NEC Changes For #14 Ampacity

[mbrooke]

yeah, i now about the N-M calcs, it summarizes the physics into a neat equation accounting for heat loss. the issue with it is this, N-M math uses heat loss variables. my math eliminated the heat loss and closely follows worse case scenario of say NM being sandwiched between insulating batts, or in the case we saw some posts back, the wire migrated into the bay that was full of insulation. as long as there exists a TA lower than conductor temp there will always be heat flow, but the flow flux will varying depending on insulation properties which has impact on if the wire itself is compromised. notice in the N-M equation that Rca is in denominator, make that # high and ampacity goes down.

Correct.

all that said, N-M equation allows NEC to define ampacity using known property values of wire types, and, the result is per NEC definition of "ampacity"

, the current in amperes a conductor can carry continuously under the conditions of use without exceeding its temperature rating.

Correct, and therefore in order to take into account most scenarios tables must air on the side of conservative.

the wire temp rise of #14 under load-end full short on a std 40A ocpd @ 75ft (two ccc's) seems like it could be (has possibility) in that hazard zone. if you test 1,000 ocpd's in this test scenario perhaps just one of them takes the full 3sec to trip. isnt it this 1/1000th scenario the exact thing NEC tries to account for in the verbiage? in this instance its not a unknown thing, its a real world thing and, at least by math, seems like a known hazard.

I would say so, the NEC is imo concerned about that 1 in 1000. However the 40amp OCPD seems to be a curve ball. It could take 3 seconds to trip if the short circuit current is low enough, thus in theory the wire would heat less, but none the less I really think this should be tested as well under varying currents ie, 75 amps, 100amps, 150 amps, 250 amps ect to see how much each scenario heats the wire.

all this however stemmed from the post about ocpd's, which is tied back to NEC exceptions, which may need to be modified. my original Q is around #14 ampacity and why has CEC gone back to 20A?

Im just as clueless as you to be honest. My theory is that under conservative conditions the continuous 24/7 ampacity of #14 is around 17 to 18.75 amps at 60*C. Thus, code making panels can either round up or round down that value, especially when other factors already restrict the current like 240.4 D and an AC MCA already having 125% factored in.

There is the exception of electric heat in Canada which lets NM go to 20amps, but in theory that value would never exceed 20 amps because hard wired heat is a fixed load. The breaker is only for short circuit protection while overload protection comes for the inability to add more to the circuit unlike receptacle outlets. Thus the CEC may have decided to round 18.7 to 20 in order to allow electric heat to take advantage of that. No harm would occur as most terminals are 75*C anyways.

My theory behind 240.4 D is simply from an OCPD stand point coupled with the fact outlet circuits are the most chronically abused in terms of overloading. Most trip curves start at around 125%, and under NEMA max trip thresholds a breaker can hold 134% indefinitely. Thus, a 15 amp breaker can take 18.75 amps to 20.1 amps. Thus, it is possible to have a 15 amp receptacle circuit carry 18 or even 20 amps forever 24/7. The breaker is on the verge of tripping, but because it does not, the wire will heat at 20amps, not 15.

Therefore, when people make the argument "if #14 is good for 20amps, why not a 20amp OCPD?" That would make sense at first glance, but here is what would happen in reality: A 20 amp breaker needs 25 amps to trip at 125%, and 26.8amps to trip at 134%. Thus a #14 outlet circuit could be loaded to 26 amps (above the 90*C column) indefinitely without ever tripping a breaker. In open air 26 amps will not bother #14 to much, but if this is dense insulation those 6 extra amps will make a difference. 6 amps may not seem like much, but when dealing with small wire a few amps is actually a substantial % increase.


In any case when dealing with small wire, a person will always need to factor in 134% to make room for worse case OCPD tripping.

Think of it like this: a 15 amp breaker is really a 20amp breaker, and a 20 amp breaker is really a 25 amp breaker.

