Thermal Overload Relay Engineering Guide

What is a thermal overload relay? A thermal overload relay is a motor-protection device wired between a contactor and a motor that opens the control circuit when sustained current above the motor's full-load current (FLC) threatens to overheat the windings, per IEC 60947-4-1. Left uncaught, that sustained overload cooks insulation and shortens motor life long before a fuse or breaker sees enough current to react at all. This guide covers bimetallic and electronic construction, the four IEC trip classes, dial setting and star-delta sizing, phase-loss and ambient-temperature behavior, hand vs auto reset, and Type 1/Type 2 coordination with the contactor and short-circuit protective device (SCPD).

What Is a Thermal Overload Relay, and Where It Sits in the Starter

An overload relay carries the full motor current on its way to the motor. It does not switch that current — the contactor does that — it monitors it. Once the current stays above the set point long enough to match the relay's thermal curve, an internal trip bar releases a normally-closed auxiliary contact, usually numbered 95-96, wired into the contactor coil circuit. The coil drops out, the contactor opens, and the motor stops. A separate normally-open contact, 97-98, is available for an alarm or PLC input.

What it does not do is clear a short circuit. A locked rotor or a phase-to-phase fault inside the motor produces current in the hundreds or thousands of amps, and a thermal element sized for tens of amps of running current cannot interrupt that safely or fast enough. That job belongs to the SCPD upstream — a fuse or a motor protection circuit breaker (MPCB). Confusing the two roles is the single most common motor-protection mistake we see: an oversized fuse "for starting margin" leaves the motor with no short-circuit protection at all, while an overload relay set to trip fast enough for a fault would nuisance-trip on every normal start.

A conventional motor starter is three devices, not one, and each covers a failure mode the others do not: the SCPD (fuse or MPCB) clears short circuits fast, the contactor switches normal running current thousands of times over its life, and the overload relay clears sustained overcurrent that stays below the SCPD's own trip threshold. Remove any one leg and a failure mode goes uncovered. A motor with only a contactor and overload relay — no upstream SCPD — has no defined short-circuit rating at all, which is why panel builders publish coordination tables for the full three-device combination, not for the relay alone.

Some builders substitute a single combination device: a motor protection circuit breaker with an integral thermal-magnetic trip replaces the fuse-plus-overload-relay pair in one housing, and a manual motor starter folds the same function into a compact, non-remote-operated package. Both still need a contactor if the application requires remote start/stop; see our motor protection circuit breaker engineering guide and the comparison of MPCB vs manual motor starter for when each makes sense. For the classic three-piece build, our note on MPCB, contactor, and motor starter combinations covers device pairing rules in more depth. This article focuses on the third leg — the thermal overload relay itself, in both its bimetallic and electronic forms.

How a Bimetallic Overload Relay Trips

Inside a bimetallic relay sit three bimetal strips, one per phase, each wrapped with a heater winding or carrying the phase current directly on larger frames. As current flows, the strips heat and bend — two dissimilar metals bonded together expand at different rates, so the strip curves toward the side with the lower expansion coefficient. All three strips push against a common trip bar. When the bend on any strip, or the differential bend between two strips, exceeds the mechanical set point, the trip bar releases and the 95-96 contact opens.

The set dial calibrates this mechanism in amps and is turned to the motor's nameplate FLC. What makes the relay usable in practice is that its response follows an inverse-thermal curve, sometimes written as an I²t characteristic: a small overload, say 110% of FLC, takes minutes to trip, while a large one, 600% or more, trips in single-digit seconds. That curve is deliberately shaped to track how the motor itself heats, which is also why the relay tolerates the motor's own starting inrush — typically 6-8x FLC for a few seconds — without tripping on every start. Set it wrong and the motor cooks; set it too tight and the relay trips on every start.

What we see in the field is confusion between the relay's set point and its trip point — they are not the same thing. The dial marks the current the relay is calibrated to carry indefinitely without tripping; the trip point is wherever the current lands on the inverse-thermal curve above that. A relay dialed correctly to FLC will still ride through a legitimate starting transient because the curve, not the dial number alone, governs the response.

Learn more in our dedicated piece on what a thermal overload relay is and how it works.

