Where the ampacity chart stops answering the question

An ampacity chart answers one question. How much current can this conductor carry continuously without its insulation cooking.

For a short run inside a house, that is the question. For anything long, a shed at the end of the garden, a well pump, a pier light, a van conversion, it is the wrong one, and sizing from the chart alone gives a conductor that is perfectly safe and still fails to do its job.

Two different limits, two different answers

Ampacity is a heat limit. Current flowing through resistance makes heat, and the chart in NEC Table 310.16 is where the conductor and its insulation stop coping. That limit does not care how long the run is, because every foot of it is shedding heat the same way.

Voltage drop is a distance limit. The same resistance that makes heat also consumes voltage along the way, and that loss scales with length. Double the run and you double the drop. Nothing about the conductor’s temperature changes, but what arrives at the far end does.

Where the crossover happens

On a 120 V branch circuit the two limits trade places somewhere around 75 feet of run. On 240 V it is nearer 125 feet, because the same power moves at half the current and drop scales with current.

Past that point the chart stops being the governing figure, and sizing by it means arriving at the far end short. The usual target is to keep drop under 3% on a branch circuit, and under 5% for the branch and feeder combined. In the NEC those appear as Informational Notes rather than mandatory rules, so they are accepted practice rather than law, which is worth knowing when a supplier insists the chart is the only thing that matters.

What it looks like in practice

Take 20 A at 120 V running 100 feet. On ampacity alone, 12 AWG copper handles 20 A and the chart is satisfied. Work the voltage drop and 12 AWG loses well over 3% across that distance, so the sizing moves up, often two steps, to hold the drop inside target.

That is not a safety failure, it is a performance one. The wire will not overheat. The far end simply gets less than it was designed for, and the motor at the end of it pays for that for the rest of its life.

The wire gauge calculator runs both checks and reports which one is binding, in AWG or in square millimetres, so you can see whether you are being pushed up a size by heat or by distance.

The things that quietly make it worse

Three factors move the number and none of them appear on the chart.

Length is measured as the round trip, out and back, not as the distance to the load. People consistently calculate half the real length because the cable only looks like one run on a drawing.

Temperature raises resistance. A conductor running hot in a sunlit conduit has more resistance than the same wire in a cool basement, so the drop at the end of a summer afternoon is worse than the one you calculated.

Starting current dwarfs running current. A pump with a locked rotor draw several times its running figure will momentarily drop the voltage far more than the steady-state calculation shows, which is why a motor at the end of a long run sometimes hums and fails to start while everything else on the circuit behaves.

Low voltage is the extreme case

All of this gets dramatic at 12 or 24 V, which is why van and boat wiring looks absurdly oversized to anyone used to mains work.

The same 3% of 12 V is 0.36 V, where 3% of 240 V is 7.2 V. There is simply no headroom, so a 12 V run of any length needs conductors that look wildly out of proportion to the current. A 20 A circuit at 12 V over 20 feet can want a size that would carry 100 A on a chart, and that is correct rather than paranoid.

The practical approach

Work out the ampacity size first, because it is the floor and it is a safety limit. Then calculate the drop at that size over the real round trip length. If it fails, go up until it passes.

Where the result is marginal, going one size larger is usually cheap insurance. Copper is a one-off cost paid at installation, while a chronically underfed motor is a recurring one, and pulling a replacement cable through a conduit that is already buried is the most expensive version of this problem there is.