NEC Changes For #14 Ampacity

[FionaZuppa]

2hr @~20A yields equilibrium at 19.82A T_NM=183.2F, T_Cu=183.6F, 66F ambient
right-on wayne, post #591 :thumbsup:
this has obviously exceeded 75C(167F), but NM is rated 90C

now cooking at 25A

2hr @~25A yields equilibrium at 24.91A T_NM=250.0F, T_Cu=247.6F, 66F ambient
has now exceeded 90C rating

now cooking at 30A

 

[mbrooke]

2hr @~20A yields equilibrium at 19.82A T_NM=183.2F, T_Cu=183.6F, 66F ambient
right-on wayne, post #591 :thumbsup:
this has obviously exceeded 75C(167F), but NM is rated 90C

now cooking at 25A

2hr @~25A yields equilibrium at 24.91A T_NM=250.0F, T_Cu=247.6F, 66F ambient
has now exceeded 90C rating

now cooking at 30A

I think its safe to say now we now (possibly) know why 240.4 (D) was put into the code. Under worse case the wire is just below 90*C at the edge of tripping the breaker which makes sense. Its interesting to note that BS7671 and other codes like it place the wire even lower, but that might be because T&E may not be rated 90*C. The fiberglass test will probably tell us just as much.

On a side note, I have a theory that #12 might actually reach or even exceed 90*C at 25 amps.

 

[FionaZuppa]

2hr @~25A yields equilibrium at 24.91A T_NM=250.0F, T_Cu=247.6F, 66F ambient
has now exceeded 90C rating

now cooking at 30A

2hr @~30A yields equilibrium at 30.29A T_NM=276.4F, T_Cu=266.1F, 62F ambient

hmmm, as the temps went up the sheath seems to be hotter than the bare copper. not sure why yet. will do final temp calibration of TC's at the end.

gonna let it cook @25A for a day or so.


and yeah, NM looks like it needs some restrictions. what those are is up for debate.

 

[mbrooke]

and yeah, NM looks like it needs some restrictions. what those are is up for debate.

In this case by all means

 

[FionaZuppa]

I think its safe to say now we now (possibly) know why 240.4 (D) was put into the code. Under worse case the wire is just below 90*C at the edge of tripping the breaker which makes sense. Its interesting to note that BS7671 and other codes like it place the wire even lower, but that might be because T&E may not be rated 90*C. The fiberglass test will probably tell us just as much.

On a side note, I have a theory that #12 might actually reach or even exceed 90*C at 25 amps.

so, now we get into the chit-chat on the reasoning. this seems to be a battle of ocpd performance. most will trip well before 125%. from a statistical view, very few will be near that +25% zone, and from the risk perspective it doesnt matter if the ocpd is a 5A or 20A, if the item is faulty it can allow any amps.

and now i start to wonder, does the reasoning around #12 & #14 (and "NM") follow suit into say #8 @50A? i have issue when things become inconsistent.

 

[FionaZuppa]

so to normalize back to 30C ambient i will bring amps closer to 30 so that wire temps are normalized. will let the wire cook there for some days to see how the damage (if any) occurs.

 

[FionaZuppa]

In this case by all means

NM restricted to 60C column under these conditions

• installed in exterior walls
• installed in any space where NM and thermal insulation co-exist
• installed in ceilings or attic spaces where NM is covered/surrounded/touching by insulation for more than 12"
• installed in walls, ceilings, attics such that the NM is enclosed on both sides to form a "closed bay" AND the framing members that form that bay are less than 1.5" in minor dimension
• passes through any wood member where the path of NM is in wood for more than 3"

reasons why?
1) most exterior walls are insulated, if not now then likely later will be
2) takes care of those interior walls that call for insulation for sound or other reasons, or might get insulation later (this one has potential issue because a contractor called in to add insulation will likely not know about the NEC rules, etc)
3) the 12" rule, allow the NM to pass across/through insulation section in order to reach destination. this has to be a cumulative rule so that installers cant bob the NM up/down through insulation, like under for 10", then out into air for 1", back under for 10", etc etc.
4) this cavity rule so that NM cant be enclosed in very small space that has no insulation, providing enough air in the bay to keep temps down. maybe turn this into a fill % # which would also handle the case of more than one NM in same bay/cavity.
5) for things like top plates were it might be more than two 2x4's, sistered members to form a post, 4x4, etc


i guess if we didnt have CT in the breaker panel we could have access to the power with lower temps on the wires. hmmmmm, who has done this before?

 

[mbrooke]

so, now we get into the chit-chat on the reasoning. this seems to be a battle of ocpd performance. most will trip well before 125%. from a statistical view, very few will be near that +25% zone, and from the risk perspective it doesnt matter if the ocpd is a 5A or 20A, if the item is faulty it can allow any amps.

