Why is a sub*panel not "effectively" bonded when connected to main panel by emt?

Status
Not open for further replies.
C

cowdr4

Member
Location
Pomona, IL, USA
Why is a sub*panel not "effectively" bonded when connected to main panel by emt?

Specific situation: 200 amp main service panel with bonded neutral bar and ufer ground. Main panel has 100 amp 240V breaker feeding 2 #3 Cu wires that are connected to 100 amp main disconnect of 100 amp sub-panel. Additional #3 Cu wire connecting sub-panel and main service panel neutral bars and a #6 bare Cu wire connecting sub-panel ground to neutral bonded bar in main 200 amp panel. Sub-panel is approximately 3 feet from main panel, and is connected by EMT conduit. Sub-panel neutral bar is not bonded to panel chassis with a bonding screw.

All of the above meets NEC, but even without the bonding screw, isn't the sub-panel effectively bonded to the chassis via 2 separate paths? The EMT conduit between the main and sub-panel provides a direct path for electricity from the sub-panel neutral to main panel neutral via the neutral wire, then to the main panel chassis via bonding screw, then to EMT, and finally to sub-panel chassis. Additionally, the bare Cu ground can provide a direct path for current from sub-panel neutral to the chassis by contacting either the chassis or EMT directly. Why is a 1/4 inch path through a bonding screw an issue when neither of the above 3-6 foot long pathways appear to be a concern? Thanks!
 
I'm approving this thread even though the OP is not in the electrical industry. I believe the answer to this would benefit other readers. Please keep all replies to the question asked and not any "how to" answers.
 
cowdr4 said:
Specific situation: 200 amp main service panel with bonded neutral bar and ufer ground. Main panel has 100 amp 240V breaker feeding 2 #3 Cu wires that are connected to 100 amp main disconnect of 100 amp sub-panel. Additional #3 Cu wire connecting sub-panel and main service panel neutral bars and a #6 bare Cu wire connecting sub-panel ground to neutral bonded bar in main 200 amp panel. Sub-panel is approximately 3 feet from main panel, and is connected by EMT conduit. Sub-panel neutral bar is not bonded to panel chassis with a bonding screw.

All of the above meets NEC, but even without the bonding screw, isn't the sub-panel effectively bonded to the chassis via 2 separate paths? The EMT conduit between the main and sub-panel provides a direct path for electricity from the sub-panel neutral to main panel neutral via the neutral wire, then to the main panel chassis via bonding screw, then to EMT, and finally to sub-panel chassis. Additionally, the bare Cu ground can provide a direct path for current from sub-panel neutral to the chassis by contacting either the chassis or EMT directly. Why is a 1/4 inch path through a bonding screw an issue when neither of the above 3-6 foot long pathways appear to be a concern? Thanks!
Click to expand...

I am a bit confused by your train of thought in the second paragraph. Maybe its just too late for me. The sub panel should have neutrals and ground separated, is that what you have? Indeed the installer could have used the EMT as the EGC and skipped the wire EGC. There is nothing that prohibits multiple Equipment grounding conductors. Using an approved raceway as an EGC has fallen greatly out of favor in the last 30 years or so, despite no history or data showing it is less safe.
 
cowdr4 said:
Specific situation: 200 amp main service panel with bonded neutral bar and ufer ground. Main panel has 100 amp 240V breaker feeding 2 #3 Cu wires that are connected to 100 amp main disconnect of 100 amp sub-panel. Additional #3 Cu wire connecting sub-panel and main service panel neutral bars and a #6 bare Cu wire connecting sub-panel ground to neutral bonded bar in main 200 amp panel. Sub-panel is approximately 3 feet from main panel, and is connected by EMT conduit. Sub-panel neutral bar is not bonded to panel chassis with a bonding screw.

