Automatic location of transient power quality disturbances (Disturbance location)

The problem of electric power quality has received growing interest in recent years due to the proliferation of sensitive electronic loads in the power system. Transient power quality disturbances such as voltage sags and capacitor switching transients, once considered minor disturbances, can now be very damaging and costly problems. As a result, electric utilities and their customers have begun to monitor the power quality in the system as a first step towards eliminating power quality problems.; An important part of the monitoring process is locating the sources of the problem-causing disturbances. Locating the sources can be useful for the diagnosis of specific problems as well as for assigning responsibility to parties who cause power quality problems seen by others connected to the power system. This dissertation presents a method by which the disturbance sources may be approximately located through a determination of the relative direction from a metering point to the disturbance source.; Transient disturbances on the power system cause the power and energy flows in the system to change from the steady-state values. In this work, a set of four rules is developed that examines characteristics of the changes in the power and energy flows in order to determine the direction to the disturbance sources. The rules are developed and tested on recorded capacitor switching waveforms. The overall accuracy rate for each of the rules for the test data set ranges from approximately 80% to a perfect 100%. The evidence provided by each of the rules may be combined using the Dempster-Shafer theory of evidence. The combined evidence allows for the accurate determination of direction to both computer-simulated and recorded capacitor switching disturbances as well as simulated voltage sags, correctly identifying the direction to the source in 100% of the cases tested. The Dempster-Shafer theory is also capable of providing a quantitative measure of confidence in the decisions made. Finally, a wavelet-based method capable of performing accurate time localization of transient disturbances (in this case, voltage sags), important for the accurate calculation of the disturbance power and energy characteristics, is presented.