Electronic vs Thermal Overload Relay: Accuracy, VFD Duty and High-Inertia Loads

Thermal overload relays have protected motors for a century by modelling heat with a bimetallic strip, and they still do it adequately on a direct-on-line starter in a temperate panel. Put the same relay downstream of a variable frequency drive, or on a high-inertia load that takes forty seconds to accelerate, and the model stops matching reality: the relay trips on waveform content it was never designed for, or fails to trip because the motor’s cooling has changed while the relay’s assumption has not.

This guide compares thermal (bimetal) and electronic overload relays on the three factors that decide the choice in practice — trip accuracy, inverter duty, and high-inertia starting — and then sets out where each one belongs. It assumes the motor protection basics are already understood; for the bimetal mechanism itself see how a thermal overload relay works, and for setting trip classes see NEMA class 20 versus IEC class 10.

Quick Answer: What Changes and Why It Matters

The short version: thermal relays model heat physically; electronic relays measure current and calculate heat. The physical model is cheap and robust but it inherits the limitations of the physics — ambient, frequency, waveform. The calculated model is accurate and flexible but needs a measurement circuit, and often an auxiliary supply.

Where the Thermal Relay’s Model Breaks Down

A bimetal relay trips when its strip reaches a temperature, and that temperature comes from two sources: the current through the heater, and the surrounding air. That is the root of every limitation that follows.

• Ambient sensitivity. In a panel at 50 °C, a relay calibrated at 20 °C trips earlier than intended; in a cold outdoor cabinet it trips later. Ambient-compensated versions add a second bimetal to offset this, but the compensation is a curve fit, not a measurement.
• Frequency sensitivity. Heater elements are designed for 50/60 Hz. Above a few hundred hertz, skin effect and proximity effect raise the effective resistance of the heater, so the same RMS current produces more heat and an earlier trip. This is why standard thermal relays are not suitable on the output of a PWM drive.
• Harmonic heating. Even where the fundamental current is at rating, harmonic content adds heating that the bimetal responds to but the nameplate current does not describe. Motors on distorted supplies run hotter than their current reading suggests.
• Fixed trip class. The class is set by the heater and relay combination. Changing from class 10 to class 20 means changing hardware.
• Cooling assumptions. The relay assumes the motor cools as it would at rated speed with its shaft-mounted fan running. At low speed on a drive, that assumption is wrong — sometimes badly wrong.

None of these is a defect. They are the consequences of modelling thermal behaviour with a heated strip, and on a direct-on-line application running near rated speed in a normal panel, they all fall inside the tolerance that IEC 60947-4-1 allows.

What the Standard Actually Requires

It is worth knowing what the trip requirement really is, because it explains why “accuracy” matters less than most buyers assume on DOL duty — and more than they assume on everything else.

That 1.05-to-1.20 window is the legally permitted scatter band. A thermal relay sitting anywhere inside it is compliant, which means two units from the same batch can trip at meaningfully different currents and both be correct. On a motor with generous thermal margin, that scatter is harmless. On a motor running close to its thermal limit, or on a motor whose cooling is compromised, it is the difference between nuisance trips and a burned winding.

An electronic relay tightens that window considerably by measuring current directly. It also lets you separate the trip threshold from the alarm threshold, which is the more useful capability in practice: an alarm at 95% of thermal capacity lets you intervene before the trip, and no bimetal relay can do that.

VFD Duty: Where the Choice Is Made for You

Variable frequency drives change the question in three distinct ways, and the placement of the relay matters as much as its type.

Relay on the drive output

Downstream of a drive, the motor current is a PWM waveform with a high harmonic content and a carrier frequency in the kilohertz range. A bimetal heater sees this as significantly more heating than the fundamental current implies, and the error grows with carrier frequency and cable length. Most manufacturers either prohibit this arrangement or require specially rated heaters. The practical alternatives are:

• Use the drive’s own electronic overload function, which is already a thermal model of the motor and is usually the correct answer.
• Where a separate device is required, use an electronic overload relay specified for inverter duty, with true-RMS measurement and a frequency range that covers the application.

Relay on the drive input

On the supply side, the current is near-sinusoidal but carries the rectifier’s harmonic spectrum. A thermal relay here is less badly affected than on the output, but it still measures heating rather than fundamental current, which matters on a drive with a high total harmonic distortion. On the input side the relay is usually protecting the drive, not the motor — so the more relevant protection is the drive’s own input protection rather than a motor thermal model.