All of the above meets NEC, but even without the bonding screw, isn't the sub-panel effectively bonded to the chassis via 2 separate paths? The EMT conduit between the main and sub-panel provides a direct path for electricity from the sub-panel neutral to main panel neutral via the neutral wire, then to the main panel chassis via bonding screw, then to EMT, and finally to sub-panel chassis. Additionally, the bare Cu ground can provide a direct path for current from sub-panel neutral to the chassis by contacting either the chassis or EMT directly. Why is a 1/4 inch path through a bonding screw an issue when neither of the above 3-6 foot long pathways appear to be a concern? Thanks!
Click to expand...

The clear distinction, if you follow the current paths for both cases, is this:
If you bond with the screw at the sub panel normal neutral current has an available path back to the main through the EGC as well as through the neutral wire.
If you only have a bond at the main panel all normal neutral current can flow only on the neutral wire. There is no reason for current to flow down the neutral to the main panel, then through the bond and EGC back to the sub panel. There is nowhere for it to go at the sub panel as the parallel path is not complete.

In both cases an ohmmeter reading will be very close between the two cases, but the current flow will be very different.

Note that from the main panel (service disconnect) back to the POCO transformer there will be a parallel path through the earth, but that is not considered obectionable and is also not part of the jurisdiction of the NEC.
 
cowdr4 said:
Specific situation: 200 amp main service panel with bonded neutral bar and ufer ground. Main panel has 100 amp 240V breaker feeding 2 #3 Cu wires that are connected to 100 amp main disconnect of 100 amp sub-panel. Additional #3 Cu wire connecting sub-panel and main service panel neutral bars and a #6 bare Cu wire connecting sub-panel ground to neutral bonded bar in main 200 amp panel. Sub-panel is approximately 3 feet from main panel, and is connected by EMT conduit. Sub-panel neutral bar is not bonded to panel chassis with a bonding screw.
Click to expand...

This is code compliant, you have a redundant EGC with both the EMT and #6 conductor which is not a problem because as Ef stated you can have more than one EGC (#6 and EMT) even if they're in parallel with each other.

cowdr4 said:
All of the above meets NEC, but even without the bonding screw, isn't the sub-panel effectively bonded to the chassis via 2 separate paths? The EMT conduit between the main and sub-panel provides a direct path for electricity from the sub-panel neutral to main panel neutral via the neutral wire, then to the main panel chassis via bonding screw, then to EMT, and finally to sub-panel chassis. Additionally, the bare Cu ground can provide a direct path for current from sub-panel neutral to the chassis by contacting either the chassis or EMT directly. Why is a 1/4 inch path through a bonding screw an issue when neither of the above 3-6 foot long pathways appear to be a concern? Thanks!
Click to expand...

Since the subpanel neutral is not bonded the part in bold is incorrect.
 
To put it simply, at no point beyond the point in the main disconnect (usually) where the EGC system (enclosures, GECs, etc) is bonded to the neutral, should the neutral ever be grounded again. It should be insulated from anything grounded as if it were an energized conductor.
 
LarryFine said:
To put it simply, at no point beyond the point in the main disconnect (usually) where the EGC system (enclosures, GECs, etc) is bonded to the neutral, should the neutral ever be grounded again. It should be insulated from anything grounded as if it were an energized conductor.
Click to expand...
And a major reason for that is the grounded conductor (often also a neutral conductor) can carry current during normal operation. When a conductor carries current there is a voltage drop across that conductor, may not be much at all in some instances but there is a voltage drop. If you connect exposed conductive objects to that conductor you can be exposed to that voltage potential. If there were a 3 volt drop on the grounded conductor (and everything else bonded to it) when you touch one of those bonded objects and are standing on "earth" you will be subjected to that 3 volts.

Then there is the fact that all possible paths back to the source will carry some of the current - each will also have a voltage drop across them. By not bonding beyond the service or first disconnect of separately derived systems, we don't create parallel paths for (grounded) conductors that were intended to carry current in normal operation.
 
cowdr4 said:
Specific situation: 200 amp main service panel with bonded neutral bar and ufer ground. Main panel has 100 amp 240V breaker feeding 2 #3 Cu wires that are connected to 100 amp main disconnect of 100 amp sub-panel. Additional #3 Cu wire connecting sub-panel and main service panel neutral bars and a #6 bare Cu wire connecting sub-panel ground to neutral bonded bar in main 200 amp panel. Sub-panel is approximately 3 feet from main panel, and is connected by EMT conduit. Sub-panel neutral bar is not bonded to panel chassis with a bonding screw.

