[jcaron2]

I recently installed some outdoor receptacles for stock tank de-icers for a horse farm. They are regular 120V receptacles, but since I have to power six 500-watt heaters at six different locations, I chose to run them on two different circuits. Otherwise I would have had to use really large (and really expensive) wire to put them all on one circuit. I ran a single 12/3 wire to all six receptacles and staggered the hot connection (so the first one used black, the second red, the third black, etc.). They all share a common neutral. Thus, three of them are on one pole, and the other three are on the other pole.

Obviously I want these outlets to be GFCI protected (not to mention that code demands it). With a shared neutral, of course, I can't just use two single-pole GFCI breakers. I either have to use separate GFCI receptacles at all six locations or use a two-pole 120/240 GFCI breaker in the panel (which is a Cutler-Hammer BR series, by the way). The cost is about the same in both cases, so I chose the second option.

I got everything installed, flipped the breaker on, verified that all the outlets work, made sure the "test" button on the breaker caused it to trip, and gave myself a pat on the back for a job well done. Then later in the day, I was in the barn where the electrical panel is located and I happened to turn on the fluorescent overhead lights. Just as I did, I thought I heard a breaker trip in the panel, but the lights were still on, so I figured I must have been hearing things. Nevertheless, the next time I walked by the panel, I checked and the 2-pole breaker for the tank de-icers was tripped! So I reset it and tried to duplicate the behavior. Sure enough, after a couple of tries I got it to trip again by turning the fluorescent lights on. That left me really scratching my head, because they're on completely different circuits. Clearly the lights couldn't have been causing current to flow in either hot leg or the neutral in the de-icer circuit!

Then later in the day, the breaker tripped again at some point when nobody had been in the barn to turn lights on or off (though there's a hot-water heater on a different circuit that cycles on and off). And to make matters even more confusing, there was ANOTHER breaker tripped. This time it was a GFCI/AFCI single-pole breaker for some receptacles in the barn. The only thing plugged into those outlets was a 200W heat tape that keeps some pipes from freezing.

So now I'm really confused! At first I thought maybe the two-pole GFCI breaker was just over-sensitive and needed to be exchanged for a replacement (even though I still can't figure out how flipping a switch on a completely different circuit could trip it). But with this second breaker now tripping, I can't make heads or tails of what could be causing this. The only possible explanation I can come up with is that there's an electric fence nearby the trench where the 12/3 wire runs. There are often times throughout the day where the fence contacts the ground, so I suppose it's possible that return current from fence is somehow coupling onto the 12/3 line underground and getting back to the panel and causing various wacky behavior. That really seems like grasping at straws though.

So does anybody have any experience with oversensitive GFCI breakers? And can one breaker tripping cause another to trip? Any other ideas as to what could be going on here?

Thanks for your help!

[tkrussell]

GFI devices are as sensitive as they need to be, not over or under, unless they are defective.

Using a two pole breaker may not have been the best choice, as if, or better, when, there is a problem, you lose both circuits, and no easy method of trouble shooting, or narrowing down what is having a problem.

Any 2 pole 120/240 volt breaker has an internal trip, if one pole has a problem, both will trip.

Had you used GFI receptacles, would have helped narrow down any problem any one heater would have.

But, with other loads on other circuits causing this 2 P CB to trip, the problem is elsewhere.

I suspect you have either a ground loop in the panelboard, or something is not wired properly regarding the neutral and grounding of the panel and/or neutral.

Any chance you can take some photos of the open panelboard? And any grounding of this panel?And the feeder for this panel?

I believe using the 2 P GFI breaker, while it should work fine for you, has caused something that exists to now become apparent.

I personally would not have used a 2 pole breaker, but had you used GFI receptacles, this underlying issue may not have surfaced.

[jcaron2]

It just occurred to me that the single-pole AFCI breaker tripping is probably unrelated. The heat tape that's plugged into it uses a simple bi-metalic cantilever type of thermostat. I can't imagine a scenario more favorable to arcing, so that's probably what's tripping it. I guess I'd better install a dedicated GFCI receptacle for it on a non-AFCI circuit.

