Could someone please direct me to a written code requirement outlining the bonding of metal studs when non-metallic electrical boxes and wiring are installed within the walls?

I don't think that 250.104 [C] or D, [2] applies here, since these metal studs are partitions, they are not really structural and are also normally faced with drywall so are not exposed.

I remember this subject being brought up at a code seminar several years ago, and it was recommended that if NM boxes were used, a tail should be extended from a branch circuit equipment grounding conductor in one of the boxes and attached to one metal stud in the wall with a lug or machine screw, but I can't recall if there was an actual reference to this in the NEC though. Obviously this is less of an issue when metal boxes are used.

Thanks

It has not made the NEC as far as I know unless you want to get fast and loose with "metal that may become energized" but the Florida building code addresses it directly.



This is one of the few additions to the NEC they have.

The issue is a timely one ...

At the Home Inspector forums, they're asserting that such 'light steel framing' MUST be bonded, and tell all manner of tales.

I agree that the NEC does NOT require such to be bonded. To me, ?structural steel' has a red primer and is at least 1/4" thick.

My experience has shown this type of framing to be something used only in commercial settings - where, oddly enough, practice has been to not use NM or plastic boxes.

Otherwise, as with suspended ceilings, I've always doubted it was possible to make one without it somehow being bonded.

On balance, I suppose it'snot such a bad idea to have some manner of bonding; I just don't want someone to go overboard, and require listed bonding jumpers all over the place! The last thing wer need is someone worrying about 'sheet metal screws being used ... to connect sheet metal!

Interesting topic john (oops I mean Reno )

Last years code revue/con-ED, we discussed the need to bond 'Victolic' fittings on the sprinkler systems (and others).

PS: On those HI forums,do they always argue like that? what a bunch of.....Lack of suitable word.

This may be why Florida has a specific requirement for bonding metal studs.

http://www.clickorlando.com/money/7302637/detail.html

Sad article Big.
But a few questions remain; What are the wiring standards there? Metal boxes,MC etc? Wood frame metal studs?

Here in the Northeast,I find it hard to find any metal framed wall not to be bonded.Not that it can't happen (as this article shows) But in proper construction,accidental bonding is far better than I could provide.

The typical 1 & 2 family here is wired with Romex in plastic boxes with steel framing unless it is a load bearing wall. All plastic piping for water and DWV.
They avoid wood because in our wet climate it is hard to find a straight 2x4 and these days mold is also a big issue.
If the structural inspector sees a moldy 2x4 on the FEPAC you are not hanging drywall tomorrow, you are getting a mold mitigation certification. It was so bad my wife was just getting her trusses sprayed and certified as soon as the roof was black, whether they needed or not. It is not a cheap sticker to get.

I said long before the guy got killed up there that they should use at least one steel box in each isolated wall section. I have even heard of a plastic box with a bonding screw to the metal mounting strap. That would work too.
In most houses, if you bonded the wall in the kitchen with the range box, the laundry with the dryer box and got the bathrooms with a metal box you would have most of the hazards covered. That is where that metal framing might get connected to something you could touch. In the case of the guy up north it was the dryer vent box. It could also easily be the medicine cabinet in the bathroom.
I also think metal sinks should be bonded but that may just be me. Usually, I suppose the disposal does that but I haven't seen anything in the listing to prove it.

This is similar to what I see around here as well, especially when a basement get finished off.
Sometimes when the metal stud wall is attached to the concrete floor with PAT fasteners, that is as close to grounding as they will get without any metal boxes being used.
At least your state has addressed the issue, I'm not sure that my state's building code has any reference to this at all yet, but I will continue to look.

I have seen alot of what you describe Kjay, I always use metal boxes on the metal studs. Perhaps just my Industrial/commercial background. Habit more than anything else.

Lee, do you recall seeing anything regarding bonding metal stud walls in the MA building Code? I was at a seminar several years back and this came up, but I haven't heard anything about it since.

TKX

Nope. Last I heard was the big to do over "Victolic" fittings. That vanished as well.

I think the language adopted by Florida in re: bonding metal studs would be a good addition to the next NEC revision. Pretty simple, pretty cheap, and would eliminate the risk of energizing a metal studded wall section. Just require one metal, grounded box attached to each partition/section, thus grounding the whole section. That should be pretty cheap to implement, and, as the news story about the tragic death of the dryer installer demonstrates, this is an issue occasionally.

