Reviewed by: Alfonso Cordova, Sales Automation and Michael Ishlove, Technical Manager
Last reviewed: July 2026
Torque specification is one of the most consequential decisions in industrial motor selection and one of the most frequently mishandled. Get it wrong in either direction and you're looking at startup failures, stalled loads, burned windings or a motor that costs far more to run than it should.
The pressure on Canadian engineers to get this right is real. As more facilities upgrade their motor-driven systems, the margin for specification error as it relates to industrial motor performance is shrinking, not growing.
Yet the mistakes keep happening. Under-specified motors stall under load and over-specified motors waste capital and inflate operating costs. High-inertia startups get sized on steady-state figures that have no bearing on what happens in the first few seconds of operation. Duty cycles get assumed rather than calculated.
This guide is for engineers who want to close that gap. At VJ Pamensky (WEG Canada), we'll walk through the torque types that actually drive specification decisions, show you how to calculate required torque for common industrial applications and cover the judgment calls that separate a solid specification from a costly one.
Under-specified torque doesn't just prevent a motor from starting, it also triggers a cascade of mechanical and electrical stress that shortens equipment life across the entire drivetrain.
When an induction motor lacks sufficient starting torque to accelerate the connected load to rated speed, it lingers in the high-slip zone longer than designed. Current draw spikes. Winding temperature climbs. If the motor does eventually reach speed, it arrives already thermally stressed. Repeat that pattern a few times per shift and you've cut insulation life dramatically.
The mechanical side is just as punishing. A motor struggling against an under-torqued startup applies uneven force through the coupling, gearbox and driven shaft. Bearings experience shock loading. Couplings fatigue faster. On conveyor systems with loaded belts, the belt itself can slip or jerk, creating wear patterns that compound over time.
Three failure modes show up most often in under-torqued applications:
Startup stall. The motor draws locked-rotor current but cannot develop enough torque to break the load away from rest. Thermal protection trips within seconds. The root cause is almost always an application where static friction was underestimated, particularly common on conveyors that sit loaded overnight and develop cold-weather stiction.
Speed hunting. The motor starts but oscillates around an unstable operating point, never reaching full-load speed. This is characteristic of a motor whose breakdown torque is too close to the actual load torque. The motor is operating near the peak of its torque-speed curve with no margin left.
Progressive insulation degradation. This one is invisible until it isn't. A motor running near its torque limit draws elevated current, raising winding temperature. Every 10°C rise in winding temperature above the rated insulation class value cuts insulation life roughly in half. A motor that looks fine today may fail unexpectedly in six months.
Conveyors are the most common source of torque specification errors in Canadian industrial facilities and for good reason. The load is rarely constant and the starting conditions are almost never what the nameplate scenario assumes.
A fully loaded belt conveyor at startup can require two to three times the torque of the same system running at steady state. The combination of static friction in the belt, bearing drag at cold temperatures and the inertia of the material already sitting on the belt creates a worst-case starting demand that never appears in a basic torque calculation.
Variable-load conveyors add another layer. If material feed rates fluctuate, as they do in most bulk handling and aggregate applications, the motor must handle peak load torque without stalling, then efficiently run at partial load the rest of the time. A single rated-torque figure won't cover both ends of that range.
The fix: size for the worst-case starting condition, not the average running condition. Calculate starting torque demand based on a fully loaded belt, cold bearings and maximum incline angle. Then verify that breakdown torque provides at least 150% of that peak demand, as recommended by ScienceDirect's engineering reference on pull-out torque thresholds.
High-inertia loads are a well-known challenge, large fans, centrifuges, flywheels and heavy rotating equipment, but the specification error is almost always the same: engineers calculate steady-state torque correctly and ignore the acceleration demand entirely.
The acceleration time formula t = (WR² × Nm) / (308 × Ta) makes the relationship explicit. The exact calculation method will vary depending on the motor data, application and units being used. The key takeaway is that startup conditions and load inertia must be evaluated together because a motor that appears correctly sized during normal operation can still be under-specified during acceleration. Acceleration time is directly proportional to load inertia (WR²) and inversely proportional to the average accelerating torque (Ta). When load inertia is high and the motor's available accelerating torque is modest, the motor spends an extended time in the high-slip zone during startup. That means elevated current and heat for the entire acceleration period, which, on a large fan or centrifuge, can stretch to 20-30 seconds or longer.
Across-the-line starting with high inertia loads can generate 600-700% of full-load current during acceleration. Most standard motors are not designed to absorb that thermal load repeatedly. The solution is either a motor specifically rated for high-inertia duty or a soft starter that limits inrush current while maintaining sufficient torque for controlled acceleration.
An application that cycles 10-15 times per hour may look torque-adequate on paper while systematically destroying the motor in practice. The issue isn't the individual start, it's the accumulated thermal load.