Which (hopefully) means that my problem with the two-pole GFCI circuit really could be just an over-sensitive breaker.

[tkrussell]

Why an Arc Fault breaker for a barn? Not needed. As you can see, any arcing will trip the AFCI CB.

The GFI breaker is doing it's job, assuming there are no wiring issues.

As I said, I would not use a 2 pole 120/240 volt GFI CB for two 120 volt circuits, even though it is designed for the purpose.

I would expect problems.

If you think it is over -sensitive, and insist on using a 2 pole breaker, you may try to return it for credit and try another one. This breaker should have cost you in the range of $130.00.

GFI 20 Amp receptacles, spec grade, should cost about $10- $15 each.

If only one heater has an issue, you only lose that one heater, not all six, as when the breaker trips.

[jcaron2]

Thanks, TKR!

When I say "over-sensitive", I really do mean defective; that's a much better word choice. Maybe I've just been unlucky in the past, but on more than one occasion I've installed a GFI receptacle brand new out of the box only to have it trip at the drop of a hat. When I swapped it out for another new one, the problem went away. I've always assumed they were just defective receptacles, and this strikes me as analogous behavior. Also, in the course of reading other posts regarding 240/120 GFCI breakers, I saw a few different people mention that these Cutler-Hammer breakers are notorious for this sort of defective behavior. Unfortunately, however, I had to special-order this thing so swapping it out isn't exactly straightforward.

Hindsight being 20/20, I definitely should have gone with the separate GFCI receptacles at each location for the reasons your mentioned. At this point, my plan is to do exactly that, and to change back to a regular two-pole breaker (or, better yet, two single-pole breakers). Then I can return the expensive 2-pole GFCI breaker.

I'm still really suspicious of the GFCI breaker because even if there was some sort of ground loop at the box (which I'm pretty confident there is NOT, but I'll definitely post some pictures for you when I get a chance if it comes to that), that shouldn't have any impact on the current flowing through the neutral line through the breaker unless the ground and neutral were somehow tied together downstream (which is the very definition of a ground fault). The breaker, of course, has no ground connection at all.

Anyway, thanks again for your input. I'll let you know how things work after I swap out the receptacles and the breaker.

[jcaron2]

I think our replies are a little out of phase with each other - but yes, I agree with every single sentence of what you just said. :-) I definitely chose poorly.

[Handyman2007]

I did a 200 amp installation in an automotive repair garage a few years ago. I needed all of the outlets to be ground fault. I used ground fault breaker for each bank of 6 Duplex outlets. I also used a dedicated neutral to each leg although the conduit carried each line. There has never been a problem with any of these breakers tripping. I believe the dedicated neutral is the key.

[jcaron2]

By the way, I chose to use arc-fault for all the receptacles inside the barn because there's a lot of flammable dust in a barn, along with a number of very precious horses who have no way of telling us that their barn is on fire while my wife and I are asleep 500 feet away. Add the possibility of insulation-chewing rodents or malfunctioning heaters, box fans, etc. and I thought it was cheap insurance against the unlikely event of an arc starting a fire.

Should I have just stuck with GFCI?

[jcaron2]

Handyman, a dedicated neutral would definitely have been easier and more conventional, but two long runs of 12/2 wire was a lot more expensive than one run of 12/3. Plus, with the shared neutral, the return currents from the two poles cancel, effectively eliminating the resistance in the neutral line, thereby cutting the voltage drop by about half when the loads are balanced between the two poles.

[tkrussell]

OK, I hear and like your thinking on the AFCI, however, you can see that these can cause problems.

Most times it is good to exceed NEC requirements.

If the wiring is done properly, then any arcing should be minimized and/or contained, and not be a fire hazard.

Any cord connected equipment is more likely subjected to physical damage and arcing, and needs to be closely monitored.

You may want to review Article 547-Agricultural Buildings of the 2008 NEC.