It is even easier than that. Just change the listing standard for plastic boxes, meant for metal studs to include a bonding lug for the metal mounting strap.
It could be as simple as a "stab" into a lip that goes into the box.

I guess I'm not concerned with the metal studs inside a wall. At the risk of suffering the wrath of the members of this group, and given the subjective words "likely to be energized" I would not have a contractor bond the metal studs at all. Is there any track record of a problem? I say if the wiring method is installed properly, the likelihood of a problem is not there.

I would say that the unfortunate dryer installer's death constitutes a compelling 'track record of a problem.' While the problem (electrifying the studs) requires some very unusual circumstances to manifest itself (errant drywall screw piercing an improperly installed piece of romex), the risks are sufficiently grave to justify a regulatory response. All metal studs should be grounded. This is easily done, either with a metal box or jumper of some sort. From a cost/benefit perspective, it is a low cost/high benefit situation.

George, If this segment of wall was truly isolated, with no way anyone could contact an energized part of it I might agree there is little hazard, until someone drills into a stud with a screw to hang something heavy. At that point you have brought the hazard into the room.
In the Florida incident it was the metal dryer vent enclosure that was screwed into the steel studs. It could have easily have been a metal medicine cabinet or a sliding glass door/window frame.

If you click the link and click on the 'video' tab you will see the news segment that ran on TV about the dryer installer's death. In the segment, there is an electrician testing the medicine cabinet and getting a reading of 120V. The little dry wall screw that barely pierced the romex created quite a hazard. I would think, code requirement or not, it would be a 'best practice' in the trade to bond a metal stud wall with electric wires running through it.

Did you guys happen to notice the word "Bazaar" or did you notice that the wires were not properly installed? I'm saying that things like this happen no matter what the code says or how we try to avoid them. We have inspectors who want the I beams bonded because they are "likely to be energized" because wires pass over them. I think bonding the metal studs inside a wall is right there with paranoia. Send in a proposal and see how far it goes.

It was quite a 'bizarre' accident, and unlikely to be repeated in exactly that way. However, it certainly does prove it is a potential hazard to leave metal studs ungrounded. Regrettably, improperly installed wiring is all too common in residential construction, and all too commonly overlooked at the inspection stage. The improper step here appears to be the romex was not clipped or connected as it should have been under code, thus was too close to the drywall. When the screw was driven in, just the point pierced the romex, thus creating the short that energized the studs. This wiring error would be easy to overlook, and evidently was overlooked at the rough-in inspection. To further compound matters, it was passed even at the final inspection, two weeks after the man was killed. Hard to know what make of that one.

In any event, quite a few code provisions are included to deal with unlikely events. It seems to me that since there is very little, if any, cost associated with bonding a metal stud partition, and the risk of electrocution certainly is present in worst case circumstances, why shouldn't it be a code requirement?

You might want to contact the widow of the man killed and ask her if perhaps it is paranoid for anyone to be concerned about this hazard.

My,my, the rhetoric is getting thick!

While I'm sorry the guy died, let's not overlook thefact that several things 'went wrong,' and that existing rules were violated.Passing new rules only affects those who follow them!

Track record? I think not ... 'freak accident' seems more appropriate.

There is a limit to "practical" safeguarding.

Just a thought...
Would a combo AFCI have tripped under the conditions of this accident?

"Track record? I think not ... 'freak accident' seems more appropriate.

There is a limit to "practical" safeguarding."

You are correct, insofar as there is a limit to 'practical' safeguarding. To me, it is very practical to require bonding a metal stud wall. What's it cost? 40 cents worth of wire? How long does it take? 30 seconds or less. What does it do? Protect against a one in a million fatality, of which one occurrence is documented. To me, that's reasonable regulation.

Here is what would be unreasonable regulation: suppose the drywall screw was the only element energized by the romex. Suppose a painter or drywall man touched it and got killed. That, to me, is a totally freak accident, and completely impossible to regulate against in any practical sense. That is a one in a billion accident.

Potentially energizing a metal stud wall is easy to prevent and a prudent regulatory measure, in my opinion. Besides, suppose you take the position that it isn't necessary to require that a metal stud wall be grounded, regardless of the death. Then, six months later, it happens again. What to you tell the victim's family?

