PRESENTED BY
Mathieu Lachance
OMICRON electronics Canada
Medium voltage (MV) distribution switchgear are a critical part of almost any industrial facilities. Their main functions can include interrupting fault currents, isolating feeders or section of a distribution network and providing metering data to different processes. Because of this, the reliability of those equipment is crucial. Unfortunately, little importance has historically been given to the assessment of the insulation of complete switchgear assemblies during maintenance [1] and at commissioning.
There is only a limited number of published surveys available regarding the causes of failure in MV metal-enclosed distribution switchgear. Probably the most notable example of such survey was published in the appendices of IEEE 493-1997 and dates back from the 1970s. However, some data from a smaller survey performed in 2008 was summarized a few years ago [2] and showed very similar results, at least for the identified causes of failure. In both surveys, approximately 25% of failures were deemed to be caused by partial discharges (PD) for MV switchgear. In addition, in the original survey, approximately 20% of failures were attributed to improper handling or improper installation as the principal root cause.
Conventional partial discharge measurement is a well-established sensitive measurement to detect anomalies during routine tests for almost every medium and high-voltage electrical apparatus. For metal-enclosed switchgear, guidance is provided by IEEE Std C37.301, IEC 62271-200 and CSA C22.2 No. 31-18. However, PD measurements are deemed optional for both IEEE and IEC standard, while it is a stringent requirement for CSA for switchgear that have a rated voltage of 15kV and above. For an unknown reason, the switchgear industry seems to rely mainly on voltage withstand test to assess the insulation of complete switchgear assemblies,
Usually, at the end of the factory tests, the switchgear is disassembled and shipped to site in different sections, where it is reassembled. Once again, the usual test procedures include an insulation resistance test and an AC voltage withstand test. Unfortunately, these will only detect major defects. If PD is present in the switchgear, deterioration will occur at a fast pace in organic insulation systems. Sharp edges from questionable assemblies, subtle damages from transportation and handling in addition to objects left behind such as tools and hardware can create surface and corona discharges. While the former can lead to failure over time, the by-products of corona discharges are corrosive and will deteriorate the switchgear mechanical parts.
This paper presents the results of an experiment that was performed in collaboration with a switchgear manufacturer in order to compare the sensitivity of conventional AC hipot and partial discharge measurements to detect anomalies in complete metal enclosed switchgear assemblies. Common onsite problems were simulated, and the switchgear was tested with both an AC Hipot and a PD measurement. The results are used to expose the advantage of using conventional PD measurement during the onsite commissioning of metal enclosed switchgear.