What is the general policy for derating NM cable conductors when they are pulled through wooden studs or basement ceiling joist? Are the inspectors asking that the derating apply after more than 3 NM cables are pulled through the same set of holes? NEC 310.15(B)(2)

George they do ask for that here in MA.

More than 9 current carrying conductors through the holes in studs and joists will need derating.

However we also have less a less restrictive derating table.

I understand there is much controversy about this and many areas do not ask this. I can sure see it going either way.

Not much help.

Bob

I think this gets down to the "bundling" question. If they fan out between bored holes this is a pretty thin beef. If the installer was very neat and they run in military precision I see the *potential* problem with the ones in the middle.
Rake them out with your fingers. Then they are not bundled. Twisted and kinked is a plus in this case.

It seems strange at first but "neat = heat".

The same thing is true in a panel when you see those "aircraft like" wire harnesses running down the gutters, tywrapped, with breaks at each device. It looks beautiful but derating can rear it's ugly head.

As I said I can see both sides and my feeling is the NEC is on the same page as our local inspectors.

Consider the following exception to the derating rules.


To me this exception demonstrates that the NEC considers MC supported on bridle rings every 5' to be bundled.

If they call this bundling I don't see how we can not call NM through studs bundling.

MC hung on bridle rings always 'fans' out some.

Really what I believe and I know of someone putting in a proposal along these lines is this whole issue should go away for NM in dwelling units as you will never have all the conductors fully loaded for any length of time.

Here in MA we still use a lot of NM in commercial jobs and I have found some pretty warm bundles of NM in these locations.

What about the "without maintaining spacing"?
If you have the NM fanned out so air can freely flow around them they are not bundled for greater than 24". MC and AC may not maintain the spacing because of flexibility, shape and weight, but NM will pretty much stay where you put it in a joist bay between stud bores.

I guess a related question is how big a hole are they boring if they are getting more than 3 NM in here (9 conductors).

We do not 'drill' many holes for NM, generally we punch them and use plastic bushings in the holes. The bushing must be about about 1" ID.

I don't find it is convenient to run to many cables together, I would rather just make more holes.

Where I see the bundles of NM is usually the last 5' to 10' into the electrical panels.

Ever see a bundle of NM with about a 12" OD?

George specifically addressed stud bays and joists but I certainly agree that "bundle" going in the top of the panel can be troubling. I haven't seen 12" bundles but 10 or more RX shoved in a 2" nipple is not unusual.

Here is one from the HI site



[This message has been edited by gfretwell (edited 09-06-2005).]

Does it meet 310.15(A)(2) Exception?

Thanks Greg- I agree the "stove piping" happens all the time without any regard for the fill or bushings. Can I pick on your picture for a minite? Let me get your permission first

It's not my picture. I robbed it from the home inspector BB.
There are plenty of things going on there. I was just pointing out the 2" full of RX

I read somewhere recently (IAEI Mag?)
that the next code cycle is going to require derating NM that passes through bored holes that are fire stopped or otherwise "sealed".
Can anyone help me remember where I read this?

Redsy- your information is in 334.80 2005 NEC.

Now help me find the information that told us how to install "stove piping" of NM cables. I can't seem to find it. Probably was removed and no dot in the margin. I hate it when they do that.

Greg- I'm not going to pick on your graphic, since it's not your work. You, as an inspector and probably see the same things I see and we'll just leave it at that.

You get to see the 12" OD bundles of NM on jobs that have multi panel switch gear and someone thinks it's a good idea to enter all the cables for 8 or 10 - 42 circuit panels in one location.

If this is a 200amp service, there is no need to derate below 20amp for #12, 30amp for #10, 40amp for #8, and 50amp for #6.

I cannot pack NM close enough to keep the heat in. (I drill 3" holes in I-joists and pull a mix of #6 for subpanels and #12 for general circuits.)

312.5(C) exception? It survived into 2005 unchanged

Thanks Greg- I looked for about half an hour and had given up. Now I can sleep tonight.

When I read this code reference I'd say it is usually a violation on most of the jobs where they use "stove piping" and most contractors aren't up to doing the calculations. I personally don't think it's the most dangerous problem out there. If there were a number of other problems on the job I'd have a contractor fix the over fill on the stove pipes but if the stove pipes were the only problem I guess I'd pass the job. But don't tell anybody.

Can you explain this?

The size of the service has no bearing on the rules of 310.15.

