Grrr... intermittent breaker trippage

[ericsherman37]

So a couple weeks ago I was dispatched to troubleshoot a tripping circuit breaker at a local parcel delivery company's distribution center/warehouse. The breaker is supplying exterior lights.

There are actually two circuits supplying outside lights. The first one (20A 120V) supplies a pair of 250W HPS wallpacks. It also provides power for the control circuitry - photocell and contactor. This breaker has reportedly never tripped and has always functioned just fine. The second breaker (also 20 A 120 V) was the problem. It supplies a total of (4) 250 W HPS wallpacks, (1) 70W HPS Wallpack, and one MH flood (don't recall the lamp wattage, but the sticker inside the fixture said it should pull roughly 4 amps at 120V).

Occasionally when the delivery crew showed up in the morning, the lights supplied by the 2nd breaker would be off, and they'd find the breaker tripped. They would reset it, and no problems for a few more days. Then, poof, magically tripped overnight again. That's when they called.

The first time I went out, it was toward the end of the day and I didn't have a whole lot of time to troubleshoot stuff. The breaker wouldn't trip for me. There were no indications of ground fault or short. The circuit pulled about 15 amps, which is about what I expected. Being short on time, I replaced the rather old breaker with a brand new one off the van.

About a week later, they called back and said it was tripped again. I went out (with more time) and opened up all the exterior lights, looking for things like moisture infiltration or burnt-up wire nuts that might be banging around against the inside of the grounded fixture when it got windy or something. Found nothin' except one of the HPS ballasts was looking fairly cooked and crispy. I replaced it. Still no ground faults or anything on the whole circuit, and it was still pulling about 15 amps or so. Nothing definitive yet.

Now just today they called AGAIN saying their breaker was tripped. I'm going there first thing Monday morning. It was pretty windy and rainy last night and that may have had something to do with it. But I'm pretty sure the weather was nice and clear the last few times I've been out.

At this point I don't really know where to go next, short of hooking up my recording meter for a night or two and seeing what exactly is going on when the breaker trips. So does anyone have any suggestions or ideas for me? Sorry for the long post.

 

[kwired]

Verify there is no other load besides the luminaires you mentioned.

Meg the luminaires to ground.

Read amp draw after lamps are at full brightness if you haven't done so.

HPS lamps draw more current as they age, they may have drawn an acceptable current level when new now that all the lamps have aged they draw more current. This is why they cycle on and off when they are near end of life, they draw more current than the ballast can supply and eventually the arc is extinguished - the lamp cools a little and the process starts over.

Has anyone noticed if any of the lamps are cycling because they are near end of life?

 

[ericsherman37]

Verify there is no other load besides the luminaires you mentioned.

There is not. The conduit is exposed inside and I visually traced out from the panel and contactor box all the way to each light's wall penetration. The lights I mentioned are the only ones on the circuit.

Meg the luminaires to ground.

Haven't done that yet.

Read amp draw after lamps are at full brightness if you haven't done so.

I have done so. Every one of the fixtures was pulling exactly (well, within a couple tenths of an amp) what their label stickers said they should pull. With the sole exception of the one with the crispy crunchy ballast that I replaced - it was pulling about half an amp more than it should have been. Now it's fine with the new ballast.

HPS lamps draw more current as they age, they may have drawn an acceptable current level when new now that all the lamps have aged they draw more current. This is why they cycle on and off when they are near end of life, they draw more current than the ballast can supply and eventually the arc is extinguished - the lamp cools a little and the process starts over.

I knew that HPS lamps tend to cycle on and off when they're getting worn out, but it's good to know exactly why they do that! Thanks for the info.

Has anyone noticed if any of the lamps are cycling because they are near end of life?

Well, no one is usually present at the building when the lamps are actually running. Everyone goes home in the afternoon before dark, and when they get there in the morning they're all inside sorting out packages and stuff. When I was there the 2nd time, I taped off the photocell and let the lights all run for a good hour and a half while I poked around and took current measurements and stuff. I did not see any lamp cycling at all.

 

[Jlarson]

If it was me I would show up real early on Monday and turn them on an poke around more while I had a logger on it.

 

[ericsherman37]

If it was me I would show up real early on Monday and turn them on an poke around more while I had a logger on it.

Something about "real early" and "Monday morning" just doesn't mesh with me :roll:

 

[Jlarson]

Something about "real early" and "Monday morning" just doesn't mesh with me :roll:

Drink a lot of caffeinated liquid.

 

[dbuckley]

... short of hooking up my recording meter for a night or two and seeing what exactly is going on when the breaker trips.

I think you are already at the point where this is the necessary next step, having excluded al the low hanging fruit possibilities.

You need to know if this is an little overload or a big overcurrent.

You'll probably end up fitting fuses in the fixtures...

 

[hardworkingstiff]

Something about "real early" and "Monday morning" just doesn't mesh with me :roll:

OK OK, how about Tuesday? :grin:

 

[ericsherman37]

I think you are already at the point where this is the necessary next step, having excluded al the low hanging fruit possibilities.

You need to know if this is an little overload or a big overcurrent.

You'll probably end up fitting fuses in the fixtures...

Yeah, it seems like it... A 20A breaker (especially a brand new one) shouldn't be tripping at 15 amps even when it's running all night long every night. And given that the rainstorm we had the other night coincided with the breaker being tripped in the morning, that kind of gives me an inkling what might be going on.

 

[Speedskater]

Maybe we should ask: What time of day does the breaker trip?
Get an old fashion analog clock and connect it to the circuit.
Maybe the controller does something strange at sun-rise.
Or has leakage to another circuit.

