Mike,

You and the other newsletter members might find this interesting - it is a 75 kVA, 480Y/277V-208Y/120V, General Purpose, shielded, dry-type transformer that is being used in a normally open maintenance bypass circuit for a UPS.

The facility in which it is located experienced an open-phase condition on the primary of the electric utility delta-wye transformer. The UPS recognized the anomaly and transferred to batteries; however, due to a discharged battery, the standby generator failed to start, which kept the automatic transfer switch in the Normal (utility) position. Although the transformer secondary breaker was open, and the primary windings experienced over-voltages, the primary breaker never experienced a sufficient overcurrent to trip.

Smoke was observed emanating from the transformer, and it was "hot to the touch" (yes, a technician actually touched the top of the transformer enclosure while it was smoking and humming). When the fire department arrived several minutes later, power was removed from the building, and the transformer was allowed to cool.

I was contacted by the owner's insurance company to perform a claims analysis of the "failed" transformer to justify its replacement as part of an insurance claim. My observations of the interior and subsequent tests I ordered (winding insulation and turns ratio) confirmed that the windings and core were not affected by the over-voltage on the primary windings (tested and looked, on the inside, as if it was new).

The factory and I surmised that the burn marks around the cover and enclosure bolt holes, both front and rear, were most likely due to an unbalanced electromagnetic field (one open winding), of higher than normal magnitude, which induced charges on both the front and rear cover.

Since the covers and enclosure were all painted before assembly, there was enough paint between the covers and enclosure and between the bolt heads and the covers, to create an insulated barrier between the covers and the bonded enclosure (basically a couple of capacitors) - if the technician would have touched a cover as he touched the top of the transformer while it was "cooking", he most likely would have received a "shocking" surprise.

The charges on the covers appear to have built up until they overcame the resistance of the paint and started flashing over to the bolts and enclosure, creating heat and the unusual but very uniform burn marks - there were no other signs of overheating or arcing anywhere on or in the enclosure, or in the core, coils, terminations or wiring

Perhaps the usual requirement to remove paint from equipment surfaces before attaching a ground lug should also apply to equipment manufacturers as they assemble their enclosures!

Regards,

Thomas C. Montgomery, PE
tom.montgomery@hendersonengineers.com
Director of Electrical Engineering
Henderson Engineers, Inc.
www.HendersonEngineers.com
Lenexa, KS 66214
Phone: (913) 307-5381

P.S.: I also did inform the owner that the triple-lugged grounding connection, with splayed strands, was not acceptable.

ike Holt's Comment: Tom, I have no clue what your explaining, but I'm sure lots of others do and this newsletter will be helpful for others.