You know, I'm kinda hackish, and that panel looks messy to me.

Real code issues aside.

The NEC should have an exception for single family for de-rating, just like the state of Wisconsin.

iwire ---

I just do electrial engineering. I am not responsible for the tables, the assumptions used in making the tables, or the errors in the tables.

One assumption that the tables make is that there is enough power from the service to allow all the circuits to draw their rated amps.

On a 200amp/220volt service, 4 50amp/220volt circuits draw a total of 200amps (average or 50), 8 50amp/220volt circuits draw a total of 200amps (average of 25).

The service derates the conductors. So if you pull either the 4 or 8 circuits through a single hole in the joists you will get about the same temperature rise.

I could do an example using 20amp circuits, but that requires actual math.

George regardless of your profession this is an NEC forum and if you are not basing your response on a code section you should state plainly that you response is an opinion.

People come here to learn the NEC.

FWIW It is possible to fully load twenty 120 volt 2 wire circuits from a 200 amp 240/120 service. Therefore you could have 40 fully loaded current carrying conductors in a single raceway.

Just My Opinion, Bob

Bob, I see where George is going and I tend to agree with the sentiment. From what Tom is saying, Wisconsin does too.
Load diversity in residential circuits does make a lot of this derating somewhat moot.
If you really had 20 fully loaded 20a circuits, the 2 ga al service conductors would burn up.

All this said your point is still valid.

We have to go to war with the code we have not the code we wish we had.

Greg I agree that in residential circuits some changes could be made.

I am still going to be a stickler for NEC facts in a forum called NEC & other Code issues

Everyone is more than welcome to post their opinions on NEC issues all I ask is do not present an opinion as code fact.

That is IMO

Bob

Bob- I'm with you. I have to enforce the code. I can't afford the insurance needed to enforce my opinion on code issues. Maybe I don't think it's necessary to run conduit when feeding a swiming pool panel I think you should be able to run SER. - I have to enforce the installation of conduit. Derating applies to all. No exception. When I review plans I expect to see atleast minimum code.

iwire ---

The question concerned NM run through holes in joists. I suspect that the question was asked because the code is silent on what constitutes bundling.

My response was that I could not bundle NM tight enough for derating to be necessary.

You are not allowed to make up a definition of bundling and enforce it as part of the NEC.

Actually this would also go against real world work ethics. You cant run that tight of a bundle through a hole in joists without pulling individual wires. That tight would burn the insulation on the NM. Also would make it alot harder ( I am referring to running many NM's though a single hole)

Stovepiping is the bain of every Sparky out there doing service changes. It is absolutely stupid.

Am I the only sparky that uses yellow #77 to help reduce cable sheath friction burn?

310.15(B)(2) says "stacked or bundled" so the degree of tightness becomes a non issue IMHO. NM cable pulled through 2x4 studs whether steel or wood could qualify for derating.

again in real world conditions...you would NOT run that many NMs through a joist. You would drill more holes just to make it easier. I would NEVER have to lube an NM cable to get it through a hole...grab a drill and drill the joists.

Yes a joist is considered a conduit...just remember the next sparky after you, leave him/her a nice place to work in.

This means:
an open/not filled conduit, an easy way to run wire, a CLEAN job, a nice cut in, ACCURATE information...etc

I am in the real world. You'd be shocked at the bundles of wires I see run through the same holes. That's what triggered this post.

George Little ---

Tightness does matter ...

I expect you are commenting on 3 NM cables going through a 1" hole.

I expect you allow the same 3 MN cables to be run through a 10"x14" joist bay.

Will you allow 3 NM cables though a 4" hole? A 3" hole? A 2" hole?

When the code contains a section that states a specific spacing for cables, I will change my position as to what the code allows. Until then anyone who requires derating because cables pass through the same hole is simply wrong.

George


That is simply your opinion, it is not a fact.

As you stated there is no definition of bundling, that being the case it is up to the AHJ or Inspector to make the call on what is or is not a bundle.

Either way the AHJ is not 'wrong' it is no different than "exposed to physical damage" it's an inspectors call.

Bob


[This message has been edited by iwire (edited 09-08-2005).]

iwire ---

When the NEC is made law, there is no "inspector's call."

In any case we have forgotten that bundling for less than 24" does not require derating. At worse the holes in joists bundle wires for 3" or so. (I guess that makes my "opinion" right and yours "wrong".)


[This message has been edited by George (edited 09-08-2005).]

