Ground Currents in Parallel Feeder Equipment Grounds

[HYDL]

I have an 800 amp feeder between the building main where the ground and neutral are bonded. There are 3 parallel 350 Kcmil and 3/O equipment ground per EMT conduit. With the equipment grounds connected at both ends, does this result in some circulating currents due to induced current in each of the equipment grounds.

I measure around 20 amps on each equipment ground individually. If the amp meter measures them collectively, there is around 12 amps.

 

[davidbeach]

Possibly some circulating ground currents. Not having the ground conductors connected at both ends would cause more problems than that circulating current.

 

[RalphChristie]

Yes, there will be some circulating currents induced if the cable carry AC and the metalic sheath is grounded at both ends, especially on single core cables. The magnitude of this circulating current depends on the current in the conductor and the resistance of the loop formed by this layer and the ground path.

See:

http://www.nexans.ca/egy/tecdoc/7.htm

This circulating current can be quite high sometimes. On one of our installations (at that time new) we've meassured over 400A circulating current when grounded at both ends. It stressed the cable a lot and eventually it failed - not even two weeks after energized.

Check also:
ANSI/IEEE std 575-1998 – “IEEE Guide for the application of sheath bonding methods for single-conductors cables and calculation of induced voltage and current in cable sheaths.”
IEC 287-1-1 1995.8 “Electric cables – Calculation of current rating, part 1: Current rating equations (100% load factor) and calculation of losses, section 1: General”


[red]Failure seldom stops us, it is the fear for failure that stops us - Jack Lemmon[/red]

Make the best use of Eng-Tips.com
Read the Site Policies at FAQ731-376

 

[rbulsara]

Before jumping to conclusions, it is not clear from your post, where is this 'feeder'? You say its between the building main and what??

And also exlain "collective" measurement by amp meter? Is this done using a flex around all conductors?

Equipment ground condutcot is always connected at both ends to the eqipment enclosure, ground bus or the ground lug.

What is the current in phases? What does it serve?

 

[HYDL]

Sorry about the lack of info. This feeder is between the main disconnect outside of the building and a main panel inside the building. The Neutral - ground bond is at the panel outside. The actual current from the main panel to the ground rod is fairly low (around 2-3 amps). However, each individual equipment ground in the conduits has around 20 amps. If you put the amp meter around all three at the inside panel, there is around 12 amps.

So, there is an equipment ground running in each of the three conduits and they terminate in the ground bus at each side along with bonding to the conduit at each end.

 

[rbulsara]

I suspect you have N-G bond at the interior panel as well (0r some where downstream) and your netural current is being shared by the G and N between the two equipment.

Is there a neutral between the two equipment? Do you have L-N loads?

Measure the current in the neutral that is coming in the outdoor disconnet 'before' the N-G bond. That would equal the the current in all the grounds plus the N at the interior panel.

The ground rod should also be at the outdoor switch where service entrance N-G bond exist. Ground rod at interior panle is just a supplemental rod, although wont hurt, it is not required by Code.

Have someone qualified check out the installation.

 

[cuky2000]

There is a good possibility to reduce the circulating current grounding the cable at only one end only without creating excessive voltage rise at the opposite end provide that the length of the feeder is less than 2,000 ft (~600m).

Detail calculation is recommended to verify the above statement.

 

[davidbeach]

cuky2000, I'm not quite sure what you mean about "grounding the cable at only one end". If you mean a connection to a grounding electrode, I agree. If you are somehow suggesting a discontinuity in the equipment grounding conductor, since it is "ground", then, no, that cannot be done. Saying "grounding the cable" suggests to me grounding cable shields and it sounds from the OP that the circuit is all unshielded single conductor cables.

I think rbulsara may be on to something, if there is a downstream N-G connection it will cause current to flow on the ground, but that current should add rather than cancel when all three ground conductors are measured together.

Since the circulating currents tend to cancel when taken in total, I'd look for a loose connection somewhere that is causing the phase currents to be unbalanced. This unbalanced (net current) will drive current one direction in one ground conductor and the other direction in the other two ground conductors.

 

[HYDL]

To Answer rbulsara's questions. There is a neutral in each conduit and the building does have L-N loads. We will have to measure the current in the Neutral before the N-G bond.

We did identify several N-G bonds in building panels and have removed them, which has helped. However, we still have 14-16 amps on each of 3 equipment grounds. I'm wondering if we are missing a bond somewhere yet.

Back to davidbeach: It seems like I should not be getting induced current on the equipment ground conductors in these parallel feeders. And, yes I thought they should all add rather than cancel out. That's why we've been wondering about some circular currents being induced somehow.

 

[HYDL]

Some more info.

The Neutral currents in the 3 parallel feeders are:
13.4a - 14.4a - 14a
With amp meter around all 3 = 21.7a

The equipment ground currents in the 3 parallel feeders are:
17 - 16 - 14
With amp meter around all 3 = 12a.

Even though the N-G bond is at the Main Disconnect, the equipment grounds are basicaly bonded together at both ends via the grounding bus.....the length here is around 100 feet.

 

[itsmoked]

What if some of the final equipment is N-G bonded. If you had panel N-Gs then it wouldn't surprise me that some of the equipment is also N-G bonded.

 

[GordS]

Currents in equipment ground conductors are normal. I think they result from the ground conductor not being equidistant to all of the phase conductors. This means the magnetic flux due to each phase that links the ground loop does not cancel out. Faradays law says a voltage is induced in any closed loop proportional to the change of enclosed magnetic flux. This will cause circulating current.

 

[waross]

Here in the third world, I often see installations with "A" phase in one conduit, "B" phase in the second conduit and "C" phase in the third conduit. I also see random distribution of phases in the parallel conduits. This causes induction heating of the conduits, and the conduits are usually too hot to hold comfortably. Re-arranging the conductors per code results in the conduits cooling down.
The point is, have you measured the currents in the phase cables? If for some reason the currents are not balancing properly the unbalance can cause induced currents. With an unbalance condition, the induced currents in each conduit will not be in phase with each other. This will account for the currents not adding up. Check the current on the ground cables leaving the panel. You could have ground currents originating from equipment. I would suspect a combination of factors; 1> ground currents originating from in plant equipment.
2> Grounding cables shareing the neutral current.
Please let us know what you finally discover. Thanks.