NEC Changes For #14 Ampacity

Status
Not open for further replies.
mbrooke said:
The OCV may be to low, but it might work out just fine. The rest sounds ok, especially the thermopaste.
Click to expand...

well, ocv doesnt really matter, its a short circuit, we only care about amps. i ran into a slight snafu, with the variac at lowest setting just turning the power on gets me 8.1amps, very little control from the dial, etc. i might put another step down between the variac and mot. not a big issue.

so, i just ran a few tests for open air:
9" length of bare #14 copper (from the Romex i will test with), 0.064"dia
ambient air 70F (big room, no air movement)

i did two tests and let it soak into equilibrium, i'll list the amps here, you guys can guess the temps, since there were some bets going. my CM660 has a TC feature (calibrated with 32F ice water), so i used that to take temp readings by letting the tip of the probe rest on the copper with just a tad of weight to hold them together.

equilib point #1 13.90A temp ?
equilib point #2 26.85A temp ?

post your temp guesses, i be back in just a few.


open_air.jpg
 
looks like i have sparked some interest on this topic. 13k+ views !
good_topic1.jpg
 
FionaZuppa said:
looks like i have sparked some interest on this topic. 13k+ views !
good_topic1.jpg
Click to expand...



You have! :D To be honest (though as always my statements are humble) I dont think anyone on Mike Holt has ever sat down to test or verify the ampacity tables in the NEC. I think this thread is destined to be a classic :thumbsup:
 
FionaZuppa said:
well, ocv doesnt really matter, its a short circuit, we only care about amps. i ran into a slight snafu, with the variac at lowest setting just turning the power on gets me 8.1amps, very little control from the dial, etc. i might put another step down between the variac and mot. not a big issue.
Click to expand...

It can matter, all wire has resistance. If the OCV is not high enough it cant push enough amps. That is actually how the variac works, the secondary OCV controls the amount of current through the wire's natural impedance. It may be well under 1 ohm, but still over zero so ohms law still holds weight so to speak.


Some wild guesses

equilib point #1 13.90A temp ? >>> 75*F?

equilib point #2 26.85A temp ? >>> 91.5*F?
Click to expand...
 
1v is plenty to push big amps on shorted copper wire that is only ~60" round-trip. on a 9" length of #14 1.9v is good for ~950A. i am working in the area x10^-3 ohms, etc

@13.90A the wire was 80F (26.66C)
@26.85A the wire was 94.1F (34.5C)

far below 60C column, way far below the rated insulation temp of 90C, and way above the NEC's restricted max amps of 15.

but, do these #'s make sense? about 2x on amps = 4x on joules, yet temp only increased 14.1F, does that make sense?

sandwich test will be interesting.
 
Last edited:
FionaZuppa said:
1v is plenty to push big amps on shorted copper wire that is only ~60" round-trip. on a 9" length of #14 1.9v is good for ~950A. i am working in the area x10^-3 ohms, etc
Click to expand...

You know this better then I do :)


@13.90A the wire was 80F (26.66C)
@26.85A the wire was 94.1F (34.5C)
Click to expand...


I was close in terms of temp differences between the two, but wrong in actual numbers.

far below 60C column, way far below the rated insulation temp of 90C, and way above the NEC's restricted max amps of 15.
Click to expand...

Id give more length because the leads can draw heat away from just a short sample, but I do agree with you.


but, do these #'s make sense? about 2x on amps = 4x on joules, yet temp only increased 14.1F, does that make sense?

sandwich test will be interesting.
Click to expand...

Makes sense. I think its safe to say the open air tables should be revised, and 240.4(D) eliminated for them.
 
my open air test is not accurate enough to warrant any changes just yet. i can run the math but i suspect the temps might be a few degrees higher for a couple of reasons, 1) as you mention, the ends are a good heat sink, and 2) the temp probe is metal and is likely acting as a heat sink passing heat to the open air.

that said though, even if you add a few degrees, still way low relative to NEC table in context of temp vs amps, etc.
 
FionaZuppa said:
my open air test is not accurate enough to warrant any changes just yet. i can run the math but i suspect the temps might be a few degrees higher for a couple of reasons, 1) as you mention, the ends are a good heat sink, and 2) the temp probe is metal and is likely acting as a heat sink passing heat to the open air.

that said though, even if you add a few degrees, still way low relative to NEC table in context of temp vs amps, etc.
Click to expand...