The cooling problem that no relay type solves alone

At reduced speed, a shaft-mounted motor fan delivers far less cooling. The motor can overheat at currents well below its nameplate rating. Neither a bimetal relay nor a basic electronic relay knows the speed, so both are blind to this — the standard remedies are to oversize the motor, fit a separately powered fan, or use a relay that accepts a thermistor or a speed signal. That last option is only available on electronic devices, which is one of the strongest arguments for them on inverter-driven loads.

High-Inertia Loads: Long Starts and the Class Question

A high-inertia load — a large fan, a centrifuge, a loaded conveyor, a ball mill — takes far longer to accelerate than a normal pump. A class 10 relay will trip during a legitimate start. The conventional fixes are class 20 or 30, or a larger relay, both of which are available in thermal form.

Electronic relays do the same job more precisely: the class is selected by DIP switch or parameter, the trip curve is repeatable, and the device can distinguish a genuine long start from a locked rotor using current level rather than elapsed time alone. The practical difference on a start that lasts 30 to 60 seconds is a relay that trips at the intended point every time, rather than one whose trip time drifts with panel temperature.

Where the load is high-inertia そして inverter-driven — a common combination on large fans — the electronic relay plus the drive’s own model is effectively the only arrangement that protects the motor through both the acceleration and the low-speed running phase.

Selection Matrix

Practical Selection Notes

• Power supply. Many electronic relays are self-powered from the load current and behave much like a thermal relay; others need 24 V DC. That single difference decides whether a retrofit is a drop-in or a wiring change.
• Direct or CT-based. Direct-connect versions pass motor current through the device and are limited in frame size; CT-based versions use separate transformers and cover larger motors without heavy current paths.
• Reset behaviour. Manual, remote and automatic reset options differ; on a motor that trips because of a genuine overload, automatic reset produces repeated restarts into a hot motor, so it should be used only where the cause is understood.
• Contactor coordination. The relay’s trip contact is typically in series with the contactor coil, or wired to a PLC input. Which one you choose determines whether the trip stops the motor directly or depends on control logic — the same design decision described in contactor terminal numbering for terminals 95/96.
• Coordination with the short-circuit device. The overload relay does not interrupt fault current; the upstream breaker or fuse does. Type 1 and Type 2 coordination determine what survives a fault, and that is a separate selection from thermal versus electronic.

Where the relay is being chosen as part of a starter rather than as a standalone device, the overall arrangement — contactor, relay and short-circuit protection — is covered in contactor selection for motor starters, and the alternative of a single manual motor starter combining both functions in overload relay versus MPCB.

Six Selection Mistakes

よくある質問

Is an electronic overload relay more accurate than a thermal one?

Yes, in two senses: it measures current directly rather than inferring it from heater temperature, and it is largely immune to panel ambient. Thermal relays are permitted a scatter band between 1.05 and 1.20 times the set current under IEC 60947-4-1; electronic devices hold a much tighter window.

Can I use a thermal overload relay with a VFD?

Not on the drive output with standard heaters. PWM output contains harmonic content that overheats the bimetal heater and causes false trips. Use the drive’s internal electronic protection, or a relay explicitly rated for inverter duty.

Do electronic overload relays need a separate power supply?

Some are self-powered from the load current and need no auxiliary supply; others require 24 V DC for the measurement and communication circuits. Check this before ordering, because it determines whether the device is a drop-in replacement.

Which trip class should I select for a high-inertia load?

Class 20 or 30 for a start that takes longer than a normal DOL acceleration, with the exact class set from the manufacturer’s starting-time data. Electronic relays allow the class to be selected in the field, which makes adjustment after commissioning practical.

Why does my overload relay trip when the panel is hot?

Because a non-compensated thermal relay responds to ambient temperature as well as current. Above roughly 40–45 °C the effect becomes significant. Options are a compensated thermal relay, an electronic relay, or derating the setting — the last is least satisfactory because it reduces protection.

Can one electronic relay replace both the overload and the phase-failure relay?

Usually yes. Most electronic overload relays measure all three phases and provide phase-loss, phase-imbalance and often ground-fault protection, which removes the need for separate phase-failure devices and simplifies the panel.

The Practical Takeaway

Thermal relays are not obsolete; they are simply a physical model with physical limits, and those limits fall outside the application on inverter duty, high ambient, long starts and anything where you want an alarm before a trip. Choose thermal for straightforward DOL duty where the motor has thermal margin and the panel is normal. Choose electronic where the waveform is distorted, the cooling assumption is wrong, the ambient is high, or the trip needs to be visible before it happens — and remember that on a VFD output the drive’s own thermal model is often the correct primary protection. For thermal relays covering the standard DOL range, including the CR2 series from 0.1 A to 630 A, see the CR2 thermal overload relay page, and the AC contactor category for the matching contactors.