Service w/ approx 70 GFCI's tripping; need help thinking through.

Status
Not open for further replies.

brantmacga

Señor Member
Location
Georgia
Occupation
Former Child
Restaurant, 1000/3 service, 208Y/120.

Square D I-Line MDP w/ 8 225A QOB panelboards. AP-1 through 6, LP-1 Lighting, CP-1 Computer (POS) panel. Also an EP-1 panel that's mounted inside a table. I'm going to attach a pic to give a reference of the layout. All of the AP panels are in a modular panelboard assembly. The LP & CP panels are separate, but they don't have any GFCI's.

There are approximately 60+ 20/1 GFCI circuit breakers, 10+ 20/2 & 30/2 GFCI circuit breakers, and 14x 50/3 GFCI circuit breakers. This store has been open for maybe 8-10/wks. I've been told that "almost" every GFCI circuit breaker within the AP panels except for the 50/3 breakers has tripped. There is no pattern; some days they might have no breaker trip, some days 4-5, and days like today where 12 breakers tripped.

Some of the 20/2 breakers that have tripped are 208v circuits. The 50/3 are all 208v circuits, and as I said none of those breakers have ever tripped, but I wanted to mention that the 50/3 breakers have no connection point for a load neutral wire. They are rated at 208v only.

Almost all of the circuits that have tripped are in EMT conduit; a couple that have tripped are in PVC through the slab landed in a steel box. I have not been on-site to troubleshoot. I've sent a couple of people down but none were able to determine anything. I spoke w/ an engineer at the utility's PQ division, and was told as far as they can tell, nothing is wrong on their end.

I've never experienced anything like this; here are my thoughts so far. I'm writing them out for help thinking this over.

A) Is it possible there is a neutral/ground bond somewhere downstream of the MCB that is causing this? Troubleshooting would involve shutting down the site in the middle of the night and check each circuit grounded conductor for a N/G bond. Why would it affect all circuits and not the circuit w/ the downstream bond? Why wouldn't it trip that breaker immediately every time? Why would it affect 208v circuits?

B) Is it possible leakage current from a piece of equipment is causing this? How am I going to find that if its not constant?

C) Is it possible I have a damaged conductor somewhere that is putting voltage on the neutral and causing this? Like in the mV range; but again why not the 50/3 breakers?

I'm having a lot of trouble wrapping my head around why the 20/2 208v circuits are also tripping.... where is that current going and why isn't it causing a dead-short? Why is it NOT affecting the 50/3 breakers? The only difference between these CB's and the 20/2 - 30/2's is the lack of a load-neutral connection.
 

Attachments

  • IMG_0533.png
    IMG_0533.png
    1.3 MB · Views: 119
  • IMG_0534.png
    IMG_0534.png
    1.2 MB · Views: 112

LarryFine

Master Electrician Electric Contractor Richmond VA
Location
Henrico County, VA
Occupation
Electrical Contractor
Is it possible the main bond is missing? Have you checked neutral/ground voltage, both at the main and in the panels in question?
 

brantmacga

Señor Member
Location
Georgia
Occupation
Former Child
Is it possible the main bond is missing? Have you checked neutral/ground voltage, both at the main and in the panels in question?

1fcfcebef54090b0d43b8b337a555952.jpg


It’s there. As for NG voltage, I can’t swear to it, but I would think I did. I was at this job when it was energized, and that’s part of my standard check.


Sent from my iPhone using Tapatalk
 

texie

Senior Member
Location
Fort Collins, Colorado
Occupation
Electrician, Contractor, Inspector
I'd have to check for sure, as it has been a while since I came across something similar, but I seem to recall issues with GFCI breakers being used on certain combinations of circuits in 208/120 Y systems as opposed to 120/240 systems.
 

synchro

Senior Member
Location
Chicago, IL
Occupation
EE
I can only speculate, but perhaps noise on the neutral bus is causing false tripping. All of the GFCI breakers need a connection to the panel's neutral bus to operate their sensing circuitry, and so that might be an entry point for interference to cause false tripping in multiple breakers.

In addition to sensing ground fault leakage, GFCIs have a "grounded neutral" detection function which can also trip the breaker. Since the 3-pole breakers don't have a load neutral output, it's very possible that this "grounded neutral" function is not present or it's disabled in 3-pole breakers. So maybe noise on the neutral bus could be false triggering the grounded neutral detection in 1 and 2-pole breakers and then tripping them, but not in 3-pole breakers if they don't have grounded neutral detection.

Now, what could cause such noise on the panel's neutral buses in your situation? Maybe a marginal connection on a neutral conductor upstream of the AP panels. Or perhaps nonlinear loads, switching supplies, motors, etc., particularly those that have higher current draws.

Hope it turns out to be something simple.
 

brantmacga

Señor Member
Location
Georgia
Occupation
Former Child
Are you certain that there are no mixed up neutrals between circuits?

As certain as I can be. I feel that would be limited to only the affected circuit and not system wide. If any neutrals were crossed on a loaded circuit it would trip immediately.

These breakers are tripping, resetting, and holding for up to days at a time before tripping again.


Sent from my iPhone using Tapatalk
 

brantmacga

Señor Member
Location
Georgia
Occupation
Former Child
Now, what could cause such noise on the panel's neutral buses in your situation? Maybe a marginal connection on a neutral conductor upstream of the AP panels. Or perhaps nonlinear loads, switching supplies, motors, etc., particularly those that have higher current draws.

Hope it turns out to be something simple.

This store is a prototype build; the one thing that stands out as being significantly different is that the RTU’s have VFD’s on the compressors and fans.


Sent from my iPhone using Tapatalk
 
Status
Not open for further replies.
Top