The Job That Looks Like Cheating
A 400-amp feeder is a heavy thing. A single copper conductor sized to carry 400 amps is roughly 600 kcmil — a cable as thick as your thumb, stiff as rebar, and miserable to pull around a 90-degree bend in a full conduit. So the trade does something that looks, at first glance, like getting away with something: it splits the load across two smaller conductors run side by side. Two 3/0 coppers in parallel share the 400 amps between them, each carrying about 200. They pull easier, bend easier, and terminate on the same lug configurations you already own.
The National Electrical Code allows this in 310.10(H). But it attaches a list of conditions that reads almost like superstition: the paralleled conductors must be the same length, the same conductor material, the same size in circular mils, the same insulation type, and terminated in the same manner. Every one of those has to match. Change one and the arrangement is no longer code-compliant — and, more importantly, no longer safe.
The rules feel fussy until you understand that they are not really rules about wire. They are rules about how current decides where to go.
Current Follows the Path of Least Impedance
When you give current two paths to the same destination, it does not split evenly out of politeness. It splits in inverse proportion to the impedance of each path. The easier path takes more; the harder path takes less. This is just Ohm's law running in parallel, and it is relentless about the arithmetic.
Impedance in an AC conductor has two parts. There is resistance, which depends on the conductor's cross-sectional area, its material, and its length. And there is reactance, which comes from the magnetic field the current builds around the conductor and depends on the geometry of the run — how the conductors are spaced, how they're arranged in the raceway. For the DC-style intuition, resistance is the part most people picture: a longer, thinner, or higher-resistivity wire is a harder path.
Now imagine two conductors that are supposed to share 400 amps, but one of them is eight feet longer than the other because the installer looped it the long way around a junction box. That extra length is extra resistance. The longer conductor is now the harder path, so it carries less than its half. The shorter conductor — the easier path — carries more than its half. It has to. The load still demands 400 amps, and if one wire refuses to take 200, the other one makes up the difference.
Where the Danger Actually Lives
Here is the part that makes this more than a physics curiosity. Both conductors were sized as 3/0, rated for roughly 200 amps in that installation. The breaker protecting the feeder is a 400-amp breaker, because from its point of view it is protecting a single 400-amp circuit. It has no way to know how the current divides between the two conductors downstream of its terminal.
So picture the mismatch playing out. The short conductor is now carrying 240 amps through a wire rated for 200. It is overloaded by 20 percent. It runs hot — not fault-hot, not trip-the-breaker hot, but continuously, quietly warmer than its insulation was designed to tolerate. The 400-amp breaker sees 400 amps total and is perfectly content. Nothing trips. Nothing announces itself. The insulation on the overloaded conductor simply ages faster than it should, embrittling over months and years until one day it fails in a place no one is looking.
That is the whole hazard in a sentence: unequal impedance forces unequal current, and the protective device is blind to the imbalance because it only measures the sum. The breaker guards the circuit. It cannot guard either individual wire.
This is why the identical-length requirement is not bureaucratic. Length is the variable most likely to drift on a real job site, and it maps directly onto resistance, which maps directly onto how the current divides. Same material matters because copper and aluminum have different resistivity — a copper and an aluminum conductor in parallel would never share evenly. Same size, same insulation, same termination method: each one closes off another way for the two paths to become unequal.
Why You Can't Parallel Small Wire
There's a companion rule that puzzles apprentices: 310.10(H) generally forbids paralleling conductors smaller than 1/0 AWG. Why set a floor? If paralleling is such a clever way to handle big loads, why not parallel two 12-gauge wires to make a cheap 20-amp circuit?
Part of the answer is that paralleling exists to solve a problem small conductors don't have. The whole point is to tame conductors so large they become unmanageable. A 20-amp load is trivially served by a single #12 — there is nothing to solve, so the Code doesn't invite you to introduce the current-sharing risk for no benefit. But there's a physical edge to it too. In smaller conductors, a small absolute difference in length or termination resistance is a large percentage of the total, so the imbalance between two paralleled small wires can be proportionally worse. The bigger the conductor, the more its bulk resistance dominates and the more forgiving the split becomes. The 1/0 floor keeps paralleling in the size range where it behaves.
The Detail That Bites Good Electricians
The requirement applies per phase, and this is where careful people still get caught. If you parallel phase A, you must equally parallel phase B, phase C, and the neutral. And every conductor within a given phase set must match its twin — but you're allowed to make phase A copper and phase B aluminum if you truly wanted to, because the matching is within each parallel set, not across them. In practice nobody mixes materials, but understanding that the rule is about balancing each set against itself is what keeps you from over- or under-applying it.
The other quiet trap is the raceway. Reactance depends on physical arrangement, so conductors of the same circuit run in separate parallel raceways must keep those raceways electrically and physically comparable — same length, same conductor arrangement — or the magnetic geometry alone can unbalance the sharing even when every wire is cut to the same length. Two identical conductors in two differently routed pipes are not guaranteed to be identical paths.
The Habit Underneath the Rule
Strip away the Code language and paralleling comes down to a single discipline: if two conductors are going to share a load, you must make them electrically indistinguishable, because current will exploit any difference you leave behind. It doesn't read the label that says 3/0. It reads impedance, and it divides itself accordingly, and the breaker will never tell you when the division went wrong.
That's a lot to hold in your head at a lug at the end of a long day — the circular-mil match, the length match, the per-phase balance, the ampacity of each leg checked against how the total actually splits. It's exactly the kind of calculation that's easy to fudge when you're tired and the deadline is real, and exactly the kind that produces a callback, or worse, when it's fudged. Voltly keeps the ampacity tables, the conductor sizing, and the derating math in your pocket offline, so you can size each leg of a parallel run and confirm it holds up before you torque a single connection — no signal, no drawer full of tables, just the numbers when the panel is open in front of you. If you'd rather check the wire than trust the breaker to catch your mistake, give it a look.