 

[FionaZuppa]

Therefore, when people make the argument "if #14 is good for 20amps, why not a 20amp OCPD?" That would make sense at first glance, but here is what would happen in reality: A 20 amp breaker needs 25 amps to trip at 125%, and 26.8amps to trip at 134%. Thus a #14 outlet circuit could be loaded to 26 amps (above the 90*C column) indefinitely without ever tripping a breaker. In open air 26 amps will not bother #14 to much, but if this is dense insulation those 6 extra amps will make a difference. 6 amps may not seem like much, but when dealing with small wire a few amps is actually a substantial % increase.


In any case when dealing with small wire, a person will always need to factor in 134% to make room for worse case OCPD tripping.

Think of it like this: a 15 amp breaker is really a 20amp breaker, and a 20 amp breaker is really a 25 amp breaker.

so, i wouldnt say #14 is good for 20A, i'd say from N-M the wiring scenario chosen its good for 35A, and a 20A ocpd is fine because we expect the ocdp to really trip at 26.8amp which is well below the N-M #.

i am not saying make this a blanket allow, but i am saying seems silly to do a blanket "15a max ocpd for #14" and then throw in some exceptions that seem to carry hazard, the restriction should be NM (or the like) restricted.

next is ocpd rating. if ocpd's trip at about 134% of rating, then why not mark the this # on the breaker, or, make a breaker that when loaded at 134% that # matches the printed rating on the device. if you mark it 20A then the internals should trip after exceeding 134%*14.9A, etc. ~50yrs since we landed on the moon and we are still dealing with having to adjust printed #'s for ocpd's to get a "real" #, doesnt make sense. mark the darn thing for what it does, a ocpd is a trip device, so mark it when it trips, not some % # less than.

and just to note, the example i gave for temp rise, that's a hard short of the #14 itself, no device there, just clamp the hot to the neutral. this type of short is a real world possibility and is regardless as to what an AC motor has for protection.

 

[FionaZuppa]

Therefore, when people make the argument "if #14 is good for 20amps, why not a 20amp OCPD?" That would make sense at first glance, but here is what would happen in reality: A 20 amp breaker needs 25 amps to trip at 125%, and 26.8amps to trip at 134%. Thus a #14 outlet circuit could be loaded to 26 amps (above the 90*C column) indefinitely without ever tripping a breaker. In open air 26 amps will not bother #14 to much, but if this is dense insulation those 6 extra amps will make a difference. 6 amps may not seem like much, but when dealing with small wire a few amps is actually a substantial % increase.


In any case when dealing with small wire, a person will always need to factor in 134% to make room for worse case OCPD tripping.

Think of it like this: a 15 amp breaker is really a 20amp breaker, and a 20 amp breaker is really a 25 amp breaker.

so, i wouldnt say #14 is good for 20A, i'd say from N-M the wiring scenario chosen its good for 35A, and a 20A ocpd is fine because we expect the ocdp to really trip at 26.8amp which is well below the N-M #.

i am not saying make this a blanket allow, but i am saying seems silly to do a blanket "15a max ocpd for #14" and then throw in some exceptions that seem to carry hazard, the restriction should be NM (or the like) restricted.

next is ocpd rating. if ocpd's trip at about 134% of rating, then why not mark the this # on the breaker, or, make a breaker that when loaded at 134% that # matches the printed rating on the device. if you mark it 20A then the internals should trip after exceeding 134%*14.9A, etc. ~50yrs since we landed on the moon and we are still dealing with having to adjust printed #'s for ocpd's to get a "real" #, doesnt make sense. mark the darn thing for what it does, a ocpd is a trip device, so mark it when it trips, not some % # less than.

and just to note, the example i gave for temp rise, that's a hard short of the #14 itself, no device there, just clamp the hot to the neutral. this type of short is a real world possibility and is regardless as to what an AC motor has for protection.

 

[mbrooke]

so, i wouldnt say #14 is good for 20A, i'd say from N-M the wiring scenario chosen its good for 35A, and a 20A ocpd is fine because we expect the ocdp to really trip at 26.8amp which is well below the N-M #.

Well, if NM is in open air sure, but consider worse case scenarios.

i am not saying make this a blanket allow, but i am saying seems silly to do a blanket "15a max ocpd for #14" and then throw in some exceptions that seem to carry hazard, the restriction should be NM (or the like) restricted.

