320/400 Dual Panel Service Bonded at the Panels - GEC Current

[kwired]

Not necessarily. The neutral/ground of each enclosure will not "itself" create a low impedance parallel path between enclosures. If there is no other low impedance path between enclosures, and the GEC taps are removed because the GEC is now connected in the trough, the only low impedance path is the neutral feeder between each panel and the trough. The trough becomes a "true" single point bond, trough connection notwithstanding.

The grounded conductor still needs bonded at the service disconnecting means. You may not have parallel path via a GEC, but your separate EGC's serve no purpose if the grounded conductor is bonded at the disconnect and still creates parallel path for neutral current between the gutter and the the disconnect, and a third path is created if you have metal raceway.

 

[JPinVA]

The grounded conductor still needs bonded at the service disconnecting means. You may not have parallel path via a GEC, but your separate EGC's serve no purpose if the grounded conductor is bonded at the disconnect and still creates parallel path for neutral current between the gutter and the the disconnect, and a third path is created if you have metal raceway.

The issue I'm concerned about isn't the parallel paths with a panel's neutral to the trough. The issue I'm concerned about is a parallel path that runs from the neutral in Panel 1 through the neutral in Panel 2 and vice versa.

 

[JPinVA]

The last picture is sort of useless without a red wire shown, but I get what you are trying to say. However, I'd point out the worst case scenario is simply maximum unbalanced current on one wire of one meter (the larger if they are different) and zero current on the other meter. The balancing of current on different phases between the two different meters can't make it any worse than that.

This is critical. I'm showing total current on the neutral and GEC. The color I used is representative of current flow as combined/cancelled as the case may be. Note the 90A imbalance of current arriving on panel 2 is opposite phase to the 60A imbalance on panel 1. We are taught that when this happens, the net is 30. And the net is 30....back to the transformer. All is well. What I"m showing is that while the 90/60 gives us 30 back to the transformer, there is 50 on the GEC taps between the neutrals.

The imbalance between phases in panel 1 and panel 2 is reconciled across the GEC taps!!!!! A perfect balance back to the transformer can result in substantial current on the GEC taps between the neutrals. For example, with my previous assumptions, 100 on panel 1 and 100 on panel 2 (split phase to panel 1) will result in a net 0 back to transformer. Hang a current meter on the GEC tap and it will read 67A!!!!

 

[jaggedben]

Not necessarily. The neutral/ground of each enclosure will not "itself" create a low impedance parallel path between enclosures.

If the trough and nipples are metal and bonded per code then they most certainly will create an additional low impedance path.

If there is no other low impedance path between enclosures, and the GEC taps are removed because the GEC is now connected in the trough, the only low impedance path is the neutral feeder between each panel and the trough. The trough becomes a "true" single point bond, trough connection notwithstanding.

Nope. The nipples and trough are another parallel path, unless you use PVC parts somewhere along the line.

This is critical. I'm showing total current on the neutral and GEC. The color I used is representative of current flow as combined/cancelled as the case may be. Note the 90A imbalance of current arriving on panel 2 is opposite phase to the 60A imbalance on panel 1. We are taught that when this happens, the net is 30. And the net is 30....back to the transformer. All is well. What I"m showing is that while the 90/60 gives us 30 back to the transformer, there is 50 on the GEC taps between the neutrals.

The imbalance between phases in panel 1 and panel 2 is reconciled across the GEC taps!!!!! A perfect balance back to the transformer can result in substantial current on the GEC taps between the neutrals. For example, with my previous assumptions, 100 on panel 1 and 100 on panel 2 (split phase to panel 1) will result in a net 0 back to transformer. Hang a current meter on the GEC tap and it will read 67A!!!!

Assuming negligible difference in impedance between the two paths (and pretending that you are using PVC between the discos and trough), here is a corrected version of your diagram.



Now take away the 60A load on the left. Note that the magnitude of the problem hasn't changed a bit. I don't believe that phase balancing between the two discos makes the issue worse.



The neutral (and trough) are likely to have a slightly lower impedance than the GEC tap route, so if anything, slightly more current will flow the 'correct' way, although the difference may be negligible.

 

[JPinVA]

Assuming negligible difference in impedance between the two paths (and pretending that you are using PVC between the discos and trough), here is a corrected version of your diagram.

I was assuming equal impedance from neutral to neutral through the GEC route, being equal to each route from neutral to trough. Hence, current flowing from a neutral through the other neutral via the GEC tap would see 2/3 of the impedance on that route vs 1/3 on the route from neutral to trough. In my example, the 90A portion from panel 2 would split 60/30. The 60A path splits 40/20. It's the phase difference that adds 30 to 20 resulting in 50 on the GEC.

Both of us are correct under our stated assumptions.

 

[JPinVA]

In the "olden days" before plastic water piping, it was not uncommon for a poor service neutral to remain undiscovered . . . until a water meter was removed.