George


We will have to agree to disagree on that.



What is "subject physical damage"?

What is "adequately protected"

Who makes those calls?

You won't find a guy that dislikes 'inspectors calls' more than myself, however even I have to admit that there are items in the NEC that are left up to the judgment of a single individual.

Bob

Example to ponder:
Hall with 12 light fixtures, each controlled by its own switch. All switches are located in a single (very ugly) multigang box, all fixtures are fed from a single raceway. Each fixture has a ballast rated at 1.3A. All fixtures are supplied by a single 20A circuit. The raceway from the switch box to the first fixture is carrying a total of 14 conductors: 12 switched 'hot', 1 grounded, 1 equipment ground.

For purposes of derating, how many current carrying conductors are considered to be in the raceway, and is 12ga THHN suitable for this install?

If we assume single phase service 240V/120, and replace the above install with a multiwire circuit, where 6 of the switches are supplied by one supply leg, 6 by the other supply leg, with a single shared neutral, does the 'official' number of current carrying conductors go up or down?

In this latter case, is the more or less chance of overheating the conductors?

-Jon

Jon given the installation you describe you have 13 current carrying conductors.

This means a 50% downward adjustment of the conductor ampacity resulting in the need for 10 AWG 90 C conductors.

In the second version of the circuit the current carrying conductor count remains the same as the majority of the load is non-linear.

That is my opinion of the NEC rules.

From a practical stand point no derating should be needed in this installation.

You would derate based on all 14 conductors but since the load on each is 1.3a you are well within the acceptable range.

Bob why would I need to increase the wire size if the applied load is only 1.3a each?
You might be able to argue that the grounded conductor should be #10.

Greg the load is not the issue with the NECs present rules on derating.

Let me ask this.

If this was only one 1.3 amp fixture on a dedicated circuit could I run 14 AWG to it but still use a 20 amp OCP?

No, because the rules of 240 require a conductor for this application to be rated higher then the OCP.

Now when we derate the ampacity of the conductor we must not use an OCP above that adjusted rating.

Thirteen 12 AWGs in a raceway have a rating of 15 amps at best, so we would need to drop the OCP down to 15 or increase the wire size to 10 AWG.

Bob

[This message has been edited by iwire (edited 09-08-2005).]

"Now when we derate the ampacity of the conductor we must not use an OCP above that adjusted rating."

Do you have that code section? I may be missing somerthing ... it happens

Greg - Wire has to be protected at it's ampacity 240.4 and if their is an adjusted ampacity the overcurrent must match that new ampacity and sometimes you can round up.

I think a more accurate statement than that would be "conductors aren't allowed to melt", citing section 310.10.

Citing 240.4 as an all-encomposing provision is little bit untrue, when you consider the applications of 240.4(A), (E) and (G)

Ryan- I like that (310.10), can't say as I've used that before in my travels. That's the beauty of this forum, you can always learn something. even an old geezer like me.

How could the wire possibly "melt" if the maximum current spread across 12 conductors is 20a.
If this was 12 20a branch circuits I would see your point.

If we didn't enforce the code, what would happen when the fixtures are changed?
The load could be higher than the load at the time of installation.

I posted the example of 12 'switched hot' conductors as a thought experiment.

I agree with iwire: it is a violation of code not to derate these conductors, but isn't really a problem to protect them at 20A.

The worst case scenario for heating is if all of the current flows in a single one of these 12 wires. All other conditions result in lower I^2R heating. Clearly having multiple conductors sharing the current should be allowed whenever a single conductor is allowed (from a heating point of view....there are current sharing issues that would make it a bad idea to use small conductors as actual parallel conductors; but having them run in parallel to _different_ loads does not cause such problems).

IMHO a reasonable additional exception to the rules used in counting current carrying conductors would be: For the purpose of counting current carrying conductors, a set of two or more conductors supplied by the same branch circuit and carrying net current in the same direction shall be considered a single conductor.

I think that the wording would need some cleanup in order to avoid switch loops counting as a single conductor, but you get the drift.

I doubt that this is actually a frequent enough situation to actually warrant a code change, however.

-Jon

I also have a problem with a residential situation where several cables leave a multi gain switch box that is fed by the same circuit. There can be no more heat at this point than the breaker itself will allow.

I was invited into a situation a couple of months back where I knew both the electrical contractor and the inspector. The electrician was turned down on rough and 334.80 were citied as the violation. The electrician trying to repair the problem got in trouble with the building inspector for destroying the top plate.