There is a Sticky Note at the beginning of this forum that has a link to read the NEC on line, minor registeration needed, and it is cumbersome to use, if you do not have your own copy.

[tkrussell]

And, I never use shared neutrals, other than the service entrance feeder, that I have no choice but to use a shared neutral.

Any branch circuit I install will have it's own neutral. An open shared neutral on a 120/240 volt system, or any system for that matter, can cause so much damage to items plugged in and on.

Now throw a 2 pole GFI in the mix, more potential problems.


I will only use a 2 pole GFI if the unit in need of power needs 240 volts, never for two 120 volt circuits.

I also hear you on cost of the wire, but one added conductor does not cost that much more.

Now I wonder how long these feeders are?

GFI breakers can nusinance trip due to long wire runs.

[Handyman2007]

Handyman, a dedicated neutral would definitely have been easier and more conventional, but two long runs of 12/2 wire was a lot more expensive than one run of 12/3. Plus, with the shared neutral, the return currents from the two poles cancel, effectively eliminating the resistance in the neutral line, thereby cutting the voltage drop by about half when the loads are balanced between the two poles.

Did you run "Romex" or use conduit? I would have used EMT or PVC because of the chance that the wiring could be damaged in the type of environment.
I am a commercial electrician and do everything better than code and NEC... I play very safe.

[jcaron2]

I ran it all in PVC conduit.

[jcaron2]

TKR, I never even thought about the potential damage that could be inflicted if the neutral line gets cut! That's a really good point.

The last receptacle is about 700 feet out from the panel, so, yes, these are very long runs.

[tkrussell]

700 hundred feet?

Well, that is all the more reason to use GFI receptacles, and the wiring should be larger than #12.

A 500 watt load at 120 volts 700 feet needs a min of #6 copper and result in a voltage drop of 2.17%. Max recommended is 3% for a branch circuit.

120 volt does not like long distances such as this. Neither does GFI breakers.

By the time I done, you may need to rewire everything.

No offense, but this is what happens when laypeople try to do their own electrical work. Good time and materials gets wasted.

[Handyman2007]

Yes. That is way under wired. If running conduit, it is worth running the dedicated neutrals AND grounds in that situation. And running 12/2 is not acceptable to NEC. I would have run MINIMUM #8 hot and neutral with #10 ground wire. For that size wire you will need 1 1/2" conduit.

[tkrussell]

Handyman, I did accurate VD calcs, and min wire size will be #6 for the load and distance given by JCaron.

And, I would like to point out that whenever ungrounded wire is increased to reduce voltage drop, or for any reason for that matter, that the equipment grounding conductor must be increased proportionally.

So, #8 for the circuit and #10 for the EGC is not accurate advice.

[Handyman2007]

Ok. I'll buy that... you are correct... went back and did the same calculations.

[jcaron2]

I hear you and can appreciate why you'd be rolling your eyes. However, I actually thought long and hard about voltage drop when deciding on the wire gauge. These receptacles are dedicated for very simple resistive heating elements. In my area of the country I really only need about 250W max on the coldest nights of the year, which means I could actually tolerate as much as a 30% voltage drop. If one heater is running at the outlet 700' from the source, it's voltage would be dropped by about 9% (about 4 times as much as your calculation since my wire is half the diameter), so I'd only be getting a little over 410W of power out of it. The very worst case scenario would be three heaters running simultaneously at the three outlet locations on the longer of the two circuit legs, with none running on the other leg. In that case, the voltage on that last heater 700 feet out would be dropped by almost 17%. That means my 500W heater would only be putting out about 345W. That's a very, very significant drop, but for my application, that's perfectly fine.

[tkrussell]

Jcaron, well thought out explanation and theory.

Not often do I see someone install something that does not need to operate at full capacity, and is able to deal with the lower output of the heater that has lower voltage.

I am glad you got back with your explanation and reasoning.

I wanted to be sure that anyone reading this understands that under most situations voltage drop is not a good thing, and in your case, your able to allow it and still get satisfatory results.

Now if you can just keep the damn thing running! Lol.