objectionable current?

[wankster]

Im in a commercial building with a slightly older 208v 3 phase system. outside the building there is a 200a main breaker that feeds a sub panel inside the building. the sub panel has a #6 ground wire going from the neutral lug in the main disconnect box directly to the main neutral bus of the sub panel. The neutral bus is not bonded to the chassis. I'm assuming this was once legal, and as such is "grandfathered." i can measure about 2-5a of current on the ground wire when there is about 25-30a flowing through the neutral. is this an objectionable current i should concern myself with? is it worth my while to install a ground bus to the chassis of the box and bring the sub up to current standards? all of the existing circuits are currently only using conduit as ground conductor and if i end up adding circuits i will most definitely be including a grounding conductor in my wire runs. that being said, i would much rather attach to a ground bus than the neutral bus of a sub panel.

what does the NEC say about adding circuits to an existing panel that was wired as such?

 

[infinity]

Im in a commercial building with a slightly older 208v 3 phase system. outside the building there is a 200a main breaker that feeds a sub panel inside the building. the sub panel has a #6 ground wire going from the neutral lug in the main disconnect box directly to the main neutral bus of the sub panel. The neutral bus is not bonded to the chassis. I'm assuming this was once legal, and as such is "grandfathered." i can measure about 2-5a of current on the ground wire when there is about 25-30a flowing through the neutral. is this an objectionable current i should concern myself with? is it worth my while to install a ground bus to the chassis of the box and bring the sub up to current standards? all of the existing circuits are currently only using conduit as ground conductor and if i end up adding circuits i will most definitely be including a grounding conductor in my wire runs. that being said, i would much rather attach to a ground bus than the neutral bus of a sub panel.

what does the NEC say about adding circuits to an existing panel that was wired as such?

So the EGC and the neutral are connected together on both ends but the current is divided as in the BOLD text?

 

[wankster]

So the EGC and the neutral are connected together on both ends but the current is divided as in the BOLD text?

yup. both go from neutral lug in main disconnect to neutral bus in sub panel. the current numbers are rough estimates, i dont recall the exact reading at present but it was somewhere in that range. how recently did code allow the EGC to bond to the neutral bus and not the box chassis? this leaves only the conduit as the grounding conductor for the chassis. there is another sub panel that is closer to current standards, it has a ground wire that connects to a bonding bushing on each side of conduit run and the "b" sub panel has a ground bus mounted to the chassis. :roll: its been a while, but i seem to remember measuring about 40 ohms of resistance between the ground wire at the bonding bushing and the neutral bus in the "b" sub panel.

 

[ptonsparky]

I can remember that I first became aware of separating the neutral & EG to subpanels back in the mid 70s.

40 ohms between the neutral bar and a grounding bushing is ALOT. I don't care how long the conduit run is. This would not be an effective grounding path. Even 4 ohms would be on the high side. I hope your memory is poor.

 

[Dennis Alwon]

If a sub panel were wired illegally I would not add to it without correcting the problem. How can you justify adding an illegal circuit?

 

[ptonsparky]

I would be more inclined to inform the owners that a correction should be made. I cannot limit my work to only places that have no code violations whatsoever, I would be out of business in short order.

 

[wankster]

If a sub panel were wired illegally I would not add to it without correcting the problem. How can you justify adding an illegal circuit?

This is the part that confuses me. assuming this work was at one time done to code, need i update it before adding circuits? I do not want to run grounds to a neutral bus of a sub panel, so i guess i just answered my own question.

my main concern was really about the current on the ground wire in the "A" sub panel. i really don't see how its even functioning as a ground, but, rather an extra neutral wire. i suppose that answers my own question again. i should add a ground bus to that sub and re-route that ground conductor to the bonded ground bus.

 

[roger]

When was the building and this electrical service originally permitted?

As Tom said, your situation has not been propper as far back as the 70's.

Roger

 

[wankster]

When was the building and this electrical service originally permitted?

As Tom said, your situation has not been propper as far back as the 70's.

Roger

probably 70's. Im not sure about panel B, however. as stated, it does have a separate ground bus, although, the method of bonding the ground conductor is not very impressive.

I'm wondering if there is/was an exception for a short run sub panel allowing it to wired like main panel. it cant be more than 6' from the main disconnect box, and it is the main source for branch circuits, as the main disconnect has only provisions for a single breaker.

guess its time to revisit the grounding/bonding section of my NEC book.

 

[wankster]

the plot thickens

the plot thickens

so i was out there today and did a lil more investigating.

the ground connection to panel B is actually nothing more than a #6 wire from a bonding bushing on the supply conduit to a ground bus on the chassis. i measured the resistance again between the EGC at the bonding bushing and the neutral bus and my meter read about 1.6 M ohms! i put the meter on manual ohms and it was reading just over 600 ohms! voltage to the ECG is good. i have witnessed first hand a ground fault on a circuit fed by panel b and there was quite a bit of fireworks. the circuit was GFCI protected and the GFCI tripped instantly but the breaker did not trip. judging by the amount or arc-welding that went on i would assume the breaker would have tripped had the GFCI not been there first. next time I'm out there I'm going to check the entire conduit run and tighten up all the set screws. much of the conduit has been painted over, but i cant imagine there is paint up inside the fittings that would be hindering continuity.

and back to panel A and the main disconnect box....

i did some investigating in the main meter/disconnect panel and initially it looked as though the neutral was floating and not even bonded to the chassis! upon further inspection i found the source of bonding: a single 10-32 machine screw that goes through the insulated neutral bus and into the chassis of the box. :roll: The resistance between the EGC and the main neutral bus is 5 ohms, not exactly ideal, i guess. there is an ECG that goes straight from a bonding bar on the chassis to the grounding electrode and does not have any direct bonding to the neutral bus other than through the aforementioned 10-32 screw.

 

[ptonsparky]

Does your meter have the option of "zeroing"? I have had to replace tips and leads before. Usually after I start getting eratic readings and am 20 miles from replacements.

 

[wankster]

Does your meter have the option of "zeroing"? I have had to replace tips and leads before. Usually after I start getting eratic readings and am 20 miles from replacements.

i dont think my meter has a zero option, but i could be mistaken. its a fluke 115.

one of my probes does have a pretty hefty chunk blown out of it and i should probably replace it.

it seems to read voltage just fine, and when i measure resistance from two points that i know have good continuity it reads as such.

 

[erickench]

There usually is objectionable current between the disconnect and the meter because the meter has the neutral bonded to the chassis. However any neutral current flowing through an EGC between the disconnect and a downstream subpanel should not be allowed.

 

[brian john]

There usually is objectionable current between the disconnect and the meter because the meter has the neutral bonded to the chassis.

That depends upon the conduit or cable type you utilize between the meter and Service Disconnect.

 

[wankster]

so today i installed a ground bus in the A sub panel. OC is gone on the ECG. now that the A sub panel has a good bond with the ECG i decided to check on the ground to neutral resistance at panel B, as it is fed by panel A. now it tells me i have 90 ohms resistance, rather than the 600+ it was reading before. i know its still not good enough, but its a start. i still need to check all the set screws on the couplings. hopefully i find some loose ones. if that doesn't work i may be forced to run an ECG from panel a to panel b. that would be a serious PITA, but i guess i gotta do what i gotta do.

is it legal to run a dedicated conduit for the ECG to panel b, rather than trying to get it through the conduit that feeds the supply wires?