EGC in PVC conduit to plasic light

Status
Not open for further replies.

Bismarck

Member
I am a Master Electrician. For 18 months I did electrical inspections for the U.S. military in Iraq. Now for the past month and a half I have been doing electrical inspections in Afghanistan for the U.S. military. We have several British inspectors here also inspecting to the NEC 2008 code. A point of disagreement over the code has arisen and I was hoping that you could straighten this out for us as I consider you a foremost expert on the code.
Here is the current situation. We have circuits in structures that are not considered residential. These are containers for sleeping that are electrically fed from a transformer to a MDP (5 wire)where bonding occurs, to a SDP (5 wire) and then to a CDP (120v/208v 4 wire, 2 phase conductors, a neutral and an EGC)in the individual units. There is a 120v branch circuit run from the CDP panel through PVC conduit (distance of 4 feet but length does not really matter) to a light fixture that is made of plastic with absolutely no metal except the contact for the ungrounded conductor and a contact for the grounded conductor which is attached to the socket. This fixture is not required to be bonded and cannot be bonded as it is an all plastic non-conductive surface. But some of the inspectors contend that a EGC is required to be laid with the circuit conductors in the PVC conduit not attached to anything at the fixture end but attached to the ground bar in the panel. I said that was preposterous and asked them to cite what code article they were using for that requirement. They claimed 250.4 (A) (5). This states that there must be an effective ground fault current path anywhere on the system that a ground fault may occur. I told them you cannot have a ground fault in PVC conduit as it is not a conductive surface and that you could only have a short circuit. They then claimed that a short circuit and ground fault are the same thing. I said that is absolutely false and erroneous thinking and that is their error. I told them that a branch circuit requires a ground only if there is a requirement to bond something such as a conductive surface for the ground fault path (boxes, metal conduit, panels, motor frames, pipes, etc. or a device that requires it such as a receptacle or metal light fixture and such or where specifically required elsewhere in the code. I said NEC 2008 articles 250.4 (A) (2), 250.4 (A) (3), and 250.4 (A) (4) tell you what must be bonded and in Part V of article 250, Article 250.90 states that bonding shall be provided where necessary to ensure electrical continuity and the capacity to conduct safely any fault current likely to be imposed.
So if you could, could you please state whether I am correct or they are and what code articles back that position. It would mean so much to the safety of the soldiers out here if we could be sure we are inspecting to the NEC correctly. I anxiously await your opinions.

PS. In article 200.3 there is a mistake in the code book. It states for any grounded conductor (should be ungrounded conductor) you require a corresponding grounded conductor. A little slip up I guess. Can we notify somebody about that?
 

K8MHZ

Senior Member
Location
Michigan. It's a beautiful peninsula, I've looked
Occupation
Electrician
I am a Master Electrician. For 18 months I did electrical inspections for the U.S. military in Iraq. Now for the past month and a half I have been doing electrical inspections in Afghanistan for the U.S. military. We have several British inspectors here also inspecting to the NEC 2008 code. A point of disagreement over the code has arisen and I was hoping that you could straighten this out for us as I consider you a foremost expert on the code.
Here is the current situation. We have circuits in structures that are not considered residential. These are containers for sleeping that are electrically fed from a transformer to a MDP (5 wire)where bonding occurs, to a SDP (5 wire) and then to a CDP (120v/208v 4 wire, 2 phase conductors, a neutral and an EGC)in the individual units. There is a 120v branch circuit run from the CDP panel through PVC conduit (distance of 4 feet but length does not really matter) to a light fixture that is made of plastic with absolutely no metal except the contact for the ungrounded conductor and a contact for the grounded conductor which is attached to the socket. This fixture is not required to be bonded and cannot be bonded as it is an all plastic non-conductive surface. But some of the inspectors contend that a EGC is required to be laid with the circuit conductors in the PVC conduit not attached to anything at the fixture end but attached to the ground bar in the panel. I said that was preposterous and asked them to cite what code article they were using for that requirement. They claimed 250.4 (A) (5). This states that there must be an effective ground fault current path anywhere on the system that a ground fault may occur. I told them you cannot have a ground fault in PVC conduit as it is not a conductive surface and that you could only have a short circuit. They then claimed that a short circuit and ground fault are the same thing. I said that is absolutely false and erroneous thinking and that is their error. I told them that a branch circuit requires a ground only if there is a requirement to bond something such as a conductive surface for the ground fault path (boxes, metal conduit, panels, motor frames, pipes, etc. or a device that requires it such as a receptacle or metal light fixture and such or where specifically required elsewhere in the code. I said NEC 2008 articles 250.4 (A) (2), 250.4 (A) (3), and 250.4 (A) (4) tell you what must be bonded and in Part V of article 250, Article 250.90 states that bonding shall be provided where necessary to ensure electrical continuity and the capacity to conduct safely any fault current likely to be imposed.
So if you could, could you please state whether I am correct or they are and what code articles back that position. It would mean so much to the safety of the soldiers out here if we could be sure we are inspecting to the NEC correctly. I anxiously await your opinions.

