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

[petersonra]

I'm thinking NO connection to ground.

That would not trip a gfci functioning properly.

I am guessing there are some pieces of equipment somewhere plugged in that leak a little current, maybe through some kind of L-G surge protector that now and then gets above 5 mA. I would start by looking for power strips with surge protection built in.

 

[ptonsparky]

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

Bing! Bullseye!

Dairies have fought stray currents for years. One of those sources has been found to be VFDs.

 

[StarCat]

See if you can give us a total log of all the hours that this problem burns up after it has been sorted.
This is a testament to what is passing for Engineering in this day and age.

 

[ptonsparky]

Whose RTUs? Mfg?

 

[Cow]

This would be a good one for Jraef, now that we know vfd's are involved.

My first thought was, I wonder if the rtu's have any kind of line reactor or filter on the front end?

I can tell you we've completely remodeled a couple of very large dairies in the last 2-3 years with 75+ vfd's in each one from 1.5HP-250HP, all the 20+ feeder breakers have GFP, and the whole dairy is filled with Levition GFCI receps. Some of the VFD's have TCI 3% KDR line reactors and MTE RFI filters with shielded VFD cable in front of them. 60+ of the smaller 1.5-3HP Lenze drives only have integral internal filtering with shielded cable, no line reactors.

We have not had to reset any GFCI's. The gear is all Siemens.

 

[hillbilly1]

Security camera in that room?

Reminds me of a call a few years ago, two breakers in two different panels would randomly “trip”, but it would be at the same time. Also, they would be in the “off” position, not “tripped” position. I put a recorder on one, and locked the panel door. While I was out in the parking lot on a conference call, the manager came running out saying it did it again, but this time it was only one breaker, the one not locked up...... We put a hidden camera in, they caught one of the pharmacists going back there and turning it off. Turns out, they were moving the store further away from her house, and was mad about it. Yeah, she got fired!

 

[ptonsparky]

The dairy we have is all 480, no 277 loads. VFDs are 480.
Step down transformers feed all the lower voltage.

OP: Are the RTU 208?

 

[kwired]

I connected several VFD driven compressor units in a school a few years ago. (Trane / American Standard). Never had any GFCI issues come up that I know of. These were all on one feeder coming from the main distribution panel. Don't think they had line reactors within the units, I never added any myself.

Something else to consider for OP is how long is the circuit lengths on the GFCI circuits - long runs can have more capacitive leakage which could be a problem.

 

[brantmacga]

RTU’s are 208/3; Kitchen RTU is CaptiveAire, the others are Lennox Prodigy.

Furthest run of a GFCI circuit is approximately 80’.

The tech rack has PDU’s with TVSS, and there is a Square D integral TVSS.

There is a BAC system that is DC powered that connects to the RTU’s and lighting controls.


Sent from my iPhone using Tapatalk

 

[Jraef]

VFDs create lots of common mode noise (referenced to ground) and if not properly grounded and shielded, can cause havoc all over. Something worth looking into. Did you install the VFDs or were they done by an HVAC contractor? Those guys are notorious for curtting corners, especially on things like grounding and shielding.

Also, are they using any sort of radio communications? I have personally witnessed where someone keys up a mic and a nearby GFCI CB trips.

 

[hillbilly1]

VFDs create lots of common mode noise (referenced to ground) and if not properly grounded and shielded, can cause havoc all over. Something worth looking into. Did you install the VFDs or were they done by an HVAC contractor? Those guys are notorious for curtting corners, especially on things like grounding and shielding.

Also, are they using any sort of radio communications? I have personally witnessed where someone keys up a mic and a nearby GFCI CB trips.

I wonder if power line carrier communications might be an issue? Haven’t seen that technology used lately, maybe they were having issues?

 

[brantmacga]

VFD’s are built into the RTU’s. We supplied & connected power and BAC wiring.

I’ve spoken with three people are Square D that were involved with procurement of the gear package; none have any ideas other than power quality issues. It has been discussed internally between them and engineering also.

An engineer from the utility called me earlier; they are willing to meet on-site next week and verify the transformer connections are torqued properly, but aren’t keen on any power monitoring. They said being it’s a pole mounted transformer, they cannot leave the monitor connected outside, and they are not allowed to install in customer equipment. May not matter anyway.


Sent from my iPhone using Tapatalk

 

[ptonsparky]

I spoke to my daughter that works for T____. She searched through their technical bulletins and found nothing.

 

[ptonsparky]

VFD’s are built into the RTU’s. We supplied & connected power and BAC wiring.

I’ve spoken with three people are Square D that were involved with procurement of the gear package; none have any ideas other than power quality issues. It has been discussed internally between them and engineering also.

An engineer from the utility called me earlier; they are willing to meet on-site next week and verify the transformer connections are torqued properly, but aren’t keen on any power monitoring. They said being it’s a pole mounted transformer, they cannot leave the monitor connected outside, and they are not allowed to install in customer equipment. May not matter anyway.


Sent from my iPhone using Tapatalk

Do you have something else in mind or are you going get your own in there?

 

[kwired]

Is this design-build or engineered? If you used specified items and followed design, you should get the designer involved as well.

 

[brantmacga]

I spoke to my daughter that works for T____. She searched through their technical bulletins and found nothing.

Thanks for that attempt.


Sent from my iPad using Tapatalk

 

[brantmacga]

Is this design-build or engineered? If you used specified items and followed design, you should get the designer involved as well.

No it’s an engineered system. I know the engineering manager and chief engineer for the brand, and I have spoken with the project EE’s in the past; I would like to have a few more ducks in a row before calling the brand engineers.

I’m looking at Monday or Wednesday making the trip down to the site to meet with a utility engineer. I’m hoping to have some ideas from Square D by then also. Once I’ve been on-site to get a better account of what is happening I can start bringing in corporate engineers.


Sent from my iPad using Tapatalk

 

[brantmacga]

Do you have something else in mind or are you going get your own in there?

No I just meant maybe I’ll find something during a meeting with the utility or get some answers from square d that would negate needing a monitor.


Sent from my iPad using Tapatalk

 

[synchro]

I noticed the your drawings have an inductive loop detector, probably for sensing vehicles at a drive-up. I believe they typically operate and are resonant somewhere in the 10-200KHz range. It's a stretch, but perhaps there's interference being conducted or magnetically coupled from the loop detector that's causing the GFCIs to trip.

Also, it might be worth cycling the RTUs between different temp settings, etc. to see if you can induce any GFCI trips that way.

 

[hillbilly1]

I wonder if in order to save money, they are using the same circuit board in the breakers as the arc faults? Just changing the software (or maybe not) No need to change the assembly line when switching back and forth during production? We all know the issues arc faults have.