What an Overcurrent Device Is Actually Protecting

Why this matters

A tech opens a panel, sees the breaker holding, and reports the circuit healthy. Ten weeks later the motor is dead and the customer wants to know why nobody caught it. The breaker was never watching the motor. It was watching the wire. Most arguments that end in a callback start with an assumption about what a protective device is for, and the assumption is almost always one size too generous: the reader believes the device is a health monitor for everything downstream of it, when it is a thermal limit on one specific thing.

The device is sized to the conductor, not to the machine

An overcurrent device exists to open the circuit before sustained current damages the insulation on the conductors it feeds, and before a fault current does mechanical and thermal damage to the raceway, the terminations and the enclosure. That is the whole job.

The number on the handle is derived from conductor ampacity, which comes from the tables in NEC Article 310 in the edition your authority having jurisdiction has adopted, then corrected for ambient temperature and adjusted for the number of current-carrying conductors bundled together. The NEC is not federal law and binds only through that adoption, so the edition matters every time you quote a rule from it. Nothing in that chain of numbers asks what the load is or how it is doing.

That is why a 20 A branch circuit will happily feed a small appliance drawing 2 A that is quietly destroying itself. Two amps is 10 percent of the device rating. The device has no opinion.

The negative space: six conditions the device is not watching

Naming what a protective device ignores is more useful in the field than naming what it catches, because the things it ignores are the ones the customer is calling you about. Six of them, and all six will produce a running complaint with a breaker that never moves:

• Internal health of the load. A motor with a dragging bearing, a compressor with worn valves, a heater with one open element. Some of these lower current rather than raise it.
• Connection quality. A loose lug or a corroded terminal turns a small resistance into local heat. The heat is at the joint, the current through the circuit barely changes, and a thermal element calibrated in tens of amps cannot see a joint problem measured in watts.
• Leakage to ground below the device rating. Insulation degrading to earth can carry current that is lethal to a person and invisible to a device rated in tens of amps. That is a different device's job, and the sibling article on ground faults versus overloads covers the split.
• Voltage. A motor fed low voltage draws more current to make the same shaft power, and a motor fed high voltage over-fluxes its iron. The device measures current only, so one of those two shows up eventually and the other never does.
• Phase loss or imbalance on a three-phase supply. A motor running single-phased pulls high current on the two surviving conductors and zero on the lost one, and depending on how loaded it was, the two surviving legs can sit under the device rating while the winding cooks.
• Duty cycle. Something that starts every ninety seconds is punishing itself with repeated inrush. Each start is far too short for a thermal element to accumulate against.

Two jobs, and on a motor circuit they are two devices

A general-purpose branch circuit combines both protective jobs in one thermal-magnetic breaker because the load is unknown and could be anything the customer plugs in. A motor circuit deliberately splits them, and the split is the clearest illustration of what an overcurrent device is for.

• Short-circuit and ground-fault protection is the breaker or fuse at the head of the branch circuit. It has to ride through starting current, which for an across-the-line start on a standard induction motor typically runs 5 to 8 times full-load current for a few seconds, with the nameplate code letter as the authority for any specific motor. So it is set well above the conductor's continuous rating.
• Overload protection is a separate element, sized close to the motor's full-load current, that protects the motor windings from sustained modest overcurrent.

The percentages come from the tables in NEC Article 430 in the adopted edition. In broad terms that article permits an inverse-time breaker for motor branch-circuit short-circuit and ground-fault protection at up to 250 percent of full-load current, with a higher figure allowed where a motor genuinely will not start otherwise, while overload protection sits at 115 or 125 percent of full-load current depending on the motor's service factor and temperature rise. Read the tables rather than the percentages here; the point is the ratio between them, not the digits.

The curve is the specification, not the number on the handle

Two elements act inside a common thermal-magnetic breaker, and they answer two different questions.

The thermal element is inverse-time: the further above rating the current sits, the faster it opens, and near rating it may take many minutes or not open at all. The magnetic element is effectively instantaneous and responds to a multiple of rating, so a bolted fault clears in a fraction of a cycle. State both ends or you have not stated a direction: at 1.2 times rating you are in minutes and the thermal element is doing the work; at 10 times rating you are in milliseconds and the magnetic element is.

That is why "it held" tells you only that current stayed inside the curve for that duration. It is a boundary, not a measurement.

Worked example: the circuit where every device was right and the motor still failed

A three-phase motor with a nameplate full-load current of 28.0 A and a service factor of 1.15. Sizing it off the general section above:

• Conductors sized at 125 percent of full-load current, so a minimum corrected and adjusted ampacity of 35.0 A.
• Overload protection at 125 percent of full-load current, permitted by Article 430 for a service factor of 1.15 or higher, so 35.0 A.
• Branch-circuit short-circuit and ground-fault protection with an inverse-time breaker at 250 percent, so 70.0 A, which is a standard rating and needs no rounding up.

