How a Protective Device Decides to Open

Why this matters

"What amps does it trip at" is a question with no answer, and every nuisance-trip call that ends with a bigger breaker starts there. A protective device does not compare current against a number. It responds to how much heating a current has delivered over how long, through two separate mechanisms that live in completely different time domains. Once you can read that, you can predict before you touch anything whether a legitimate starting transient will survive, which device is the one actually operating, and whether upsizing anything is going to move the failure somewhere with no protection at all.

Before you reset or replace one

A protective device that opened did so because something happened. 29 CFR 1910.334(b)(2) is explicit that after a circuit has been de-energized by a protective device, it may not be manually re-energized until it has been determined that the equipment and circuit can be safely energized. That is a rule, not a habit, and it is the answer to "just reset it and see."

When you do operate a device, do the arc-flash risk assessment first under NFPA 70E-2021, 130.5, keep the enclosure door closed where the equipment allows operation that way, stand to the hinge side rather than square in front of it, and keep your body out of the line of the door. Never hold a breaker handle closed against a fault, and never wedge or defeat a device to keep equipment running.

A device that has interrupted a high fault current may be damaged internally even though it resets, so follow the manufacturer's guidance on replacement after a fault interruption. Replacing a fuse means replacing it with the same class and rating: class markings describe interrupting rating and time-current behavior, and a fuse that physically fits is not the same as a fuse that belongs there.

Any work inside the enclosure is done de-energized, locked and tagged under 29 CFR 1910.333(b)(2), or 29 CFR 1926.417 on construction work, and proved dead with the live-dead-live sequence at NFPA 70E-2021, 120.5. Live current or voltage readings taken to identify which device is operating fall under the troubleshooting exception at 29 CFR 1910.333(a)(1) and need instruments rated for the system under 29 CFR 1910.334(c)(2).

A device does not have a trip point

What a thermal element measures is accumulated heat, and heat is current squared times time. That is why the relationship between current and operating time is inverse: double the current and you deliver four times the heating rate, so the device gets to its operating temperature in roughly a quarter of the time. Every published time-current curve is that relationship drawn out, on log scales in both directions because the useful range spans milliseconds to minutes and single multiples to fifty times rating.

Two things follow immediately. A current with no duration attached cannot be evaluated. And a device that "held at 300 percent" tells you nothing until you say for how long.

Two mechanisms, two jobs

A thermal-magnetic breaker contains two independent trip mechanisms sharing one set of contacts:

• The thermal element is a bimetal strip heated by the load current. It bends as it heats and eventually trips the mechanism. It works in seconds to minutes and it is what responds to overload. It also carries thermal memory: it does not start cold if it has been warmed by recent load.
• The magnetic element is an electromagnet that pulls the mechanism open directly when current is high enough. It works in milliseconds and it is what responds to a short circuit. Below its pickup level it does nothing at all.

Fuses do the same two jobs in one element: a time-delay fuse is built with a mass that carries a starting transient, and a fast-acting fuse is not. Substituting across that line is the same error as fitting the wrong breaker, and the class marking on the fuse body is what tells you which one you are holding.

Motor overload relays are a third family and they are sized to a trip class rather than a curve you read by eye. A Class 20 relay trips within 20 seconds at 600 percent of its current setting; Class 10 and Class 30 are the same statement with different numbers. Match the class to the machine's acceleration time, not to a preference.

The coordination worksheet

Reading a curve is not a browsing exercise. Fill these fields in and the curve answers a specific question three times.

The worksheet filled in

A three-phase motor, nameplate full-load current 22 A, measured starting current 132 A (six times full load) with a measured acceleration time of 6.5 seconds. Overload relay dialed to the motor's nameplate full-load current of 22 A, Class 20, the relay itself selected within the percentage limits Article 430 of the electrical code places on overload protection. Branch device an inverse-time breaker at 60 A, sized under the motor-circuit rules in Article 430 of the electrical code, which permit a branch short-circuit and ground-fault device to be a large multiple of full-load current precisely so it does not see the start. Conductor ampacity after correction: comfortably above the overload setting and below the branch device rating. Available fault current at the equipment: 3,200 A.

