NEC Changes For #14 Ampacity

[mbrooke]

And your point is? I2R losses are real and that's what you discount when you want to reduce ampacity to the bare minimum.

I could wire with #10 for 20amps and if my run is long enough voltage drop will still get me.

Good for them. We're not in Canada so it doesn't affect me or how I wire things.

So what?

But Canada has proven it wont burn down someones home, which I think the OP was wondering about.

 

[mbrooke]

From the 1918 NEC:

http://antiquesockets.com/NEC-LIB/NEC-1918.pdf

 

[FionaZuppa]

i am not concerned bout why CEC changed ampacity for #14 wire, i am want to see the evidence as to why NEC changed it between NEC2002 and NEC2008 and why NEC has the 15A OCD restriction.

i see some are digging into the 125% bit. but let me clarify, if the OCD's dont actually trip until 125% of their rating (which also makes no sense) and NEC restricts ampacity of romex to the 60C column, then putting a #12 onto a 20A OCD would allow the #12 to run above the 60C ampacity #, which by the NEC is a violation (not a wire issue, just a violation by NEC verbiage).

as for I^2R, #14 @ 100ft @ 20A is 1W/ft which is a fairly low density exothermic thing. within say romex cable the wire insulation is subject to 1W/ft directly and 1W/ft indirectly and that heat is in a fairly poor heatsink (romex sheeting), yet the cable is rated 90C. given that NEC allows >15A OCD on #14 wire via special exceptions, why is I^2R an issue?

i am basically just looking for the evidence that supports the NEC verbiage, etc. making the rules on the "just because" or "runs cooler" basis doesnt make sense to me.

NEC 240.4 Protection of Conductors

Conductors, other than flexible cords, flexible cables, and
fixture wires, shall be protected against overcurrent in
accordance with their ampacities specified in 310.15, unless
otherwise permitted or required in 240.4(A) through (G).

this 240.4 keeps throwing me off. is Romex a "flexible cable" ?


and post #22, oddly the amps per circular mil seems to go down as the wire size goes up. ~3.6mA/cmil for #14 down to ~1.6mA/cmil for #5 wire.

 

[jim dungar]

So you're both saying it's acceptable to load #14 wire to 20 amps or more?

I simply made a statement about the insulation not being degraded when it is operating below its listed temperature rating.

Given allowances for 'non-general purpose' circuits (e.g Table 240.4(G)) and ambient temperature adjustments, it is possible to find NEC compliant applications that permit #14 wire to be loaded to 20A or more.

 

[mbrooke]

i am not concerned bout why CEC changed ampacity for #14 wire, i am want to see the evidence as to why NEC changed it between NEC2002 and NEC2008 and why NEC has the 15A OCD restriction.

i see some are digging into the 125% bit. but let me clarify, if the OCD's dont actually trip until 125% of their rating (which also makes no sense) and NEC restricts ampacity of romex to the 60C column, then putting a #12 onto a 20A OCD would allow the #12 to run above the 60C ampacity #, which by the NEC is a violation (not a wire issue, just a violation by NEC verbiage).

as for I^2R, #14 @ 100ft @ 20A is 1W/ft which is a fairly low density exothermic thing. within say romex cable the wire insulation is subject to 1W/ft directly and 1W/ft indirectly and that heat is in a fairly poor heatsink (romex sheeting), yet the cable is rated 90C. given that NEC allows >15A OCD on #14 wire via special exceptions, why is I^2R an issue?

i am basically just looking for the evidence that supports the NEC verbiage, etc. making the rules on the "just because" or "runs cooler" basis doesnt make sense to me.


this 240.4 keeps throwing me off. is Romex a "flexible cable" ?


and post #22, oddly the amps per circular mil seems to go down as the wire size goes up. ~3.6mA/cmil for #14 down to ~

If I remember correctly it had something to do with an equation from the IEEE book that was used to derive it.

But you have to keep this mind, the NEC wire tables are very conservative assuming worse case scenario with a safety factor on top of that. Whether its 15 or 20amps at 60*C is more relative and arbitrary then anything else.

Trust me, I wonder the same behind the change.

 

[FionaZuppa]

sorry couldnt make edit in time, here's more

and post #22, oddly the amps per circular mil seems to go down as the wire size goes up. ~3.6mA/cmil for #14 down to ~1.6mA/cmil for #5 wire. a significnat decrease in allowed current, but kinda makes sense as heat is function of I^2. but oddly the table allows for about 1W/ft in #5 @55A, but in #14 (restrcited to 15A) its only 0.57W/ft (@100ft). allowing 20A on #14 would be almost identical heat density, thus why i am baffled as to why #14 is so restricted.

the ampacities (and thus allowed OCD sizes) should be chosen so that the allowed heat density is about the same across all wire sizes. 1.02W/ft @100ft seems like a good std. if the wire is subject to more than 1.02W/ft @100ft then next wire size should be used. this basically defines max OCD to be used, thus the table would look something like

#14 20A
#12

 

[mbrooke]

this 240.4 keeps throwing me off. is Romex a "flexible cable" ?

