Is anyone aware of a requirement in the NEC or any local code (this is in New England) to bond the metal siding and roofing of a wood framed building to the grounding electrode system of the building?

Generally no. The closest thing I can find in the NEC to mentioning this is the FPN to 250.116.
I don’t see anything the MA Amendments either, but you may have some additional local requirements where you are.
Likely to become energized wouldn’t seem to hold much weight here IMO, unless maybe you have electrical equipment attached directly to the surface.

However, if there where a pool or hot tub within 5-feet then IMO, the bonding requirements of 680.26,B [7] and/or possibly 680.43, [4] would apply for the metal siding and gutter downspouts, etc… but probably not the metal roofing itself, at least if it’s over 12-feet above the max water level.

The real question is whether it is "likely to become energized" and then you bond it to the EGC of the circuit likely to energize it.
The interpretation I have seen is when you have a penetration for an outside light or receptacle, you be sure the EGC to box grounds the metal sheathing. Since that can't be a sheet metal screw it might be a little more complicated than just screwing the box on it.

I really don’t want to pick a fight with the IAEI, but IMO, if the electrical equipment and wiring attached to the metal siding is properly installed, then "likely to become energized" is for all intents and purposes, a moot point.
The listed equipment, boxes and enclosures that are mounted directly to the siding, when properly installed in accordance the NEC and the manufacturers listed instructions, are bonded to the equipment-grounding conductor of the branch circuit or circuits likely to energize it through listed means, meaning grounding screws, lugs, etc.

All of the shock hazards that I have seen or heard of where energized metal siding was involved, were the direct result of either improper installation or a screw being driven into a wire inside a wall that was not part of a circuit likely to energize the metal siding.

The point is, even if the bonding was only a 14 ga, it would still operate an overcurrent device unless we were talking about a big feeder.
Florida is pretty serious about bonding since we are the ones who killed an appliance installer with energized steel studs and we have had problems with energized siding. Usually a metal box, securely attached to the metal siding, will accomplish this.
If this is a metal building it is a slam dunk 250.104(C)

The only thing I wonder about though, with wafer thin resi siding… is if you could actually get the two-thread engagement required to mount a lug on it.
Seems that tearing of the siding might be a potential problem when using a nut and bolt to securing a lug somewhere down at the bottom of the first course panels.

I guess if they are considering the siding panels as the "structural frame" of the building, then 250.104[C] would apply. If they are having issues in the area, I don’t see how it could hurt to bond them.
In my area, usually the steel building frame itself is bonded and the siding panels are coincidentally bonded through the screws used to fasten the panels to the metal building frame.

Please, please, please ... don't open that can of worms, KJ!

If you do, someone is sure to notice that the entire building is assembled using sheet metal screws, and will want a 'proper' lug and jumper between each and every piece that makes up the building!

Greg:
A utility meter surface mounted to the structure could also be a 'bonding' point; neutral bonded to meter pan, meter pan screwed to/thru al. siding?

Should we not consider that the service entrance conductors are the largest conductor/circuit that may energize the siding?

With 2008, 250.94, another optional bonding point? Depending upon the mfg of the 250.94 termination device, IF it's screwed to.thru the siding?

As to getting a 'two thread' tap for a lug, I really don't see that happening. WE have bonded al. downspouts, installing lugs (rivited).

I have seen an electrical job where the EC bonded the metal siding of the house to the hot tub because it was close enough to touch. He wanted to be safe and bond the 2 together.

If you have metal siding near the pool/spa it must be bonded along with screen cages, window frames or whatever, just like the ladder.
There was a lot of confusion about this sheet metal screw thing but I think that really only applies to how you fasten a lug or bond wire, not the connection of each piece.

In Florida there are not a lot of "siding" houses, at least not here but you do see it a lot in trailers and they use aluminum gable ends and soffit. The meter pan is as likely to be against block as anything and in a trailer it is in a post.

Again, a while ago, when the Prodigy SS guys were talking a lot I did an experiment and found my Ecos ground tester was happy with the incidental bonding between the gable ends at each end of my house (42') merely by the soffit between them.
The exhaust fan on one end effectively bonded that side and a test from the HVAC service outlet on the other end to the gable showed <1 ohm under load. (the Ecos test parameter)
If you made an effort to connect any metal you penetrated to the circuit you penetrated it with, I think you are good to go. Usually the box is mounted right on that metal (much liker John's meter can example).
I would want to see something more substantial with a pool/spa tho. I do believe if you get one piece, the way aluminum siding snaps together, you have them all.