The code is based on the idea that you need several things to go wrong before you present a hazard. In this case there was really only one mistake, either not maintaining the 1.25" setback or using a 2" screw. Take your pick.
The picture I saw showed this in the corner of the framing and it was hard to figure out how far it was from the surface of the closest framing.

John, I see no reason why the AFCI would have tripped until the victim had taken 30ma of fault current.

According to the article, it was a 1 1/8th inch screw, but looking at the picture, I think it was a 1 1/4 inch screw. In any event, it certainly was a rare event.

I agree, but I think it's too late for the next code cycle anyway, that is unless someone has already submitted a similar proposal.
So now that I think about it... wouldn't this mean that even if the NEC wanted to address this issue at all, it would have to wait until 2014 anyway?

My how the wheels of change grind slowly along in a bureaucracy... unless of course, it comes to something like handing out taxpayer dollars like a giant Pez dispenser or changing the tax code, which can seemingly be done overnight. But that's another bedtime story.

Some think the NEC is revised too frequently. Of course, it isn't a government mandated or even created code. It is simply adopted by many government agencies (as it should be) because it is the most comprehensive and thorough set of model regulations governing electrical installations. Regardless of whether it is required or not, I think it is simply good sense to bond a metal stud partition. It costs little, and can mean so very much.

I agree, but I think it's too late for the next code cycle anyway, that is unless someone has already submitted a similar proposal.
So now that I think about it... wouldn't this mean that even if the NEC wanted to address this issue at all, it would have to wait until 2014 anyway?


[/quote]

Great example to promote Gregs' view of a 5 or 10 year code.
The new code is being written and half the country doesn't even know the current one.

Lee:

"Great example to promote Gregs' view of a 5 or 10 year code.
The new code is being written and half the country doesn't even know the current one."

I had a student ask last nite...'why are you (me) basing this class on the 2008 NEC, when NJ has not adopted it yet'

Last I heard from DCA the '08 was still pending adoption'.

Greg:

"John, I see no reason why the AFCI would have tripped until the victim had taken 30ma of fault current."

Would not a combo AFCI have detected this situation as a series fault, as it appears the screw hit the 'hot'?? I'm under the impression that a series arc (break/nick) in one conductor will cause a trip, and a parallel arc between two conductors will cause a trip, hence "combo". Is my summation correct, or am I not seeing something? Please remember, AFCI's are not required here in NJ yet...

Final thought from me on this subject. It is of course a tragedy that anyone would get hurt or worse killed. It could have been prevented possibly and there is no way of knowing of the numerous things that can happen. As an inspector, I can not ask a contractor to do something that appears to be a good idea. As a contractor I am hesitant to do something different than code because I think it is a good idea. If the code panel is convinced by way of substantiation that the code needs to be changed, it will be changed. The code is a reactive document not a proactive one.

George, that is true. In Florida the code "reacted" to a dead guy.
There is nothing that prevents a contractor from bonding steel studs. In fact it may even be implied if you have a wide view of what is "structural" or "likely to be energized".
This would not even be a question if the family sued the box manufacturer for not providing a bonding means on the box.
They would all have one as soon as the line could be turned around. Then it would be a 110.3(B)

John, until there was a fault path there was no arc to detect. AFCIs look for current anomalies. Even if the wall did provide a significant load, a power driven screw may have made a clean "make" of that circuit. Your best hope would be that it was a 30ma or greater ground fault. If this was tapconned/shot pinned into a concrete slab, tied to the Ufer, it probably would trip a AFCI or GFCI but if it was on a wood deck, maybe not so much. You might not even get a trip if this was stem wall construction with a floating slab, sitting on visqueen. Usually it is on notched block and poured with the stem wall cell pour tho, in Fla, so it would be integral with the Ufer.

There are several ways to approach this.

The first is to say 'too bad' and moveon. You can't prevent everything.

Then there's the old practice of megger testing your circuits before you fire them up.

If you believe that there is a need for some 'common sense' to enter the code, where to start? Romex in metal studs seems to be, well, wrong (and I don't care that they sell bushings for the openings!). I'd like to see how the stuff is stapled! Add to this using plastic boxes - I don't see how those boxes were mounted securely - well, I guess I've lived a sheltered life. Must be from growing up in Chicago At any rate, here's an ideal opening for local amendments to simply say: No Romex in metal framing."

