Locked-rotor protection and starts-per-hour protection for motors — ANSI 51LR and 66

The motor overload relay trip class guide covers how long a motor overload relay may allow an elevated current, expressed as trip class 10A/10/20/30. This article covers two additional, more specific protection functions that an ordinary trip class setting does not cover: the distinction between a normal start and a locked (stalled) rotor (ANSI 51LR), and the limitation of the number of starts per hour (ANSI 66).

Why trip class alone is not enough

A trip class setting is calibrated for the normal starting time of the motor plus driven load — for example, a trip class 20 allows an inrush current of typically 6× the set current for up to 20 seconds. This works well as long as the motor actually accelerates to operating speed within that time. If the rotor stalls during the start itself — for example due to a jammed drive, a seized pump, or a mechanical fault — the current stays at the full inrush level without speed increasing, and the protection needs to recognise this faster than the normal trip-class time would allow, because heat generation in the winding under a fully stalled rotor is more concentrated than under a motor that is actually accelerating.

ANSI 51LR — locked-rotor protection

The 51LR function (Locked Rotor, time-dependent overcurrent protection) distinguishes a stalled start from a normal start by comparing speed or actual starting time against an expected curve:

On motors with a speed sensor or a current-/voltage-based slip calculation, the protection relay can detect whether the motor is actually accelerating.
If the starting time exceeds a preset threshold without speed following the expected curve, 51LR trips faster than the normal trip-class time would allow — specifically tuned to the fact that a fully stalled rotor loads the winding thermally faster than a motor that is actually coming up to speed.
For motors without a speed sensor, a simplified form is often used: a separate, shorter tripping time that applies specifically when the current remains at the full starting level beyond the normally expected starting time for that specific installation.

Note: the exact detection method (speed sensor, slip calculation, or a simple fixed time limit) varies considerably between relay manufacturers and motor applications — this article covers the principle, not one specific implementation.

ANSI 66 — starts-per-hour protection

Every start of a three-phase motor causes a brief, intense heating of the winding due to the high inrush current (see also the motor starting methods guide for how this inrush current differs per starting method). Between two starts, the winding needs time to cool down. If too many starts occur within too short a time — for example on a control system that repeatedly switches a motor on/off, or an operator who tries again immediately after a failed start — the heat from successive starts builds up faster than the motor can cool, even if no individual start by itself exceeds the trip-class time.

The 66 function limits this by:

Setting a maximum number of starts within a fixed time period (for example, a maximum of 3 starts per hour).
Enforcing a minimum waiting time between two consecutive starts, so the winding has adequate time to cool.
Often distinguishing between a cold start (motor at ambient temperature) and a hot start (motor already warmed by a recent start), with a correspondingly stricter limit for consecutive hot starts.

Note: the exact parameters (number of starts, waiting time, cold/ hot distinction) are not fixed standard values, but are derived from the thermal data of the specific motor (often supplied by the motor manufacturer) — this guide gives the principle of the limitation, not a universal setting value.

Relationship to thermal image protection (ANSI 49)

The thermal image protection guide (ANSI 49) covers a continuous, computational model of winding temperature based on measured current over time. 51LR and 66 are specific additions to this, aimed at two particular situations (a fully stalled start, and a series of too-frequent consecutive starts) that a generic thermal model does not always recognise with the same speed or accuracy — in particular because a stalled rotor produces a different, more concentrated heating pattern in the winding than the normal, more uniform starting profile the thermal model is often calibrated on.

Practical relevance

This is relevant for motors driving loads with a realistic risk of mechanical jamming (conveyors, pumps prone to clogging, fans prone to stalling from debris build-up), and for installations with a control system that switches a motor frequently (for example cyclic processes, or a feedback loop oscillating close to the on/off threshold). When assessing a motor protection scheme, it is important to check whether, in addition to the trip-class setting, a locked-rotor and a starts-per-hour limit are also present, specifically for applications where these two risks are realistic.

Common mistakes

Assuming the trip-class setting covers all scenarios — a fully stalled rotor during the start needs a faster, specific response than the normal trip-class time.
Not applying a starts-per-hour limit on an installation that switches a motor repeatedly, resulting in thermal build-up that no individual start by itself would cause.
Not distinguishing between a cold and a hot start when setting the starts-per-hour limit — a motor already warm from a recent start has less thermal margin left for a subsequent start than a cold motor.
Confusing 51LR/66 with the ordinary trip-class setting of the overload relay — these are additional, not replacement, protection functions aimed at specific failure scenarios.