After meeting with the inspectors and contractors we all agreed that 334.80 at a switch was a little ridiculous. This home was on a pad and had two four gain switch boxes in one stud bay. The outside bracket lights was the only single pole switch.

Made my heart feel good that I was able to mitigate this situation to a peaceful and understanding end with all parties in agreement.

"If we didn't enforce the code, what would happen when the fixtures are changed?
The load could be higher than the load at the time of installation"

You still are limited by a single 20a O/C device. How much bigger can the load get?

BTW the code does not protect against the next stupid thing an installer might do. You can put in a 400a fused disconnect with 225a fuses and you only have to serve it with a 225a rated conductor.
Sure someone could come along and shove in 400a fuses. That is not addressed.

(editted, found answer to my dilema)


Dnk......

[This message has been edited by Dnkldorf (edited 09-10-2005).]

Table 310.15(B)(2)(a) doesn't have "qualifiers", that permit field modification of the rule, so derating should apply.
It seems to me that derating these conductors to 50% (which would result in an ampacity of 15 amps), would be similar to using #14 wire as the switch legs to the lights on a 20-amp ckt. That is something we all have agreed that would be prohibited.

There _are_ qualifiers to 310.15(B)(2), in the form of the exceptions that permit certain conductors, which are clearly expected to carry current, but which are not _counted_ as current carrying conductors. Eg. the permission to not count a neutral which only carries the imbalance current in a multiwire circuit, with the qualifiers that insure this is only used when the neutral actually doesn't carry current.

The example set of 12ga wires (which under current code _must_ derated to 15A) is different from a 14ga switch loop in a very important fashion. If any _one_ of these conductors is actually carrying 20A, then the other conductors cannot carry much current or the OCPD will trip. In other words, if any one of these conductors is actually carrying the full circuit current, the other conductors are _not_ carrying current.

If I stuff a conduit with a single 'hot', a single 'neutral', and 11 unused conductors, then under current code none of the conductors would be derated.

-Jon

I understand that exceptions exist, but the exceptions wouldn't pertain to this example.
I believe that current carrying conductors are current carrying conductors, regardless of the amount of current they are carrying.

Redsy,

I completely agree. 1) Under present code this exception does not exist. 2) I believe that this situation is a safe situation and thus a reasonable exception to add to the code 3) I believe that this situation is unlikely enough as to not be worth the time to add such an exception.

This was more of a thought experiment that goes back to an earlier discussion in this thread, and a much more common situation: if you have a panel with limited feeder capacity, does this feeder capacity in any way influence the derating of bundled conductors fed by this panel. The answer is that there is no provision in the code to adjust the derating to account either for load diversity or feeder capacity. The question then becomes: is there a way to describe a _safe_ exception to 310.15(B)(2) that would account for such load diversity.

I don't know that there is a good way to build such an exception without restricting it to a few number of extreme cases as to make the exception essentially useless.

To get back to load diversity in residential panels with 'stove pipes' (big bundles of romex coming off the panel), perhaps a way to calculate the 'effective number of current carrying conductors in the worst case' would read as follows: 1) take the result of the service calculation as per article 220, 2) allocate this current to circuits operating at their rated current, starting with the smallest capacity circuits, 3) count the number of conductors used in this 'worst case' situation, 4) use this number for determining derating using 310.15(B)(2)a

Again, this is _not_ code; this is a vague outline of a suggestion for an exception to add to the code.

To work the above example, consider a residence with 200A service 240V/120 , a calculated load of 77A and a 40 circuit panel with a bunch of 20A general lighting and receptacle circuits. If all of the conductors were stove-piped, there might be 60 current carrying conductors in that single pipe. The worst case scenario for current heating would be if the load current were concentrated in a minimum number of circuits, so 8 120V 20A circuits being used to capacity.

I believe that it would be 'reasonable' (but not code) to count the above as 16 current carrying conductors for 310.15(B)(2).

There are clearly holes in the above; with a 200A service and calculated 77A load, what happens when the load increases (as it most certainly will)?

-Jon

When in doubt of the mind of the CMP and NFPA, write a proposal.

I like George's idea. ... Where multiple switched legs of a single branch circuit occupy the same raceway they may be allowed to be considered a single conductor for the purposes of derating ... or words to that effect.

How about derating based on the quantity of circuits?

How about an exception for single family dwellings, like the state of Wisconsin
has in their state code.