Motor Insulation Testing: The Complete Field Guide

Motors fail. The question is whether you catch it during a scheduled outage or at 2 AM on a Sunday. Insulation testing is the tool that decides which one.

This guide covers everything you need to run a complete motor insulation testing program — from a single 5 kW pump motor to a 2,000 kW medium-voltage drive. It’s not a basic how-to (we have that). It’s the full program: which tests to run when, how to interpret results against IEEE 43-2013, how to tell surface contamination from insulation failure, and how to decide when a motor needs rewinding.

All values come from IEEE 43-2013, cross-referenced with IEEE 95-2002 and industry practice.

Why Motor Insulation Is Different

Motor insulation takes a beating that most other electrical insulation doesn’t.

Thermal cycling — A motor running intermittently heats and cools repeatedly. Thermal expansion and contraction of the windings against the slot liners creates mechanical stress on the insulation.

Vibration — Unlike a transformer or a cable in a tray, a motor’s windings physically move during operation. Coil movement against the stator slots abrades the insulation over time.

Dielectric stress — Operating voltage continuously stresses the insulation. For medium-voltage motors, this stress is significant. For VFD-fed motors, the stress is even higher due to fast-rising voltage pulses.

Environmental exposure — Motors often operate in harsh environments: humid pump rooms, dusty cement plants, oil-soaked machine shops. Contamination accumulates on the windings over years.

Moisture absorption — Motors that sit idle for weeks or months absorb moisture. This is the #1 cause of insulation failure on spare motors and equipment after plant shutdowns.

The result: motor insulation degrades in ways that cables and transformers don’t. The testing program needs to account for all of these failure modes.

Form-Wound vs Random-Wound: Why It Matters

This distinction drives almost every interpretation in IEEE 43-2013. You need to know which type you’re testing before you can evaluate the results.

Random-wound motors

Round magnet wire dropped into the stator slots in a random pattern. Insulation is primarily the enamel coating on each individual wire, plus slot liners between the coil and the iron core.

Typical applications: Motors below 200 HP, motors rated below 1 kV, most industrial process motors (pumps, fans, compressors).

Key characteristic: The insulation volume is small. Absorption current decays quickly — often within 1–2 minutes. This means the PI test may not give meaningful results on random-wound motors.

Form-wound motors

Rectangular conductors pre-formed into coils, wrapped with mica tape and epoxy resin, then fitted into the stator slots. Each coil is individually insulated with multiple tape layers.

Typical applications: Motors above 200 HP, all motors rated above 2,300 V, most medium-voltage motors used in utilities, mining, and heavy industry.

Key characteristic: Much larger insulation volume. Absorption current decays slowly — takes 10 minutes or more. The PI test gives meaningful results on form-wound motors.

How to tell which one you have

Check the motor nameplate. If the motor is:

• Rated above 2,300 V → almost certainly form-wound
• Rated 460 V or lower and below 200 HP → almost certainly random-wound
• In between → check the manufacturer’s documentation, or assume random-wound as the conservative default

Why IEEE 43-2013 treats them differently

IEEE 43-2013 Clause 12.1 sets different minimum IR values for each type:

Motor Type	Minimum IR (at 40°C)
Form-wound stators, built after 1970	100 MΩ
Random-wound stators, form-wound <1 kV	5 MΩ
DC armatures, pre-1970 windings	(kV + 1) MΩ

The PI test is considered reliable for form-wound motors and less reliable for random-wound motors (where absorption current decays so fast that the 10-minute reading often equals the 1-minute reading, giving a PI near 1.0 even on healthy insulation).

The Four Tests You Should Run

A complete motor insulation program uses four tests, each telling you something different.

1. Insulation resistance (spot reading)

Apply test voltage for 60 seconds. Record the reading. Quick, simple, and reveals gross failures.

Use it: Every time. It’s the baseline test. Limitation: Single number, heavily temperature-dependent.

2. Polarization index (PI)

Apply test voltage for 10 minutes. Calculate PI = IR at 10 min ÷ IR at 1 min.

Use it: On form-wound motors, critical equipment, motors after water damage. Limitation: Not reliable on random-wound motors. Not meaningful when IR > 5 GΩ.

3. Dielectric absorption ratio (DAR)

Apply test voltage for 60 seconds. Calculate DAR = IR at 60s ÷ IR at 30s.

Use it: On random-wound motors (where PI doesn’t work well), quick routine screening. Limitation: Shorter time window gives less diagnostic depth than PI.

4. Step voltage test

Apply increasing test voltages in steps, recording IR at each step.

Use it: On aging motors, after overvoltage events, when spot readings are declining. Limitation: Requires multiple voltage levels. Not needed on new or clearly healthy motors.

A standard testing program: run IR and DAR (or PI for form-wound) at every test. Run step voltage annually on critical motors.