I disagree, breakers can easily trip at 125% under various conditions like manufacturing tolerance and temps. The time current curves in most circuit breakers are intentionally engineered to start above 100%. 24 amps of #12 is possible without the breaker ever tripping. Where the rarity comes in is tripping at 134%, nearly all breakers trip well before this value.

and now i start to wonder, does the reasoning around #12 & #14 (and "NM") follow suit into say #8 @50A? i have issue when things become inconsistent.

It does, its interesting too. In the NEC #8 is still restricted to 60*C, but (and yes I know you don't care about the CEC), #8 is allowed to be used at the 75*C column. Most ranges are fed via 8/3 NMD-90 on a 50amp breaker where we go with 8/3 40 amp breaker or 6/3 50 amp breaker.

 

[mbrooke]

NM restricted to 60C column under these conditions

• installed in exterior walls
• installed in any space where NM and thermal insulation co-exist
• installed in ceilings or attic spaces where NM is covered/surrounded/touching by insulation for more than 12"
• installed in walls, ceilings, attics such that the NM is enclosed on both sides to form a "closed bay" AND the framing members that form that bay are less than 1.5" in minor dimension
• passes through any wood member where the path of NM is in wood for more than 3"

reasons why?
1) most exterior walls are insulated, if not now then likely later will be
2) takes care of those interior walls that call for insulation for sound or other reasons, or might get insulation later (this one has potential issue because a contractor called in to add insulation will likely not know about the NEC rules, etc)
3) the 12" rule, allow the NM to pass across/through insulation section in order to reach destination. this has to be a cumulative rule so that installers cant bob the NM up/down through insulation, like under for 10", then out into air for 1", back under for 10", etc etc.
4) this cavity rule so that NM cant be enclosed in very small space that has no insulation, providing enough air in the bay to keep temps down. maybe turn this into a fill % # which would also handle the case of more than one NM in same bay/cavity.
5) for things like top plates were it might be more than two 2x4's, sistered members to form a post, 4x4, etc


i guess if we didnt have CT in the breaker panel we could have access to the power with lower temps on the wires. hmmmmm, who has done this before?

I agree, in this case NM has a good foundation to be restricted 60*C. However, I think more testing is still needed. We still do not know the temp rise of fiberglass insulation. Your test has proved spray foam insulation could be an issue, but fiberglass could yield different results.

Im going to be bluntly honest... I think the NEC does not take spray foam insulation into account. My theory is these tables are based on fiberglass. My hunch is #12 will reach or exceed 90*C under the same test conditions, and even then table 310.15 B 16 says 3 current carrying conductors rather then two, so if three existed here we might reach 90*C or even exceed it.

 

[mbrooke]

And oh, more food for thought.

I think I found a loop hold in the NEC. It is possible to take an electric air handler and place it in the attic, run 75*C sized THHN conductors (in conduit) from the panel into a splice box in the basement, use 90*C connectors splicing into 90*C rated wire in conduit, up through 3 stories of dense spray foam insulation leading to another splice box in the attic where 90*C sized conductor goes to ones sized at 75 or 60*C leaving into a disconnect and then electric heat unit.

The question here is, am I protected? Why am I not restricted to 60*C? Is this a safe scenario despite being code legal? Could the 125% rule play a role here? For that matter why isnt conduit restricted to 60*C but NM automatically is?


BTW, here is what I am referring to in terms of splicing to smaller conductors in order to use the 90*C column:


http://static.schneider-electric.us/docs/Power Management/0110DB9901.pdf

7:19

https://www.youtube.com/watch?v=WL81zOzNagM


Also somewhere in here but I have to find it:

https://www.youtube.com/watch?v=OLIHEiiY_Rs

 

[FionaZuppa]

I disagree, breakers can easily trip at 125% under various conditions like manufacturing tolerance and temps. The time current curves in most circuit breakers are intentionally engineered to start above 100%. 24 amps of #12 is possible without the breaker ever tripping. Where the rarity comes in is tripping at 134%, nearly all breakers trip well before this value.

It does, its interesting too. In the NEC #8 is still restricted to 60*C, but (and yes I know you don't care about the CEC), #8 is allowed to be used at the 75*C column. Most ranges are fed via 8/3 NMD-90 on a 50amp breaker where we go with 8/3 40 amp breaker or 6/3 50 amp breaker.

where is #8 restricted , i only see 14/12/10 that has "**" on them in the ampacity table. if its NM #8 then its restricted by NM and not by #8. am i reading that wrong?

i also found what might be a discrepancy in my ambient temps. the metal probe that came with my CM660 shows ~+12F over my IR heat gun. oddly both show 32F for a bucket of ice water. after i cook this #14 NM in the sandwich for a day or so i will use one of the TC's from the sandwich with my CM660 to see how that compares to both the CM660 oem probe and my IR heat gun.

as for the 125% mark, the breakers might get there, but under their trip graph. if i recall correctly, exponentially faster as you enter that 100-125% zone. am i correct?