Each startup event generates a burst of heat in the rotor and stator windings. At low cycling rates, the motor dissipates that heat between starts. Push the cycle rate high enough and heat accumulates faster than it dissipates. The motor's thermal rating, typically expressed as a time rating (S1 through S9 duty), defines how much of this it can handle. An S1 motor specified in an S3 or S4 duty application will overheat even if the torque figures look correct at first glance.
The NEMA MG 10-2001 guidelines are clear on this: allow sufficient rest time between starts for heat buildup to dissipate. Where that's not operationally possible, the motor must be derated or upgraded to a duty class matched to the actual cycle profile. Torque adequacy alone is not enough, thermal capacity at the specified duty cycle is the real constraint.
Duty-cycle mismatch is the quietest of the four torque traps because it produces gradual degradation rather than dramatic failure. A motor running at 120% of its rated torque for 30% of each cycle may technically stay within its thermal class or it may not. The answer depends on the specific cycle pattern, ambient temperature, enclosure type and ventilation.
Engineers sometimes treat the service factor as a permanent operating buffer, running a 1.15 SF motor at 115% continuously, assuming the nameplate covers it. It doesn't. Operating at service factor continuously elevates winding temperature and every 10°C increase above the rated insulation class cuts thermal life in half. Service factor is a short-duration margin, not a license for sustained overload.
Understanding the torque-speed curve of an induction motor is the foundation of correct specification. Three points on that curve matter most.
Starting Torque (Locked-Rotor Torque). This is the torque developed at zero speed. The moment the motor attempts to break the load away from rest. For applications where the load is present at startup (loaded conveyors, positive-displacement pumps, screw compressors), starting torque must exceed the breakaway friction of the entire mechanical system. A lower starting torque may be acceptable for centrifugal fans and pumps, where the load at zero speed is near zero. For everything else, starting torque is the number to size against first.
Pull-Up Torque. This is the minimum torque the motor develops while accelerating from zero to the breakdown torque point. On the torque-speed curve, it's the dip between starting torque and peak torque. If pull-up torque drops below the load torque at any point during acceleration, the motor will stall before reaching operating speed. This matters most in high-inertia applications where acceleration takes several seconds, the motor must sustain torque throughout that window.
Breakdown Torque (Pull-Out Torque). This is the maximum torque the motor can develop before it stalls. Pull-out torque must exceed the maximum torque of the driven equipment, typically by at least 150% of full-load torque, to prevent stalling during load transients. In applications with variable loads, peak load torque can occur unpredictably. Breakdown torque is the last line of defense.
From an engineering standpoint, torque is the work a motor must deliver to effectively operate its load. Faults in machine design or improper motor sizing based on actual load requirements create a compounding problem, one that doesn't always surface immediately but becomes a recipe for drivetrain failure over time.
When using drives to operate motors across different speed ranges, two load types drive every sizing decision. A constant torque load does not change as a function of speed. The motor must deliver the same torque whether it's running at 20 Hz or 60 Hz. A variable torque load changes as a function of speed. The torque demand rises and falls with the operating speed of the driven equipment. Understanding which load type your application presents is the starting point for every torque calculation that follows.
For applications operating above base speed, the motor enters a region called constant power. In this range, the motor can no longer produce its nominal torque. Torque begins to drop as speed increases above base speed, because the motor is working beyond its designed operating ceiling. Any application requiring above-base-speed operation must account for this torque reduction explicitly in the sizing calculation. This is one of the most commonly overlooked specification mistakes[j] when retrofitting existing equipment or increasing production throughput without re-evaluating motor requirements.
A practical way to monitor load in real time is through amperage. Current draw directly reflects the torque demand the system is placing on the motor, as load increases, amperage rises. Tracking amperage gives engineers a live indicator of whether the motor is operating within its designed torque envelope or being pushed toward an overload condition.
To avoid overload, the motor must be sized to deliver the required torque not only up to base speed, but also above it, matching the full load profile across the entire operating speed range. A motor sized only for its base-speed torque output will be under-specified the moment the application demands above-base-speed operation.
Centrifugal pumps are among the more forgiving applications from a starting torque perspective. Because centrifugal load torque scales with the square of speed, the starting torque demand is low, typically 15-20% of full-load torque. The motor doesn't need exceptional starting torque; it needs reliable, sustained torque through the acceleration curve.
The calculation starts with shaft power: P = (flow rate × differential head × fluid density × gravitational constant) / pump efficiency. Convert shaft power to torque using T = (9550 × P) / N, where P is power in kilowatts and N is rated speed in RPM. Apply a 1.15-1.25 service factor to the calculated torque to account for pump wear, fluid viscosity variation and minor system inefficiencies.
