A conductor is governed by more than the number printed beside its size in the ampacity table. There are three separate ceilings on the same copper, set by how hot the insulation may get, how hot the thing on the end of it may get, and how long the circuit is. Voltage drop is the third, and it is the only one that does not care how much current the conductor can carry. It cares how far the current has to travel.
That distinction explains most arguments about wire size. Two people size the same circuit, both read the table correctly, and land on different conductors.
The Ampacity Table Answers a Heat Question
Every value in the ampacity table is a heat balance. The table asks how much current the insulation can pass before its temperature rating is exceeded, given how it is installed and how much air reaches it. That is the entire content of the table, and it says nothing about the far end of the circuit, because the table does not know where that is.
Two details trip people up. The same conductor appears at three temperatures and the highest column is not automatically available. And ampacity is a multiplication, not a lookup: the table value is the starting point, and ambient correction and conductor adjustment factors are applied to it. The code now calls the bundling factor ampacity adjustment and the ambient factor correction.
Those factors can be substantial. Three 6 AWG THWN-2 conductors in a raceway ¾ inch above a roof take a 40 °F adder for sunlight, so at a 90 °F ambient a 75 A table value becomes 57 A. Nothing about the conductor changed. The roof did.
The Termination Caps the Table
Section 110.14(C) is where most designs actually land, and it is routinely skipped.
A conductor may be insulated for 90 °C and be perfectly capable of running at that temperature. The breaker it terminates into is not. Terminations are rated 60 °C or 75 °C, and the conductor’s usable ampacity cannot exceed that. Conductors with insulation ratings above the terminal rating may be used for the adjustment and correction arithmetic, but the answer still cannot exceed the temperature limitation of the equipment it lands on.
The 90 °C column is mostly an arithmetic device. It is not a license to run a conductor at 90 °C into a 75 °C lug.
The Small-Conductor Rule Caps It Again
Below 10 AWG there is a third ceiling, and it overrides the table outright.
Overcurrent protection must not exceed 15 A for 14 AWG copper, 20 A for 12 AWG copper and 30 A for 10 AWG copper, with 15 A for 12 AWG aluminum and 25 A for 10 AWG aluminum. The rule applies after adjustment and correction, so you cannot calculate your way past it.
A 10 AWG conductor’s real ceiling on an ordinary circuit is therefore 30 amperes, whatever the table says.
Voltage Drop Is a Question About the Run
Voltage drop is not a property of a cable. It is a property of a circuit, and the product of current, distance and the impedance of the conductors carrying it.
The first is the resistance-only model most calculators use: drop equals two times a constant, times current, times length, over the circular-mil area. The constant is 12.9 ohm-circular-mil per foot for copper and 21.2 for aluminum, at 75 °C. It is a direct-current resistance model.
The second is the one the code intends for alternating current. It takes the conductor’s alternating-current resistance and its inductive reactance from the chapter of tables, combines them at the load’s power factor, and applies the same length factor. The two disagree in a way that matters.
Why the Formula Doubles
Current has to come back, and the drop is measured around the whole loop. For a single-phase circuit the drop over a one-way run of L feet at I amperes is two times I, times the effective impedance, times L over 1000. For a balanced three-phase circuit it is the square root of three instead of two.
That is why the entered length is the one-way distance. A round-trip distance doubles a drop that has already been doubled.
Where the drop is measured
Two Calculators, Two Answers
The two methods agree closely on small conductors and part company badly on large ones, and the direction of the error flips with size.
Reactance barely changes with size while resistance collapses. On 12 AWG copper in PVC conduit the tabulated reactance is 0.054 ohms per 1000 feet against a resistance of 2.0, so reactance is under 3 percent of the resistance and ignoring it costs almost nothing. On 500 kcmil copper in steel conduit the reactance is 0.048 against a resistance of 0.029. Reactance is now larger than resistance, and a model that omits it has thrown away most of the answer.
That case at 0.85 power factor gives an effective impedance near 0.050 ohms per 1000 feet. The resistance-only model returns about 0.026. It answers with roughly half the real number, and in the optimistic direction.
Set against a small conductor the shortcut errs the other way. A 12 AWG at 0.85 power factor has an effective impedance near 1.73 ohms per 1000 feet against the model’s 1.98, making it conservative by roughly 14 percent.
So the shortcut is not simply wrong. It is wrong differently at each end of the size range, which is worse: a rule of thumb that is safe on one job is unsafe on the next. On large feeders in steel raceway, treat any resistance-only result as a floor.
Two calculators, two answers
Reading the Impedance Table
Resistance is tabulated per 1000 feet. Reactance is tabulated for three single conductors in a raceway at 60 Hz and 75 °C, with separate columns depending on what the raceway is made of.