Electronic (Solid-State) Overload Relays

Electronic overload relays replace the bimetal strip with current transformers or shunts feeding a microcontroller that runs a thermal model of the motor in software. The practical gain over bimetal is range: an electronic relay commonly covers a 1:3 to 1:4 setting ratio against roughly 1:1.5 for a bimetallic unit, so fewer frame sizes cover the same current spread. Trip class becomes a selectable parameter — 10, 20, or 30 on the same unit — rather than a fixed mechanical characteristic baked into the strip geometry.

The electronics also add protections a bimetal strip physically cannot provide: true phase-loss detection independent of load current, phase-imbalance monitoring, stall and locked-rotor protection with its own timer, and on some models ground-fault detection and a thermistor input for direct winding-temperature sensing. Thermal memory is the other meaningful difference — an electronic relay retains its calculated thermal state through a power cycle, where a bimetal strip cools and resets as soon as current stops, which matters if a motor is stopped and restarted quickly after a near-trip.

Electronic relays are also less sensitive to their own ambient temperature than a bimetal element, because the trip decision is computed rather than mechanically derived from the sensing element's own temperature. That does not remove the need to respect the relay's rated operating temperature range — it just removes ambient drift from the trip-point calculation itself. Compare the two families in more depth in thermal vs electronic overload relays.

Trip Classes: Matching Class to Motor Start Time

IEC 60947-4-1 defines trip class by a single test: apply 7.2x the current setting from cold and measure the trip time. Class 10A trips in 2-10 seconds, Class 10 in 4-10 seconds, Class 20 in 6-20 seconds, and Class 30 in 9-30 seconds. The number is not arbitrary — it is a bound on how long the relay tolerates heavy current before opening, and it must be picked so the motor's actual run-up current stays under the relay's trip curve for the whole start.

Class 10 and 10A cover most standard pumps and fans, where the motor reaches full speed in a second or two. Class 20 and 30 exist for high-inertia loads — large fans, centrifuges, crushers — whose rotating mass takes much longer to accelerate, so the motor draws near-locked-rotor current for a stretch that a Class 10 relay would read as a fault and trip on. Pick the class so the run-up current tracks under the trip curve for the full acceleration time, or the relay nuisance-trips on a healthy start. Full breakdown in overload relay trip classes 10A, 10, 20, and 30.

Setting the Relay: FLC, Service Factor, and the Dial

The dial is set to the motor's nameplate FLC, not to the relay's own maximum range and not to the motor's locked-rotor current. Service factor and duty cycle shift the margin around that number in practice: a motor with a 1.15 service factor can tolerate the dial set slightly above nameplate FLC within the manufacturer's allowance, while a motor on intermittent duty may need a lower setting to account for less time to cool between starts. Neither substitutes for reading the nameplate first.

A relay set below FLC trips under normal running load; one set above FLC leaves the motor exposed to a real overload before the relay reacts. On multi-speed or dual-voltage nameplates, use the FLC value that matches the motor's actual connection and speed in service, not whichever number happens to be listed first. See how to select and set an overload relay for a motor for the full walkthrough, including how to handle a nameplate with more than one FLC value.

Sizing for Star-Delta Starters

In a star-delta (wye-delta) starter, the overload relay is normally installed in the delta leg, not the incoming line, and it sees only the phase current inside the delta winding — roughly 58% of the motor's line FLC, or line FLC divided by the square root of three. Set the dial to that reduced figure, and the relay's own current range needs to cover it; a relay whose minimum setting sits above the delta-leg current cannot be dialed low enough and the wrong frame size gets ordered.

This depends on where in the circuit the relay actually sits — a small number of star-delta designs put the overload relay in the line ahead of the star-delta contactors instead of the delta leg, in which case it sees full line FLC and the divide-by-root-3 step does not apply. Confirm the wiring diagram before dialing the setting; the two topologies look similar on a panel layout but call for different numbers on the dial. Full walkthrough, including contactor sizing, in how to wire a contactor for star-delta motor starting and overload relay sizing for star-delta starters.

Phase-Loss, Ambient Compensation, and Reset Modes

Lose one incoming phase and the motor does not stop — it keeps running on two, and the remaining two windings pick up roughly 1.7x their normal current to make up the torque, overheating fast. A plain bimetal element eventually trips on that extra current, but phase-loss-sensitive relays add a differential mechanism that reacts to the imbalance between phases directly, tripping faster than the thermal element alone would and cutting winding damage from single-phasing. More detail in phase-loss and single-phasing protection in overload relays.

