ECN Forum Forums General Topics of the Trade General Discussion Area Locating a grounded phase

Any suggestions with respect to locating a ground fault on a 240V delta where they do not want to turn off machines?

Old building complex, mix of wiring methods including knob and tube, buss duct, EMT, and rigid conduit. Cranky old CNC machines that do not like to be power cycled.

This 240V delta system supplies about 1/2 of the building load. Ground fault appeared sometime in the last 2 years. Was not present during last power quality audit. Building has a history of ground faults according to the current electrician.

With all of the metal piping, was not sure if a thumper signal injector would work. Utility feed is 480V 1500 KVA. 240V is created by 1000 KVA transformer.

Switching parts of the distribution system off is the only way that I know of to find a ground fault.
Just last week we found a ground fault on a Delta system that we had been looking for since last October.

Unfortunately if you have multiple ground faults on the same phase you may never be able to track it down (it's like the old series connected Christmas tree lights where any single bulb going out kills the whole string).

One of the reasons that the Delta system is good is that the equipment will continue to operate if a ground fault occurs.
One of the reasons that the Delta system is not so good is that the system will fail in a particularly spectacular fashion if a 2nd ground fault occurs on another phase.

Have they already measured the grounded phase voltage to earth in several places?

I have, and that got me wondering If it could be isolated to the bus with the lowest phase to earth voltage. We are measuring it with a DVM and initially I thought it was just leakage, but the ground lights in the switchgear are definitely lighting up.

If we can isolate it to a buss, perhaps we can selectively power down certain machines and monitor to see if the fault changes.

Well, my favorite expression is, "All the world's a voltage divider." All the wires and conduits are resistors. The lower the voltage, the closer you should be to the contact point with earth. You might try using a clamp-on ammeter on the conduits to see if you pick up fault current.
The last time I had to track down a ground fault, it was an FACP signal circuit. The culprit zone was the one that measured the closest to ground. Lifting the wire shifted the fault to the other rail, through the faulted device.
In your case, you should probably verify that your ground detector isn't the source of your ground fault: Snubbers, MOVs, gas discharge tubes, Et,c.
Joe

Anything interesting?
Joe

Was just told Wednesday is my last day. Company is downsizing. The electricians will have to deal with it. I'll give them the advice and wish them luck.