But keep in mind in all cases outside of 240.4D and Canada's blanket rule small conductors are protected from long term overload.

A motor might load #14 to 16 amps, and in Canada electric heat will certainly put 20 amps of load on #14, however the load does not go over 20amps, that is the difference. Receptacle circuits are different. The NEC has no way of knowing what will get plugged in, and in some cases people will load a circuit up until it trips, learning they can plug in X amount of appliances before that circuit pops. That X amount of appliances can easily be over the handle rating but just under the actual trip point.

Take a 20 amp circuit with #12. A person could plug in 2 1,500 watt heaters translating to 24 amps without the breaker ever tripping. Same for a 15 amp circuit with one heater (12amps) and 5 amps miscellaneous load. A person with no knowledge of electrical would not think twice the circuit is overloaded since nothing has tripped.


Personally table 310.15 should just list 15, 20 and 30amps respectively with 240.4D making exceptions for motors and electric heat.

next is ocpd rating. if ocpd's trip at about 134% of rating, then why not mark the this # on the breaker, or, make a breaker that when loaded at 134% that # matches the printed rating on the device.

This is a good question and a valid one. The answer is that thermal magnetic circuit breakers are very crude devices. The bimetal strip in a circuit breaker is effected by both ambient temperature as well as inadvertent manufacturing variants. Ambient temperatures will either raise or lower a breaker's trip point. Breakers are ambient compensated to some degree, but that can only do so much. In fact this is why the 80% rule exists. The 80% rule off sets heat buildup in panel boards that could otherwise shift the 125% trip curve under to 100%. Further the lack of true precision that accompanies electro-mechnical devices plays a role. A set of identical (make and model) breakers can all trip slightly lower or slightly higher then intended for the same environmental conditions due to minuscule manufacturing variants. If manufacturers did not consider this and set the trip curves to begin at 100% some breaker would be ok tripping between 100 to 105% while other would fail tripping at under 100%, say 95%.

Thus to offset these two uncontrollable variables manufactures intentionally start their trip curves around 125%. This will take care of both ambient differences and manufacturing variants which will provide good assurance that breakers will not trip under their handle rating. To further offset this possibility the NEC requires circuits running over 3 hours to be loaded to only 80%.

if you mark it 20A then the internals should trip after exceeding 134%*14.9A, etc.

True, at 40*C ambient and ideal internals. In reality that breaker could trip atb 18 amps or 22 amps. If I have a panel where half the breakers are tripping under 20amps I technically do not have 20amp breakers.

~50yrs since we landed on the moon and we are still dealing with having to adjust printed #'s for ocpd's to get a "real" #, doesnt make sense. mark the darn thing for what it does, a ocpd is a trip device, so mark it when it trips, not some % # less than.

Only doable where high precision with true compensation for all environments can be achieved. Microprocessors is about the only way to do it.

and just to note, the example i gave for temp rise, that's a hard short of the #14 itself, no device there, just clamp the hot to the neutral. this type of short is a real world possibility and is regardless as to what an AC motor has for protection.

Im with you, Im puzzled on this one.

 

[FionaZuppa]

mbrooke, the 1st line you wrote in last post, N-M is the Neher-McGrath Calculation, not NM as in cable, thus why i said 35A is ok, but perhaps NEC restricts the NM(cable) ampacity due to it's large use.

and you mention that today's ocpd are crude devices. they have been crude devices for a long time. we need better ocpd's, etc.

you think #14 THHN w/ 20A ocpd that is a 75ft run (150ft, two ccc's) in 1" buried pvc conduit is a hazard?

you also say "overloaded" in the 2x1500 watt outlet example you give, but that's based on the NEC ampacity #'s, not N-M equation results.
does NEC apply a N-M * 0.5 rule to come up with its ampacity #'s ??

 

[mbrooke]

mbrooke, the 1st line you wrote in last post, N-M is the Neher-McGrath Calculation, not NM as in cable, thus why i said 35A is ok, but perhaps NEC restricts the NM(cable) ampacity due to it's large use.

