Another Neutral Ground Bond Question

[mcmanumi]

I apologize in advance as I know this subject has been discussed exhaustively, but I feel my application is a bit unique, although similar to a thread started by superdave02 in September 2021 titled "Louisiana Emergency Generators". The majority of our business is with local electric utilities... We provide temporary power via mobile diesel generators anywhere from 20kW to 2MW for either emergency outages or planned ones (i.e.-system upgrades). I would estimate that 75% of the time we are connecting our generators to either the secondaries of a pad transformer or overhead directly to the secondary crib (other 25% of the time were going directly to the customers gear). In the case of a pad transformer connection, we usually connect the generator 1-of-3 ways...

1. connect the generator leads directly to the secondary bushings (primary elbows/terminators stood-off/lifted)
2. lift the secondaries off the bushings and connect the generator leads back-to-back on the secondaries
3. connect the generator leads directly to the secondary bushings, but leave the primaries connected so we can backfeed the transformer via the secondaries and pick up multiple transformers connected to the same primaries

In all cases, with the exception of an extremely rare 240 delta or 480 delta, we are providing 120/240, 120/208 or 277/480V (depending on the transformer). The majority of our generators are neutral ground bonded & most of the time the load is bonded as well. Not considering code, should we lift the generator bond? Like I mentioned above, our application is purely temporary until the electric utility can establish "normal" supply. In most cases, it isn't possible to simply pound a rod in the middle of a city intersection or a sidewalk and finding a dedicated ground is nearly impossible. Thanks in advance...

 

[LarryFine]

Not considering code, should we lift the generator bond?

Welcome to the forum.

Yes. In theory, even without code the reasoning is the same.

Otherwise, why run an EGC at all? Just run lines and neutral.

 

[mcmanumi]

Welcome to the forum.

Yes. In theory, even without code the reasoning is the same.

Otherwise, why run an EGC at all? Just run lines and neutral.

Thanks LarryFine... Just to clarify, you recommend removing the bond between the neutral & ground on the generator when the neutral & ground are bonded at the load? Can you go into detail of why this is necessary & what the possible downfalls are if the bond is in place at both the generator & the load? Again, in my scenario, we do not have a ground conductor between the generator & the load, nor do we have a ground conductor from the generator to a dedicated ground (i.e.-ground rod). Thanks again for your input... This has been a discussion in our company for many years now & I'm looking to get a concrete answer.

 

[zbang]

Are you operating under the NEC or under the utility's NESC? (Not sure it matters, but it might.)

Double-bonding is a general issue, not specific to this installation-
When the N and G are bonded twice, part of the return current that should all flow on the neutral will also flow on the grounding conductor (per Kirchoff's law about parallel paths) which will create a voltage drop on the grounding conductor. With that being bonded to equipment cases, it opens up the possibility of a non-zero voltage between those cases and creates a shock hazard.

That said, you still need a separate grounding conductor from the generator to the tie-in to carry fault current if there is a fault-to-ground in the leads from generator to tie-in.

Grounding electrodes...... if the generator is a "structure" (not on wheels or some other exceptions), you need one there. And if it's only connected to the generator frame but not also bonded to the neutral, it's only able to make an inspector happy, doesn't increase safety. Also remember that the earth is a terrible conductor and can't/won't trip an overcurrent device when there's a fault.

Or... since I'm being chatty-
Look at all the possible fault locations and current paths, are there any which won't trip the OCPD?
Look for places where voltage drop might be enough to create a shock hazard. (Also, check these forums for discussions of "step potential".)

 

[ActionDave]

I would leave the neutral bonded at the generator. You didn't say how you are connecting to the power company transformers but I would guess it is flexible cord of some kind not pulled in any metal conduit.

 

[mcmanumi]

I would leave the neutral bonded at the generator. You didn't say how you are connecting to the power company transformers but I would guess it is flexible cord of some kind not pulled in any metal conduit.

The majority of the time we use 4/O or 2AWG extra flexible Class K cable with Eaton Cam-Lok connectors. When landing the cables at the load, we connect a number of different ways depending on the termination point (i.e.-bare copper, compression terminals, 3M Stem Connectors).

 

[mcmanumi]

Are you operating under the NEC or under the utility's NESC? (Not sure it matters, but it might.)

Double-bonding is a general issue, not specific to this installation-
When the N and G are bonded twice, part of the return current that should all flow on the neutral will also flow on the grounding conductor (per Kirchoff's law about parallel paths) which will create a voltage drop on the grounding conductor. With that being bonded to equipment cases, it opens up the possibility of a non-zero voltage between those cases and creates a shock hazard.

That said, you still need a separate grounding conductor from the generator to the tie-in to carry fault current if there is a fault-to-ground in the leads from generator to tie-in.

Grounding electrodes...... if the generator is a "structure" (not on wheels or some other exceptions), you need one there. And if it's only connected to the generator frame but not also bonded to the neutral, it's only able to make an inspector happy, doesn't increase safety. Also remember that the earth is a terrible conductor and can't/won't trip an overcurrent device when there's a fault.

Or... since I'm being chatty-
Look at all the possible fault locations and current paths, are there any which won't trip the OCPD?
Look for places where voltage drop might be enough to create a shock hazard. (Also, check these forums for discussions of "step potential".)

Thanks zbang... All of our generators are trailer mounted, so yes, on wheels. So, if I understand you correctly, you recommend removing the Neutral Ground bond at the generator & that's it (assuming the load is N G bonded)? And since it's on wheels, no grounding electrode / rod is necessary at the generator? I've attached a simple sketch (277/480V scenario) of what I believe you're saying...