Positive-displacement pumps are a different story. They present load torque at zero speed and require starting torque that matches or exceeds the pressure resistance in the discharge line. Don't apply centrifugal pump sizing logic to a piston or gear pump application, they are mechanically opposite in their startup behaviour.
Reciprocating compressors load the motor at startup with both inertia and pressure resistance. If the discharge valve is open and line pressure is present, the motor must develop enough starting torque to move the piston against that back-pressure from the first revolution.
Compressors may need more robust pull-up torque to handle variable pressure loads. That translates directly into specification: for reciprocating compressors, verify pull-up torque against the worst-case cylinder pressure at mid-acceleration, not just starting torque at zero speed.
Rotary screw compressors are somewhat more predictable, but they still present a meaningful starting load. Size for at least 130% of full-load torque at startup and confirm that pull-up torque clears the load torque curve throughout acceleration.
The fundamental torque calculation for a belt conveyor uses load force and drive pulley radius: T = F × r, where F is the total belt tension in Newtons and r is the drive pulley radius in metres. Total belt tension includes the weight of the material being conveyed, belt weight, idler friction and incline loading if applicable.
Friction load is the most consistently underestimated variable. Engineers often apply standard friction coefficients from clean, warm, well-lubricated conditions. Real conveyors in Canadian industrial environments, cold warehouses, dusty aggregate facilities, wet food processing lines, operate with friction loads 20-40% higher than textbook values. That gap goes directly into your torque deficit.
For material handling systems with frequent start-stop cycles, calculate the acceleration torque separately: Ta = (J × ?), where J is the total system moment of inertia (motor rotor + all rotating components + reflected load inertia) and ? is the required angular acceleration. Add acceleration torque to running torque to get total required starting torque. Then confirm the motor's starting torque exceeds that sum by a meaningful margin.
This is one of the most common judgment calls in electric motor specification, and the answer depends on the nature of the overload, not the size of it.
Use service factor when the overload is occasional, brief and predictable: A NEMA 1.15 SF motor can handle 15% above its rated horsepower for short durations without immediate damage. The operative words are short and occasional. The SF provision raises operating current by approximately 15%, which increases winding temperature. Running at service factor continuously shortens insulation life substantially. If the overload is a routine operating condition rather than a transient, service factor is the wrong tool.
Upgrade to the next frame size when any of these conditions apply: The motor runs above rated torque for more than 20-25% of its operating time; the application involves frequent cycling that limits heat dissipation; ambient temperature is elevated; or the application is critical enough that reduced motor life is an unacceptable risk.
Frame size upgrades deliver more than additional torque capacity: A larger frame provides greater thermal mass, better heat dissipation surface area and higher breakdown torque, all of which directly address the failure modes described earlier. The first cost is higher, but the lifecycle cost of a properly sized motor in a demanding application is almost always lower than the cost of repeated failures on an undersized one.
The decision framework is straightforward: if you're leaning on a service factor to cover a routine operating condition, you've under-specified the motor. Upgrade the frame.
Torque adequacy and thermal capacity interact in ways that can make a torque-adequate motor fail thermally or a thermally-rated motor fail mechanically.
The clearest example is frequent cycling. An electric motor may have sufficient breakdown torque for its application and still overheat if the starts-per-hour rate exceeds what the thermal design can handle. Each startup event, particularly on high-inertia loads, dumps a significant heat pulse into the rotor bars and stator windings. If the cycle rate doesn't allow sufficient cooling time between those pulses, winding temperature climbs until insulation breaks down.
The correct approach is to start with the NEMA MG 10 starting duty limits for the specific motor design, then calculate whether the application's required cycle rate falls within those limits. If it doesn't, the options are to extend the duty interval (operationally), select a motor with a higher duty class rating or add thermal protection that trips before damage occurs.
The torque-thermal interaction also matters in sustained peak load scenarios. A motor operating near its breakdown torque limit draws substantially elevated current. That current raises the winding temperature. Insulation life halves for every 10°C the operating temperature exceeds the rated class value. An application with frequent load spikes near breakdown torque is thermally degrading the motor between every spike.
The practical guidance: when specifying for applications with variable loads or frequent cycling, calculate both the torque margin and the thermal margin independently, then size the motor to meet the binding constraint, whichever is tighter.
Yes and it's the most common. Load inertia determines acceleration time and acceleration time determines how long the motor spends drawing elevated current during startup. A motor sized for steady-state torque with no inertia analysis is a motor that's been half-specified.
The calculation is not complicated: total system inertia (WR²) = motor rotor inertia + gearbox inertia (reflected) + load inertia (reflected to motor shaft). Acceleration torque = (WR² × ?N) / (308 × t), where ?N is speed change in RPM and t is the required acceleration time. If the required acceleration torque exceeds the motor's available accelerating torque at any point in the speed range, the motor will stall or over-heat before reaching operating speed.