Oh! My mistake :ashamed1: Correct. #14 in open air can handle 35 amps, however toss in thermal insulation and its becomes a different story. That was my point. My apologies about that.

and you mention that today's ocpd are crude devices. they have been crude devices for a long time. we need better ocpd's, etc.

In your words, what would be better?

you think #14 THHN w/ 20A ocpd that is a 75ft run (150ft, two ccc's) in 1" buried pvc conduit is a hazard?

In such a case not likely as 25 amps will not heat the wire past dangerous temps.

you also say "overloaded" in the 2x1500 watt outlet example you give, but that's based on the NEC ampacity #'s, not N-M equation results.

Correct, but my point is that the ampacity tables assume worse case where N-M does not.

does NEC apply a N-M * 0.5 rule to come up with its ampacity #'s ??

To be honest I dont know, but my theory is actual testing. The links mentioned something about a paper which was used to drive the NEC tables many decades back... hold on...

 

[mbrooke]

Here is it is:


http://www.electrician2.com/articles/t31016.htm

[FONT=Arial, Helvetica]Table 310-16 of the National Electrical Code. Rosch's original table was based on an ambient temperature of 30 degrees centigrade and a conductor temperature of 50 degrees centigrade for code grade rubber, the type of insulation used in those days. If we convert the ampacities in table XI to 60 degrees centigrade using the formula given in note 1 to tables 310-69 through 310-84, setting delta TD equal to 0 (delta TD is for high voltages: we are only concerned with 600 volts and under), and rounding off to the nearest 5 amperes, we can calculate the ampacities for 60 degree insulations as found in the first column of table 310-16. Likewise, the same calculation can determine the the ampacities for the 75 degree and 90 degree columns in table 310- 16. [/FONT]

 

[FionaZuppa]

Here is it is:


http://www.electrician2.com/articles/t31016.htm

but it goes on to say:

[FONT=Arial][SIZE=2][SIZE=2][COLOR=#0000cd]To develope a more accurate method of finding the ampacity of conductors in underground installations two cable engineers, in 1957, developed the [/COLOR][URL="http://www.electrician2.com/articles/ampacity.htm"][COLOR=#0000cd]Neher-McGrath equation[/COLOR][/URL][COLOR=#0000cd] found in 310-15(c) of the 1999 NEC.[/COLOR][/SIZE][/SIZE][/FONT]

this tells me that the results of N-M calculation is the ampacity. i calculated 35A, and that was not free air, it had a thermal R inside a 1" conduit.

i think what we are seeing is, a gross generalization by NEC on wiring. why would NEC call it a violation to run a #14 THHN bc in underground conduit when the ocpd was 20A, and for argument sake, this run is for a few lights and ceiling fans where calculated load is no more than sticker max bulb watts + motor amps.

Correct, but my point is that the ampacity tables assume worse case where N-M does not.

not exactly true, N-M does exactly that by including R values of the wired scenario. simply take the #'s i have for my N-M example and swap out the R value for a 24" thick fiberglass batt (a sandwich in real test, etc).

perhaps the definition of "ampacity" (a NEC term) should not be used in the N-M paper because its kinda confusing. there should be a dsictinction made. the NEC "ampacity" should be a corrected # whereas N-M # should be a max calculated #. see, i never read anywhere in N-M stuff that says how close to max temp the wire will get if you ran that wire for the # of amps that was just calculated for the wired scenario. N-M says "ampacity w/o exceeding max temp", well, max temp is a variable in the equation, so if i use 90C as max temp does that mean 89.9C is ok? no is the answer. N-M should just say "amps that would take the wire to max temp", then from there the NEC could apply a derating factor for various wire types and wiring scenarios. NM-B, ok, make that 15A(max) ocpd. a #14 THHN run in 1" conduit all by itself, a 20A ocpd is ok. #12 vs #14 is much more $$, etc.

 

[mbrooke]

this tells me that the results of N-M calculation is the ampacity. i calculated 35A, and that was not free air, it had a thermal R inside a 1" conduit.

