NEC Changes For #14 Ampacity

[FionaZuppa]

You said the foam board is R10. If I looked up and converted from metric properly, the R value of copper per inch is 1/2700. So the R-value of 24 in of copper is about 1/100, i.e. the copper is on the order of 1000 to 10,000 times better thermal conductor than the foam.

i get R(copper) = 0.00036
copper is 4.01W/cm*K from 25-225C, = 2782.18 BTU inch/ht *ft^2*F

1/2782 = 0.00036

so yep, R total from center is 0.00864. but remember, we are not producing heat just at the center of the wire, exothermic heat density is uniform along the whole wire given that ohms/inch is exactly the same everywhere. we are generating heat even in the wire that is exposed to ambient, so although copper is a good conductor of heat this is not a simple heatsink model. you would also need to do the analysis of heat transfer from the internal copper wires up to the sheath and then to air. for these several reasons i do not believe the exposed ends have any significance to the temp measurements being taken at the center of the wire.


my ambients were taken using IR gun (whole degree display), which was about on par with my CM660 TC sitting in ambient for an hr or so (IR gun was showing couple degrees colder than the CM660 TC probe). i can get a faster read with IR gun than the metal TC probe.

 

[wwhitney]

we are not producing heat just at the center of the wire, exothermic heat density is uniform along the whole wire given that ohms/inch is exactly the same everywhere. we are generating heat even in the wire that is exposed to ambient, so although copper is a good conductor of heat this is not a simple heatsink model. you would also need to do the analysis of heat transfer from the internal copper wires up to the sheath and then to air.

I agree.

for these several reasons i do not believe the exposed ends have any significance to the temp measurements being taken at the center of the wire.

I don't see how you get from the above to this statement. I expect the careful analysis you suggest will show that the wire outside the foam will significantly heat sink the wire in the foam. The copper outside the foam will be much better at convecting away the heat, and since copper is a good thermal conductor, it may be competitive for the heat generated in the center of the sample to conduct outside the foam and then convect away versus conduct through the foam. (Obviously both happen, it is a question of comparative magnitude.)

Cheers, Wayne

 

[FionaZuppa]

I don't see how you get from the above to this statement. I expect the careful analysis you suggest will show that the wire outside the foam will significantly heat sink the wire in the foam. The copper outside the foam will be much better at convecting away the heat, and since copper is a good thermal conductor, it may be competitive for the heat generated in the center of the sample to conduct outside the foam and then convect away versus conduct through the foam. (Obviously both happen, it is a question of comparative magnitude.)

Cheers, Wayne

if heat flows in a direction and every point along that trail (axial path) is exothermic in same magnitude, what does the heat density look like as you walk from center out to the end? There is no exposed copper outside of the foam (very very small sections for soldering eyelets, thats it).

i can wrap the exposed wire if you think it will affect temp measurements in a significant way. i dont think it will because the wire is exothermic at every point in the wire, the net R value from the copper up through the sheath is significantly larger than R(copper).

 

[wwhitney]

So in the test pictured in post 421, the large diameter MOT leads and the vice grips are undoubtedly acting as heat sinks for the short length of bare #14 copper. I think you could check for that as follows:

1) Hook up the MOT leads to each other with both vice grips and no #14 copper, assuming your setup works without the resistance of the #14 copper. Then measure the temperature at the MOT ring terminals for a given current.

2) Repeat the test with the #14 copper as originally pictured, and check the temperatures at both the #14 copper and the MOT ring terminals, for the given current. If the MOT terminals are hotter than in #1, some of the heat generated by the #14 copper is being dissipated through the rest of the system.

3) If you repeat test one and increase the current so that the MOT ring terminals get to the same temperature as in #2, then I would think that info should allow you to roughly estimate how much of the power dissipation of the #14 copper in test 2 is instead being dissipated as heat elsewhere in the system. I haven't quite worked out the details, though.

The same concern applies to the foam test. For the far end of the NM test sample shown in the second picture in post 490, it would definitely help to thermally insulate that connection, ideally more heavily than R-10.

I hope my comments are helpful, you are taking such great care with all the other aspects of the test that I think the possible issue of heat sinking deserves comparable attention.

