Sub panel bonding

Kopper

Kopper

Member
Location
Lewisburg, Pennsylvania, USA
Occupation
Electrical Contractor
It's likely that this question has been asked before in multiple ways and from multiple perspectives. I didn't find what I was looking for in the search that I did.

I have a customer who - about 15 years ago had a subpanel installed next to the main panel by another electrician. The grounds and neutrals were not kept separate in the new subpanel.

The situation is unique - and there is no "problem" or "issue" but I want to understand why and how this poses a potential safety hazard when there are pool circuits - or any other circuits for that matter - installed in the sub panel.

Here is the situation:
The 400A 240V single phase meter base is installed at the barn.
There is a 200A disconnect for the barn and a 200A disconnect for the house.
There are 3 wires running to the house from the disconnect - 2 phase wires and 1 neutral wire
There is a ground rod installed at the house and the main panel has the grounds and neutrals connected together
the 100A subpanel has a 3 wire with ground but the grounds and neutrals are connected together in the subpanel

I realize that this does not meet current code requirements but I am trying to wrap my head around the why of it all to better understand what I am doing when the grounds and neutrals are bonded together in a subpanel or the grounding is done at a panel downstream from the first disconnect after the meter.
For me it is important to understand the concept of what is being accomplished not just that it must be done this way cause code says so...

I understand that providing separate paths for ground and neutral is a good thing but what does it do differently than when the ground and neutral are together and why?

I hope my question is sufficiently clear to minimize unnecessary "bunny trail" answers that lead to nowhere!
 
What would happen to all the exposed live parts of the branch circuits connected to the 100A subpanel if its neutral fails?
 
If the main disconnects are in the barn then the feed to the house is a feeder in my view and should be 4 wire 2 hots, neutral and ground unless in metal pipe that can serve as the ground.

The sub panel is just wrong. You loose the neutral and then the smaller ground carries the neutral current and can overheat cause a fire etc.
 
Eddie702 said:
If the main disconnects are in the barn then the feed to the house is a feeder in my view and should be 4 wire 2 hots, neutral and ground unless in metal pipe that can serve as the ground.

The sub panel is just wrong. You loose the neutral and then the smaller ground carries the neutral current and can overheat cause a fire etc.
Click to expand...
I understand the what-if's.

I understand to that if a wire breaks there are problems regardless of whether it was done correctly or not. Those are the bunny trails that "get my goat" in this forum.

I am trying to understand what the technical reasons are for bonding the ground only in the first disconnect after the meter base. If it is only to minimize what-if's that's ok but it seems to me there might be "better" reasons than just that.

I have some ideas of what some of those technical things might be and keeping things separate - as code requires - makes a lot of sense but I thought maybe someone here would have better answers than "just because" and "because this might happen" etc.
 
They wired the subpanel like a separate service?

When you say grounds and neutral you will need to be more specific. The GECs to neutral or an EGC to neutral in the subpanel?

It was NEC 250.32(B) exception 1
 
Elect117 said:
They wired the subpanel like a separate service?

When you say grounds and neutral you will need to be more specific. The GECs to neutral or an EGC to neutral in the subpanel?

It was NEC 250.32(B) exception 1
Click to expand...
Is this better?


Here is the situation:
The 400A 240V single phase meter base is installed at the barn.
There is a 200A disconnect for the barn and a 200A disconnect for the house.
There are 2 ground rods at the disconnects and a bare ground wire is connecting the ground rods to each of the disconnects
There are 3 wires running from the disconnect at the barn to the house panel - 2 phase wires and 1 neutral wire
There is a ground rod installed at the house and a bare ground wire connects the neutral/ground bars in the the main panel at the house to the ground rod - the grounds/neutrals are bonded
The equipment grounding conductors and branch circuit neutrals are connected together on the same ground/neutral bars in the main panel at the house
the 100A subpanel is connected to the main panel with a 3 wire with ground but the equipment grounding conductors and branch circuit neutrals in the subpanel are all terminated on the same ground/neutral bars
 
Kopper said:
Is this better?