That is correct.

i think what we are seeing is, a gross generalization by NEC on wiring. why would NEC call it a violation to run a #14 THHN bc in underground conduit when the ocpd was 20A, and for argument sake, this run is for a few lights and ceiling fans where calculated load is no more than sticker max bulb watts + motor amps.

You are correct. Table 310.15 is based around worse case or near worse case scenarios yet applies to all installations outside of free air. The end result is amapcity values that are much more conservative then the actual scenario at hand.

IEC based codes like BS7671 have over a dozen different ampacity tables entirely based on how the cable is installed.

For example, 2.5mm2 (a bit bigger then out 2.08 14 gauge wire) is rated 13.5 amps when covered in dense insulation, 20 amps in wall, 27 amps clipped direct, 29 amps direct burial or single wire, and 36 amps mineral insulated not exposed to human touch, 49 amperes at 105*C operation temperature mineral insulated not exposed to touch and at 750 volts the prior is allowed to go to 54 amps. Other tables and de-rating factors can raise or lower these values.

Thus 2.5mm2 can be anywhere from 10 amps to 60 amps all depending on heat dissipation, dielectric insulating material, material around said conductor and exposure to human touch.


I believe the NEC should take a similar approach where its guaranteed a wire's heat dissipation will not change such as underground installations.

not exactly true, N-M does exactly that by including R values of the wired scenario. simply take the #'s i have for my N-M example and swap out the R value for a 24" thick fiberglass batt (a sandwich in real test, etc).

IS the ampacity less then table 310.15?

perhaps the definition of "ampacity" (a NEC term) should not be used in the N-M paper because its kinda confusing. there should be a dsictinction made. the NEC "ampacity" should be a corrected # whereas N-M # should be a max calculated #. see, i never read anywhere in N-M stuff that says how close to max temp the wire will get if you ran that wire for the # of amps that was just calculated for the wired scenario. N-M says "ampacity w/o exceeding max temp", well, max temp is a variable in the equation, so if i use 90C as max temp does that mean 89.9C is ok? no is the answer.

Correct. Though Id even guess the N-M equation might be conservative. One reason why you would not want conductors actually reaching 90*C is human touch. While human touch is not likely in wall, conduit at 194*F may burn someone.


My brain is a bit fried at the moment, but I will look an N-M deeper. You are clearly on to something as your train of thought is on track.

N-M should just say "amps that would take the wire to max temp", then from there the NEC could apply a derating factor for various wire types and wiring scenarios. NM-B, ok, make that 15A(max) ocpd. a #14 THHN run in 1" conduit all by itself, a 20A ocpd is ok. #12 vs #14 is much more $$, etc.

I think that would be better. Ampacity based on the condition of use is not something the NEC is known for.

 

[FionaZuppa]

I think that would be better. Ampacity based on the condition of use is not something the NEC is known for.

how about this:
1) "romex" "NM" #12/14 gets the "20/15A max ocpd" restriction.
2) some of the motor exceptions need to be re-evaluated
3) the rest then needs some ampacity tuning using test data

 

[mbrooke]

how about this:
1) "romex" "NM" #12/14 gets the "20/15A max ocpd" restriction.
2) some of the motor exceptions need to be re-evaluated
3) the rest then needs some ampacity tuning using test data

Sounds good. Although where feeding fixed loads like water heaters I think the max ocpd restriction should be removed.

 

[FionaZuppa]

Sounds good. Although where feeding fixed loads like water heaters I think the max ocpd restriction should be removed.

max ocpd only for romex, to escape the restriction another type of wiring would be needed. simple, restrictive, flexible.

the idea that we know what the end device can do is not a good approach. a better approach is to evaluate faults at the wire-device junction. as example, the junction of a 16A water heater develops and short to N and even when heater is off the circuit is allowing 8A to flow 24/7. #14 THHN in conduit on 20A ocpd will be ok and not trip.

why dont we have smart breakers? forget AFCI/GFCI stuff, how about a small flip switch, A/B setting, A=continuous circuit, B=non-continuous circuit. when on B mode if the breaker sees 50% of rated amps (i dunno some %) for more than 24hr period then a small buzzer sounds until it is cleared. or if switch is too complicated then just make separate A & B breakers, etc.