Could not the same thing be stated today, but wrt to the GEC tap? If one panel neutral fails, the tap becomes the neutral path, with current riding the GEC to the other panel neutrals and back to source. I guess one could argue having these two paths might help avoid the consequences of a failed panel neutral...provided there was enough other issues with the failure to get one's attention.

 

[jaggedben]

I was assuming equal impedance from neutral to neutral through the GEC route, being equal to each route from neutral to trough. Hence, current flowing from a neutral through the other neutral via the GEC tap would see 2/3 of the impedance on that route vs 1/3 on the route from neutral to trough. In my example, the 90A portion from panel 2 would split 60/30. The 60A path splits 40/20. It's the phase difference that adds 30 to 20 resulting in 50 on the GEC.

Both of us are correct under our stated assumptions.

That's good as far as it goes, but your assumptions are somewhat less realistic. The neutral conductors being larger, and more likely to be shorter (since who knows how far away the electrode is), that route will likely have lower impedance. The trough is an arbitrary point in the middle of one of the parallel paths; there's no reason to view it as two paths instead of one. One could just as easily view the electrode as dividing the other path in two.

With that said, I really don't think it makes a difference to the overall point under discussion. I think one would have to get out in the real world and survey a lot of currents on real installations in order to justify that it's something the code making panel would need to address. Also, as long as Dominion prevents you from having a single GEC landing in the trough, do you really not want the NEC to give you another option?

 

[infinity]

That's good as far as it goes, but your assumptions are somewhat less realistic. The neutral conductors being larger, and more likely to be shorter (since who knows how far away the electrode is), that route will likely have lower impedance. The trough is an arbitrary point in the middle of one of the parallel paths; there's no reason to view it as two paths instead of one. One could just as easily view the electrode as dividing the other path in two.

With that said, I really don't think it makes a difference to the overall point under discussion. I think one would have to get out in the real world and survey a lot of currents on real installations in order to justify that it's something the code making panel would need to address.

I agree. If this were a real world problem bonding jumpers would be burning up all over the place. I don't see the CMP changing a century's worth of wiring services this way.

 

[JPinVA]

So....in conclusion. We all know the GEC tap is code compliant. Actually, it's more than just code compliant in that it is a designated connection scheme stated in 250.64(D)(1). We recognize this connection scheme establishes a low impedance path between panel neutrals. As such, the GEC taps will be carrying current on a regular basis during normal operation of the system. This current is acceptable.

 

[jumper]

So....in conclusion. We all know the GEC tap is code compliant. Actually, it's more than just code compliant in that it is a designated connection scheme stated in 250.64(D)(1). We recognize this connection scheme establishes a low impedance path between panel neutrals. As such, the GEC taps will be carrying current on a regular basis during normal operation of the system. This current is acceptable.

Yep and it sucks. If Dominion and other POCOs let us use the damn meter base or CT cabinet as a connection for the GEC, all this could be resolved.

 

[JPinVA]

One other point of clarification. The parallel paths at the "Service" do not adversely affect the main safety mechanisms on the load side of the Service. Ground faults are still brought back to the Service via the EGC, and connected to the neutral establishing path-to-source and ultimate OCPD protection. The parallel path mantra at sub panels (or anywhere else downstream of the Service) is still valid. Those parallel paths energize the EGC and subsequently all the objects along the way, subjecting occupants to potential touch danger. So the "no parallel paths" thinking is still a good way to think. That said, this problem does not exist at the Service in the same way it exists down the line. GFCI's still work fine as the direction the neutral current bends at the panel doesn't matter. Even GFCI breakers are fine because that current is measured before the circuit neutral returns to the bar.

My original concern, perhaps alleviated mostly herein, was putting too much current on the taps. I don't have any concerns downstream.

 

[LarryFine]

Could not the same thing be stated today, but wrt to the GEC tap? If one panel neutral fails, the tap becomes the neutral path, with current riding the GEC to the other panel neutrals and back to source. I guess one could argue having these two paths might help avoid the consequences of a failed panel neutral...provided there was enough other issues with the failure to get one's attention.

I was actually referring to services in separate houses. For example, a bad neutral in your house sending its imbalance current through the GEC to the water service piping to the neighbors' houses, using their GEC's as a pathway to their neutrals. That's one reason water meters are supposed to have jumpers around them.

 

[kwired]

The issue I'm concerned about isn't the parallel paths with a panel's neutral to the trough. The issue I'm concerned about is a parallel path that runs from the neutral in Panel 1 through the neutral in Panel 2 and vice versa.

I'd be more concerned about other pathways one doesn't immediately concern themselves with instead of this relatively short path between the two adjacent panels.

Any and all bonded objects on the premises are a potential path for neutral current if the right sequence of failures occurs. In fact many of said objects already will be carrying some of the neutral current but as long as connections back at the main leads of the panels are in good condition they should still carry a very large majority of the current because that should be the lowest resistance path. A rather small amount of current is also flowing in the grounding electrode - but again it should be high enough resistance to remain a very small amount as long as the main conductors remain in good condition.