Cheers, Wayne

 

[FionaZuppa]

i am digging further on this odd heatsink model. i will gather some info and figure out some math.

and to note, i understand your suggestion of testing but i dont think my test gear is sensitive enough to measure it, which leads me to make the statement that its not a significant factor. the vice grips have a very small contact pad on that connection. surely yes, conduction mode is there, but my hypothesis is that it is not significant.

why dont i just test a "very long" piece of bare copper so that the convection mode in free air is significantly more dominant than conduction to the ends? 20ft and i measure temp at the 10ft mark (center)? if the temp is the same then conduction mode is insignificant?

 

[mbrooke]

Pure gold thus far

Anyway, as promised. What makes this table so interesting being the conductors are based on 75*C exactly as the NEC, however the ampacity varies based on how the cable is installed. Bear in mind, this is only two (not three) current carrying conductors in a raceway. There is also a table for 3 current carrying conductors and two others which are the same but use 90*C. Very eye opening.

 

[mbrooke]

The first two are from 2 current carrying conductors at 90*C, second two are 3 current carrying conductors at 75*C and the last two are 3 current carrying conductors at 90*C. Note that these arent the full tables but only part of them. In the last pic I am re-including the AWG to mm2 chart to facilitate comparison.

It is important to note, that as a whole these tables are far less conservative then the NEC at both 75 and 90*C. In fact nearly all current ratings in 310.15 B 16 fall between "partially surrounded by thermal insulation" and "completely surrounded by thermal insulation" with number leaning toward completely surrounded. This re-enforces my theory the NEC tables are based on worse case.

 

[FionaZuppa]

The first two are from 2 current carrying conductors at 90*C, second two are 3 current carrying conductors at 75*C and the last two are 3 current carrying conductors at 90*C. Note that these arent the full tables but only part of them. In the last pic I am re-including the AWG to mm2 chart to facilitate comparison.

It is important to note, that as a whole these tables are far less conservative then the NEC at both 75 and 90*C. In fact nearly all current ratings in 310.15 B 16 fall between "partially surrounded by thermal insulation" and "completely surrounded by thermal insulation" with number leaning toward completely surrounded. This re-enforces my theory the NEC tables are based on worse case.

so a few questions arise from this
1) on what basis (info) were these tables made?
2) on what basis (info) has the NEC rooted its #'s into a worse case scenario? what scenario?
tables, no matter who made them, are not "good" unless the derivation process can be well described, which is kinda what i was asking in post #1.

is a bundle of 35 NM's passing through a short chase nipple from bay to panel a worse case scenario? is NM running in a attic bay in so. TX that is covered by 14" of blown cellulose a worse case scenario? did the NEC derive a worse case scenario by studying the IRbC. i all ears.

just as a reminder to readers, we are testing #14 NM-B wire several ways to obtain a reference related to the NEC rules of NM being restricted to 60C column, and, why #14 is restricted to 15A ocpd. up till now these two restrictions dont make sense to some of us.

 

[wwhitney]

why dont i just test a "very long" piece of bare copper so that the convection mode in free air is significantly more dominant than conduction to the ends?

That will work, for the proper value of "very long".

20ft and i measure temp at the 10ft mark (center)? if the temp is the same then conduction mode is insignificant?

I'm not sure how to figure out what is long enough. One would think that for bare copper 20 ft should be long enough. You could compare results at 20 ft to results at 9 inches to see how big the difference is. For the foam test, "long enough" will be longer than bare copper, as the reduced heat flow through the foam will make the (undesired) heat sink path more competitive.

Cheers, Wayne

 

[ADub]

just as a reminder to readers, we are testing #14 NM-B wire several ways to obtain a reference related to the NEC rules of NM being restricted to 60C column, and, why #14 is restricted to 15A ocpd. up till now these two restrictions dont make sense to some of us.

Good for you. What do you plan on doing with all this valuable information once you're done accumulating it? What's the end game?


Sent from my iPhone using Tapatalk

 

[wwhitney]

OK, consider this calculation:

Suppose we have 1 foot of #14 copper carrying 20 amps and it is thermally insulated to R-10 to a constant temperature heat reservoir and there is no heat transfer longitudinally. A random web site tells me #14 copper has a resistance of 2.525 milliohms/foot (at what temperature?) So the wire has to dissipate 50.5 milliwatts through the R-10 insulation. Another web site tells me 1 watt is 3.41 BTUs/hr, meaning 50.5 milliwatts is 0.172 BTUs/hr.

