Here is a very devisive issue - if several EGCs enter a distribution panel [or service panel], can you connect two or more of them together and pigtail a properly sized conductor to the EGC terminal bus? (e.g. 3 #12s enter the panel, you connect them together and run a #12 pigtail to the terminal).

The 2002 HEC Handbook 'commentary' for 250.122(C) says: "It [the EGC] is not required to be sized for the 'composite' of all the circuits in the raceway because it is not anticipated that all circuits will develop faults at the same time." - yes, this is a commentary on properly sizing the EGC in a j-box or conduit.

At issue: If you can size a single EGC to protect several circuits in a raceway or j-box, can you not do the same thing in the panel?

384-20 speaks of the requiring the grounding of panelboards, but it does not specifically address this question and I know of no other code section that does address this specifically.

The current language of 384-20 came into the code in 1990 and the 1990 NEC Handbook does not speak directly of this issue either.

BTW - this is NOT a practice I condone or support; however it is one of those 'code issues' I would like a clear and concise answer to. Opinions are great (mine is - don't do it), but the question is: Is there a particular code section prohibiting this practice?

Anyone have copies of the ROPs/ROCs for the 1990 code cycle?

Hello Dana,

Actually you have the code reference cited in your post. 99 NEC 384.20 & 2002 NEC 408.20 are the section to find it.
The second sentence... Where the panelboard is used with nonmetallic raceway or cable or where separate grounding conductors are provided, a terminal bar for the grounding conductors shall be secured inside the cabinet.

This is telling you to terminate the EGC at the terminal bar.
The NEC and the UL White Book both stipulate that if the terninal bar is not listed for more than 1 conductor in a terminal, than all conductors shall be individually terminated.

250.122(C) does not specify splicing multiple conductors, but says if you run MULTIPLE CIRCUITS, you can use '1' EGC in the raceway or cable as long as it is sized to the largest overcurrent protective device of the largest circuit in that raceway or cable.

Splicing all the EGC in a junction box is not a recomended practice if they are of different circuits.

I hope this helps

Pierre

Pierre, You have to join all EGC in the J-box per code. Also some older panels do not have enough terminals on the neutral bar and you must do as you state.

408.20 requires the equipment grounding bus, and 110.3(B) tells us that lissted equipment shall be used as per any instructions in the listing. That means one wire per terminal unless two per terminal are permitted. No bundling of all equipment grounding conductors in a panelboard.
Creighton

Pierre

250.148 is the article that requires all EGCs to be connected together in a j box.

There is one exception to this for isolated grounds.



[This message has been edited by iwire (edited 06-13-2003).]

Thank you Creighton!! I hope this settles this question that also asked in other areas on the net, and as stated the answer can be found in the reference that you quoted!

Hello iwire,

I do respect what everyone here has to say, so I am going to ask you a favor. Read 250.148 again, I will post it here


...any separate equipment grounding conductors ASSOCIATED WITH THOSE CIRCUIT CONDUCTORS

Lets say there is a 6x6 jbox and we have multiple raceways (jbox and raceways are PVC) with multiple circuits. I am pulling an EGC with each circuit. I splice the EGC's of ONE CIRCUIT together(lets say those are 12 wire for receptacles). There is another 12 wire circuit(for lighting) in the same jbox, and I splice those associated conduuctors together and send them on their merry way. If I am reading this correctly, because they are not associated circuit conductors, and not associated circuits, I am not required to splice the EGC's of that circuit together with the other circuit EGC's. How big of a splice might that be if I had to, and how would I terminate such a large splice?

Thanks, let me know what you think,

Pierre

In the area I work it is interrupted as all grounds other than IGs must tie together.

Can you explain the need for the exception for IG circuits if your interruption is accurate.


As for how I do it, if it is more than a large blue wire nut is listed for I install a ground bar.

I can see your point if pulling separate grounds for each circuit and some circuits pass straight through with out splices I would not interrupt them at the box.

Most times in raceway we run one grounding conductor as allowed by 250.122(C) so I would have to land at each box with splices.

What do you think about a residential metal switch box with 2 or more circuits in it, how would you deal with the grounds here.

I am here to learn too, so if I am wrong straighten me out.

Bob


[This message has been edited by iwire (edited 06-14-2003).]

Joe,
I don't see any wording in the cited section that would prohibit this practice. It doesn't say that each EGC must be directly connected to an equipment grounding terminal. I don't like this idea, but still don't see any code rule that will prevent it.
Don

Hello Bob


This is good, we can hash this out, and be able to help each other out here.

About the exception: I believe this exception is trying to deal with 250.148, in a sense as to the extent that the IG is not required to be connected to the EGC, for obvious reasons, that I know you know.(that was hard for me to put in words, if it does not make sense to you, I will try to put it in other words, let me know)

I am not at all against using a ground bar, in a large enough jbox, and actually have, if you do want to connect them all together.
BTW - I am not against connecting All the EGC together, I just do not think this section is making it required.

I also run 1 EGC in a raceway sometimes when the bidding process makes that necessary, But... if I can I like to run individual EGC's

In a metal residential switch box I would probably splice them all together, only because there probably will not be too many.

Not trying to be a pain, but answering how I would do it, so as to see what you would do.

Pierre

Have to agree with Don. I don't see any working that would prohibit such a pigtail arrangement. Lots of service replacement jobs have spliced conductors in panels. It ain't pretty, but it suffices... as long as the panel has adequate room for the splicing.

my humble vote goes for the original post at the top where he says it is a devisive issue..... i'm not convinced by any of the code sections here that you can't pigtail them in the panel, but i've never seen anyone that does it... the commentary reasonoing sounds right, 1 fault at a time, so i guess if you have two-8awg's two-10awg's and two-12awg's in the right splicing device, it should be ok to use the 8awg[40a] as the pigtail, right? i wouldn't bet any money against this in a lawsuit, and if an inspector turned it down, he'd be doin' a lot of tap dancing, imho

Here's a good example of what not to do, or approve when it comes to the equipment grounding conductors that are terminated in any type of cabinet:

The rule in 384-20 or now 408.20 was added to the code a long time ago to put a stop to this method.