GFCI breakers? AFCI breakers? Once again, we're using technology to try to accomodate an inadequate wiring method.

Reno, they have plastic boxes meant for steel studs (why I think the bonding lug could be integrated quite easily).
It has a metal mounting tab. The grommet provides the horizontal support, just like Romex in a bored hole. The vertical support and within 12" of the box, is usually a tywrap.
You can get mad about plastic boxes but they are cheap and fast to install (no Romex clamp screw to tighten) so I think they are just a fact now.
I am not an engineer but it only took me a minute to come up with several ways to get from the EGC to the metal mounting strap. It could be as complicated/expensive as a green ground screw in a tab that comes in the bottom of the box to a spring stab or just a metal lip over the edge of the box you could drive a ground clip on.

I do think there is already a plastic box with a bonding means but I don't have a link to it. I think I saw it mentioned on the Fl IAEI BB when this first came up but I may be wrong about that.

At any rate, here's an ideal opening for local amendments to simply say: "No Romex in metal framing."

That's a good idea, functionally, but financially doesn't that add significant cost? I assume AC/MC/BX whatever would be the alternative, which is more expensive and problematic to use with plastic boxes.

I agree though, Romex and metal steel studs just seem...wrong. I did a walk through with a senior inspector one afternoon last summer and it was weird seeing all that yellow wire snaking through a metal stud wall.

Greg, if there are boxes intended for mounting to steel, then I am simply astounded that ther is no 'idiot proof' bonding already incorporated.

Cost? As the AFCI advocates will point out, cost is not a consideration. Any wiring method has to be suitable for it's use ... and perhaps we have to re-think this expanding acceptance of Romex. With Romex being expanded in scope to commercial buildings, there will be many more opportunities for it to encounter metal studs.

Heck, with all the fuss AFCI's have caused, perhaps we ought to re-think Romex alltogether. After all, there's all that concern about the 'errant staple.'

I'm not being critical here; rather, I'm showing my ignorance. I just can't see Romex in either metal studs or commercial uses.

I guess the issue here is metal studs are very popular for 1 & 2 family in Florida and nobody wants to make MC cable or pipe necessary there.
I tend to agree that RX may not be particularly appropriate for most commercial

Metal studs absolutely must be bonded for safety. It's no different than I-beams in a commercial building. If the manufacturer has them UL listed for screws providing adequate bonding, that's great, that's easy, just bond each section to the highest ampacity circuit it's likely to be energized with.

Regardless of what's done for safety, the metal studs in a house should be grounded for lightning protection. Not so much to protect against a lighting strike, but to prevent one; ungrounded metal can become charged during an electrical storm and attract lightning bolts. Proper grounding of electrical systems and metal building components is a lot of why homes are only very rarely struck.

Steel studs are not, and have never been, considered in the same category as structural iron - at least as it pertains to bonding.

Nor should they! The last thing we want os for somone to take the next step, and deliberately use them as a grounding conductor - as we already do when we bond a transformer to structural iron.

I think you have an 'apples and oranges' situation here, as bonding a metal stud to provide a safety ground back to the service panel in the event of a short is hardly the equivalant to using a metal stud to bond a transformer.

Bigplanz ... I agree completely. I am just at a loss trying to write code that differentiates between "structural steel" and "steel framing."

Florida's code section seems pretty straightforward and clear to me. In the 'definitions' section of the code some language could be added to differentiate between structural steel (intended to be a load bearing component of the structure) and 'steel framing' (non-load bearing, structurally designed to support interior walls or partitions).

I am still not sure why this isn't a 250.104(C) situation.

I don't know... I kind of went round and round with that same thought before I originally posted, but the reason that I didn't feel that 250.104[C] applied here was because it mentions structural metal that forms a building frame and that this framing must be bonded at the service enclosure or the grounding electrode conductor at the service, etc... and that the points of attachment must be accessible, so it doesn't sound like it was intended here. It also refers back to 250.64, which doesn't really seem like it would relevant here either. I don't think that the reference to Table 250.66 would be correct here either even though it is about bonding conductors. If it were, we would have to pull a minimum #8 back to the service to bond metal stud walls, which I hope isn't true because that would be insane.
I would think 250.122 would be more inline instead, even though it is for EGC sizing, not bonding conductors.
Another thing that I'm not sure there would be any provision or exception to allow for the use of metal box as a means for bonding the stud walls.