No, NM-B is classified as a type of building wiring method. Flexible cord would be like SJOW which is not for building wiring.

and post #22, oddly the amps per circular mil seems to go down as the wire size goes up. ~3.6mA/cmil for #14 down to ~1.6mA/cmil for #5 wire.

Larger wires dissipate less heat relative to smaller wire.

 

[mbrooke]

sorry couldnt make edit in time, here's more

and post #22, oddly the amps per circular mil seems to go down as the wire size goes up. ~3.6mA/cmil for #14 down to ~1.6mA/cmil for #5 wire. a significnat decrease in allowed current, but kinda makes sense as heat is function of I^2. but oddly the table allows for about 1W/ft in #5 @55A, but in #14 (restrcited to 15A) its only 0.57W/ft (@100ft). allowing 20A on #14 would be almost identical heat density, thus why i am baffled as to why #14 is so restricted.

Question is, restricted compared to what? Keep in mind the material around the wire plays the biggest role.

In the IEC for example 2.5mm2 twin and earth cable (similar to NM) is rated 27amps clipped direct (stapled to wood no insulation), 20amps in contact with insulation and in BS7671 13.5amps when surrounded by dense thermal insulation. This is all governed by the cable's ability to dissipate heat.

 

[jumper]

this 240.4 keeps throwing me off. is Romex a "flexible cable" ?

No. Those other types of cables/cords/wires ampacities are covered in different tables in other sections. See 400 and 402. NM/Romex is in 310.15 and 334.

 

[FionaZuppa]

again , sorry, my edit time was reached.

what i am suggesting is a std in heat density. ~1.02W/ft @ 100ft of wire seems like a good # (just as example here). however, the current ampacity table does not provide a constant heat density.

but if it were constant across all wire sizes then the table would look something like this:

(rounded up to nearest whole)
#14 20
#12 25
#11 28
#10 32
#9 36
#8 40
#7 45
#6 50
#5 57
#4 64
#3 71
#2 80
#1 92
#0 101

these max amps per wire size based on 1.02W/ft @100ft. in NEC2011 #1 is 110A in 60C column, thats 1.45W/ft @100ft, and #4 is 1.23W/ft @ 100ft. its not constant.

heatsink of insulation area only?
ok
#1 would be 10.9sq.in/ft
#4 would be 7.698sq.in/ft

thus the heatsink area density is
#1 0.133W/sq.in.
#4 0.160W/sq.in.

not even the heatsink area density is constant.

 

[FionaZuppa]

so my point is, if its insulation damage from heat that NEC is trying to protect against, then use a std based on watts/area, or watts/ft of wire to define ampacity and OCD #'s.

 

[FionaZuppa]

Larger wires dissipate less heat relative to smaller wire.

I^2R defines the exothermic heat. for same current yes, but since ampacities go up with wire size the heat generated could be exactly the same between two wire sizes with different current flow. and since the insulation type is exactly the same the temp of the insulation will be a function of insulation surface area. this is what i describe in last posts, etc. thus, the ampacity table should be rooted in a std heat density, either W/ft@100ft or W/sq.in/ft of insulation area. if you choose 1.02W/ft the actual temp of insulation will go down as wire size goes up because theres more surface area on the larger wires, an inherent bonus for the way NEC thinks, etc.

 

[don_resqcapt19]

I here that often, but the interesting thing is that ampacities actually went up in the 2012 CEC, and a special lax was made regarding electric heat where #14 could be loaded to 20amps and the breaker sized by 125% without the wire. Very peculiar.


EDIT: The two CEC tables side by side:

You hear it because that is what the substantiation for proposal 6-99 said.

Substantiation: Acceptance of this proposal will harmonize the ampacity values between NEC Table 310.16 and CEC Tables 2 and 4 with technically substantiated values. ...

60ºC copper conductor temperature – There are only two differences (5 amps) in ampacities between the NEC and the CEC and the CEC matches the rounded calculated values on these two conductor sizes. There are three values where the NEC and CEC agree but all three differ from the rounded calculated values by 5 amps. A proposal is also being submitted to correct the CEC on the five minor deviations.

90ºC copper conductor temperature – There are only two differences (5 amps) in ampacities between the NEC and the rounded calculated ampacities. The CEC is lower on all values and a proposal is being submitted to correct the CEC to agree with the rounded calculated values and the revised NEC. The CEC agrees with the values in the 1978 NEC.

60ºC aluminum conductor temperature – There are three differences (5 amps) in ampacities between the rounded calculated ampacities, and the NEC and the CEC have the same values for these three sizes (8 & 4/0 AWG and 700 kcmil). On 12 AWG, the NEC is 5 amps higher (20 vs 15 amps) than the rounded calculated value and the CEC. A proposal is also being submitted to revise the CEC on the four sizes to agree with the rounded calculated values and the revised NEC.

