3 phase delta ground detect

On 27 Apr 2004 09:26:04 -0700 bob peterson wrote: | A client has requested ground detect circuitry for an MCC. The | circuit he sent me shows a PT primary connected between each phase and | ground and a light conencted from each PT secondary to ground, with a | loading resistor across each secondary. This is apparrently Square | D's standard circuit for ground detection on delta circuits. | | I priced this up and by the time I got done it was $1000. | | Is there something I can just buy someone has used that might be less | costly and simpler?

There are those little proximity voltage sensors that flash or beep when near a hot wire. Supposedly they don't do that when near a grounded wire. I'm not sure if that is the case if there is current in that wire (since that implies some potential anyway). But if you spread the wires far enough apart, and if the grounded wire doesn't activate such a sensor, maybe that technology could be used. If simply observing the lights at the location of interest when interested (as opposed to remote sensing and monitoring) is adequate, that could be the cheapest way to go.

Try this site, do not have a clue if it is less than a grand.

Suggest not oversimplifying the task.

--s falke

all very informative, but not on point.

"...not on point."? If you are dealing with a truly ungrounded system, ignore transient-overvoltage damping at your/your client's peril.

Beeman explained it very graphically in 1955.

--s falke

| I need something the consulting engineer will approve as a permanent | installation.

Still, perhaps something along the same lines using capacitive coupling and very high impedance sensing could do it. It just depends on what all your requirements (including the engineer's requirements) are. If it is something that needs to connect to remote observation or recording devices, it's certainly going to require very specific engineering. If, however, it's just needed for periodic human checks right at the wire location, but without someone having to probe live wires with a meter, this could be your cheapest way.

What sort of voltage are we talking about here? The light bulb and load resistors should be relatively cheap, so I'm guessing that most of your $1000 is tied up in the PTs. If the system voltage is such that you need them, I don't see a cheaper solution.

For lower voltages, you could connect three load resistors (light bulbs?) in a wye, with the neutral point floating. Then connect a meter (or voltage sensing relay) between that point and ground. All components must be rated for full line-to-line voltage. Suitable time delay must be incorporated in the relay system to handle switching transients.

The coal industry often uses corner grounded delta with a resistor in series with the grounding conductor. An ammeter with an adjustable setpoint reads any leakage or fault current across the resistor and when the current meets or exceeds the setpoint it sounds an alarm and/or trips a shunt trip breaker.

Simple and relatively inexpensive.

Its a 480V MCC. The original plan from the consulting engineer was to have the utility change the service from 3W delta to 4 wire wye, but the cost was so much the end user balked. The PTs are just off the shelf 80VA xfmrs (< $30 each). Most of the money is in the labor and design time to take the $300 worth of parts and make it do what they want.

There are no relays involved, just lights. Basically the circuit the consulting engineer sent me is as you described, its just sort of hokey looking to me.

I've seen hundreds of MCCs and have never seen this type of arrangement before, and most of them were delta configurations, so I am not all that sure just what the issue is here. of course with a wye configuration a grounded phase will trip the main cb, so I guess this is just a way of indicating a grounded phase. Seems like there should be a simpler solution.

Design of the power system is not within my scope.

That's fine, but on the longer term, the end user/facility owner may have to deal with insulation damage from arcing-fault or resonant overvoltage incidents. Maybe you'll have washed your hands of the issue before the end user/facility owner has to deal with that.

WHOA.

What do you mean, "make it do what they want?" The 3 bulb system is the standard way of indicating accidentally grounded phases in an ungrounded system, but it is totally unsuited for doing anything else. Unbalanced lights tell you there is a ground, and you go out, look for it and fix it. All it does is light 3 lamps.

If you are trying to use this as a ground fault detector to do any circuit switching, you are on the wrong track.

Well, there's no way around the labor involved. You might be able to find a ready-made unit that doesn't need to be assembled (still have to pay for installation time), but it will probably cost a few hundred dollars.

I don't know about hokey. It will work as a rough indication. A single ground fault in an ungrounded system is passive. Often times, a ground alarm will prompt the maintenance folks to start looking for the fault without having to shut the plant down.

Ground faults such as these can be difficult to locate, since they don't result in significant fault currents affected branch circuit or equipment. As a result, overcurrent or differential protection won't operate.

I've seen quite a few ungrounded systems myself. And I've seen quite a few of them that may have been running with a single ground fault for years.

Its not an attempt to trip anything, just an indicator.

A premade unit that cost $300 might well end up being half the installed cost of this abomination. :-) Thats why I asked if anyone knew of something off the shlef rather than having to make something.

Not up to me to wash my hands of anything, since I am not involved in the power system design at all.

Well, we tested it today, and it worked as advertised although the sight of a 20 gauge wire with alligator clips tied from a phase to ground made the QA guy really nervous.

:-)

At times, a "puckered" QA guy evolves into a better QA guy.

--s falke