All of the above meets NEC, but even without the bonding screw, isn't the sub-panel effectively bonded to the chassis via 2 separate paths? The EMT conduit between the main and sub-panel provides a direct path for electricity from the sub-panel neutral to main panel neutral via the neutral wire, then to the main panel chassis via bonding screw, then to EMT, and finally to sub-panel chassis. Additionally, the bare Cu ground can provide a direct path for current from sub-panel neutral to the chassis by contacting either the chassis or EMT directly. Why is a 1/4 inch path through a bonding screw an issue when neither of the above 3-6 foot long pathways appear to be a concern? Thanks!
Click to expand...
If you are questioning suitability of that 1/4 inch bonding screw - you do have to use the factory supplied screw in the hole it was intended to go in. It was listed and tested for the purpose.

I have seen such bonding screws fail though. Probably because it wasn't properly tightened when installed. One was fairly recently. I had a service call for a couple breakers that were tripping. They both had short circuit/ground fault conditions and tripped immediately when you tried to turn them on. I happened to notice burn marks around the bonding screw - tried to tighten it and it wouldn't tighten -stripped threads. My guess is it wasn't tightened properly and the fault current was enough to damage threads to where they won't work right anymore. I had to use alternate means to bond this panel and of course find their problems causing the tripping breakers.
 
This is exactly why I asked this question

This is exactly why I asked this question

LarryFine said:
To put it simply, at no point beyond the point in the main disconnect (usually) where the EGC system (enclosures, GECs, etc) is bonded to the neutral, should the neutral ever be grounded again. It should be insulated from anything grounded as if it were an energized conductor.
Click to expand...

I appreciate that the neutral and ground should never be bonded beyond the main disconnect and that is why I asked this question. I know that NEC allows bare ground wire in EMT conduit, but in this case the bare copper "ground" connects the neutral bars in the main panel and the sub-panel. Common sense tells me that at some point, this bare copper ground has to "short" against the EMT conduit, creating a second (or more) point of connection between the neutral and grounding systems. If this concern is correct, am I better off having the electrician: a) remove the bare copper wire between the neutral bars in both panels or b) leave the bare copper wire, but change the connection to the grounding bars in both boxes? At this time there is no grounding bar in the sub-panel as all circuit grounds are run to the bonded neutral bar in the main panel. I asking for advice as I work in an area with no building inspectors to rely upon, and I would like to ensure that I am building a safe house. Thanks!
 
cowdr4 said:
I appreciate that the neutral and ground should never be bonded beyond the main disconnect and that is why I asked this question. I know that NEC allows bare ground wire in EMT conduit, but in this case the bare copper "ground" connects the neutral bars in the main panel and the sub-panel.
Click to expand...
Are you saying the EGC's are landed on the neutral bar in the "sub-panel?

cowdr4 said:
Common sense tells me that at some point, this bare copper ground has to "short" against the EMT conduit, creating a second (or more) point of connection between the neutral and grounding systems.
Click to expand...
It doesn't matter, the EGC's (copper wire and EMT) are going to be tied together at each end anyways. How is this going to create a second connection to the Neutral.

cowdr4 said:
If this concern is correct, am I better off having the electrician: a) remove the bare copper wire between the neutral bars in both panels or b) leave the bare copper wire, but change the connection to the grounding bars in both boxes? At this time there is no grounding bar in the sub-panel as all circuit grounds are run to the bonded neutral bar in the main panel. I asking for advice as I work in an area with no building inspectors to rely upon, and I would like to ensure that I am building a safe house. Thanks!
Click to expand...