Ambient temperature is the other variable a bimetal strip cannot ignore on its own, since the strip's bend depends on temperature, not just current. Manufacturers add a second, compensating bimetal that offsets the sensing strip's drift, holding the trip point roughly stable across the relay's rated ambient range, typically about -5°C to 55-60°C. That compensation only works if the relay sits in the same thermal environment as the motor's control gear; mount it in the hottest corner of a poorly ventilated enclosure and even a compensated unit reads warmer air as more current than is actually flowing.

Reset mode is a wiring and safety decision as much as a technical one. Hand (manual) reset requires someone to physically press a button before the contactor can re-energize, which is the default for most motors — it stops a faulted motor from restarting itself into the same fault unattended. Auto reset re-closes the contact once the bimetal cools, appropriate only where an unattended restart is genuinely safe, some pump applications being the usual example. A STOP/TEST button, alongside the separate 95-96 (NC) and 97-98 (NO) contact pair, is standard on both bimetallic and electronic relays. See manual vs automatic reset on overload relays for the selection criteria in full.

Coordination and Brand Comparison: Type 1/Type 2, Schneider, ABB, Siemens

IEC 60947-4-1 defines two coordination types for what happens to the starter after a short circuit passes through it. Type 1 permits damage to the contactor and overload relay, provided no one is endangered and the starter can be safely made ready for further service after repair or replacement. Type 2 is the tighter standard: no damage is permitted except light contact welding that a screwdriver easily separates, meaning the same starter is serviceable again after the fault with minimal intervention.

Getting Type 2 in practice is not a matter of picking any fuse or breaker that clears the fault current — manufacturers publish coordination tables that pair specific SCPD ratings with specific contactor and overload relay combinations, tested together to the declared type and short-circuit current rating. Swap one component for an equivalent-looking part from another line and the tested coordination no longer applies, even if every individual device meets its own rating. This is why panel builders keep to a single manufacturer's coordination table per starter rather than mixing SCPD, contactor, and overload relay across brands for a "best price on each" build. See overload relay coordination: Type 1 vs Type 2 and the broader IEC 60947-4-1 standards for contactors and motor starters for the full coordination-table logic, and our contactor selection checklist for pairing the contactor itself.

All three of the major European brands follow the same basic architecture: a bimetallic line that clips directly under their own contactor family, set to FLC, Class 10 or 10A, phase-loss sensitive and ambient compensated, plus an electronic line with a wider setting ratio and selectable class for larger or more demanding motors.

Where the three lines actually differ is frame breakpoints — the top current a bimetal unit reaches before the design moves to a CT-based electronic relay — plus terminal options and how tightly the relay's mounting ecosystem ties to its own contactor range. Siemens leans on the SIRIUS load-feeder concept, pairing 3RU21/3RB3 tightly with 3RT2 contactors and 3RV2 MPCBs in one coordinated system; Schneider's TeSys D range covers the widest single bimetal span up to 630 A; ABB's E-series spreads across more discrete models (EF19 through EF460) than either competitor's electronic line. Eaton's ZB/Z series and other lower-cost brands compete on price, but the three European majors lead on published coordination-table depth, which matters more than headline current range once Type 2 coordination is the requirement. Full brand-by-brand breakdowns: what a thermal overload relay is and how it works covers baseline terminology if any of the above is unfamiliar.

Conclusion

A thermal overload relay does one job — stop a motor from cooking itself on sustained overload — and it does that job correctly only when three things line up: the dial matches the motor's actual FLC (adjusted for star-delta wiring where it applies), the trip class matches how long the load takes to start, and the relay is paired with its contactor and SCPD per a published coordination table rather than assembled from compatible-looking parts. Bimetallic units cover the bulk of standard motors at lower cost; electronic units justify their higher price on wide setting range, selectable class, and protections a bimetal strip cannot provide on its own: phase imbalance, stall timing, ground fault, thermal memory. Get the fundamentals in this guide right first, then work through the linked cluster articles for setting math, trip-class selection, phase-loss behavior, and coordination tables in more depth.