Here is the situation:
The 400A 240V single phase meter base is installed at the barn.
There is a 200A disconnect for the barn and a 200A disconnect for the house.
There are 2 ground rods at the disconnects and a bare ground wire is connecting the ground rods to each of the disconnects
There are 3 wires running from the disconnect at the barn to the house panel - 2 phase wires and 1 neutral wire
There is a ground rod installed at the house and a bare ground wire connects the neutral/ground bars in the the main panel at the house to the ground rod - the grounds/neutrals are bonded
The equipment grounding conductors and branch circuit neutrals are connected together on the same ground/neutral bars in the main panel at the house
the 100A subpanel is connected to the main panel with a 3 wire with ground but the equipment grounding conductors and branch circuit neutrals in the subpanel are all terminated on the same ground/neutral bars
Click to expand...

I don't think that was ever allowed.
 
Elect117 said:
I don't think that was ever allowed.
Click to expand...
That wasn't my question! The AHJ in my area DID allow it. But you guys can't see past the technical stuff that isn't to code to discuss the deeper question of what all happens when things are done this way versus the right way and how and why does it work better. I'm not trying to get around code or disputing that code should be different. We know it works the way it was done. But what are the technical reasons that the "right" way is superior to the scenario I presented. What makes it safer? I need to understand this well so I can explain it well to my customers, so I can be sure that what I am doing, in the little nuances that code doesn't address SPECIFICALLY, will be in alignment with what is good and correct and safe. I hope I am not confusing things more! :eek::cry:
 
Before 2011 the installation you describe was legal if there were no parallel paths from the detached structure. So if this was permitted under the 2008 or earlier NEC there would be no problem.
 
The feeder to the house with 3 wire was allowed up through the 2008 NEC. The subpanel being only a 3 wire feed was never allowed, but is one of the most common errors (or allowed deviances) you see. If the sub and main are adjacent, I don't see much of a safety issue.

Why a 3 wire feed to a detached building was suddenly deemed unsafe in the 2011 code I'm not sure. It is probably the desire for a "safer" 4 wire system but the NEC can't force the utilities to run a 4th wire so you have to allow it to some point and then transition. You can ask why for many things in the NEC, like why do so many thins have to have GFCIs now without much of a history of electrocutions.
 
roger said:
Before 2011 the installation you describe was legal if there were no parallel paths from the detached structure. So if this was permitted under the 2008 or earlier NEC there would be no problem.
Click to expand...

So my question is - again: What are the technical reasons why separating grounds and neutrals is superior? Are there safety issues? How does it all work that this is so important? I'm not trying to find fault with code I am trying to understand this. If there are real issues with the way it was done then I want to be able to explain that to my customer. If I can't give him a clear answer other than the fact that code changed over the years he's gonna say "then let's not fix what ain't broke". If it is a safety issue then I want to be able to explain that to a customer well so he can understand that. I can't sell him a good upgrade just on new code requirements - although that may be a "good enough" reason. But if it really is a deeper issue then I want to know so I can care for my customers well! It's not nearly enough for me to know what code says and just do it that way blindly. I must know why it says it that way. I guess maybe I'll need to find some good electrical theory courses but I might be left hanging with a lot of unanswered questions there too.:unsure:
 
