Walk up to a motor disconnect on almost any commercial job and read the numbers. The breaker says 70 amps. The wire leaving it is 10 AWG copper — a conductor the ampacity table rates at 35 amps, and one that most electricians carry in their heads as "30 amps, period." By every rule you learned in your first year, that installation looks like a violation. The breaker is twice the size of the wire it is supposed to protect.
It isn't a violation. It's Article 430 of the National Electrical Code, and it may be the most misunderstood article in the book — because it quietly abandons the premise every other circuit is built on.
The Rule Motors Break
For an ordinary branch circuit, protection is a one-device job. The breaker is sized at or below the conductor's ampacity, and it guards against both kinds of trouble: the slow kind, where a circuit runs overloaded for minutes and the insulation cooks, and the fast kind, where a fault dumps hundreds or thousands of amps in a heartbeat. One handle, both threats. That's NEC 240.4 in a sentence, and it works beautifully for receptacles, lighting, and heat.
It fails for motors, and the reason is physics, not paperwork.
A Motor Starts Like a Short Circuit
At the instant you energize an induction motor, the rotor isn't turning. Electrically, a stalled motor is a transformer with its secondary shorted: there is no counter-EMF pushing back against the supply, so the only thing limiting current is the raw impedance of the windings. The result is locked-rotor current — typically six to eight times the motor's full-load amps, stamped on the nameplate as a NEMA code letter most people never read.
As the rotor accelerates, it begins generating counter-EMF, and the current falls off toward the running value. On a small motor with an easy load, that takes a fraction of a second. On a big motor spinning up a loaded conveyor or a flywheel, it can take several seconds — seconds at six times rated current.
Now imagine protecting that circuit the ordinary way. Our 5-horsepower, 230-volt motor draws 28 amps running, so you install a 30- or 35-amp breaker to match the wire. Every start, that breaker sees roughly 170 amps of inrush. Its magnetic trip element — the fast one, calibrated somewhere between five and ten times the handle rating — sits right in the path of that surge. The motor trips the breaker before it ever reaches speed, not because anything is wrong, but because a healthy motor start is indistinguishable, for a few cycles, from a fault.
You could keep upsizing the breaker until the starts stop tripping it. But now you've created the opposite problem: a breaker big enough to ignore inrush is far too big to notice a sustained overload. A motor grinding against a jammed load might pull 40 amps indefinitely — enough to destroy its windings in minutes, and nowhere near enough to move a 70-amp breaker. One device cannot be both patient enough for starting and vigilant enough for overload. So the Code stopped asking it to be.
Two Devices, Two Jobs
Article 430 splits protection down the middle.
The breaker or fuse handles short circuits and ground faults only. Under 430.52, an inverse-time breaker can be sized up to 250 percent of the motor's full-load current — with permission to round up to the next standard size, and to go as high as 400 percent on smaller motors if a stubborn load still won't start. Time-delay fuses get 175 percent. These numbers look reckless until you understand what the device is no longer responsible for: it exists to clear the huge, instant currents of a genuine fault, and to stay out of the way of everything else.
The overload relay, mounted in the motor starter, handles the slow kind of trouble. Under 430.32, it's set from the motor's nameplate current — 125 percent for motors with a service factor of 1.15 or more, 115 percent for the rest. Whether it's the old melting-alloy type, a bimetal element, or an electronic sensor, an overload relay is built with a deliberate thermal character: it rides through six-times current for the seconds a start requires, because its element heats at roughly the same rate the motor's windings do, then trips firmly if current sits even 25 percent high for too long. It is, in effect, a thermal model of the motor bolted next to it.
One device is coarse and fast. The other is precise and slow. Together they cover what neither could alone.
Sizing One Circuit, Step by Step
Take that 5-horsepower, 230-volt, single-phase motor and walk the Code's path.
Start with the current — and not from the nameplate. Section 430.6(A) says conductors and breakers are sized from the NEC's own tables, which for this motor (Table 430.248) give 28 amps. The tables are deliberately standardized so that when the motor burns out in ten years and its replacement has a slightly different nameplate, the wiring doesn't have to change.
Conductors, per 430.22, get 125 percent of that table value: 28 × 1.25 = 35 amps. That's 10 AWG copper in the 75°C column. And here's a detail worth savoring: the familiar small-conductor rule that caps 10 AWG at 30 amps — 240.4(D) — explicitly does not apply, because 240.4(G) exempts motor circuits and hands the decision to Article 430. The 30-amp ceiling you memorized was never absolute; it was a default with a trapdoor.
The breaker, per 430.52, may be up to 28 × 2.5 = 70 amps for an inverse-time type. Seventy is a standard rating, so a 70-amp breaker it is.
The overloads come last, and they alone use the nameplate — say it reads 26 amps with a 1.15 service factor. Maximum setting: 26 × 1.25 = 32.5 amps.
Final circuit: a 70-amp breaker, 10 AWG wire, overloads at 32 amps. Legal, inspectable, and genuinely safer than the intuitive version.
Why the Wire Is Still Safe
It helps to notice that the 70-amp handle rating never actually matters in the life of this circuit. Normal current is 28 amps. A sustained overload gets cleared by the relay at 32 amps — comfortably under the conductor's 35-amp rating. A short circuit or ground fault produces hundreds or thousands of amps, which slams the breaker's instantaneous element open regardless of what the handle says. There is no realistic current that harms the wire without one of the two devices clearing it first. The wire isn't unprotected; it's protected by committee.
The confusion, when it comes, is usually a schema problem rather than a knowledge problem. "Breaker protects wire" is the first rule the trade teaches, and rules learned first get applied everywhere — psychologists call the habit overgeneralization, and it's the same reflex that makes a child say "goed" instead of "went." The fix isn't memorizing an exception. It's understanding that motors face two different threats on two different timescales, and the Code assigned each one its own specialist.
The Two Mistakes to Watch For
Both common errors are nameplate errors in disguise. Sizing conductors from the nameplate instead of the table can leave you a size small when a replacement motor draws more. Sizing overloads from the table instead of the nameplate defeats the whole point — the relay is supposed to model this motor's thermal limits, not a statistical average. Table for wire and breaker; nameplate for overloads. That single sentence prevents most of the callbacks Article 430 generates.
None of this arithmetic is hard — multiply by 1.25 here, 2.5 there. What burns time in the field is the lookups: which FLC table for single-phase versus three-phase, which percentage for which protective device, which standard breaker size comes next, which ampacity column applies. That's exactly the drawer of dog-eared reference cards Voltly was built to replace — full-load current tables, ampacity, conduit fill, voltage drop, and box fill in one calculator that works in a basement mechanical room with zero bars of signal. If you'd rather spend your morning landing conductors than flipping pages, it's at voltly.lumenlabs.works.