Measurements and Modelling of Low-Frequency Disturbances in Induction Machines

Doctoral thesis, 1996

The thesis deals with the dynamic response of the induction machine to low-frequency perturbations in the shaft torque, supply voltage and supply frequency. Also the response of a two-machine group connected to a weak grid is investigated. The results predicted by various induction machine models are compared with measurements performed on a laboratory set-up. Furthermore, the influence of machine and grid parameters, machine temperature, phase-compensating capacitors, skin effect, saturation level and operating points is studied. The results predicted by the fifth-order non-linear Park model agree well with the measured induction machine responses to shaft torque, supply frequency and voltage magnitude perturbations. To determine the electric power response to very low-frequency perturbations in the magnitude of the supply voltage, the Park model must be modified to take varying iron losses into account. The temperature and supply frequency affect the low-frequency dynamics of the induction machine significantly while the influence of saturation, phase-compensating capacitors, skin effect and static shaft torque is of less importance to an ordinary industrial machine. The static shaft torque is, however, of importance for determining the responses to voltage magnitude perturbations. The performance of reduced-order induction machine models depends on the type of induction machine investigated. Best suited to be represented by reduced-order models are high-slip machines as well as machines that have a low ratio between the stator resistance and leakage reactances. A first-order model can predict the rotor speed, electrodynamic torque and electric power responses to shaft torque and supply frequency perturbations up to a perturbation frequency of at least 1 Hz. A second-order model can determine the same responses also for higher perturbation frequencies, at least up to 3 Hz. Using a third-order model, all the responses to torque and frequency perturbations as well as the reactive power response to voltage magnitude perturbations can be determined up to at least 10 Hz.