I agree entirely. Even 310.15 (B) 17 (free air) has 240.4(D) limitations which leads me to believe this serves no real world purpose for open air conductors.
 
so here's another open air test i will do today and i'll try to target the same amps.

i'll use insulated wire and i'll remove just a small section of insulation for the TC probe to contact the copper, i'll also insulate the TC probe except for the very tip. lets see what type of temps we get from that.
 
so, did the insulated wire open air tests. last one right at the top of the 60C column. as before, the probe just sat there until equilibrium was there for about ~10min or so.

note: in this pic the wire is about 1.25" above the table.
ambient air 60F
13.46A 70.7F (21.5C)
26.0A 84.7F (29.3C)
48.15A 137.3F (58.2C), basically feels warm to the touch.

that stated, ambient was different, but the relationship looks like 10F, so i might hypothesize normalization back to 70F ambient as:
13.46A 80.7F (27.05C)
26.0A 94.7F (34.44C)
48.15A 147.3F (64C), basically feels warm to the touch. 3.2x (2.57x more than 125% of rated ocpd) more than NEC's prescribed max amps with temp hitting ~60C. there is no way this temp has potential to start any fires, unless you bring me an example or scenario that you think might.

DSC02721.jpg
 
Last edited:
to give some reference (and recall, combustion needs O to happen):

When wood is completely dry and is not a type of artificial wood, the combustion temperature is generally 451 degrees Fahrenheit, according to the Argonne National Laboratory
Click to expand...

147F is ~3x less than 451 !! and yeah yeah, i know, these are open air tests, i am getting to that "worse case" scenario.

i thought NEC was using a 50% rule after running the N-M equation, i think i am wrong, the conservative factor seems to be much more than that.

i am leaning towards hypothesis that of the fires we have seen involving electrical wire, there was likely a failure in ocpd or there was some very odd circumstances that occurred. under normal use of properly installed wiring, and using the data presented thus far, i dont see where using 20A ocpd with #14 wire poses a hazard. feel free to argue this.
 
Last edited:
post #397, i am confused. that PDF suggests those amp #'s are the threshold points of the wire insulation in free air. my testing is well above that and i dont see, smell, feel, or measure any issues with the insulation.

the PDF table shows #14 with 90C insulation is only good for 30A, but hmmmm, i am well above that. maybe the wire insulation from this Romex wire is not 90C wire, maybe its Kapton (not likely)? my temps dont even come close to what that PDF table is suggesting.

something doesnt seem right, either my testing is providing erroneous data, or the data elsewhere is just off or that data is not being explained correctly.
 
Last edited:
Not that it makes an enormous difference, but are you testing with a 30C or 40C ambient? So far you are mainly measuring the temp differential, not the ultimate conductor and insulation temperature.
 
GoldDigger said:
Not that it makes an enormous difference, but are you testing with a 30C or 40C ambient? So far you are mainly measuring the temp differential, not the ultimate conductor and insulation temperature.
Click to expand...


the ambients were 15.55C (60F) and 21.1C (70F)

the TC was on the copper wire itself, actual conductor temp. you think having ambient at 30C and running it at 48A will make a big diff in conductor temp? for every 10F in ambient the conductor seems to go up 10F for same current.

so lets add 30F to my 15.55C test to bring it just above 30C (32C), that takes the 48.15A test to 167.3F (75C). the insulation is not melting off at 75C.
 
Last edited:
>there is no way this temp has potential to start any fires, unless you bring me an example or scenario that you think might.

60 Celsius, certainly not, but I've read about a couple of hazards well below ignition points.

Wood stove installers have to compensate for a phenomenon where sustained high heat changes the surrounding wood chemically to become easier to ignite. "High heat" means three figures.

I may never find the reference again, but there's a paper out there that claims hazardous changes to NM insulation can begin as low as 100 Celsius.
 
Last edited:
Curious nerd said:
>there is no way this temp has potential to start any fires, unless you bring me an example or scenario that you think might.

60 Celsius, certainly not, but I've read about a couple of hazards well below ignition points.

Wood stove installers have to compensate for a phenomenon where sustained high heat changes the surrounding wood chemically to become easier to ignite. "High heat" means three figures.

I may never find the reference again, but there's a paper out there that claims hazardous changes to NM insulation can begin as low as 100 Celsius.
Click to expand...
100C is above NM-B UL listing of 90C. if i am running NM-B @ ~20A and the temp is only 35C, what's the worry? @48A i am seeing 75C(extrapolated), way high on amps yet still below 100C.

but hey, your two examples, i would like for you to hold onto those until after i run the sandwich testing. then we can pull out scenarios that might impact the safety factor that is being applied.

my oven is on, lets bake the wire at 300F just to see what happens to the insulation (on wire, internal paper, outer sheath).
 
FionaZuppa said:
post #397, i am confused. that PDF suggests those amp #'s are the threshold points of the wire insulation in free air. my testing is well above that and i dont see, smell, feel, or measure any issues with the insulation.

the PDF table shows #14 with 90C insulation is only good for 30A, but hmmmm, i am well above that. maybe the wire insulation from this Romex wire is not 90C wire, maybe its Kapton (not likely)? my temps dont even come close to what that PDF table is suggesting.

something doesnt seem right, either my testing is providing erroneous data, or the data elsewhere is just off or that data is not being explained correctly.
Click to expand...

Your referencing This PDF via This wire manufacturer FZ

I think they're hiring :)

~RJ~
 
Status
Not open for further replies.
Top