There's a community up this way called "Sterling" that was a joint venture between GE and United States Steel, built in the early-mid 1960s. This planned community was meant to showcase new building technologies that would result in affordable housing. Of course, the proximity to a huge international airport flight path affected the low price tags as well.

Anyway, just about anything that could be made out of steel in these houses was just that. Steel studs, trusses, doors, siding, pretty much everything but the sheetrock. They ran a #6 aluminum wire from the panel to a lug that was screwed into the siding with a #12 sheet metal screw. These all-electric houses featured GE brand 150 amp split bus panels, plus the furnaces, water heaters and appliances. There were bonding jumpers all over these houses to tie significant parts of the structure together. I'll never understand how this actually saved any money with all of that labor.

Today, these homes are maintenance nightmares since they didn't plan on the temperature extremes between inside and out. Moisture inside the walls eventually rusted the studs and caused major mold issues.

Sorry for the side track....

Large sheets of unbonded metal are just begging for a lightning strike. I highly recommend bonding for that reason alone.

I recommend bonding for safety, as well. This metal is just as "likely to become energized" as structural steel is.

Structural steel is not, generally speaking, considered 'likely to become energized.' You don't have to bond even the largest I-beam.

If there is an appliance that might fail, and energize the steel, that is a different situation, and the bonding is accomplished at the appliance.

The only bonding issue regarding structural steel is where that steel is going to be used as the grounding electrode - for, say, a transformer. I would certainly hope no one is using sheet metal as the ground when they 'create a neutral' at a transformer!

In nearly every other circumstance, the incidental bonding that comes about in the course of normal construction is perfectly adequate; there is no need for bonding lugs or jumpers.

2008 NEC 250.104 (C) Exposed Structural metal that is interconnected to form a metal building frame and is not intentally grounded and is likely to become engerized Shall be bonded to the Service equipment encloser at the Service.
Also see 250.104 (D) 2.
250.30 (A)6.

Right ...we're right back where we started ... "likely to become energized' is NOT the same as 'gee, if I try hard enough, and enough things break, it might become energized.'

As for 'bonding.' simply screwing the panel to a column accomplishes that.

So we"re depending on the Screw supporting the Panel to bond the steel???
250.8 List Permitted Methods for grounding & bonding. I don't see Panel screws on there??
Even the service must be Bonded to the Structureal steel .

2008 Handbook page 241 Right after 250.104 (C) Explaination:
250.104(C) Requires exposed metal building framing work that is NOT intentionally OR inherently grounded to be Bonded to the service equipment or grounding Electroid syastem . This requirement applies to ALL metal framework not only to steel framework.

I'm not sure I see your objection.

250.8 mentions screws as acceptable in two forms, and the panel itself comes with a green screw for bonding the case. Ground busses attach with screws. 250.8 also mentions 'assemblies.' So, where's the problem with the panel mounting screws?

Ah, maybe you're thinking of the 'no other purpose' regarding jumpers in junction boxes. OK, the top two mounting screws are for mounting; that third one is my bond. Voila!

Perhaps you would prefer to see a little green wire poking out of the panel, then attaching it to the column with ... you guessed it ... a screw?

Well, we could weld the things together ... except that would compromise the strength of the column. Incidentally ... aren't all those bits of structural steel held together by ... screws?

As far as bonding the service, that is an entirely different matter, and is more an instance of 'deliberately energizing' than 'likely to become energized.' We're using the structure - or more correctly, it's anchors - as a grounding electrode. (Which is why we use it for grounding transformers).

Let's face it; if we were expected to bond the structural steel, we would be allowed to use it as an equipment grounding conductor. Yet, we are forbidden to do so.

Reno, your thinking may have been acceptable in 1968 but the current thinking at NFPA is if it could possibly happen, in anyone's wildest imagination we better invent a product and write a rule about it.

I've seen Metal roofing, Trailer skins, & Pole building metal engerized . So it can & does happen/
Another example they add on to a block building with structal steel and block BUT the two sections of building Old & new (steel) are not tied togerther. Thus one sided bonded other is not expect by conduits & mechical pipes. I;d say Bonding jumpers between the two seperate steel sections is required . Otherwise your depending on conduits ect to bond your steel,

Alright, but what else are we going to talk about? ... how many pieces of NM we can fit into a two-screw connector.

Got asked today; 'How many #12 THHN in 1"'

My answer: "How many guys pulling?"

That's an easy one. ALL OF THEM!