 

[mbrooke]

max ocpd only for romex, to escape the restriction another type of wiring would be needed. simple, restrictive, flexible.

the idea that we know what the end device can do is not a good approach. a better approach is to evaluate faults at the wire-device junction. as example, the junction of a 16A water heater develops and short to N and even when heater is off the circuit is allowing 8A to flow 24/7. #14 THHN in conduit on 20A ocpd will be ok and not trip.

why dont we have smart breakers? forget AFCI/GFCI stuff, how about a small flip switch, A/B setting, A=continuous circuit, B=non-continuous circuit. when on B mode if the breaker sees 50% of rated amps (i dunno some %) for more than 24hr period then a small buzzer sounds until it is cleared. or if switch is too complicated then just make separate A & B breakers, etc.

The flip switch idea sounds complicated, but when you have a GFCI breaker it will trip on a ground fault which could otherwise introduce more then 20amps of current.

 

[FionaZuppa]

The flip switch idea sounds complicated, but when you have a GFCI breaker it will trip on a ground fault which could otherwise introduce more then 20amps of current.

gnd faults are different than hot-N faults. both are real possibilities.

NEC makes a big distinction between continuous and non-continuous circuits, however, this distinction does not match any ocpd technology that i am aware of.

 

[mbrooke]

gnd faults are different the hot-N faults. both are real possibilities.

Good point.

NEC makes a big distinction between continuous and non-continuous circuits, however, this distinction does not match any ocpd technology that i am aware of.

My understanding is that continuous loads are restricted to 80% to limit heat build up.

 

[romex jockey]

This is a good question and a valid one. The answer is that thermal magnetic circuit breakers are very crude devices. The bimetal strip in a circuit breaker is effected by both ambient temperature as well as inadvertent manufacturing variants. Ambient temperatures will either raise or lower a breaker's trip point. Breakers are ambient compensated to some degree, but that can only do so much. In fact this is why the 80% rule exists. The 80% rule off sets heat buildup in panel boards that could otherwise shift the 125% trip curve under to 100%. Further the lack of true precision that accompanies electro-mechnical devices plays a role. A set of identical (make and model) breakers can all trip slightly lower or slightly higher then intended for the same environmental conditions due to minuscule manufacturing variants. If manufacturers did not consider this and set the trip curves to begin at 100% some breaker would be ok tripping between 100 to 105% while other would fail tripping at under 100%, say 95%.

Thus to offset these two uncontrollable variables manufactures intentionally start their trip curves around 125%. This will take care of both ambient differences and manufacturing variants which will provide good assurance that breakers will not trip under their handle rating. To further offset this possibility the NEC requires circuits running over 3 hours to be loaded to only 80%.

This makes sense Mr MBrooke....

Most of us have experiences heated panels under load

So the Q here is how OCPD's are tested , solo, or in such an environ?

~RJ~

 

[mbrooke]

This makes sense Mr MBrooke....

Most of us have experiences heated panels under load

So the Q here is how OCPD's are tested , solo, or in such an environ?

~RJ~

Solo is my understanding. The factory trip curves are based on a 40*C ambient, but may get hotter in panel boards.

 

[FionaZuppa]

the hotter they are (@ambient) the sooner they will trip from amps heating.

UL should require three tests, 40/50/60C just so we have an idea of how they react in a warmed panel, etc.

 

[mbrooke]

the hotter they are (@ambient) the sooner they will trip from amps heating.

UL should require three tests, 40/50/60C just so we have an idea of how they react in a warmed panel, etc.

They should. A test of panel board heat build up would also be interesting.

 

[romex jockey]

110.14

110.14

Considering 'heat' , and/or 'heat related' events along the entire circuit , OCPD, conductor ,devices and terminations seems to have had the NEC's focus on the latter>

(C) Temperature Limitations. The temperature rating as-
sociated with the ampacity of a conductor shall be selected
and coordinated so as not to exceed the lowest temperature
rating of any connected termination, conductor, or device.

Some of you might recall 110.14's past revisions , along with all the derating confusion .....

~RJ~