A systematic guide for maintenance engineers — diagnosing overheating, bearing failure, winding insulation breakdown, and voltage imbalance with megger testing, vibration analysis, and thermography.
Electric motors are the workhorses of industrial manufacturing — driving pumps, fans, compressors, conveyors, and machine tools. The EASA (Electrical Apparatus Service Association) estimates that more than 60% of all industrial electricity consumption passes through electric motors. When they fail unexpectedly, the downstream impact is immediate: lost production, unplanned downtime, and replacement costs that dwarf what a basic monitoring program would have cost.
The challenge is that most motor failures give advance warning — vibration signatures appear weeks before bearing failure, insulation resistance trends downward over months before winding failure, thermal imaging shows hot spots before breakdown. The plants that catch these signals before failure have measurably lower maintenance costs and higher uptime than those that run to failure. This guide covers the four dominant failure modes, how to test for each one, and what the test results mean for repair decisions.
Industry data from EASA and the Electric Power Research Institute consistently shows the same failure mode distribution across large motor populations:
| Failure Mode | Share of Motor Failures | Primary Cause | Typical Detection Window |
|---|---|---|---|
| Bearing failure | ~41% | Lubrication, misalignment, overloading | Weeks to months with vibration monitoring |
| Winding insulation breakdown | ~37% | Thermal stress, moisture, voltage spikes | Detectable months before failure via megger |
| External / environmental | ~12% | Contamination, moisture ingress, vibration from driven equipment | Visual and temperature monitoring |
| Rotor failure | ~5% | Broken rotor bars (squirrel cage), balance loss | Current signature analysis, vibration |
| Other (shaft, coupling, frame) | ~5% | Mechanical overload, misalignment | Visual inspection, vibration |
Bearings and winding insulation together account for nearly 80% of all electric motor failures — and both are highly detectable with standard maintenance tests. The case for predictive maintenance is direct: the tools required (vibration analyzer, megohmmeter, IR camera) pay back against a single avoided motor replacement in most facilities.
Overheating is the silent accelerator of almost every other motor failure mode. For every 10°C increase above rated winding temperature, insulation life is roughly halved (the Arrhenius rule of thumb for insulation aging). A motor running 20°C above its design temperature runs through its insulation life four times faster than one running at rated temperature.
Infrared thermography is the primary tool for motor thermal assessment. An IR camera scan of operating motors identifies hot spots on the motor frame, shaft bearings, and conduit box that indicate internal thermal problems before they cause failure. During an IR survey:
Bearing failures produce detectable vibration signatures days to months before they progress to catastrophic failure. The progression follows a consistent pattern: early-stage subsurface fatigue produces high-frequency ultrasonic signals; as fatigue advances to surface defects, vibration at bearing defect frequencies becomes measurable; as damage advances, broadband vibration increases and audible noise appears; then catastrophic failure.
Vibration analysis uses characteristic defect frequencies calculated from bearing geometry (BPFO, BPFI, BSF, FTF — outer race, inner race, ball spin, and cage frequencies) to detect bearing damage at specific locations before the overall vibration level rises significantly.
| Vibration Severity (ISO 10816-3) | RMS Velocity (in/sec) | Condition | Recommended Action |
|---|---|---|---|
| Zone A (new) | < 0.10 in/s | Good | Normal monitoring interval |
| Zone B (acceptable) | 0.10 – 0.25 in/s | Acceptable | Monitor; investigate trend if rising |
| Zone C (alarm) | 0.25 – 0.40 in/s | Caution | Reduce monitoring interval; plan maintenance |
| Zone D (danger) | > 0.40 in/s | Danger | Immediate maintenance; risk of imminent failure |
Thresholds for medium-size motors (15–300 kW) per ISO 10816-3. Large machines and machines on flexible mounts use different criteria.
Winding insulation failure is the most catastrophic motor failure mode — typically requiring a full rewind or motor replacement. Unlike bearing failures, which progress gradually, winding failures often complete quickly once insulation integrity drops below a threshold. The key is catching insulation degradation before it reaches breakdown.
A megohmmeter (megger) applies a DC test voltage between the motor winding and frame ground and measures the resulting leakage current to calculate insulation resistance. This is the standard, lowest-cost method for assessing winding insulation health.
1. De-energize motor and disconnect all power leads. Discharge any capacitors in the starter. 2. Short all motor leads together at the winding end. 3. Connect megger lead to shorted winding leads; connect ground lead to motor frame. 4. Apply test voltage (see table below) for 1 minute. Record reading at 1 minute as IR₁. For PI (polarization index), record at 10 minutes as IR₁₀. PI = IR₁₀ / IR₁. 5. Never megger a hot motor — insulation resistance drops significantly with temperature. Test at 40°C or below; apply temperature correction factor if above.
| Motor Voltage Rating | Megger Test Voltage | Minimum Acceptable IR | Investigate If Below | Critical — Do Not Energize |
|---|---|---|---|---|
| Below 1,000 V (LV) | 500 VDC | ≥ 100 MΩ | < 10 MΩ | < 1 MΩ |
| 1,000 V – 2,500 V | 1,000 VDC | ≥ 500 MΩ | < 100 MΩ | < 10 MΩ |
| 2,500 V – 5,000 V (MV) | 2,500 VDC | ≥ 1,000 MΩ (1 GΩ) | < 500 MΩ | < 100 MΩ |
| 5,000 V – 15,000 V (HV) | 5,000 VDC | ≥ 5,000 MΩ | < 1,000 MΩ | < 500 MΩ |
Per IEEE 43-2013 guidelines. Minimum insulation resistance also expressed as: IR ≥ (kV + 1) MΩ for a simplified rule. Polarization Index (PI = IR₁₀/IR₁) should be ≥ 2.0 for Class A/B insulation, ≥ 4.0 for Class F/H. PI below 1.0 indicates contamination or severe degradation.