I am with Reno on this one. The only time I would even thing about bonding it my self or requiring others to bound would be if it were likely to become energize like from a badly located service drop only because a service drop is not grounded on the secondary side. I am curious about how many cases of siding getting hit directly by lightning and surviving the strike, not to mention someone touching it at the monent of the strike.

Soares Book on Grounding & Bonding 9th Edition.
Thoses of you who have it read Pages 145/146.
Page 166. Theres even pictures .
The bible of Grounding & Bonding.
Yoopersup

The Soares book just shows the structural steel frame as it says in 250.104[C] in the NEC.
As I was saying, I don’t see the exterior metal sheathing included in this requirement, but it usually ends up bonded to the frame through the screws used to fasten it in place.

Apparently though, some states AHJ’s interpret this requirement to include the exterior sheathing as part of the structural steel the building frame, so want some other form of intentional bonding jumper attached to it. Seems redundant to me, but I guess it couldn't hurt to do it.

I can see it now ... a shiny band of copper plumbers' tape, affixed all around the building at roof line, held on with sheet metal screws.

Less enticing is the thought of doing all that from a 26 ft. ladder, on uneven ground, resting against smooth steel walls, whose indentations are not quite right to catch the ladder. All in the name of 'safety,' of course.

That’s another reason why I don’t believe that metal sheathing was intended to be included in the requirements of 250.104[C].
If the building had 750 kcmil CU service conductors, according to Table 250.66, you would need to bond the metal sheathing with a minimum 2/0 CU conductor.
I'm just wondering where and how would the lug for the 2/0 conductor get securely attached to the thin metal sheathing used on steel buildings?

Ever hear of a Pole Building ??? Wooden frame all metal sheating.
A lot of theses buildings have Electrical panels , with outlets, lights, welding outlets ect.
To me if a building has Metal frame & metal sheating screwed to & the frame is bonded its bonded. But what if like a Pole building the sheating is not bonded? What about steel beams in houses the romext is layed in them ??
Instead of sly comments lets try to be constructive here
Yoopersup

The steel beam with romex next to it was the specific reason the 'likely to be energized' reason was included! It was specifically so some idiot would not demand it be bonded.

I clearly remember that debate, long ago. Those who are trying to confuse the matter are dishonestly trying to reopen a debate that was decided more than 40 years ago.

Ditto for the steel roof, aluminum siding, and every other variation. I does not need any special effort be nade to bond it.

The honest way to approach this is to propose a change, deleting that phrase - rather than trying to parse around it.

I have seen inspectors require the steel I-beam between the 2 garage doors be bonded. They said it could become energized and failed the EC when the beam wasn't bonded to the service. Even when the I-beam sat upon all wood construction.

I'm not surprised. Remember, I'm in Nevada - where folks also insist they've seen UFO's

We could go off on a tangent as to the foibles of inspectors - we all have such stories - but that would probably not enlighten anyone.

The key here is "likely to become energized." If you want to say that's a judgement call, I won't argue; it is. The far side of the moon 'might' become energized 'if' some things happen. Let's look at the 'likely' part.

To use your garage door example, I can't picture any fault that would energize the beam without first energizing the door, the conduit, the case of the opener - and that's where there would already be a bond. If you're not already bonding it at the appliance, you have a problem.

All I can say is building codes are tombstone documents. They usually get changed because somebody was killed (... or these days, some code panel member has a product they want to sell).

Yes, I've heard that said ... or, another variation, that the code is 'written in blood.' I'm not buying it. Nice rhetoric, but to imply that anything but the most aggressively expansive, "power of the pen" code application is akin to murder is simply irresponsible.

Codes, like all laws and all exercises of government authorty, are prescribed within limits- at least, in this country. The NEC itself recognizes this in its' introduction, where it speaks of the "practical" safeguarding of electricity.

We carry that through to the current example, where the code again limits its' application to 'likely' situations. Had the code panel wanted everything imaginable bonded, they would have said so.

Is there room for professionals to have differences of opinion in their judgement calls? Of course there is. That's why they're called 'professionals.'

If any code publisher wants to make the claim that theirs is a 'tombstone' document, I fully expect every sentence of code to be followed by the FPN "This line added in memory of Lotsa Sparks, whose untimely demise would have been avoided had only this change been in effect on March 32, 1874, when he was electricuted." Less specific than that, and it's just so much kant.

I guess the Florida rule for bonding steel studs could be called the Rafael Ugalde law.

