What Happens Mechanically When a Motor Is Oversized

Why this matters

Oversizing a motor is usually discussed as an energy question, and at part load the efficiency penalty on a modern motor is genuinely small. The mechanical consequences are not small, and they are the ones nobody prices. A larger motor delivers more absolute breakaway torque into the same drive train, pushes harder into a jam before it stalls, and, most importantly, carries its overload protection with it: the trip point is set as a percentage of the motor's full-load current, so a motor sized well above the load moves the protection reference off the driven machine entirely. The driven machine loses the cheapest fault detector it had.

Whether that matters depends on one thing, and the rest of this article runs that one thing against two machines that answer it in opposite directions.

Lead with this: clearing a jam is lockout work

The consequences below are mostly about what happens when a driven machine binds. If you are going to a bound or jammed machine, the isolation is not optional and it is not a control-panel stop. Apply your lock and tag to the energy-isolating device under 29 CFR 1910.147 before any part of your body enters the machine, and understand that a jammed drive train is holding wound-up torsional energy in the shafts, belts and couplings, which releases the moment the obstruction frees. Relieve that by hand-barring the drive backwards with the machine locked out, from outside the plane of rotation, rather than by pulling the obstruction free with the train still loaded. Any current or voltage reading at the starter is energized electrical work under 29 CFR 1910.333(b)(2), taken in arc-rated PPE and voltage-rated gloves after an arc-flash risk assessment.

The gate

Does the driven machine have a fault detector that is independent of the motor's overload?

An independent detector is anything that senses a condition on the driven side and acts without reference to motor current: a relief valve, a high-temperature cutout, a flow or pressure switch, a shear pin, a torque-limiting clutch, a vibration cutout, a torque monitor set to the load's own normal band. If one exists and is proven to work, oversizing costs you efficiency, power factor and starting stress. If none exists, oversizing costs you the ability to detect that the driven machine is failing at all.

Run the gate on a positive-displacement pump

A rotary positive-displacement pump with a relief valve set to the system's design pressure, a discharge temperature cutout, and a low-level cutout on the supply side. The correct motor for the duty would run at about 0.85 of its rating. The one that was fitted, because it was on the shelf, is twice that rating, so the same duty sits at about 0.425 of the installed motor's rating.

What is genuinely worse here:

• Power factor. For a general-purpose induction motor, efficiency holds up reasonably from roughly half load to full load and falls away below that, while power factor falls much more steeply across the same range. Read the manufacturer's part-load curve for the actual machine rather than assuming a value; the point is the shape, and at 0.425 of rating you are on the falling part of it.
• Inrush. Locked-rotor current scales with the motor's own rating, so the larger motor pulls a proportionally larger starting current through the same branch circuit and produces a bigger voltage dip on every start.
• Starting torque into the same train. For motors of the same design letter and comparable speed, locked-rotor torque is expressed as a percentage of that motor's own full-load torque, so doubling the rating roughly doubles the absolute breakaway torque applied to the pump's shaft, key and coupling on every start. The percentage varies by design letter and frame, so use the motor's own data rather than a rule; the direction is what matters.
• Physical fit. A larger frame frequently arrives with a different shaft extension and a heavier sheave or coupling hub, changing the overhung load on its own bearings, and it often needs a base adapter. That is where soft foot is created, which the alignment articles in this library cover.

What is not worse: the protection story. If the relief valve is set and proven, an overpressure event is limited by the valve, not by the motor. If the temperature cutout works, a deadhead condition is caught by temperature, not by current. Those detectors do not care what motor is upstream. On this machine, the oversizing is a defect worth correcting on the next replacement and not an emergency.

Run the gate on a gear-reduced rotating machine

Now the same oversized motor on a direct-coupled, gear-reduced rotating machine: a mixer, a screw conveyor, a tower gearbox. No relief path, no independent temperature or torque sensing, nothing that slips. The motor overload is the only device in the system that has any information about the driven machine's condition.

Here the loss is quantifiable, and it is worth doing the arithmetic carefully because the intuitive version of it is wrong.

Where the trip point sits. NEC Article 430 sets running overload protection as a percentage of the motor's full-load current, with 125 percent applying to motors with a marked service factor of 1.15 or greater and 115 percent applying to others, so which number you use is gated by what is stamped on that nameplate. Take a 1.15-service-factor motor, so the overload is set near 1.25 times the installed motor's full-load amps.