At the moment, the articles I feel most closely apply, so far, are 250.4, 250.8 and 250.102, but of course, subject to change as things progress.
Who would of thought that something as simple as bonding interior metal stud walls could be such a terror to explain in NEC.

Metal studs may not normally be load bearing, but they're still structural and just as likely to become energized.

In many applications, they are load bearing. Why should we consider them differently from one application to another just because of vertical load? Loading of the members makes no difference to the electrons.

I think the real issue is "exposed"(per 250.104(C)) Electrically, as soon as you put a metal cabinet, door or window frame in there that is accessible to the user, the stud is exposed.
Florida is giving you a break by allowing 250.122 bonding based on the circuit likely to energize.

I truly hope that they are categorized differently because if not, they would be subject to 250.104[C] and 250.104[D],2 which would mean a bonding conductor brought back to the service entrance and could also allow metal stud walls to be used as a grounding electrode for a transformer, as Reno pointed out earlier, since there could be a separately derived system serving one area of a building.
The thin gauge stamped steel studs walls are held together with short self-tapping sheet metal screws and could be required to carry a tremendous amount of fault current under the wrong circumstances. This is not very safe or adequate IMO.

I am not sure 250.104(C) implies this is a 250.52(A)(2) ground electrode although it is bonded that way.
I assume they do that because they are not defining the circuit likely to energize. If you think the 60a to the heaters in the air handler or the range is a candidate, a #10 is not unreasonable. I still think if you just used a metal box for the dryer, the range and the couple required GFCI outlets you are covered. Those are generally where your hazards are. (bath, kitchen, laundry and garage/basement if there are steel studs.)
Then pick up any wall segments you haven't hit. This is the kind of thing a decent CAD program could highlight on the plan.

At the risk of being repetitive ... why not address the porblem at the source - and ban Romex from light steel construction? How many 'band-aids' will we ultimately need to compensate for an inadequate wiring method?

I doubt you will get much traction in banning RX in 1&2 family. The NFPA seems to be going the other way.

Metal studs are no different than screw-set couplings with respect to impedance and quality of bonding, yet the latter are considering adequate for grounding...

If I was writing the code book, I would require ground wires to be run with every circuit so that none of this other metal is primary fault grounding path... but I would still required it to be bonded to ground. I like the florida code, too, that allows them to be bonded to the largest circuit they're liable to be energized with. If only 20A runs are near that wall, #12 is fine for bonding, etc.

If properly run (with grommets or strain relief couplings, etc), steel studs are no more dangerous to romex that metal panel enclosures or handy boxes. Just because somebody "might" get lazy with it and ignore the rules is no reason to ban it.

Alas, Greg, I argee - the code committee seemsintent upon expanding, rather than restricting the use of Romex; not the rather recent change allowing it's use in come commercial occupancies.

Maybe that's why we need to head this off at the pass; one is far more likely to encounter metal studs in commercial construction.

Steve, as for conduit being allowed for grounding, you are only partly correct.
Conduit is considered adequate for most branch circuits, and there the use of conduit as a grounding conductor ends. For feeders, and for bonding transformers, etc., a separate ground wire is needed.
If you want to get really pedantic, I was actually referring to using structural steel as the 'grounding electrode conductor,' (as opposed to 'equipment grounding conductor.')

"Common sense" is, right now, enough to make clear that you don't use light steel studs for this purpose, while an I-Beam will work. There is still some question as to the heavier 'sheet metal' structural forms used in those quick-to-erect light steel buildings. If you thought the code debateover 'sheet metal screws' was fun, just wait until we start debating 'structural steel!'

It's already pretty hard to isolate / not ground light steel framing in the course of doing other work - the example cited in this thread notwithstanding. Sort of like suspended ceiling grids; try NOT to bond one some time! Add a metal jacket to your wiring method - as with MC - and I don't think it's possible.

Sure, simply using a metal box will also create this bonding - I can hardly wait to see the arguments over where such would be required - but three remodels later, and it's anyone's guess where the 'isolated' sections are in relation to those boxes.

The more I think about this, I conclude that IF there is a problem, it lies in the choice of wiring method.