 

[mbrooke]

where is #8 restricted , i only see 14/12/10 that has "**" on them in the ampacity table. if its NM #8 then its restricted by NM and not by #8. am i reading that wrong?

I meant number 8 NM romex, my apologies :ashamed1:

Anyway, just for future readers, here where NM is restricted to 60*C:

334.80 Ampacity.

The ampacity of Types NM, NMC, and
NMS cable shall be determined in accordance with 310.15.
The allowable ampacity shall not exceed that of a 60°C (140°F) rated conductor.
The 90°C (194°F) rating shall be permitted to be used for
ampacity adjustment and correction
calculations, provided the final derated ampacity does not exceed
that of a 60°C (140°F) rated conductor. The ampacity of
Types NM, NMC, and NMS cable installed in cable tray shall
be determined in accordance with 392.80(A).
Where more than two NM cables containing two or
more current-carrying conductors are installed, without
maintaining spacing between the cables, through the same
opening in wood framing that is to be sealed with thermal
insulation, caulk, or sealing foam, the allowable ampacity
of each conductor shall be adjusted in accordance with
Table 310.15(B)(3)(a) and the provisions of 310.15(A)(2),
Exception, shall not apply.
Where more than two NM cables containing two or
more current-carrying conductors are installed in contact
with thermal insulation without maintaining spacing between
cables, the allowable ampacity of each conductor
shall be adjusted in accordance with Table 310.15(B)(3)(a).

i also found what might be a discrepancy in my ambient temps. the metal probe that came with my CM660 shows ~+12F over my IR heat gun. oddly both show 32F for a bucket of ice water. after i cook this #14 NM in the sandwich for a day or so i will use one of the TC's from the sandwich with my CM660 to see how that compares to both the CM660 oem probe and my IR heat gun.

Sounds good The more data the better.

as for the 125% mark, the breakers might get there, but under their trip graph. if i recall correctly, exponentially faster as you enter that 100-125% zone. am i correct?

Yes and no, a breaker could hold 120% indefinitely (for days) without any tripping taking place. After a certain point the breaker will trip and that time will become faster as the current rises, but there is also a "blind spot" where the current is over 100% but not high enough to push the breaker into tripping after any length of time. This normally does not matter, but with receptacles such overloads do happen.


An example is one of my old homes. With the bathroom and hall lights going, TV, curling iron and cord & plug electric heater in the master bedroom going the circuit would frequently see 16 to 17 amps (as measured) and never trip. The same could happen in a home where 2 circuits share the same bedroom. Both rooms end up with a 1,500 watt space heater. At 3000 watts (25amps) the breaker trips. The occupants say, lets turn the heater down in the small room, anyways the room does not need that much heat. So 1,500 watts + 750 = 2,250 giving 18.75amps. At 18.75 amps the breaker may or may not trip, less likely if the panel is in a cold basement. The circuit is technically overloaded, but since the breaker is not tripping the occupants assume no risk. This scenario could go for decades on end. Therefore, it would be wise for code making members to take this into consideration. If a fire did break out, everyone would be asking "why didnt the breaker trip before hand?"

 

[FionaZuppa]

Sounds good The more data the better.

Yes and no, a breaker could hold 120% indefinitely (for days) without any tripping taking place. After a certain point the breaker will trip and that time will become faster as the current rises, but there is also a "blind spot" where the current is over 100% but not high enough to push the breaker into tripping.

if a breaker holds 1.25x rated for days, its likely faulty. Eaton molded --> UL 489

 

[FionaZuppa]

if a breaker holds 1.25x rated for days, its likely faulty. Eaton molded --> UL 489

well, perhaps i spoke too soon on that, if ambient is way low certainly the delay is longer, days longer?? not sure.

 

[FionaZuppa]

so last night i left 25A, this AM i moved it back to 30A to try and keep temps up, i started to smell something funny this afternoon, the rigid foam started to melt and the NM is starting to sink into the foam board. so we have obviously exceeded max "temp" of the foam board. the NM sank in about 3/16" thus far, and temp now =247F is an indicator that NM has pulled away from the TC's and that the air gap is leaking heat.

i quoted temp above because as i have noted before, temp alone doesnt mean so much, its transfer of energy that does. there's enough watts there to damage the foam board. pics to follow.