90ºC aluminum conductor temperature – There are four differences (5 amps) in ampacities between the NEC and the rounded calculated ampacities. The CEC is lower on all values and a proposal is being submitted to correct the CEC to agree with the rounded calculated values and the revised NEC. The CEC agrees with the values in the 1978 NEC. The 5 amp differences in ampacities between the rounded calculated values and the NEC are most likely due to rounding when the NEC table was revised
in 1981. The supporting material for that revision no longer exists.

This Proposal was generated by the NFPA/CSA NEC/CEC Ampacity Harmonization Task Group ...

 

[FionaZuppa]

so this task force is basing the changes on "rounding" and to "harmonize" the two. how about they get real data like i describe and revamp the whole table based on data. it boils down to how hot the insulation of wire gets when installed per the allowable installations and wire types. since NEC is way over conservative, using a base # like 1.02W/ft@100ft (<-- notice it is not wire specific, the table i gave a few posts back was for copper) would make the ampacity table more accurate, and the code wouldnt need all these wacky exceptions. it would also make it very simple to calculate if the next wire up is needed. i am not even sure if the derating would be needed for 4+ CCC's in the raceway if there is a max fill # to begin with. you can adjust max fill # perhaps down just a tad, this should be enough to accommodate x # of wires in raceway up to max fill @ 1.02W/ft@100ft, you would get max watts when all wires were #14 because under my insanely proposed std each wire would generate same max heat, thus putting bigger wires into the raceway would mean less wires and thus less generated heat, etc.

but sure, NEC is what it is, some of it makes no sense to me as to how/why the verbiage is the way it is. i myself like to see those lab tests, or FD reports showing real data, etc.

 

[wwhitney]

thus the heatsink area density is
#1 0.133W/sq.in.
#4 0.160W/sq.in.

not even the heatsink area density is constant.

Of the different quantities you have calculated, this is the only one that seems directly related to insulation temperature. The above numbers are close, why don't you calculate the whole series for copper wire sizes and see how much it varies?

Cheers, Wayne

 

[mbrooke]

You hear it because that is what the substantiation for proposal 6-99 said.

Thanks

I saw that but Im puzzled why as to why the tables seem to be doing the opposite? Id imagine table 2 should look like NEC's 310.15 number for number?

 

[mbrooke]

so my point is, if its insulation damage from heat that NEC is trying to protect against, then use a std based on watts/area, or watts/ft of wire to define ampacity and OCD #'s.

I think you have good idea going, however consider that a conductor's ability to dissipate heat plays a role, not just the heating itself.


Please forgive me if I am viewing it wrong.

 

[don_resqcapt19]

so this task force is basing the changes on "rounding" and to "harmonize" the two. how about they get real data like i describe and revamp the whole table based on data. it boils down to how hot the insulation of wire gets when installed per the allowable installations and wire types. since NEC is way over conservative, using a base # like 1.02W/ft@100ft (<-- notice it is not wire specific, the table i gave a few posts back was for copper) would make the ampacity table more accurate, and the code wouldnt need all these wacky exceptions. it would also make it very simple to calculate if the next wire up is needed. i am not even sure if the derating would be needed for 4+ CCC's in the raceway if there is a max fill # to begin with. you can adjust max fill # perhaps down just a tad, this should be enough to accommodate x # of wires in raceway up to max fill @ 1.02W/ft@100ft, you would get max watts when all wires were #14 because under my insanely proposed std each wire would generate same max heat, thus putting bigger wires into the raceway would mean less wires and thus less generated heat, etc.

but sure, NEC is what it is, some of it makes no sense to me as to how/why the verbiage is the way it is. i myself like to see those lab tests, or FD reports showing real data, etc.

Per the substantiation for proposal 6-99 the ampcities are based on the following technical documents.

IEEE Std 835-1994, IEEE Standard Power Cable Ampacity Tables, Section 3.4.2 Adjustment for change in maximum conductor temperature or temperature due to dielectric loss.

AIEE-IPCEA Power Cable Ampacities, 1962 (AIEE Pub. No. S-135-1 /IPCEA Pub. No. P-46-426) Volume 1 – Copper Conductors, Adjustment for Change in Parameters, page III.

 

[mbrooke]

Anyone have a link to those docs or how to get them?

 

[FionaZuppa]

Of the different quantities you have calculated, this is the only one that seems directly related to insulation temperature. The above numbers are close, why don't you calculate the whole series for copper wire sizes and see how much it varies?

Cheers, Wayne

i can generate the whole table based on total wire diameter (wire + insulation). the previous calcs where just using wire (metal) diameter, so new #'s will be slightly less than the previous ones, even better for NEC, etc. but heck, using just wire (metal) diameter is even better as it removes any differences a wire maker may have in insulation thickness, etc. there will certainly be heatsink variances based on insulation thickness, type of insulation, etc etc, so there has to be some generalization done, otherwise there will need to be sets of calculations based on the numerous types of insulation and how it is applied to wire.

i will hunt down the two docs referenced.