line to line loads on 3 wire ciruit

[sceepe]

Ok wise collegues, heres a situation I ran into. Got a group of tennis courts with total of 20, 208v, 1000W MH fixtures supplied from a 100A, 3 pole breaker, in a 4 wire 208 panel. Only 3 wires leave the panel to go out to the poles (no grounded conductor (and sadly no EGC either)).

The problem is that the 100A CB is overheating and tripping.

Questions:

1. is it allowed by the code? Circuit breaker serves as the switch and it opens all the conductors.
2. any guess why the 100A is tripping. I have access to a clamp on meter, what should I expect to measure on each phase.
3. This is not a multiwire branch circuit because no neutral. So I guess you could model as a delta connected load?

 

[charlie b]

It is definitely a delta load. If you don?t use the neutral for any of the connected loads, there is no need to run the neutral from the panel to the loads. So using only 3 wires (I presume you mean the three phase conductors) from the breaker to the various loads is fine. Not having an EGC is not fine, but it should not be related to your symptom.

20 fixtures, 1000 watts each, connected (it is to be hoped) in as balanced a way as possible, should give you about 56 amps on each leg. That should not trip the breaker. So I infer that there is a ground fault somewhere in the system. An EGC should have been there, and should have allowed the breaker to trip sooner (.e., without overheating first).

Just as a possible, though unlikely, explanation, suppose for the moment that all loads were connected between Phase A and Phase C. That would mean you would measure zero current on Phase B, and about 96 amps on the other two Phases. That might explain the overheating, and the subsequent tripping, of the breaker.

I urge you to use the greatest caution on this one. If there is a fault somewhere, you might find it in an unpleasant way.

 

[bcorbin]

There is no requirement for a neutral that I know of, but there certainly must be an EGC.

My belief as to the tripping:

The lamps are 1000W. When you add in power factor and ballast losses (some non-electronic HID ballasts are only 50% efficient), your "1000W" site fixtures could be using 1500-2000VA. That could easily push you over your 100A.

 

[petersonra]

Just as a possible, though unlikely, explanation, suppose for the moment that all loads were connected between Phase A and Phase C. That would mean you would measure zero current on Phase B, and about 96 amps on the other two Phases. That might explain the overheating, and the subsequent tripping, of the breaker.

Why would 96A even continuous current on 2 legs of a 3 pole 100A CB trip it? Its not a good idea, but unless the CB is located in a very warm area, I just don't get it.

I vote for BCorbin's explanation.

<added> I would suggest having a qualified person go measure the current.

 

[jim dungar]

The NEC now requires that breakers used for switching HID ballasts must be HID rated.

Does the breaker trip on start-up? If so then it is probably related to the ballast inrush and the breaker definitely should be rated for HID switching. Although a bad ballast could also be a factor.

If it trips after a few minutes then the breaker contacts have probably been damaged by the HID switching and the breaker needs to be replaced with an HID one.

If it trips after many minutes then you are probably dealing with an overloaded circuit.

 

[sceepe]

Thanks guys. The breaker does not trip on startup, just after a long while. I have seen a lot of things that make me believe this system needs a lot of work to get it safe. Where is Joe Tedesco with his camera when I need him?

As for the current, I am going to get some readings on each phase conductor so I can get an idea as to the degree of inbalance.

Bcorbin,
Your idea is interesting. This all started after a "group relamping" the 1000W merc vapors were replaced with 1000w MH. The ballasts were not replaced. I don't know much about old HID ballasts but do you think there could be some incompatability between the ballast and the new lamp that is increasing the current draw?

 

[coulter]

screepe -

... 1. is it allowed by the code? Circuit breaker serves as the switch and it opens all the conductors. ...

Just has to be rated for SWD, HID.

... 2. any guess why the 100A is tripping. I have access to a clamp on meter, what should I expect to measure on each phase. ...

Wore out CB from switching. Loose terminations. A bad ballast (or two). Water is a fixture (or two). I would expect 60 - 65A on one, a little less on the other two - if they are balanced.

I'd probably start with the clamp-on wrapped around all three conductors. If there is no GF, (Sum)I = 0.

As for the EGC, not even a metalic conduit? As charlie said, this is a classic for getting shocked.

Let us know what you find.

carl

 

[coulter]

... This all started after a "group relamping" the 1000W merc vapors were replaced with 1000w MH. The ballasts were not replaced. I don't know much about old HID ballasts but do you think there could be some incompatability between the ballast and the new lamp that is increasing the current draw?

Lots of stuff got posted while I had the post window open.

As for incompatibility, if someone didn't check to see if the new bulbs were compatible with the old ballasts, they really should have. I'd have no reason to suspect they would be.

carl

 

[boater bill]

The new MH ballasts are about 15% additional load per fixture. So your 1000 watt fixtures come out to 1150 each. This should not cause tripping. I'm going with incompatible ballasts or pre-existing conditions of no EGC.

Be careful troubleshooting this one!

 

[charlie b]

Why would 96A even continuous current on 2 legs of a 3 pole 100A CB trip it?

It shouldn't; I agree with you there. But then, why do we have to load a circuit to no more than 80%, when the load is expected to be continuous?

 

[EdJ]

I would think the most likely culprit is a ground fault. A few ohms in the path from the faulted ballast/wiring to the service neutral could result in over current but not enough to trip the breaker.

Take the earlier advice and measure all conductors w/ clamp and see if you get zero.

Be careful, do not touch any other metal objects while around the poles. I would even be afraid to stand on the ground and touch the pole. What would it take to add an EQC. This is a dangerous situation.

Ed Jackson, PE, Charlotte NC