Steel raises reactance. On 500 kcmil it is 0.039 in PVC raceway and 0.048 in steel, while resistance moves only from 0.027 to 0.029. Anyone specifying large feeders in steel conduit uses the higher figure whether or not they noticed the column.
Resistance is corrected to the conductor’s operating temperature and reactance is not. And the values are per conductor, so the return path still has to be accounted for.
Copper conductor | R, PVC | R, steel | XL, PVC | XL, steel |
|---|
12 AWG | 2.0 | 2.0 | 0.054 | 0.068 |
|---|
10 AWG | 1.2 | 1.2 | 0.050 | 0.063 |
|---|
8 AWG | 0.78 | 0.78 | 0.052 | 0.065 |
|---|
6 AWG | 0.49 | 0.49 | 0.051 | 0.064 |
|---|
4 AWG | 0.31 | 0.31 | 0.048 | 0.060 |
|---|
2 AWG | 0.19 | 0.20 | 0.045 | 0.057 |
|---|
1/0 AWG | 0.12 | 0.12 | 0.044 | 0.055 |
|---|
4/0 AWG | 0.062 | 0.063 | 0.041 | 0.051 |
|---|
250 kcmil | 0.052 | 0.054 | 0.041 | 0.052 |
|---|
500 kcmil | 0.027 | 0.029 | 0.039 | 0.048 |
|---|
Aluminum and Parallel Sets
Aluminum has about 61 percent of copper’s conductivity, so an aluminum conductor needs roughly 1.6 times the circular-mil area for the same drop. That is not an argument against aluminum on feeders, but the aluminum conductor is always a size or two larger than the copper one it replaces.
Conductors may be run in parallel, and each set divides both resistance and reactance, so the drop divides by the number of sets. The permission applies to conductors 1/0 AWG and larger, so below 1/0 you cannot parallel your way out of a voltage drop problem.
Are Those Figures Required?
The two numbers everyone quotes are 3 percent for a branch circuit at the farthest outlet and 5 percent for a feeder and branch circuit combined. Both come from informational notes, and informational notes are not enforceable code.
That is not a reason to ignore them, because the same code makes voltage drop mandatory in specific places.
Limit | Applies to | Reference | Status |
|---|
3% | Branch circuit, at the farthest outlet | 210.19(A) informational note | Recommendation |
|---|
5% | Feeder plus branch circuit combined | 215.2(A) informational note | Recommendation |
|---|
1.5% branch, 2.5% combined | Sensitive electronic equipment | 647.4(D) | Mandatory |
|---|
15% starting, 5% running | Fire pump circuits | 695.7 | Mandatory |
|---|
Proportional upsizing | Equipment grounding conductor, when conductors are upsized for drop | 250.122(B) | Mandatory |
|---|
Where a circuit is deliberately upsized to control voltage drop, the equipment grounding conductor has to be upsized in the same circular-mil proportion, and that part is enforceable. Sensitive-electronic circuits have hard limits roughly half the usual figures. So whether the 3 and 5 percent figures are required depends where the job is.
One Conductor, Four Ceilings
Here is where the arithmetic lands, on an ordinary circuit: 120 volts, single-phase, one-way run of 200 feet, 10 AWG copper in PVC conduit, power factor 0.85, conductors at 75 °C.
The tabulated resistance is 1.2 and the reactance is 0.050. Combined at 0.85 power factor the effective impedance is about 1.046 ohms per 1000 feet, so the drop is roughly 0.419 volts per ampere of load current.
Ceiling on this conductor | Current | Where it comes from |
|---|
Small-conductor overcurrent limit | 30 A | Overcurrent rule for 10 AWG copper |
|---|
Drop held to 5%, feeder and branch combined | 14.3 A | This run, at 6.0 volts |
|---|
Drop held to 3%, the branch-circuit figure | 8.6 A | This run, at 3.6 volts |
|---|
At the 30 amperes the overcurrent rule allows, this run drops 12.6 volts, or 10.5 percent. The conductor is not too small for the current. It is too small for the distance, and the table that would have told you so is not the ampacity table.
Three ceilings on one conductor
Where Voltage Drop Is Not the Problem
On a short run with a conductor generously sized for its current, the ampacity ceiling and the small-conductor cap both sit below the drop limit, and the drop never enters the conversation.
It also stops being the governing question where the load is not the reason for the circuit. A bonding conductor or a signal cable is sized by what it must survive and what impedance it must present, not by current.
What to Ask Before You Buy
- The one-way circuit length, not the distance between two points on a floor plan.
- The load current and whether it is continuous, because continuous loads are increased before the conductor is selected.
- The system voltage and phase, which decides whether the length factor is two or the square root of three.
- The power factor, or an honest statement that it is unknown.
- The raceway material, because steel and PVC do not give the same reactance on large conductors.