#14 copper has a diameter of 64.0 mils, so the one foot section has a cylindrical surface area of 0.0167 ft^2. Dissipating 0.171 BTUs/hr through that area gives a heat flux of 10.3 BTUs/hr-ft^2. I expect there is a correction factor for cylindrical heat flow, but otherwise to dissipate 10.3 BTUs/hr-ft^2 through R-10 insulation (hr-ft^2-degF/BTU) would require a temperature differential of 103 degree F.

So why are the measured results so much lower? Did I make a computational error? If not, there is the correction factor for cylindrical heat flow, and there is longitudinal heat flow (heat sinking).

Cheers, Wayne

 

[FionaZuppa]

That will work, for the proper value of "very long".


I'm not sure how to figure out what is long enough. One would think that for bare copper 20 ft should be long enough. You could compare results at 20 ft to results at 9 inches to see how big the difference is. For the foam test, "long enough" will be longer than bare copper, as the reduced heat flow through the foam will make the (undesired) heat sink path more competitive.

Cheers, Wayne

hmm, well, i kinda dont agree. whether there is foam or not the heatsink factor at the ends remains constant, the only diff i see will be the heat flux found in the two modes of transfer due to diffs in wire temps.

let me dig some, i can run a single wire test using a much longer wire to see if this temp is much higher than the shorter wire.

So why are the measured results so much lower? Did I make a computational error? If not, there is the correction factor for cylindrical heat flow, and there is longitudinal heat flow (heat sinking).

Cheers, Wayne

well, i didnt run the foam test yet
and, the foam board is a 2x2, 4sq.ft. (x2, so 8sq.ft), not 1.
i dont think its as simple as your calc is, the wire itself is magnitudes smaller than the size of the foam, the exothermic heat is being generated in a line, not uniform heat like you get with sunlight hitting the side of a wall, thus the heat has infinite # of radial paths to escape on and each path has a different R value (its a 2x2' x 4" thick sandwich).

 

[FionaZuppa]

well, i didnt run the foam test yet
and, the foam board is a 2x2, 4sq.ft. (x2, so 8sq.ft), not 1.
i dont think its as simple as your calc is, the wire itself is magnitudes smaller than the size of the foam, the exothermic heat is being generated in a line, not uniform heat like you get with sunlight hitting the side of a wall, thus the heat has infinite # of radial paths to escape on and each path has a different R value (its a 2x2' x 4" thick sandwich).

note, i did not account for the sides of the foam board, an extra 2.66 sq.ft., but as i noted, the R values vary, and i am not up to doing integration's right now :blink:

so as of now, the tests are/were:

all #14 btw

1. a piece of bare copper wire from NM tested in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
2. a piece of insulated copper wire from NM tested in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
3. a piece of full NM in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
4. a piece of full NM in rigid foam sandwich - TBD, waiting on Omega connectors
5. a piece of NM in a 2x R13 glass sandwich - TBD, need to get glass
6. a piece of NM in a bay on wood w/ glass and bay finished like a typical ext wall of a home - TBD, need to build it
7. a piece of NM on a piece of wood (or maybe this is #6 but w/o the glass?) - TBD, need to build it
8. wago 222 tests - TBD, waiting on 222's to arrive

 

[wwhitney]

hmm, well, i kinda dont agree. whether there is foam or not the heatsink factor at the ends remains constant

I think the heat sink rate remains the same, but if the rate of heat dissipation elsewhere is reduced (by the foam), then the proportion of the heat dissipation that goes through the heat sink increases.

i dont think its as simple as your calc is,

OK, I did a little reading on cylindrical heat transfer. It turns out the cylindrical correction factor is quite large for a very small cylinder. I'll post more on this when I have a chance, hopefully later today.

Cheers, Wayne

 

[FionaZuppa]

I think the heat sink rate remains the same, but if the rate of heat dissipation elsewhere is reduced (by the foam), then the proportion of the heat dissipation that goes through the heat sink increases.


Cheers, Wayne

i agree with your statement, which summarizes equilibrium, the question is, is the heatsink mode a significant factor for measuring temp at the center?

and so i am clear, heat sink rate and heat sink factor are different things. we expect the rate to increase if wire temps in center are higher with insulation vs no insulation simply due to fact that the rate is based on temp diffs, you describe it as "proportion". for me, the heatsink factor simply means there is a heatsink of some sort connected at the ends, and its a physical constant, etc.

the contact pad of vicegrip clamping eyelets together is on magnitude of 2sq.mm per wire (1sq.mm per side). and, the exposed wire is not bare copper, so you'll need to adjust R values and areas accordingly.