PS. In article 200.3 there is a mistake in the code book. It states for any grounded conductor (should be ungrounded conductor) you require a corresponding grounded conductor. A little slip up I guess. Can we notify somebody about that?

I read 200.3 as requiring a grounded conductor in the supply if there exists a grounded conductor in the premises.

Also, as for the EGC, why would you not want to pull one in? That would allow for, at a later time, a metal device to be safely connected at that outlet.

A ground fault could be a short circuit if you think about it. A short circuit is one wherein the current makes it back to it's source before it gets to the intended load. Most of the time that route is shorter, so it's called a short circuit. Usually it takes the neutral back, but if there is a ground fault ahead of the load, and the current takes a shorter path back to it's source via the EGC, we have both a short circuit and a ground fault.

BUT, most in the trade try to keep things simple and classify a short as hot to hot, or hot to N, and a ground fault as hot to metal not intended to carry current under normal operation.

The biggest difference between a short and a ground fault is that a short must be a complete circuit whereas a ground fault needs not to be, it can be open and still faulted.
 
The biggest difference between a short and a ground fault is that a short must be a complete circuit whereas a ground fault needs not to be, it can be open and still faulted.


Marky
That is not a correct statement. For any current to flow, there needs to be a complete circuit, whether the current is normal current or ground fault current.

The ground fault interrupts a "normal circuit" and completes a 'ground fault current path' along an unintended path. The unintended path could be either via the Equipment grounding path, or an unintended metallic path, or even people, etc...
 

jetlag

Senior Member
Marky
That is not a correct statement. For any current to flow, there needs to be a complete circuit, whether the current is normal current or ground fault current.

The ground fault interrupts a "normal circuit" and completes a 'ground fault current path' along an unintended path. The unintended path could be either via the Equipment grounding path, or an unintended metallic path, or even people, etc...

I dont believe we are suppose to plan on the type of fixture , metal or not when running the conductors, any more than you would run nm with no ground just because you knew the fixture was plastic.
 

K8MHZ

Senior Member
Location
Michigan. It's a beautiful peninsula, I've looked
Occupation
Electrician
Marky
That is not a correct statement. For any current to flow, there needs to be a complete circuit, whether the current is normal current or ground fault current.

The ground fault interrupts a "normal circuit" and completes a 'ground fault current path' along an unintended path. The unintended path could be either via the Equipment grounding path, or an unintended metallic path, or even people, etc...

As an example of what I was thinking, suppose you have metal toaster with no EGC and for some reason the hot touches the metal and energizes it. With no EGC, this would be an open, but still a fault.

What am I missing?
 

Bismarck

Member
EGC to plastic

EGC to plastic

The thing here is these are existing conditions. I can't tell anyone they have to pull an EGC into a plastic light. We do not cite someone for what may be in the future. If it were ever changed to a metal fixture they would then need an EGC. We are inspecting the here and now. There can be no ground fault in non-conducting material. Nothing to bond!
 

jetlag

Senior Member
As an example of what I was thinking, suppose you have metal toaster with no EGC and for some reason the hot touches the metal and energizes it. With no EGC, this would be an open, but still a fault.