Note what that pairing means before any fault exists: a 70.0 A breaker is sitting at the head of a conductor rated 35.0 A, which is exactly double. That is legal and correct, and it is only correct because the 35.0 A overload element is doing the conductor's overload job. Strip the overload device out and the conductor has no overload protection at all.

The service call: intermittent thermal trips on the overload after long runs, nothing found. Measured running current, taken with a clamp meter under the live-work gate described below, reads 32.0 A on all three legs, balanced.

Run that against the numbers already fixed above rather than against instinct:

• 32.0 A against the 70.0 A breaker is 46 percent of rating. The breaker will never operate on this and is not a data source.
• 32.0 A against the 35.0 A overload setting is 91 percent. The overload will not open here either, and it is right not to.
• 32.0 A against the 28.0 A nameplate is 114 percent, which sits just inside the 1.15 service factor, or 32.2 A.

Every device in the circuit is behaving exactly as designed, and the motor is still in trouble. Service factor is an allowance for occasional operation above nameplate at rated voltage and frequency, not a continuous rating, and continuous operation up in that band runs the winding hotter than the insulation system was rated for. The common rule of thumb for organic insulation systems is that each roughly 10 degree C rise above the rated hot-spot temperature halves expected insulation life; it is a rule of thumb derived for insulation aging, not a calculation you can run on a specific machine, and the manufacturer's own thermal data governs if you have it.

So the finding is not "the overload is nuisance tripping." The finding is that the motor is loaded to 114 percent of nameplate and something is asking it for more torque than it was sized for, or the driven load has changed. The overload trips on long runs because heat accumulates over a run, which is precisely what an inverse-time element is built to integrate.

The failure mode if you read this wrong: the tech concludes the overload is oversensitive and moves it up, or replaces a correctly operating protective device, and the winding fails within a season with a new part number on the last invoice. Establishing why a protective device operated has to come before treating it as the defect.

Taking the reading without becoming the fault

Everything above ends in someone putting a meter in an enclosure. 29 CFR 1910.333(a)(1) requires that conductors and parts be de-energized before an employee works on or near them unless the employer can demonstrate that de-energizing introduces additional or increased hazards or is infeasible due to equipment design or operational limitations, and a running-current reading is the classic infeasible case because the measurement does not exist de-energized. That demonstration is the employer's to make in writing, not the tech's to assume on site.

Where the live reading is justified, the approach boundary and the arc-rated PPE come from the risk assessment in NFPA 70E-2021 130.5 and the PPE requirements at 130.7, in the edition your employer's electrical safety program has adopted, since NFPA 70E binds through that program or through a contract rather than on its own. Clamping an insulated conductor still means an open enclosure and full exposure to an arc-flash event at the bus. For the isolation half of the work, note the fork: 29 CFR 1910.147 excludes exposure to electrical hazards from work on conductors and equipment in electric utilization installations at (a)(1)(ii)(C), so panel and branch-circuit lockout runs under 29 CFR 1910.333(b)(2) in general industry and 29 CFR 1926.417 on construction work, and the live-dead-live proving sequence is NFPA 70E-2021 120.5.

How to verify you got this right

Three checks, each of which catches a different wrong conclusion:

1. Compare measured current to the nameplate, then separately to the device rating. Two ratios, not one. If you only ever compute current against the breaker, you will never find the case above, because the answer is always a comfortable percentage.
2. Confirm the conductor's protection actually exists somewhere. On a motor circuit, find the overload element and read its setting. If nobody can show you one, the conductor is protected only at the branch device rating, which on the example above would be double its ampacity.
3. Ask what the device would have to see to operate, and whether that condition is physically possible here. A 70.0 A device on a load that peaks at 32.0 A running and roughly 5 to 8 times 28.0 A for a few seconds at start will only ever operate on a fault. If you are waiting for it to warn you about anything else, you are waiting on an instrument that is not connected to the question.

References

• 29 CFR 1910.333(a)(1) and (b)(2), general industry safety-related work practices for electrical work; 29 CFR 1926.417 for the construction counterpart
• 29 CFR 1910.147(a)(1)(ii)(C), the carve-out sending electric utilization equipment to Subpart S
• NFPA 70E-2021, 120.5 (live-dead-live verification), 130.5 and 130.7 (risk assessment and PPE), as adopted by your employer's electrical safety program
• NEC Articles 240, 310 and 430, in the edition your authority having jurisdiction has adopted
• See related: What a Ground Fault Is, as Distinct From an Overload; Why Inrush Current Trips Things That Should Hold