obtained with T m turbine model output are more conservative than those obtained with P m turbine model output. The paper is organised as follows. In section 2, we discuss the linearised system equations; section 3, stator voltage calculation; section 4, turbine model output; section 5, case study; section 6, conclusions. IFAC Symposium on Power Plants and Power Systems Control, Kananaskis, Canada, 2006 291