 

[LarryFine]

Thanks LarryFine... Just to clarify, you recommend removing the bond between the neutral & ground on the generator when the neutral & ground are bonded at the load?

I'm saying that either the bond should be opened in the generator, or a separate EGC should not be used.

Can you go into detail of why this is necessary & what the possible downfalls are if the bond is in place at both the generator & the load?

If the neutral and ground are bonded at both ends, the ground carries as much as half of the neutral current.

Again, in my scenario, we do not have a ground conductor between the generator & the load, nor do we have a ground conductor from the generator to a dedicated ground (i.e.-ground rod).

Then there's no parallel-conductor issue. Note that ground rods have nothing to do with this discussion.

 

[LarryFine]

That said, you still need a separate grounding conductor from the generator to the tie-in to carry fault current if there is a fault-to-ground in the leads from generator to tie-in.

If there is no genny feeder neutral, I agree. Technically, an EGC would act like a service neutral during a fault.

 

[LarryFine]

So, if I understand you correctly, you recommend removing the Neutral Ground bond at the generator & that's it (assuming the load is N G bonded)?

I've attached a simple sketch (277/480V scenario) of what I believe you're saying...

You said you were not running a ground(ing) conductor, so you should not remove the bond.

Your sketch shows both a neutral (grounded) conductor and an EGC (grounding) conductor.

If running both, break it; if running only one, leave it.

 

[ActionDave]

The majority of the time we use 4/O or 2AWG extra flexible Class K cable with Eaton Cam-Lok connectors. When landing the cables at the load, we connect a number of different ways depending on the termination point (i.e.-bare copper, compression terminals, 3M Stem Connectors).

That is what I suspected. So you don't have any metal paths between the transformer or building and the generator where objectionable neutral current is going to flow.

And there is no reason to put a ground rod in anywhere.

 

[zbang]

I think Larry and I are saying the same thing. And I will grant that the idea of "objectionable current" on an insulated grounding conductor probably doesn't matter in this case. My view is that code requires the EGC from generator to utilization and a that there be only a single N-G bond, but the reality is that it may not matter in some/most/many cases.

 

[ActionDave]

I think Larry and I are saying the same thing. And I will grant that the idea of "objectionable current" on an insulated grounding conductor probably doesn't matter in this case. My view is that code requires the EGC from generator to utilization and a that there be only a single N-G bond, but the reality is that it may not matter in some/most/many cases.

I agree with you also.

 

[mcmanumi]

I'm saying that either the bond should be opened in the generator, or a separate EGC should not be used.


If the neutral and ground are bonded at both ends, the ground carries as much as half of the neutral current.


Then there's no parallel-conductor issue. Note that ground rods have nothing to do with this discussion.

Thanks Larry... Why do you say that "ground rods have nothing to do with this discussion"? I ask because there are many times when our customers ask if we're going to install a grounding rod to connect the generator / trailer. Let me make one thing clear... I am not a licensed electrician nor do I claim to have a fraction of the knowledge that a licensed electrician possesses (i.e.-local / national codes). My education is in Marine Engineering with career experience working for & contracting with local electric utilities in both infrastructure construction & troubleshooting. This topic is continually brought up within our industry & I'm trying to find a solid answer not only for our safety, but for our customers & the public who may come in contact with our equipment while on the jobsite. Again, thanks to everyone for sharing their knowledge in this subject...

 

[zbang]

(sorry, this got longer than I expected, but the short answer is "It depends")

You probably won't find solid answers- the code is too vague in the relevant places so other principles must be extended to these situations.

The rods/etc are correctly called Grounding Electrodes (GE)and the wire(s) to them are Grounding Electrode Conductors (GEC). There's also the the Equipment Grounding Conductor (EGC) which is the green wire from the equipment, looks the same but has a different function (carrying fault current). A GE can be many things, rods are only one option.

I'd doing a bit of hand-waving here, but generally the GE itself serves about zero purpose in protecting people from electric shock. The real purpose is to minimize voltage gradients across a larger area in case of lightning (lots of articles around MH.com and other places about that).


Rereading 250.34 (Portable and Vehicle-Mounted Generators) and following Charlie's Rule (paraphrased- "read it again") I still get that the generator must be connected to a GES, but that doesn't mean that the GES must be directly located at the generator; a sufficiently-heavy wire from the generator's grounding terminal to whatever the load is using as a grounding system would usually do.

Second question-
Separate structures must have their own grounding electrode system (e.g. rods, that's in 250.32), but there is a sometimes fierce debate of whether a trailer-mounted generator is a structure. I say it isn't.


Circumstances differ....
If I parked a generator next to a building and wired that into a disconnected switchboard, I would run 5 wires in (3-phase, N, G) and use the building's GES. Would not drive a separate rod unless the customer or AHJ insisted -- billing for work is often easier than arguing about not doing it.

Flipping it, if the setup is at a farm's service point and the generator is basically in a field next to a pole, I'd drive a couple of rods to make a GES since I'm not sure what's available. I'd also probably bond the N-G at the generator to make it look like the power company. (There are a lot of "it depends" there, like whether this is connecting on the line or load side of the main service point.)

Or... Generator parked in the middle of the street and wired into a manhole by PoCo lineman.... that becomes tricky if they're connecting to the PoCo lines, which won't have a GES handy. I'd bond the N-G at the generator and ask the lineman where to find a "ground". Or make it their engineer's problem .


I hope that helps.

 

[zbang]

Relevant and recent- https://forums.mikeholt.com/threads/purpose-of-ground-rod.2566343/