Current on GEC and Bonded EMT

[dereckbc]

note the sub in question is 3 wire feeder with rigid between main panel and subpanel to a detached building with its own ground rod. No bond screw visable on neutral bus.

Do not go by visual or Ohm Meter readings. The correct way to tell is to make sure there is a good amount of load current on the Sub Panel and then measure the voltage between N-G buss bars. If you see ZERO VOLTS then you know for certain it is bonded very close by. If isolated you should see some voltage up to 1 or 2 volts. Depends on the length, size of feeder cables, and current. Current clamping of the EGC feeder cable will also tell the story as you will see about equal current on both N and G feeder cable.

 

[ardmi]

Do not go by visual or Ohm Meter readings. The correct way to tell is to make sure there is a good amount of load current on the Sub Panel and then measure the voltage between N-G buss bars. If you see ZERO VOLTS then you know for certain it is bonded very close by. If isolated you should see some voltage up to 1 or 2 volts. Depends on the length, size of feeder cables, and current. Current clamping of the EGC feeder cable will also tell the story as you will see about equal current on both N and G feeder cable.

I forgot to mention, In the subpanel , I measured .4 vac between the gec and the neutral bus bar with a 10A load with a low impendence meter. I also measured 0 current on the gec with the 10A load. To test my meter I was getting 0vac and continuity alarm when both probes were on the neutral bus.

I assume that means the neutral and ground are not bonded in the subpanel.

 

[GoldDigger]

I assume that means the neutral and ground are not bonded in the subpanel.

Absolutely correct.

Now, one test that remains is to put the clamp on the conductors themselves (bundled) for that branch where you measured current by clamping the whole pipe. That should tell you very conclusively what is happening, just not where.
(Sorry for diverting the thread temporarily from the the other of the two original questions.)

,

 

[ardmi]

Absolutely correct.

Now, one test that remains is to put the clamp on the conductors themselves (bundled) for that branch where you measured current by clamping the whole pipe. That should tell you very conclusively what is happening, just not where.
(Sorry for diverting the thread temporarily from the the other of the two original questions.)

,

Yes indeed, that was my first line of thought, unfortunetly I wasn't able to get into a junction with permission to disassemble joints.

What do you think I would likely see if I clamped the individual conductors outside of the conduit?

Also I found the following thread that dismisses clamping an entire conduit as an accurate measurement... (note I am not second guessing you Golddigger, just curious what I might be missing here) http://forums.mikeholt.com/showthread.php?t=90372

 

[ardmi]

Come to think of it, there are likely some 3way travelers and switch legs in the raceways that play into this.... how do they play into unbalanced raceways and canceling out etc?

If a current was flowing on a conduit, would it be likely this current would transfer to nearby junction boxes and other raceways connected to them? How much of this would be fed to the rest of the bonded raceways adjacent?

For instance, theres a few conduits in close proximity to the circuit in question that are showing some current as well with the same method of clamping around the entire raceway. The current on the surround conduits are in the 3-5 mA range at times.

 

[ardmi]

Hi - I haven't had the chance to get back and recheck, its out of my hands. Meanwhile I am still trying to understand.

Heres another scenario.

PANEL>JUNCTION1>JUNCTION2 etc

I did get a non-zero current reading between the panel and junction 1 but know for certain there is an even # of conductors in that conduit So two circumstances.

There are 3 circuits in the conduit. Between panel and junction 1 (before other conductors enter the conduits downstream (ie switch legs) there are 3 hots, 3 neutral.

If it was wired correctly there would be no inbalance on that portion of conduit. ie: power in/power out for each of the 3.

However, I believe 1 neutral is part of a MWBC. That would leave 2 neutrals run in parallel for the third circuit. If that third circuit now had two returns, would that cause the non-zero current readings?

PLEASE NOTE: In all cases, opening circuit 2 in the drawings below, makes the non-zero current readings disappear on the conduit:

Heres a crude drawing of what I am describing above:



Heres another drawing of if all 3 circuits returns were tied together, which is yet another scenario. What would occur in this case (the neutral is overheated correct?)



I apologize if I am not making much sense at this point, but I do appreciate all the great advice Golddigger and the others have provided.

 

[ardmi]

PLEASE NOTE: In all cases, opening circuit 2 in the drawings below, makes the non-zero current readings disappear on the conduit:


.

This is an error. Actually opening any combination of the 3 circuits had no effect on whether current was being detected on the conduit when the other circuits were under load.

 

[ardmi]

This is for golddigger who apparently is probably the only guy actually following along.

I spoke to homeowner (my neighbor down the block actually) and they had another electrician out.

He removed the redundant wire from the mwbc that was in that conduit. Apparently thinking that neutral and ground was tied together, the electrician disconnected the neutral of the circuit in question from the busbar and tested no continuity between box(egc) and neutral in the downstream homerun. Concluded neutral/ground were not bonded downstream.

Wrote that the ungrounded conductors were ran with the respective neutrals however there was many travelers from lighting circuits also in same raceway with these conductors.

Said that he had never used a clamp meter around an entire conduit as we were doing. But decided "the current reading are induced because of the magnetic field of the 3 circuits inside of the conduit (1 mwbc and another standard circuit as well as numerous travelers), that it posed no safety issue"....

Is induction a fair assessment?

 

[GoldDigger]

I would rather say that the most likely cause for the meter reading was current in one or more travelers IF any of the lights in question were on during the testing.
The only way to properly check for current in the raceway metal itself is to open that path and wire in an ammeter or find a way to clamp just the raceway or just the wires.
I am inclined to agree that it is not a hazard given that the wire configuration has been confirmed, but I do not buy "induction" as a cause.

Tapatalk!

 

[ardmi]

The only way to properly check for current in the raceway metal itself is to open that path and wire in an ammeter or find a way to clamp just the raceway or just the wires.
.

Tapatalk!

On the invoice the electrician noted that he separated the conduit at a couplings and measured .090VAC under load between the two disconnected conduits with a low-z meter. The homeowner indicated that was one of the electricians reasonings that the reading was a product of the wires inside and not current on the conduit and felt it was not a shocking hazard.

Does that miniscule voltage reading indicate anything? I assume some difference of potential is normal in such a test?

I will let this thread die in peace after this last inquiry... the information here is more than unwieldy.But I thank you all and esp. you GOLDDIGGER.

 

[gar]

140402-2106 EDT

ardmi:

Don't disconnect the two ends of the conduit, but measure the voltage drop between the two ends of the conduit directly from the conduit with a high impedance meter, and with a filter tuned to 60 Hz at the meter input. Keep one test lead tight against the conduit to the other end. Twist the test leads coming from the conduit to the meter.

Determine the resistance of a piece of equivalent conduit of the same length. See my conduit resistance measurement method in photo P21 at http://beta-a2.com/EE-photos.html .

A slightly different aspect. Three tests. 10 A 60 Hz thru a single wire with the wire thru two nipples, one brass and one steel. I could not see any significant magnetic shielding effect of the magnetic field seen by the clamp on current probe between either nipple or plain wire. The nipples had no current thru them. The only current was in the wire threaded thru the center hole in the nipples.

.