- The termination rating, because it caps the ampacity before correction is applied.
- Whether the run is upsized for drop, because that decision carries the grounding conductor with it.
The Short Version
One conductor carries three or four separate ceilings, set by the insulation temperature, the terminal temperature, the small-conductor overcurrent rule, and the geometry of the run. The ampacity table answers only the first. Voltage drop answers the last, and on a long run it binds long before the current limit does: on a 200-foot 120-volt circuit, a 10 AWG conductor capped at 30 amperes by the overcurrent rule is limited to about 8.6 amperes by a 3 percent drop.
The two common ways of calculating that drop disagree, and the disagreement is not a rounding error. A resistance-only model is mildly conservative on small conductors and optimistic by roughly half on large conductors in steel raceway, because reactance does not shrink with size the way resistance does. And the 3 and 5 percent figures are a recommendation in most places and a requirement in a few, so the jurisdiction decides.
We build the cable, and we can supply conductor resistance figures for a specific construction, so the drop arithmetic rests on measured data rather than a table entry for a nominal size. Tell us the circuit, the run and the environment.
Sources
- Nordix, "Voltage Drop Calculator" technical documentation — the voltage-drop formula for single-phase, three-phase and direct-current circuits and the effective-impedance form of it; the published limits table distinguishing the 3 percent branch-circuit and 5 percent combined figures in the informational notes to 210.19(A) and 215.2(A) from the mandatory limits in 647.4(D) for sensitive electronic equipment, 695.7 for fire pumps and 250.122(B) for proportional upsizing of the equipment grounding conductor; the statement that informational notes are not enforceable under 90.5 but that energy codes and local amendments make the 3 and 5 percent figures binding in some jurisdictions; the constant of 12.9 ohm-circular-mil per foot for copper and 21.2 for aluminum at 75 °C and the observation that this is a resistance-only model that ignores reactance and skin effect; the excerpt of alternating-current resistance and reactance values for copper conductors at 75 °C in PVC and steel raceways, which is the table reproduced above; the worked example of a 240 volt single-phase 32 ampere load on 150 feet of 8 AWG copper at 0.85 power factor giving 0.690 ohms per 1000 feet, 6.63 volts and 2.76 percent, which is the arithmetic behind the statement that short runs are not drop-limited; the statement that aluminum has about 61 percent of copper’s conductivity and needs roughly 1.6 times the circular-mil area for the same drop; the treatment of parallel sets dividing resistance and reactance by the number of sets and being permitted only for 1/0 AWG and larger; and the notes that resistance is corrected to conductor operating temperature while reactance is not, that the length entered is the one-way distance, and that the tabulated reactance rises in steel raceway. This is a voltage-drop calculator published by a software vendor, and its method section, not its tool, is what is cited here — https://www.nordixhq.com/tools/voltage-drop-calculator
- EC&M, "Conductors for General Use" — the requirement in 110.14(C) that conductors with insulation temperature ratings above the termination rating may be used for ampacity adjustment and correction but that the resulting ampacity still cannot exceed the temperature limitation of the equipment termination; the observation that the code now uses the term ampacity adjustment for conductors installed without maintaining spacing and the term correction for ambient temperatures other than 86 °F, and that the older term derating has largely been retired; and the rooftop example in which three 6 AWG THWN-2 conductors in a circular raceway ¾ inch above a roof in direct sunlight take a 40 °F ambient adder, so that at a 90 °F outdoor ambient the conductor is corrected against 130 °F and a 75 A table value becomes 57 A. This article documents the 2011 revision of Article 310, so its table and section numbers are the 2011 ones, in which the ampacity table is numbered 310.15(B)(16) and is explicitly noted as formerly 310.16 — https://www.ecmweb.com/national-electrical-code/code-basics/article/20896215/conductors-for-general-use
- EC&M, "Article 240: Overcurrent Protection" — the small-conductor rule under 240.4(D), under which overcurrent protection must not exceed, after ampacity adjustment and correction, 15 A for 14 AWG copper, 15 A for 12 AWG aluminum, 20 A for 12 AWG copper, 25 A for 10 AWG aluminum and 30 A for 10 AWG copper; the next-size-up allowance for overcurrent devices above the conductor ampacity, limited to devices not exceeding 800 A and not applicable to feeder tap or transformer secondary conductors; the requirement that above 800 A the conductor ampacity after adjustment must be not less than the device rating; and two reference values from the ampacity table, being 500 kcmil copper at 380 A and 600 kcmil copper at 420 A at 75 °C. The article is by Mike Holt and dates from 2007, and it cites the ampacity table by its then-current number, Table 310.16, which is also the number current references use — https://www.ecmweb.com/national-electrical-code/code-basics/article/20885916/article-240-overcurrent-protection