 

[LarryFine]

Random thoughts (as if I had any other kind ):

Something on a photocell that doesn't use current until you leave the site.

A supposedly-disconnected photocell whose load wire is somehow grounded.

Troubleshooting idea: disable one fixture at a time until the breaker holds.

 

[rt66electric]

Start inthe middle

Start inthe middle

Disconnect half of the load on the circuit and see if the situation persists ??
This may be time consuming, but it narrow down the intermitant fault.

 

[ericsherman37]

Random thoughts (as if I had any other kind ):

Something on a photocell that doesn't use current until you leave the site.

It's possible, but I traced out the entire circuit in question and it only feeds the lights that I expected it to feed.

A supposedly-disconnected photocell whose load wire is somehow grounded.

It's a pretty simple building: just a big ol' pole barn kind of thing with high bay lighting, a conveyer belt and loading docks, and a small office. I've been all around the building and haven't seen another photocell. Maybe I'll jump up on the roof on Monday and see if there's anything up there.

Troubleshooting idea: disable one fixture at a time until the breaker holds.

I might be stuck doing that, but it would take forever... one fixture at a time for 6 fixtures would be over a week of fiddling with that. I might go with the individual fuseholder idea in each one... I can do them all at once and then go back after a few days and see if any fuses have blown. I wonder though, how should I go about sizing the individual fuses? The 250W HPS lights, for instance, draw about 2.5 amps each. So if I got, say, a 5 amp fuse for each of those, would that be low enough to account for overloading? Maybe 4 amp? I don't know.

 

[dbuckley]

I wonder though, how should I go about sizing the individual fuses? The 250W HPS lights, for instance, draw about 2.5 amps each. So if I got, say, a 5 amp fuse for each of those, would that be low enough to account for overloading? Maybe 4 amp? I don't know.

You don't know (yet) because you haven't connecting your recoring meter.

Thats how you will know the answer to your question

If you have a recording meter, connecting it up is the first thing to do with this sort of a problem - it is very informative to know what the fault condition is, even if you don't know where or why.

 

[ceb58]

There are actually two circuits supplying outside lights. The first one (20A 120V) supplies a pair of 250W HPS wallpacks. It also provides power for the control circuitry - photocell and contactor. This breaker has reportedly never tripped and has always functioned just fine. The second breaker (also 20 A 120 V) was the problem. It supplies a total of (4) 250 W HPS wallpacks, (1) 70W HPS Wallpack, and one MH flood (don't recall the lamp wattage, but the sticker inside the fixture said it should pull roughly 4 amps at 120V).

Have you checked the contactor? You could have a set of bad contacts.

 

[ericsherman37]

You don't know (yet) because you haven't connecting your recoring meter.

Thats how you will know the answer to your question

If you have a recording meter, connecting it up is the first thing to do with this sort of a problem - it is very informative to know what the fault condition is, even if you don't know where or why.

Okay you sold me on that one - right now I'm kind of grasping at straws, so having some solid data would suit me well.

Have you checked the contactor? You could have a set of bad contacts.

How would that make the breaker trip?

 

[dbuckley]

Okay you sold me on that one

Excellent.

If your recording meter has clamps for both hot and neutral, do measure both, as that'll determine if your overcurrent is phase - neutral or phase - ground.

You're actually lucky (or perhaps just well equipped) to have a recording meter, they are reasonably rare in the contracting community, yet without one, tracing this type of mischief is harder than it need be.

And having reread the original post, I do not agree that the breaker is tripping "by magic"; theres always a reason for these things, albeit sometimes the reasons are a tad difficult to come by

 

[ceb58]

How would that make the breaker trip?

Could be the same as having a loose wire. The breaker could be going out on thermal overload.
I had a situation where another elect. wanted to replace a panel that was only 1 year old. Every thing worked fine until the heat strips in the HVAC unit would run for about 20min. then the main would trip. PhA in the main was loose. It wasn't loose enough to arc but was loose enough to cause the main to heat up and trip on thermal. Torqued it down and it has been good for three years now.

 

[kwired]

Could be the same as having a loose wire. The breaker could be going out on thermal overload.
I had a situation where another elect. wanted to replace a panel that was only 1 year old. Every thing worked fine until the heat strips in the HVAC unit would run for about 20min. then the main would trip. PhA in the main was loose. It wasn't loose enough to arc but was loose enough to cause the main to heat up and trip on thermal. Torqued it down and it has been good for three years now.

That makes sense - the loose connection is on the terminal of the device that is tripping, but how does a loose connection on a contactor create enough heat in the breaker to cause it to trip?

 

[ceb58]

There are actually two circuits supplying outside lights. The first one (20A 120V) supplies a pair of 250W HPS wallpacks. It also provides power for the control circuitry - photocell and contactor. This breaker has reportedly never tripped and has always functioned just fine. The second breaker (also 20 A 120 V) was the problem. It supplies a total of (4) 250 W HPS wallpacks, (1) 70W HPS Wallpack, and one MH flood (don't recall the lamp wattage, but the sticker inside the fixture said it should pull roughly 4 amps at 120V).

That makes sense - the loose connection is on the terminal of the device that is tripping, but how does a loose connection on a contactor create enough heat in the breaker to cause it to trip?

This is a stab in the dark. From the OP I would infer that the lights on the breaker that is tripping are running through the contactor that is controlled by a photo cell by another circuit. The contactor is a link between the breaker and the lights. If the contacts are worn/pitted they could, over the course of the night, act as a loose connection causing the circuit to overheat.