 

[FionaZuppa]

so, i have no idea at what temp this melting foam issue started. safe to say foam board is not worse case scenario, the melting is a good thing as it creates a air pocket thus reducing R value, and, wire is no longer in contact with the insulation. is this less or more a fire hazard than NM in glass or up against wood? dunno.

also safe to say that the measured temps may or may not be accurate. the air pocket that was created certainly was still hot and the ends were closed for quite some time, but there was a air gap between wire and TC's, so i think i can say the TC's did not stay where they were supposed to stay, likely affecting measured temps. likely the reason why early on i saw the two TC's at equal temps, but later as amps went up the temps started to separate from one another. due to ambient being ~65F i had set the amps to 30 to bring wire temps back to ambient normalization of 30C (so basically 25A @ 30C ambient).

the glass sandwich is probably the worse case scenario, glass wont melt at these temps. i also question if this blue DOW foam board is even acceptable having a NM wire on it or through it.

here's some pics:

 

[FionaZuppa]

from DOW site
"Polystyrene products will melt when brought into direct contact with high temperature heat sources. For STYROFOAM™ Brand Insulation from Dow, the recommended maximum continuous operation temperature is 75°C. Do not expose the boards to flame or other ignition sources."

75C=167F, got past that temp when soaking in @20A, so my guess is temps were accurate up until then. supporting evidence is that after the soak @20A the temps of the two TC's were no longer the same. this melting can be a issue even when ocpd is 15A because ocpd can go beyond that, AND, i havent even normalized my measured temps back to 30C ambient !!

forgot to mention in last post (this 15min edit rule is for the birds btw), the THHN insulation did not appear to be cracked anywhere, and its "softness" in the heat stressed zone seemed to be about the same as the THHN at the very ends of the wire.

does Europe run into these heat problems running 240v everywhere? same power needs = 1/2 the amps = 1/4 the heat !!

 

[FionaZuppa]

and just for documentation purposes.

the correction factor to TC_measured is

TC_adjsuted= +1% TC_measured (^oK)

and i verified ambient. my IR gun is -4^oF compared to the TC's used for taking the wire temps. both TC's confirm same #.

 

[mbrooke]

if a breaker holds 1.25x rated for days, its likely faulty. Eaton molded --> UL 489

Which doc though?

 

[mbrooke]

from DOW site
"Polystyrene products will melt when brought into direct contact with high temperature heat sources. For STYROFOAM™ Brand Insulation from Dow, the recommended maximum continuous operation temperature is 75°C. Do not expose the boards to flame or other ignition sources."

That would make sense now, the melting of the foam caused the thermal couples to shift.

75C=167F, got past that temp when soaking in @20A, so my guess is temps were accurate up until then. supporting evidence is that after the soak @20A the temps of the two TC's were no longer the same. this melting can be a issue even when ocpd is 15A because ocpd can go beyond that, AND, i havent even normalized my measured temps back to 30C ambient !!

Thinking about it in depth I do not believe 310.15 takes worse case scenario into account, rather what used to be worse case scenario (R-17 fiber glass bats with NM stapled to wood stud). This proves 310.15 normalized temps exceed those shown but also two other pieces of info further my belief: 1. Ive seen reports claiming that NM will over heat in dense spray foam insulation. 2. See my reply below in the answer to Europe.

forgot to mention in last post (this 15min edit rule is for the birds btw), the THHN insulation did not appear to be cracked anywhere, and its "softness" in the heat stressed zone seemed to be about the same as the THHN at the very ends of the wire.

I think THHN can handle higher temps then listed just so the product is guaranteed under all manufacturing conditions, but I would never drive THHN at such temps for a variety of reasons.

does Europe run into these heat problems running 240v everywhere? same power needs = 1/2 the amps = 1/4 the heat !!

Believe it or not European and Australian codes do take this into account. The 240 volt operating voltage has nothing to do with it, in fact circuits can and are loaded to code specified tables like there are in the US, the only difference being more power per circuit. Rather the codes take this into account via there ampacity tables. Unlike the NEC European codes list the ampacities of each conductor based on the insulation (or lack there of) around the conductor. The values vary dramatically, being far less to far more conservative when compared to 310.15.

As an example, BS7671 which applies to the United Kingdom: 2.5mm2 (slightly larger then 2.08mm2 14 gauge) Twin and Earth (Euro equivalent to romex) is rated 27 amps when stapled to a wood stud having no insulation, 20amps when one side of the cable is resting against insulation and 13.5 amps with the cable completely covered in insulation. The conductor operating temp is assumed to be 70*C. If we were to compare these values to 14 gauge wire, we would roughly get 25 amps, 18.5 amps and 10.5 amps. Thus, the European codes require NM wire to be kept at current ratings lower then our NM wire when completely covered in insinuation and higher rating when not.


These, which are based on BS7671:

http://www.lightwiring.co.uk/lighti...-cables/twin-and-earth-cable-current-ratings/

http://www.batt.co.uk/upload/files/currentratingstable4d5_1312876143.pdf


The Australian/ New Zealand codes are even better since the same concept is applied to 75*C conductors. Like BS7671, where the wire is completely enclosed in thermal insulation the values are much more conservative.