| Summary: | Voltage stability problems originate when a generation-transmission system is not able to supply a load connected remotely from the generation centers. Knowledge of the load characteristics and the load composition are necessary tools for voltage stability assessment (VSA). This thesis presents methods and results to evaluate local voltage stability conditions, considering the actual static load characteristics.
Simulation of a modern load, verified by experimental tests, such as a variable frequency drive feeding an induction motor (IM), shows that the real power characteristic is very similar to the directly-connected motor, while the reactive power characteristic is different. The effect on voltage stability is described.
A small-scale voltage stability test is performed for a single IM under increasing mechanical load, fed by a source and series reactance. A set of slip dependent PV curves, which show variable power factor (PF) behavior, is obtained and compared with the assumption of constant PF loads for VSA.
A comprehensive case study is performed, which presents an experimentally obtained IM load characteristic, resulting in variable PF as expected from the equivalent circuit model. The IM is aggregated by simulation in a bus fed by a Thevenin network, and a numerical method is proposed to compute a local PQV curve that considers the actual load characteristic. It is demonstrated that traditional PV curves for constant PF loads do not describe properly the aggregation of induction motor loads in a bus. A graphical approach of network and load PV characteristic intersections for this variable PF load, confirmed with a time domain simulation, shows that the point of matching impedance, typically assumed as the voltage stability limit, is not the power transmission limit (“nose”), and the latter is not the static voltage stability (loadability) limit.
The methods and results developed in the case study are extended to other motor, heating and lighting types of loads. Their experimental characteristics are later combined by simulation in one distribution bus of a university building to perform static local VSA.
Finally, some implementation ideas for on-line load characteristic estimation and PQV curve computation are described as part of the tools for local VSA. === Applied Science, Faculty of === Electrical and Computer Engineering, Department of === Graduate
|
|---|