If you just want to just talk theory and not NEC thats fine. I'd suggest to evaluate it in IEC terminology;
they break it down into a simple system in 3 steps;
  1. The relationship between the power supply (generator or transformer) and Earth.
    If the power supply is grounded, Call the system "T" for Terra.
    If the transformer is completely isolated and ungrounded, I call it "I" for Isolated.
  2. The relationship between exposed-conductive-parts of the installation and Earth.
    These are metal parts that could become energized if a fault occurs. In most systems, they are bonded to the point on the power supply that is connected to Earth, typically the transformer neutral. Use "N" if the neutral is solidly connected to a grounding electrode system.
    Some systems use a grounding method that’s independent of the power supply, such as resistance grounding or completely ungrounded (floating) systems, use "T" for Terra for these less common systems, since they don’t rely on a solid connection to the power supply’s grounded neutral.
  3. The 'protective conductor' or the wire used for protection against lightning, surges, or accidental contact with higher voltage systems, and to ensure an effective ground fault path.
    "S" when the neutral and protective conductor functions are provided by separate conductors.
    "C" when both functions are combined in a single conductor.
So in IEC terms the system you’re describing is a TN-C system, also known as a Multi-Grounded Neutral (MGN). It’s used for outdoor distribution to limit voltage from lightning, surges, or accidental contact with higher-voltage lines. In this system, the combined neutral/protective conductor is grounded at multiple points, the further you get from one of the grounding points the more chance the protective conductor in your case the neutral will create a Earth - Neutral voltage.
Hence with TN-C a open neutral to the pool subpanel will create a deadly voltage drop across the pool and with a TN-C-S (what the NEC requires) it wont.
 
suemarkp said:
The feeder to the house with 3 wire was allowed up through the 2008 NEC. The subpanel being only a 3 wire feed was never allowed, but is one of the most common errors (or allowed deviances) you see. If the sub and main are adjacent, I don't see much of a safety issue.

Why a 3 wire feed to a detached building was suddenly deemed unsafe in the 2011 code I'm not sure. It is probably the desire for a "safer" 4 wire system but the NEC can't force the utilities to run a 4th wire so you have to allow it to some point and then transition. You can ask why for many things in the NEC, like why do so many thins have to have GFCIs now without much of a history of electrocutions.
Click to expand...
Maybe I'm asking a question that I shouldn't even ask. I guess I'm just curious to know if there are reasons behind it. It helps significantly to understand as many "whats" and "whys" as possible when wading through nuances that are existing or not specified. I'm ok with "I don't know" or "there's no good reason". That's much more of an answer than "just because code says so". I am very happy to do something "just because code says so" but I will always strive to understand "why code says so" so I can be more effective in my work than just blindly and happily doing it "just because code says so".
 
You’re looking for an ultra-simple explanation for something that’s too nuanced to offer one. Don’t get upset because you’re unhappy with the attempts to provide a satisfactory answer.
 
tortuga said:
If you just want to just talk theory and not NEC thats fine. I'd suggest to evaluate it in IEC terminology;
they break it down into a simple system in 3 steps;
  1. The relationship between the power supply (generator or transformer) and Earth.
    If the power supply is grounded, Call the system "T" for Terra.
    If the transformer is completely isolated and ungrounded, I call it "I" for Isolated.
  2. The relationship between exposed-conductive-parts of the installation and Earth.
    These are metal parts that could become energized if a fault occurs. In most systems, they are bonded to the point on the power supply that is connected to Earth, typically the transformer neutral. Use "N" if the neutral is solidly connected to a grounding electrode system.
    Some systems use a grounding method that’s independent of the power supply, such as resistance grounding or completely ungrounded (floating) systems, use "T" for Terra for these less common systems, since they don’t rely on a solid connection to the power supply’s grounded neutral.
  3. The 'protective conductor' or the wire used for protection against lightning, surges, or accidental contact with higher voltage systems, and to ensure an effective ground fault path.
    "S" when the neutral and protective conductor functions are provided by separate conductors.
    "C" when both functions are combined in a single conductor.
So in IEC terms the system you’re describing is a TN-C system, also known as a Multi-Grounded Neutral (MGN). It’s used for outdoor distribution to limit voltage from lightning, surges, or accidental contact with higher-voltage lines. In this system, the combined neutral/protective conductor is grounded at multiple points, the further you get from one of the grounding points the more chance the protective conductor in your case the neutral will create a Earth - Neutral voltage.
Hence with TN-C a open neutral to the pool subpanel will create a deadly voltage drop across the pool and with a TN-C-S (what the NEC requires) it wont.
Click to expand...
That comes as close to answering my question as I've seen. And I can find that in the IEC book? To understand why a system is designed as it is is invaluable to me in "getting it right". I'm happy to follow code instructions but to know why they provide those exact instructions can be invaluable when wading through nuances that are existing or not specified. Thank you very much!!!
 