Question 1: does the start survive? At 132 A the motor is at 600 percent of the overload setting, and a Class 20 relay must permit that for up to 20 seconds. Acceleration takes 6.5 seconds, which is about a third of the permitted time, so the relay holds with real margin. At the branch device, 132 A is 132 divided by 60, or 2.2 times its rating. Read the published curve for that specific device at 2.2 times rating: the thermal band there is measured in tens of seconds at minimum, far beyond 6.5 seconds. The magnetic element on a device of this type picks up at a much higher multiple, and 2.2 times is nowhere near it - read the actual published curve, because instantaneous pickup varies by design and is adjustable on some devices. Both devices hold. The start is coordinated.

Question 2: what protects the motor from a modest sustained overload? Suppose the machine settles at 26 A continuous, 18 percent above nameplate. At the branch device that is 26 divided by 60, about 0.43 times rating. It will never trip. Not in an hour, not in a year. The overload relay sees 26 against its setting, roughly 1.18 times, and its curve puts that in the minutes range, so it operates. Only the overload device can protect the motor here, and no adjustment to the branch breaker would ever change that.

Question 3: what clears a bolted fault? At 3,200 A the overload relay is useless twice over: its bimetal is far too slow, and its contacts are not rated to interrupt that current at all. The branch device sees 3,200 divided by 60, about 53 times rating, deep in its magnetic element's range, and opens in milliseconds. Only the branch device can do this job.

Three questions, three different answers, two devices. That is what coordination means in practice, and it is why a motor circuit carries separate overload and short-circuit protection rather than one device sized in between. A single device sized to catch an 18 percent overload would open during every start; a single device sized to survive the start could never see an 18 percent overload.

What the worksheet says about upsizing

Now run the common field response through it. The complaint is nuisance tripping, and the reflex is a larger branch device.

If the overload relay is what is operating, replacing the branch breaker changes nothing whatsoever, because the branch device was never involved. That is diagnosable before you buy anything: an overload relay latches and needs a reset, and it is a different physical action than resetting a breaker.

If the branch device is what is operating during starts, at 2.2 times its rating, the worksheet says that should not happen on a device with a normal curve. So either the acceleration time is far longer than 6.5 seconds, which is a machine or load problem covered in a sibling article on locked-rotor conditions, or the device is not the type assumed, or it has been thermally stressed and its band has shifted. Upsizing hides all three.

And upsizing has a cost the worksheet makes explicit. The branch device's rating is bounded not only by the motor rules but by the corrected ampacity of the conductor it protects, and a conductor's thermal limits are set by its installation, which a sibling article covers. Push the device rating above what the conductor supports and the conductor is now protected by nothing. Conductors do not reset, do not announce, and fail inside walls and raceways where nobody sees the process.

What shifts the curve

Three conditions move real operating times away from the published line, and all three explain nuisance trips that look inexplicable.

The curve is a band, not a line. Manufacturing tolerance means the device may operate anywhere within the published band, so a load sitting right at the edge of the band will trip some days and not others. A load that needs the band's optimistic edge to survive is not coordinated, it is lucky.

Ambient temperature is read as load. A thermal element is a heat-operated device and it cannot distinguish heat from its own current from heat in the enclosure around it. Thermal overload relays are commonly available in ambient-compensated and non-compensated forms, and a non-compensated relay in an enclosure much hotter than the motor's own surroundings will trip on the enclosure's heat. Check which type is installed before adjusting a setting to stop nuisance trips, because raising the setting on a relay that is reporting a real ambient problem removes the machine's protection instead of fixing anything.

Thermal memory across repeated events. Each start leaves the thermal element warmer than it was. A machine that starts cleanly cold and trips on the third start in ten minutes is not getting worse each time; the device is starting each attempt from a higher point on its own heating curve. The permitted starts per hour in the equipment documentation exist for exactly this reason.

References

• NFPA 70 (National Electrical Code), Article 430 for motor-circuit overload and branch short-circuit and ground-fault protection, and Article 240 for overcurrent protection generally - use the edition adopted in your jurisdiction
• 29 CFR 1910.334(b)(2) - prohibition on manually re-energizing a circuit after a protective device has opened until it is determined safe to do so; 29 CFR 1910.334(c)(2) for instrument ratings
• 29 CFR 1910.333(a)(1) and (b)(2) - the troubleshooting exception and safe work practices for electrical work; 29 CFR 1926.417 for the construction counterpart
• NFPA 70E-2021, 120.5 (establishing an electrically safe work condition) and 130.5 (arc flash risk assessment)
• Manufacturer published time-current curves and overload relay trip-class data for the specific devices installed
• See related: What a Locked Rotor Condition Actually Is; How Heat and Current Relate in a Conductor; The Difference Between an Overload and a Short Circuit