Insulation thermal classes and thermal ageing — Montsinger's rule of thumb (10°C halves the life)

The motor derating for altitude and ambient temperature guide (IEC 60034-1) and the polarisation index/DAR insulation test guide both reference the "insulation class" of a winding without explaining that concept itself. This article covers what that class rating precisely means, and why a winding that structurally runs a few degrees too hot does not cause an acute problem, but does cause a slow, predictably shortened service life.

The thermal classes of IEC 60085

IEC 60085 categorises electrical insulation materials based on the maximum continuous operating temperature at which the material achieves an acceptable, standardised service life:

Class	Maximum temperature
Y	90°C
A	105°C
E	120°C
B	130°C
F	155°C
H	180°C
C	>180°C

This temperature is not the ambient temperature, but the total winding temperature — ambient temperature plus the temperature rise that the winding itself produces through its own losses. A motor or transformer with class F insulation (155°C) is therefore designed to reach, at reference ambient temperature and rated load, a winding temperature that stays below the 155°C limit, with a built-in thermal margin that manufacturers often use to apply a higher insulation class than strictly required for the machine's temperature-rise class (see also the motor derating guide).

Note: class F and class H are often used interchangeably as "higher quality" insulation, but they are two separate limit temperatures (155°C and 180°C respectively) — a winding with class F insulation that is operated according to a class H temperature-rise limit ages faster than intended, even though the machine appears to function normally.

Montsinger's rule of thumb

American engineer V.M. Montsinger showed in the 1930s that the ageing of organic insulation materials approximately follows an exponential relationship with temperature, which in practice is summarised as a simple rule of thumb: every 10°C above the insulation class's limit temperature roughly halves the expected insulation life (and, conversely, every 10°C below it roughly doubles that expected life).

Concretely, this means that a winding structurally operated 10°C above its class limit does not age "a bit faster" but roughly twice as fast — a winding with a normal expected life of 20 years at the correct temperature then, in practice, ages to the same point of insulation degradation in roughly 10 years. At 20°C above the limit, this is roughly 5 years, and so on.

Note: Montsinger's rule of thumb is a practical approximation useful for estimation and trend analysis, not an exact physical law — the actual ageing rate depends on the specific insulation system, moisture, vibration, chemical exposure, and other environmental factors besides temperature alone.

Why this does not look like an acute problem

The deceptive part of thermal ageing according to Montsinger is that no acute failure occurs when the limit temperature is exceeded — the motor or transformer simply keeps running, appears to function normally, and the temperature itself does not cause a fault in the short term. Only after months or years of structurally elevated temperature is the insulation so brittle and degraded that an otherwise harmless event — a minor overvoltage, a vibration, moisture ingress — causes the insulation to break down. The underlying cause-and-effect — structurally excessive winding temperature months earlier — is then often not recognised, because the failure itself occurs at a different time and via a different mechanism.

Practical relevance

This explains why a seemingly minor, chronic cause of elevated winding temperature — a blocked cooling channel, an ambient temperature in the switchgear enclosure that is just slightly too high, a fan running slower than specified, harmonic distortion causing extra losses (see the harmonics/THD guide), or simply a motor structurally loaded at the edge of its rating — matters so much more than the temperature itself appears to indicate. Thermographic inspection is precisely intended to signal such structural, chronic temperature deviations early, before the cumulative insulation ageing leads to an actual breakdown. For large motors and transformers, periodic trending of insulation resistance and the polarisation index (see the PI/DAR guide) is an additional way to track the consequences of cumulative thermal ageing, even when temperature itself is not continuously logged.

Common mistakes

Confusing the insulation class with the machine's temperature-rise class — the insulation class is the limit of the material itself; the temperature-rise class (often lower, with a built-in margin) is the design limit of the specific machine.
Assuming a few degrees above the limit temperature is negligible — according to Montsinger's rule of thumb, this has a significant, cumulative effect on expected life, even without an acute fault.
Only reacting to an acute temperature alarm without attention to structurally, chronically elevated operating temperature that does not trigger an alarm but does age the insulation prematurely.
Applying the 10°C rule of thumb as an exact formula for life prediction, while it is a practical approximation that can differ per insulation system and environmental factor.