As I see it, the issue is the possibility (likelihood) of creating a ground/neutral loop due to the presence of multiple N-G bonds, one in each panel. Which will happen if the EGC system connected to the main panel and the EGC system connected to the subpanel are ever interconnected elsewhere. Which would definitely happen if any single box contains a device on a circuit from one panel along with a device on a circuit from another panel, but could easily happen even without that, due to each EGC system being bonded to a some common non-electrical piece of metal, say.

Once you have a ground/neutral loop, with two different connections between neutral and ground, then all the neutral current that needs to flow between those two points will have an alternate path, through the other side of the loop on the ground system (EGCs et al). So current will divide in inverse proportion to the impedance of the two paths.

Now with the main panel and the subpanel next to each other, the impedance of the short stretch of neutral conductor between them is probably quite a bit less than the impedance of the path through the ground system, so not much current will flow through the ground system. Nonetheless, the EGC system is designed so that no normal current flows through it, so a little current is still too much.

Edit: to clarify, when I refer to "ground system" I'm not including the earth in that term.

Cheers, Wayne
 
Last edited:
retirede said:
You’re looking for an ultra-simple explanation for something that’s too nuanced to offer one. Don’t get upset because you’re unhappy with the attempts to provide a satisfactory answer.
Click to expand...
I won't get upset at any attempt to answer my question. I realize too that I may not communicate my question well. What upsets me is when there are all sorts of "answers" that don't even attempt to understand or answer the real question - I see it all the time on this forum. Everything gets picked apart that isn't even a part of the question and there is no answer to the real question. I do appreciate too that there is effort made to be sure someone isn't being stupid!
 
wwhitney said:
As I see it, the issue is the possibility (likelihood) of creating a ground/neutral loop due to the presence of multiple N-G bonds, one in each panel. Which will happen if the EGC system connected to the main panel and the EGC system connected to the subpanel are ever interconnected elsewhere. Which would definitely happen if any single box contains a device on a circuit from one panel along with a device on a circuit from another panel, but could easily happen even without that, due to each EGC system being bonded to a some common non-electrical piece of metal, say.

Once you have a ground/neutral loop, with two different connections between neutral and ground, then all the neutral current that needs to flow between those two points will have an alternate path, through the other side of the loop on the ground system (EGCs et al). So current will divide in inverse proportion to the impedance of the two paths.

Now with the main panel and the subpanel next to each other, the impedance of the short stretch of neutral conductor between them is probably quite a bit less than the impedance of the path through the ground system, so not much current will flow through the ground system. Nonetheless, the EGC system is designed so that no normal current flows through it, so a little current is still too much.

Edit: to clarify, when I refer to "ground system" I'm not including the earth in that term.

Cheers, Wayne
Click to expand...
That makes a lot of sense! Which is why in a pool situation it could viably introduce current into the pool as a result!
 
Kopper said:
I won't get upset at any attempt to answer my question. I realize too that I may not communicate my question well. What upsets me is when there are all sorts of "answers" that don't even attempt to understand or answer the real question - I see it all the time on this forum. Everything gets picked apart that isn't even a part of the question and there is no answer to the real question. I do appreciate too that there is effort made to be sure someone isn't being stupid!
Click to expand...
Well you should have just asked the question "why are EGC's isolated from neutrals downstream of the service equipment" and left the real life scenario out of your question. You muddied the water talking about a "most likely" code compliant installation.
 
You must log in or register to reply here.
Top