A single megger reading tells you little. Insulation resistance values vary significantly with temperature, humidity, and winding history. What matters is the trend over time. A motor that read 2,000 MΩ last year and now reads 80 MΩ at the same temperature is telling you something — even if 80 MΩ exceeds the minimum. Record every megger test with date, temperature, and humidity, and plot the trend. A declining trend over 3–4 test intervals warrants increased testing frequency and evaluation for proactive replacement.
Voltage imbalance is a supply-side problem that causes motor-side damage. An imbalance of just 3.5% in three-phase supply voltage produces approximately 25% additional heating in the motor — because the negative-sequence current from voltage imbalance circulates in the rotor creating heat with no useful torque contribution. NEMA derates motor output capacity for voltage imbalance: at 5% imbalance, the motor should be derated to approximately 75% of rated output.
Calculate percent voltage imbalance as: % Imbalance = (Maximum deviation from average / Average voltage) × 100
Example: Three-phase supply measures 478V, 480V, 487V. Average = 481.7V. Maximum deviation = |487 – 481.7| = 5.3V. % Imbalance = 5.3/481.7 × 100 = 1.1% — acceptable. If instead the readings were 470V, 480V, 490V: average = 480V, max deviation = 10V, imbalance = 2.1% — investigate the supply source.
| Symptom | Likely Failure Mode | Primary Diagnostic Test | Corrective Action |
|---|---|---|---|
| Motor trips on thermal overload repeatedly | Overloading, voltage imbalance, or insufficient cooling | Check FLA vs. nameplate; measure voltage imbalance; IR camera scan | Reduce load, correct voltage supply, clean cooling fins |
| High vibration — primarily 1× running speed | Rotor unbalance, bent shaft, or coupling unbalance | Vibration spectrum analysis; separate motor from load to isolate source | Balance rotor; replace bent shaft; balance or replace coupling |
| High vibration — 2× running speed dominant | Shaft misalignment (angular or parallel) | Vibration spectrum; laser alignment check | Precision laser alignment of motor-to-load shaft |
| High vibration at bearing defect frequencies | Bearing defect (inner race, outer race, or rolling element) | Vibration spectrum with BPFO/BPFI overlay; ultrasonic bearing scan | Replace bearings; address root cause (lubrication, alignment, fluting) |
| Insulation resistance declining trend | Winding insulation aging, moisture, or contamination | Megger test (500V–5,000V depending on motor class); polarization index | Dry out motor (bake or space heater); plan rewind if trend continues |
| Low IR at test; motor runs hot | Severe insulation degradation — failure imminent | Megger confirms; IR camera shows hotspot | Remove from service; do not re-energize; rewind or replace |
| Audible growling or grinding noise | Advanced bearing damage; debris contamination | Ultrasonic bearing scan; vibration broadband level | Immediate bearing replacement; check seal condition; investigate contamination source |
| Motor runs slow / doesn't reach full speed | Broken rotor bars, single-phasing, or severe voltage drop | Current signature analysis (MCSA); check all three phases for voltage and current balance | Motor current analysis for rotor faults; check supply connections; check fuses/contactors |
| Ground fault trip | Winding-to-frame short (phase-to-ground fault) | Megger phase-to-ground; insulation resistance near zero confirms | Remove from service; rewind or replace |
The economics of motor rewind versus replacement depend on motor size, age, and failure mode. As a general guideline based on industry practice:
A poorly executed motor rewind can reduce motor efficiency by 1–3% if the rewind uses the wrong wire gauge, changes the number of turns, or uses improper coil pitch. Specify rewinding to EASA rewind standards and request the rewinder's documentation that burnout temperature did not exceed 350°C during stripping. Core losses increase permanently if burnout temperature is exceeded.
ProcessIQ guides you through structured root cause analysis for electric motors, drives, and rotating equipment — based on your specific failure symptoms and operating conditions.
Try AI-Powered Diagnosis Free →Electric motors fail for predictable reasons and give advance warning through predictable signals. The maintenance gap between facilities that catch these signals early and those that run to failure is measured in downtime days and replacement budgets, not just maintenance labor. Megger your windings, trend your vibration, and align your shafts — the tools are not expensive and the data is unambiguous.
Related: Bearing Failure Root Cause Analysis • Compressed Air System Leak Detection • Hydraulic System Troubleshooting • Steam Trap Failure Diagnosis • CNC Machine Alarm Codes • Root Cause Analysis in Manufacturing