What am I missing?

its not a ground fault untill someone touches it and the current takes a wrong path . Not sure thats a correct statement but I couldnt resist :grin:
 

K8MHZ

Senior Member
Location
Michigan. It's a beautiful peninsula, I've looked
Occupation
Electrician
I see nothing in the code that would require the installation of an EGC for this application.

I agree that there is no requirement under the given set of circumstances.

I do think it's still a good idea to pull one in, albeit not required.

ON EDIT:

I just re-read and noticed that the circuit in question is only 4 feet long and through pipe. If no EGC was pulled it would be an easy matter to pull one in at a later date. So I take a change in stance, no big deal to leave it out in a case where pulling in one in if needed at a later date would be that easy.
 
Last edited:

jetlag

Senior Member
The thing here is these are existing conditions. I can't tell anyone they have to pull an EGC into a plastic light. We do not cite someone for what may be in the future. If it were ever changed to a metal fixture they would then need an EGC. We are inspecting the here and now. There can be no ground fault in non-conducting material. Nothing to bond!

GFCI receptacles can be wired with no EGC in certain cases , they will trip without a GFCI , water and other things can cause a path without metal and cause a ground fault if all the current doesnt return via the neutral
 

don_resqcapt19

Moderator
Staff member
Location
Illinois
Occupation
retired electrician
I agree that there is no requirement under the given set of circumstances.

I do think it's still a good idea to pull one in, albeit not required.
...
But the question was not "what is a good installation practice, or what you would do?". The question was if the code rules required an EGC for this installation.
 

hurk27

Senior Member
Wasn't there an allowance for switches where they were all plastic including the cover screws, and a EGC wasn't required? kind of would be similar, other then one being an outlet?

But then 2011 requiring neutrals at switches who knows what the future might bring.:grin:
 

jetlag

Senior Member
GFCI receptacles can be wired with no EGC in certain cases , they will trip without a GFCI , water and other things can cause a path without metal and cause a ground fault if all the current doesnt return via the neutral

I made a typo I meant to say " the GFCI will trip without an EGC "
 
I dont believe we are suppose to plan on the type of fixture , metal or not when running the conductors, any more than you would run nm with no ground just because you knew the fixture was plastic.

I did not address that portion of the question.

As an example of what I was thinking, suppose you have metal toaster with no EGC and for some reason the hot touches the metal and energizes it. With no EGC, this would be an open, but still a fault.

What am I missing?


My comment stated that for fault current to flow, there needs to be a complete path.
 

Bismarck

Member
In response to the ground fault question, nobody has to touch some energized metal for it to be a ground fault. If a ground fault occurs, for example a hot shorting to a metal enclosure, as long as there is a low impedance ground fault path back to the source, current will flow and the breaker will trip. The same principle goes for a short circuit which is not a ground fault. A hot will accidently connect with another hot or a neutral, current will flow and the breaker will trip. These are two distinctly different events. If you have a GFCI breaker and a short circuit occurs, the GFCI function of the breaker sees that as nothing wrong. It measures current going out and current returning and sees it as equal and therefore no tripping at 6ma. But the over current protection sees the short circuit from the massive current flow and that trips the OCPD and not the GFCI portion of the breaker. It is very important to understand the difference to apply code correctly. And thank you very much for verifying that you can find no code requiring the EGC mentioned in this thread.
 

kwired

Electron manager
Location
NE Nebraska
I dont believe we are suppose to plan on the type of fixture , metal or not when running the conductors, any more than you would run nm with no ground just because you knew the fixture was plastic.

If you don't plan on the type of fixture how do you know what size of conductors to run, size of overcurrent device, etc?

I have run many totally non metallic fixtures in livestock buildings through non metallic raceway and not run an EGC. I don't feel there is a need to consider the fact that it will be replaced with a metallic fixture in these cases because if that happens it is likely not a fixture suitable for the location anyway and should not be installed.
 

acrwc10

Master Code Professional
Location
CA
Occupation
Building inspector
One way to look at this question, is to post the code section(s) that require the EGC installation in regards to the OP's thread.

There are only 822 pages in the '08 NEC, it should not take toooooooooo long to find it. ;)

Ya but not all of them are about grounding,and grounded. So it should be even easier.

Just curious on the OP, is there an EGC to the switch?
 
Status
Not open for further replies.
Top