Where the running current sits. This is the part people get wrong. Current is not proportional to load, because the magnetizing component of the current is roughly constant regardless of shaft load. On a general-purpose induction motor, a shaft load near 40 percent of rating commonly draws somewhere in the region of 55 to 60 percent of full-load amps rather than 40 percent; the exact value comes off that motor's own current-versus-load curve. Read this one at 0.58 of full-load amps, labelled as a curve reading rather than a calculation.

The detector loss. The overload sits at 1.25 and the machine runs at 0.58, so the demand from the driven machine has to rise by a factor of 1.25 divided by 0.58, about 2.2 times, before the overload has anything to say. With the correctly sized motor at 0.85 of rating, the same curve reads current near 0.88 of full-load amps, because current tracks load much more closely at high load, and the trip point sits at 1.25 divided by 0.88, about 1.4 times.

So the correctly sized installation reports a driven-machine problem when demand rises about 40 percent. The oversized installation stays silent until demand has more than doubled. Both figures rest on part-load current readings from a motor curve, so treat them as the right order of magnitude rather than as settings, and note that the comparison is only fair because both were read off the same kind of curve.

What that buys the failure. A gearbox with a degrading bearing, a conveyor picking up material load, a mixer with product setting up in it, all raise demand gradually. On the correctly sized machine that shows up as a trip while the gearbox is still repairable. On the oversized machine the demand climbs past the point where the gear teeth are being damaged and the overload never operates, and because there is nothing in this train that slips, the torque goes somewhere: the key, the shaft, the coupling, or a gear tooth.

The belt-drive contrast worth knowing. Had this been a belt drive rather than a direct-coupled one, the belts would have slipped and burned as demand rose, which is a bad outcome that is also a loud, cheap and early one. A direct-coupled train has no such release. That difference is a genuine argument for keeping a belt drive on a machine that is prone to binding, and it belongs in the conversation before a direct-drive conversion.

What to do when you cannot right-size today. The gate says the problem is the missing detector, so the fix is a detector, not necessarily a motor. A shear pin or torque-limiting clutch sized to the driven machine's own rating, or a current or torque monitor with its alarm band set from the load's normal running value rather than from the motor's rating, restores the signal the oversized motor removed. Right-sizing on the next failure is still the durable answer.

What would change the call on either machine

• A relief or cutout that has never been proven is not an independent detector, it is an assumption. A relief valve that has not been exercised or verified against its set point puts the positive-displacement pump into the second case, not the first.
• Frequent starting moves the starting-torque penalty from a footnote to the main event. A machine that starts many times per hour applies that doubled breakaway torque many times per hour, and keys and couplings fail from repetition, not from magnitude alone.
• A variable-frequency drive on the motor changes several of these at once: it removes most of the inrush and the fixed starting-torque problem, it gives you a torque or current signal referenced to whatever you configure, and it introduces its own considerations, including that a service factor generally is not available on a motor that is not rated for inverter duty.

How to verify you got this right

• Answer the gate in writing for the machine in front of you. Name the independent detector and say how you know it works. "There is probably a relief valve" is not an answer.
• Compare the running current to the overload setting, not to the nameplate. The ratio between those two numbers is the detector's sensitivity, and it is the single most useful number this article produces.
• Confirm you read part-load current off a curve rather than scaling it linearly from the load fraction. Linear scaling understates current at part load and makes the detector look better than it is.
• Check the mechanical fit consequences physically: shaft extension, sheave or hub weight and position, base fit and foot flatness. An oversized motor that introduced soft foot has created a second, independent problem.
• On any machine that binds, confirm what is intended to give way and that it is the cheap element. If the answer is "the gearbox," the train has no designed weak point and needs one.

References

• NFPA 70 (National Electrical Code) Article 430, motors and motor circuits, for running overload protection percentages and their dependence on marked service factor
• NEMA MG 1, Motors and Generators, for design-letter torque characteristics expressed relative to full-load torque, and for service factor conditions
• 29 CFR 1910.147, control of hazardous energy, for isolation before clearing a jam and for stored torsional energy in a bound drive train
• 29 CFR 1910.333(b)(2), electrically safe work practices for energized measurements; NFPA 70E-2021 for arc-flash risk assessment and PPE
• See related: What a Service Factor Actually Buys You; Why Soft Foot Defeats an Alignment; How to Decide Whether a Drive Is Undersized or Misapplied