Skipping this calculation is not a conservative choice, it's a gap in the specification.
Friction loads in real industrial environments are consistently higher than standard engineering coefficients suggest. Cold temperatures increase grease viscosity in bearings and gearboxes. Contamination increases belt drag on conveyors. Corrosion increases breakaway resistance in centrifugal pump seals. Each of these adds to the effective torque demand at startup without appearing in a clean calculation.
The fix is straightforward: add 15-25% to your calculated friction load for industrial environments with temperature variation or contamination exposure. For cold-start applications, common in Canadian facilities that operate through winter, model your friction load at the lowest expected ambient temperature, not at room temperature.
Motor sizing is frequently done at steady-state rated conditions because that's where the published data is most accessible. Startup conditions are harder to characterize as they involve transient torque demands, inertia, friction and sometimes, external loading that doesn't exist in the steady-state.
Ignoring startup conditions produces a motor that runs fine once up to speed and fails or stalls every time it starts. In applications with frequent cycling, this means the motor spends a significant fraction of its operating time under the most stressful conditions, exactly where it was never designed to operate.
Other recurring mistakes engineers should verify in every specification:
Torque specification mistakes fall into two camps and both are expensive. Under-specification produces startup failures, stalled motors and premature drivetrain damage, often in the most demanding applications, at the worst possible time. Over-specification inflates capital costs, reduces part-load efficiency and adds system complexity without operational benefit.
The path between those two failure modes is disciplined, application-specific analysis. Calculate starting torque against real worst-case conditions, loaded conveyors, cold bearings, maximum friction. Evaluate breakdown torque against peak load demand, not average running load. Quantify load inertia and model the acceleration torque demand. Assess thermal capacity against the actual duty cycle. Then make the service factor versus frame size decision based on the nature of the load, not the size of the number.
Getting this right the first time is worth the effort. Motors that are correctly specified for their application run longer, fail less often and cost significantly less over their service life than motors that are repeatedly stressed beyond their actual design intent.
If you're specifying industrial motors for Canadian facilities and want to ensure your torque profiles match your actual application demands, our technical team is here to help.Contact VJ Pamensky today to find the right motor for your application.
Reviewed by: Alfonso Cordova, Sales Automation and Michael Ishlove, Technical Manager
Last reviewed: July 2026
Starting torque is the torque an induction motor develops at zero speed, the force available to break the load away from rest. Breakdown torque, also called pull-out torque, is the maximum torque the motor can develop at any point before stalling. Starting torque determines whether the motor can initiate movement; breakdown torque determines whether it can sustain load without stalling during operation or under sudden load spikes.
Common indicators include startup stalls or trips on overload protection, motors that take longer than expected to reach rated speed, abnormally high motor temperatures after each start and premature bearing or coupling failures. If your motor draws elevated current during startup and never fully clears it before the next cycle, the torque margin is likely insufficient for the duty being demanded.
Use service factor for occasional, short-duration overloads where the motor has adequate time to dissipate heat between events. Upgrade to a larger frame size when the overload is a routine operating condition, the cycling rate is high, ambient temperatures are elevated or the application is critical enough that reduced motor life is unacceptable. Running a motor continuously at its service factor rating shortens insulation life and is not an acceptable substitute for proper sizing.
Load inertia determines how long the motor must sustain accelerating torque to bring the load to operating speed. High inertia means a longer acceleration period, which means the motor spends more time drawing elevated startup current and generating heat. A motor sized only for steady-state torque without accounting for inertia may have adequate running performance but fail thermally or stall during every startup event.
Centrifugal pumps have very low starting torque requirements because load torque scales with the square of speed. At zero speed, the load is near zero. Positive-displacement pumps present significant torque resistance at startup because they must move fluid against system pressure from the first revolution. These two pump types require fundamentally different specification approaches and applying centrifugal logic to a positive-displacement application will produce a consistently under-specified motor.
Each startup event generates a heat pulse in the motor's rotor and stator windings. In applications with frequent start-stop cycles, heat accumulates faster than the motor can dissipate it if the cycle rate exceeds the motor's duty class rating. A motor may be torque-adequate for the load while still overheating thermally due to excessive cycling. Both torque margin and thermal capacity must be evaluated independently and the specification must satisfy whichever constraint is tighter.
Cold Canadian winters increase bearing drag and grease viscosity, raising breakaway friction at startup by 20-40% compared to standard conditions. This directly increases starting torque demand, particularly for applications that sit idle overnight or through cold weekends. Engineers specifying motors for Canadian industrial applications should model startup friction loads at the lowest expected ambient temperature, not at standard room temperature conditions.
