ServicesHow We WorkSectorsAcademyLibraryAboutContactOpen the Toolbox →
Rotating Equipment · Motors · Overheating

Motor winding overheating: why motors die of heat

A motor almost never wears out — it cooks. The dominant failure mode of an electric motor is insulation breakdown, and the dominant driver of that breakdown is temperature. Get the heat right and the winding lasts decades; run it 10 °C too hot and you halve its life, quietly, every hour it runs. Here is the physics of thermal ageing, what pushes a winding hot, and how to catch it at P instead of F.

IEC 60034-1IEC 60085NEMA MG-1IEEE 43ISO 14224 · OHE
⚡ TL;DR

Heat is the enemy. Roughly a third to a half of motor failures are insulation failures, and temperature is what ages insulation. Overheating is the mechanism that degrades a winding until a turn-to-turn short, phase fault or ground fault finally trips or burns out the motor.

The number to remember is the 10 °C rule: insulation life roughly halves for every ~10 °C above its rated hotspot — and doubles for every 10 °C below. That single exponential governs how long a motor lives.

Because the damage is a smooth function of a signal you can measure, overheating is a prime target for predictive maintenance: trend the winding temperature, project it to the insulation limit, and act inside the P–F window — before the thermal trip.

1 · The 10 °C rule: why heat is destiny

Winding insulation is an organic/polymeric system, and like all such materials it degrades chemically at a rate that rises exponentially with temperature. The governing physics is the Arrhenius equation — reaction rate ∝ e−Ea/kT — which in the motor world becomes a rule every reliability engineer should carry:

Montsinger’s 10 °C rule. Insulation life roughly halves for every ~8–10 °C the winding runs above its rated hotspot temperature, and doubles for every 10 °C below it.

The consequences are stark — and this is why a motor that runs “a little warm” for years can fail years early with no dramatic event. The damage is cumulative, silent, and exponential. Set the operating temperature below and watch the life multiplier move:

Interactive — Insulation life & the 10 °C rule

Live model

Set how far the winding hotspot runs above (or below) its rating, and the design life you’d expect at rated temperature. The 10 °C rule converts that into a life multiplier and the effective years you’ll actually get.

How far above (or below) the rated hotspot the winding actually runs
Expected insulation life if operated exactly at its rated temperature
Life multiplier
×
vs. life at rating
Effective life
yr
at this temperature
Hotspot vs. rating
°C
+ = over the limit
Verdict
 
Relative insulation life vs. temperature
Life = 2−ΔT/10 · solid dot = your operating point
Life curveRated (1×)< ½ life (rapid ageing)
Model: the Montsinger approximation, life multiplier = 2−ΔT/10. Real thermal-ageing curves come from the Arrhenius chemistry of the specific insulation system (IEC 60216) and vary in slope, but the halving-per-10 °C rule is the field-usable form — and the direction is never in doubt: heat costs life, exponentially.

2 · Insulation classes & the hidden margin

Insulation systems are graded by the maximum hotspot temperature they tolerate for a normal service life. Under IEC 60085 (mirrored by NEMA MG-1):

ClassMax hotspot °CTypical use
A105Legacy / small machines
E120Older general-purpose
B130Common rating point
F155Modern industrial standard
H180Severe-duty / high-temperature

That hotspot budget is spent in three parts: ambient (rated at 40 °C by IEC 60034-1) + average temperature rise under load + a hotspot allowance for the hottest spot above the average.

The reliability engineer’s favourite detail: most modern motors are wound with Class F insulation but rated (“used”) at Class B temperature rise — a deliberate ~25 °C thermal reserve. Run such a motor at its Class B rating and, by the 10 °C rule, you’re banking roughly a 4× life margin. Let the cooling foul or the ambient climb, and you quietly spend it.

3 · Rating & temperature rise

A nameplate rating assumes 40 °C ambient and ≤1000 m altitude (IEC 60034-1). Two field corrections matter:

Service factor (SF) is a short-term overload allowance (SF 1.15 permits 15% overload) — but running continuously in the service-factor band uses thermal margin and shortens life. It’s headroom for transients, not a licence for steady overload.

4 · What makes a winding run hot

Overheating is a symptom; the discipline is finding the cause. Group them:

Electrical

Cooling & environment

Duty & mechanical

5 · The P–F curve for overheating

Thermal degradation is a textbook P–F story. Somewhere on the decline the condition becomes detectable — that is P, the potential failure. Left alone it proceeds to F, functional failure: insulation breakdown → turn-to-turn short → phase or ground fault → trip or burnout.

The P–F interval — the usable maintenance window — depends entirely on the mechanism and the detection technique:

The lesson mirrors the rest of condition monitoring: the earlier your detection sits on the curve, the larger your window. A thermal-trip relay detects at F (lead time ≈ 0). A temperature trend detects at P — and buys you the whole interval.

6 · Detecting it early

TechniqueWhat it seesOn the curve
Embedded RTDs / thermistors (PT100, PTC)Actual winding temperature — the primary signalContinuous; the anchor for trending
Temperature trending & rate-of-riseA slow climb, or rising temp at constant loadEarliest P — ideal for prediction
Infrared thermographyExternal hotspots: terminals, connections, bearing housingsPeriodic P
MCSARotor/thermal-related electrical faultsNon-intrusive P
Insulation Resistance & PI (IEEE 43)Cumulative insulation health, moisture, contaminationOffline health trend
Voltage-unbalance / current monitoringThe upstream cause before temperature even movesLeading indicator

The highest-value signal is winding temperature at constant load. A motor whose winding runs hotter this month than last — at the same load and ambient — is telling you its cooling or its electrical health is degrading, long before any absolute limit is reached.

7 · The predictive play: catch it at P, project to F

This is where continuous monitoring changes the economics. Instead of waiting for the thermal trip:

  1. Stream winding temperature from embedded RTDs.
  2. Trend it — the rate of rise, normalised for load and ambient.
  3. Project the trend to the insulation limit to estimate the remaining useful life — the projected P–F window.
  4. Act within the window — schedule the cleaning, the unbalance fix, the load review — as planned work, not a 3 a.m. breakdown.

Crucially this is deterministic and explainable: the “days to limit” number is a straight-line extrapolation of a slope you can see on a chart, not a black-box score. A rising winding-temperature trend that projects to cross the Class-F limit in, say, eight days is a work order you can justify — and defend — today. (See Predictive Maintenance & RUL.)

Bluestream Predictive Maintenance

Turn a winding-temperature trend into a dated work order

Bluestream’s predictive-maintenance platform does exactly what this article describes: it streams temperature (and vibration, current and more) from the asset, trends it, projects the time-to-limit, and raises a work order — distinguishing a predicted alert at P from a measured alarm at F, and showing the P–F window you have to act in.

Overheating is the ideal first case: a slow, trendable signal with a long P–F interval and an expensive failure. Catch it at P.

Talk to us about predictive maintenance →

8 · Remediation & prevention

Cheapest and highest-impact first:

Key takeaways

Across the Academy