You are entitled to your opinions, and I respect them, but the bottom line is that you will get a "RED TAG" from me if you do it this way in a new panelboard!

What's so hard about installing a simple terminal bar in an existing panel?

Joe,
In your last picture, there are code violations in the use of the split bolts. They are not listed for use with that many conductors and you could cite that violation. The panel also does not have an equipment grounding terminal bar which has been required for many years by 384.27 and later by 384.20 or 408.20.
The rule was placed into the code to prevent the direct connection of EGC to the panel enclosure. Nothing in the current code section says that every EGC must be directly landed to the equipment grounding bar. In the '87 code section 384.27 required that "all" of the feeder or branch circuit EGCs to be connected to the equipment grounding bar. When the section was rewritten and relocated to 384.20 in the '90 code the word "all" was deleted from the section. The code no longer clearly requires that all of the EGCs be directly connected to the equipment grounding terminal bar. There was no reason given in the TCR for the deletion of this word. It may have been unintentional. The word "all" should be put back into the section, but as the code stands now, it does not support the opinion that all of the EGCs must be directly landed to the terminal bar.
Nothing is hard about this and this is the way that the installation should be done, but this thread is not about how things should be done. It is about the absolute minimum requirements that can be supported by the current code language.
Don

Don:

OK, Ynow you need to show me what you are thinking about.

Please, supply a picture or drawing.

I am surprised that you are defending the strange way in which this is, or has been interpreted here just because of a simple word!

Most electrical inspectors would disagree with you, unless they are supporting a rule in the Chicago code?

What does the IBEW Apprentice program have to say about this type of installation?

Joe,
I am not defending the practice, I am only saying that the current cosde wording does not support a "red tag" for the installation. There is a big difference between how I would like to see the installation made and the code minimum requirements. The CMP made a mistake in this section in the '90 code when the word "all" was left out.
The above quote is the orginal question in this thread. All I am saying is if the installer brings in #12 EGCs and pigtails them to a single EGC and lands that EGC on the terminal bar, there is no code violation as long as the splicing device is installed per its listing as to the number of conductors. Should each EGC be directly connected to the equipment grounding bar? Of course they should be. Is there a code section that forces the installer to do that? I don't think that in the current code there is.

Don

The words are here -- all of the grounding conductors are required to be secured to a "Terminal Bar" not "Terminal Bars" or to each other, or to separate lugs, or bundled!

Joe,
Nothing in that section requires any of the branch circuit equipment grounding conductors to be connected to the equipment grounding terminal bar. That is certainly implied, but not specified in the wording. The only code required connection to this bar is the EGC that is installed with the panel feeder. This is just very poor code wording. We all know how it should be installed, but the actual code wording does not require us to make the installation in the correct way. In the '87 code this section (then 384.27) said that "all feeder and branch circuit EGCs shall be connected to the equipment grounding terminal". This wording was lost in the revisions for the '90 code. The old wording needs to be returned to the code. This would be a good section for a 2008 proposal.
Are you saying that the code wording prohibits more than one equipment grounding terminal bar? Many large panels require more than one bar.
Don

Don:

The NEC 1989 TCR, on Page 352 for Proposal 9-142, Section 384-27, Log #1039 was "Accepted in Principle" and included in the substantiation were the following words:

"The second sentence is changed to eliminate superfluous wording"

The committee also changed the Section number from 384-27 to 384-20.

The way I see it is that the committee CMP-9 never agreed to what you interpret, and the only reason for this change was to eliminate wording "beyond what is required or sufficient."

I will stand fast with my interpretation and will look for your proposal in the 2008 NEC.


[This message has been edited by Joe Tedesco (edited 06-16-2003).]

Joe,
In my opinion, CM-9 screwed up. They deleted the meat of the section. They removed the only words that commanded the installer to connect the branch circuit EGCs to the equipment grounding terminal bar. The words that they left in the section only command you to install the terminal bar, but they make no command that you actually use it. Of course, the use of this bar is implied in the code section, but implications cannot be enforced. If I remember, when it comes time for the 2008 proposals, I will submitt one to put the '87 wording back into the section.
Sometimes errors in the code wording are missed for many code cycles before they are brought to the surface and a proposal to make a correction is submitted. How long did the code say that you could install conduit with 360° of bends between each fitting before the error was caught and the wording changed? The use of the word "fitting" actually permitted 360° of bend between each coupling!! Was this ever the intent of the code? No way. It was just an error that was not caught for many code cycles. I think that the section being discussed in this thread is another error of this type. Everyone really knows what the code wants the installer to do, but the words just don't actually say it.
Don

"I will stand fast with my interpretation and will look for your proposal in the 2008 NEC."

...which is exactly what this boils down to. The AHJ has the final word right or wrong. (a devisive issue indeed) I happen to agree with Don for whatever that is worth. I don't think that because you find one bad worst-case example (photo) that you throw the baby out with the bathwater and declare prohibition on splicing EGC's in a panel. I don't think a EGC needs to be continuous like GEC's. If I'm on a job and don't have a ground bar, I would like to think I am not jeopardizing practical safgaurding by splicing a couple of grounds together because I don't want to double up grounds on the bar.

And I thought interpretation of nuclear radiogical controls policies and practices was bad. I feel your pain.

Would the case of 'terminal bar' vs 'terminal bars' for grounding conductors of 408.20 be better served if the wording was " grounding terminal bar(s)" ?