The neutrals and EGC's must tie together in the service equipment (main Panel or disconnect) and they must be isolated in the sub-panel.



Roger
 
cowdr4 said:
I appreciate that the neutral and ground should never be bonded beyond the main disconnect and that is why I asked this question. I know that NEC allows bare ground wire in EMT conduit, but in this case the bare copper "ground" connects the neutral bars in the main panel and the sub-panel. Common sense tells me that at some point, this bare copper ground has to "short" against the EMT conduit, creating a second (or more) point of connection between the neutral and grounding systems. If this concern is correct, am I better off having the electrician: a) remove the bare copper wire between the neutral bars in both panels or b) leave the bare copper wire, but change the connection to the grounding bars in both boxes?At this time there is no grounding bar in the sub-panel as all circuit grounds are run to the bonded neutral bar in the main panel. I asking for advice as I work in an area with no building inspectors to rely upon, and I would like to ensure that I am building a safe house. Thanks!
Click to expand...
Reply to blue: That conductor should be insulated (and identified with white or gray color that designates it as a grounded conductor) so it can't make inadvertent contact with non current carrying conductive components.

Reply to red: They should probably have installed a grounding bar in the second panel and landed those conductors there.

Though the enclosure for a panelboard is technically called a cabinet, 250.148 more less says if those grounding conductors pass through a box - they need to all bond to one another and to the box. Not sure where your circuits come to. If they enter the first panel but not enough spaces and you had nipple to the second panel to get them to a breaker, then landing those grounds in the first panel isn't a problem. If they pass through the second panel we then must question whether 250.148 is supposed to apply here as well. Most people would put a ground bar in the second panel and land them there though with little additional thought about it.


​Is there neutral conductors landed on the neutral bus of the second panel?
 
cowdr4 said:
I appreciate that the neutral and ground should never be bonded beyond the main disconnect and that is why I asked this question. I know that NEC allows bare ground wire in EMT conduit, but in this case the bare copper "ground" connects the neutral bars in the main panel and the sub-panel. Common sense tells me that at some point, this bare copper ground has to "short" against the EMT conduit, creating a second (or more) point of connection between the neutral and grounding systems. If this concern is correct, am I better off having the electrician: a) remove the bare copper wire between the neutral bars in both panels or b) leave the bare copper wire, but change the connection to the grounding bars in both boxes? At this time there is no grounding bar in the sub-panel as all circuit grounds are run to the bonded neutral bar in the main panel. I asking for advice as I work in an area with no building inspectors to rely upon, and I would like to ensure that I am building a safe house. Thanks!
Click to expand...

The subpanel needs a separate ground bar. The bare #6 needs to land on this instead of the neutral bar. You seriously need to get the electrician to check this or get another electrician.
 
ECG lands on neutral bar in sub-panel

ECG lands on neutral bar in sub-panel

roger said:
Are you saying the EGC's are landed on the neutral bar in the "sub-panel?

It doesn't matter, the EGC's (copper wire and EMT) are going to be tied together at each end anyways. How is this going to create a second connection to the Neutral.


The neutrals and EGC's must tie together in the service equipment (main Panel or disconnect) and they must be isolated in the sub-panel.
Roger
Click to expand...

1. Yes, the bare so-called (by the electrician) "ground" wire from main to sub-panel connects to the neutral bar in the sub-panel.
2. Concern in this case is that the EGC is tied into the neutral bar in sub-panel, which will "short" against EMT creating a second connection between neutral and EGC.
3. Sounds like I need to have the electrician remove this wire.
 
cowdr4 said:
3. Sounds like I need to have the electrician remove this wire.
Click to expand...

I agree and will add that it would be money well spent to hire an independent inspection service to inspect the whole job, the cost isn't that much for peace of mind.

Roger
 
I think the question has been effectively answered. It's getting to the point that it's hard to answer any further without getting into DIY. So I am closing the thread.
I recommend getting the electrician to check this or if he doesn't agree, get another one that knows the code.
 
Status
Not open for further replies.
Top