A unique and innovative method for the characterization of high-frequency micro-discharges or partial discharges and their relation to physical defects exhibiting high electric field stress while in the time domain.

摘要

Electrical insulation is the simplest yet most important element in an electrical system. An exposed conductor, due to damaged insulation, can give rise to arcs, shorts, electromagnetic emission and interference [1]. As power and performance demands increase, insulation reliability becomes of utmost concern due to increased operating frequencies, voltages, and multifactor stresses. The ability to accurately detect and analyze in real time, in a nondestructive fashion, the high-frequency microdischarge behavior in integrated insulation architectures is severely limited by many factors. The following research will describe the newly developed high-frequency discharge detection technique that is applied to the partial discharge detection equipment here in the Energy Systems Institute. The very high level of sensitivity and extremely low level of the Biddle system noise summarize the uniqueness of this piece of equipment, making it the right choice for an implementation platform.;Partial discharges, their mechanisms and detection schemes will be discussed as a way of understanding and introducing the current capabilities and limitations of existing detection methods and technologies. From this discussion it was found that the best detection method was using circuitry; however, high-frequency limitations exist. Research has found that space charge can have detrimental effects when accumulation occurs at a void, defect, triple point or any other anomaly in the insulation architecture. Eventually when the accumulation has surpassed the electric field strength, there will be a violent destructive release which could initiate partial discharges. In order to assist in the research and development of modern dielectric systems, an effort has been made to create a customized partial discharge detector that can accurately monitor fast partial discharge events over a wide frequency bandwidth. This detection network is referred to as the high-frequency network. Its design, implementation, evaluation and research applications are documented in the following chapters. Note that in addition to the measurement of fast partial discharges we are now able to measure and display discharges in the time domain as well as conventional partial discharge detection using the conventional power separation filter shock excited LRC network that is displayed on a phase resolved crt.