 

[FionaZuppa]

i have some prelim info regarding the heatsink mode, and what i am getting is, heatsink mode is not significant beyond a couple of insulating diameters. as of now this would put it at about ~8" in from the ends. center of wire is 12" in from the end. however, i need to qualify all this as "dia" doesnt fit well with a square sandwich. my bare copper TC in the foam sandwich sits right at the edge of this #.

 

[mbrooke]

so a few questions arise from this
1) on what basis (info) were these tables made?

They are the current carrying capacity tables in the Australian/New Zealand electrical code. How they were derived I have no idea but the code mentions calculations based on IEC 60287

(g) Current-carrying capacities for cables installed in wiring enclosures have been
recalculated according to IEC 60287.

(h) The values for all current-carrying capacities have been expressed to the nearest
ampere to align with current IEC practice.

2) on what basis (info) has the NEC rooted its #'s into a worse case scenario? what scenario?
tables, no matter who made them, are not "good" unless the derivation process can be well described, which is kinda what i was asking in post #1.

To be honest I am just a clueless as you, if not more. But using rudimentary equations, computer models and comparing foreign codes I can only reach the conclusion the NEC is based around raceways within thermal insulation material like dense fiberglass or spray foam.

is a bundle of 35 NM's passing through a short chase nipple from bay to panel a worse case scenario? is NM running in a attic bay in so. TX that is covered by 14" of blown cellulose a worse case scenario? did the NEC derive a worse case scenario by studying the IRbC. i all ears.

In my opinion a worse case scenario is NM or conduit in a wall full of insulation for dozens of feet. An attic covered in insulation would be another.

just as a reminder to readers, we are testing #14 NM-B wire several ways to obtain a reference related to the NEC rules of NM being restricted to 60C column, and, why #14 is restricted to 15A ocpd. up till now these two restrictions dont make sense to some of us.

I cant wait to see how real world testing compares

These restrictions dont make sense to me outside of worse case (and even then, the CEC lets you use NM on the 90*C column in theory), certainly not in free air or where wire is not touching thermal insulation.

 

[romex jockey]

Gentlemen, i applaud your tenacity here

I would wager testing labs less discriminate.....

~RJ~

 

[FionaZuppa]

all #14 btw

1. a piece of bare copper wire from NM tested in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
2. a piece of insulated copper wire (TC touching bare copper) from NM tested in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
3. a piece of full NM in free air - done, but some are questioning if the clamped ends have significant heatsink affect on temp
4. a piece of full NM in rigid foam sandwich - TBD, waiting on Omega connectors
5. a piece of NM in a 2x R13 glass sandwich - TBD, need to get glass
6. a piece of NM in a bay on wood w/ glass and bay finished like a typical ext wall of a home - TBD, need to build it
7. a piece of NM on a piece of wood (or maybe this is #6 but w/o the glass?) - TBD, need to build it
8. wago 222 tests - TBD, waiting on 222's to arrive

i am re-doing test #2 @26A, but instead of a 9" piece of copper its 84" long and measuring temp at the 42" mark. as before, just a small section of the insulation was removed, about 1mm worth so that the TC can contact the copper directly. ambient now is ~55F. i can safely say that due to metal TC probe the measured temp will be a few degrees lower than actual.

 

[FionaZuppa]

i am re-doing test #2 @26A, but instead of a 9" piece of copper its 84" long and measuring temp at the 42" mark. as before, just a small section of the insulation was removed, about 1mm worth so that the TC can contact the copper directly. ambient now is ~55F. i can safely say that due to metal TC probe the measured temp will be a few degrees lower than actual.

open air test.
well, meter says 83.2F, was sitting there for past couple of hours. if i normalize it back to 90F the #'s are:

114.7F (9" of same wire @26A)
118.2F (84" of same wire @26A)

there's a 3.5F (3.1%) diff after normalization, i dont see this diff as significant, which leads me to believe the heatsink affect on the clamped ends are not significant.

but now lets extrapolate out the % diff per inch = 3.1/75 = 0.0413%/inch, for all intensive purpose this in %/in between heatsink affecting the temp vs infinite long wire (no heatsink affect).

the wire in foam experiment is 48", my free air test of NM was 24", do the diff is 24"
24x0.0413 = 0.9912%

so thats my crude adjustment factor, 1%, regardless of any other error. if i measure 100F we will tack on 1F,........... not significant.