When I posted this I didn't expect this much discussion~
Let me ask a couple other questions...
Has anyone seen a Listed method or product, technical reference, specification, engineering drawing, standard, or anything similar that details how to bond together perhaps several hundred individual pieces of siding or roofing such that you have can quantify the fault current carrying capacity? Say you are attempting to bond the siding and roofing of a large wood frame building. How many bonds are required and where do they have to be placed? If you bond only once in the area of the service is that adequate to clear a fault perhaps a couple hundred feet away in perimiter distance? What fault current is the design value? How about the case where it is a preexisting building and you don't know how the siding/roofing sections are joined or fastened?
Just curious.

I am not aware of any system specifically designed for this purpose.

The best I can think of is using a #10 bare copper wire, attached to the panels with the little spring clips that you sometimes see used on metal boxes; I hope they would grab the thinner metal.

For heavier metal, such as I-beams, I suppose a lug mounted to a beam clamp would work.

I say #10 for mechanical strength, and I'd run it inside, near the roof line for protection. I would not worry about a larger ampacity wire; #25 sheet metal can't carry all that many amps - any circuit larger than 30 amps is going to have to rely upon it's own ground wire.

We have a Lot of Pole buildings in our area. As some contain livestock ect we require the metal siding to be bonded. The usual Method is when the # 6 is run to the ground rod a Lug is fastened to the siding and that wire run thur the lug on the way to the grounding Electroid.
We do not require each individual sheet of siding to get a connection. We've found that to work quite well. If you want to comment on this renosteinke please make it a Polite and Possibly Constructive comment. I;ve been in the Trade over 45 yrs & formally carried Masters licenses in nine states I feel I;ve earned that!!
Thank you

Yoop:
FWIW, the method you describe above would be OK in my opinion as an inspector. Alum. siding is not very prevanant here (NJ), but I have run accross a few metal skin structures lately. Where the 'skin' mechanical connection to the metal framing was compromised, lugs and #6 were installed. I commented earlier (page 1) as to my opinion, and I'll stick with that comment also.

I don't see any general requirement to bond the metal siding, roofing, or incidental structural steel in the NEC. Indeed, by saying 'likely to become energized' the NEC seems pretty clear in saying "no need to bond."

Granted, that's a general requirement. Other parts of the NEC -pool areas and some agricultural buildings being examples- the NEC has another requirement that all metal be bonded. Those are special circumstances, and the requirement is limited to them.

I am curious: where do we get the #6 requirement? After all, we're not talking about the water bond - where the incoming water pipe might also become a grounding electrode.

# 6 Came from 250.66 (A) as the wires going to A rod.
Yoopersup

Reno:
What I said above was a 'spec' by the project engineer & architect for that structure. BTW, commercial garage, maintenance shop for a gov't branch.

Basically, if the EE/PE/Arch specs 3/0 Cu to 10' x 3/4" rods, 25' apart in a triangle, that's what they get. No, it's not NEC required. & I don't believe I said that anywhere.

I have no problem with a 'spec' .... and I have a confession to make:

I almost always use the 'quick reply' box. This seems to specifically address my comments to the previous poster ... when that is not always my intent.

In this case, two posters mentioned #6; I thank both of you for responding, and I understand both approaches.

For general discussion purposes, if the intent is to cause a breaker to trip should someone get a shock off the siding, I'm not sure I'd approach the matter the same way. I don't think there's anything in the code that would require a wire larger than #14, I don't think you need to use lugs, and I would want the wire to land on the bussbar in the panel. That's how it looks to me, anyway. I don't see the metal siding as being part of the grounding electrode system at all.

Reno:
The siding is NOT any part of the grounding electrode system, the discussion is regarding bonding the siding, or that's what I thought.

Bonding conductors are sized based on the OCP of the circuit that is likely to energize the conductive material. Hence, as I said on page 1, the largest OCP IMHO would be the service main, and in actuality the OCP at the utility primary. With the al. siding 'floating'...it could become energized by a fault, ranging from a 15 amp branch, to the 100+ service main, or the utility primary. Now, bond the al. siding, a fault trips the branch feeder, or...the al. siding vaporizes.

Sticking with a resi job, IMHO other that the service, the largest OCP would likely be the HVAC, or possibly a hot tub at 50 amps, hence a #10, or good old #6 for a 200 amp service.

Again, this scenario is not (to the best of my knowledge) addressed within the NEC directly, but..."likely to become energized" could be interperted many ways.

I follow you now. Another poster mentioned running the wire to a ground rod ...