Test Voltages by Motor Rating (IEEE 43-2013 Table 1)

Motor Winding Rating	DC Test Voltage
< 1,000 V	500 V
1,000 V – 2,500 V	500 – 1,000 V
2,501 V – 5,000 V	1,000 – 2,500 V
5,001 V – 12,000 V	2,500 – 5,000 V
> 12,000 V	5,000 – 10,000 V

Choosing within the range

For each band, the IEEE standard gives a range. Use the lower end for:

• Older motors with aged insulation
• First-time testing of an unknown-condition motor
• Routine maintenance testing

Use the upper end for:

• New motor acceptance testing
• Motors with a history of healthy readings
• Step voltage tests

Common practice: Use 500 V DC for motors up to 480 V. Use 1,000 V DC for 2,300 V motors. Use 2,500 V DC for 4,160 V motors. Use 5,000 V DC for medium-voltage motors above 6.9 kV.

Minimum Acceptable Values

Insulation resistance minimums (IEEE 43-2013, Clause 12.1)

Motor Type	Minimum IR at 40°C
Random-wound stators, form-wound <1 kV rated	5 MΩ
Form-wound AC stators (after 1970), DC armatures	100 MΩ
Windings built before 1970	(kV + 1) MΩ

Example for a 460 V motor (modern, random-wound): minimum = 5 MΩ at 40°C.

Example for a 6.9 kV motor (form-wound): minimum = 100 MΩ at 40°C.

Polarization index minimums (IEEE 43-2013, Table 3)

Insulation Class	Minimum PI
Class A (105°C)	1.5
Class B (130°C)	2.0
Class F (155°C)	2.0
Class H (180°C)	2.0

Most modern motors are Class F. Minimum PI = 2.0.

Critical exception: when PI is invalid

IEEE 43-2013 Clause 12.2.2: If the 1-minute IR exceeds 5,000 MΩ (5 GΩ), the PI may not be meaningful. The leakage currents at that resistance level are in the nanoampere range, where small variations in voltage supply, humidity, or test connections greatly affect the calculated ratio.

What to do when PI is invalid: Trust the absolute IR value. A motor reading 8 GΩ with a PI of 0.9 is in excellent condition. The low PI is measurement noise, not insulation condition.

When PI doesn’t apply (random-wound motors)

On random-wound motors, absorption current often decays to near-zero within 1–2 minutes. The 10-minute reading equals the 1-minute reading, giving a PI of 1.0 — even on healthy insulation. For random-wound motors, use the DAR or the absolute IR value instead of PI.

The Complete Test Procedure

Before the test

1. De-energize the motor and verify with a voltage tester
2. Apply lockout/tagout to the motor starter and the disconnect
3. Disconnect the motor from its feed cables (at the motor terminal box)
4. Disconnect from any VFD or soft starter — physically remove the cable lugs
5. Record the conditions: ambient temperature, winding temperature (if available from RTDs), humidity, motor history
6. Short all terminals (T1, T2, T3) together and to ground for at least 60 seconds to discharge

Test connections

For a three-phase motor tested winding-to-ground (all phases together):

• LINE lead → all three phase terminals connected together
• EARTH lead → motor frame / ground terminal
• GUARD lead → not used for this test

For phase-by-phase testing (preferred by IEEE 43-2013):

• LINE lead → phase under test (e.g., T1)
• EARTH lead → motor frame
• GUARD lead → other two phases (T2 and T3) connected together

Running the test

IR spot reading:

1. Apply test voltage
2. Record reading at 60 seconds
3. Total test time: ~1 minute

PI test (form-wound motors):

1. Apply test voltage
2. Record reading at 30 seconds (for DAR calculation)
3. Record reading at 1 minute
4. Continue applying voltage for additional 9 minutes
5. Record reading at 10 minutes
6. Calculate: DAR = IR₆₀ / IR₃₀ and PI = IR₆₀₀ / IR₆₀

Phase-by-phase procedure:

1. Test phase T1 (T2 and T3 grounded or on guard)
2. Discharge for 1 minute
3. Test phase T2 (T1 and T3 grounded or on guard)
4. Discharge for 1 minute
5. Test phase T3 (T1 and T2 grounded or on guard)
6. Compare readings — all three should be within 10–20% of each other

After the test

1. Discharge critical. For a 10-minute PI test, discharge for at least 40 minutes
2. Verify voltage below 50 V with a voltmeter before disconnecting leads
3. Reconnect motor cables
4. Remove lockout/tagout
5. Record all results in the maintenance log

Phase-by-Phase vs Bulk Testing

Bulk testing (all three phases together)

Connect T1, T2, T3 together, test to ground. Single measurement.

Advantage: Fast. One test instead of three. Disadvantage: You only see the overall insulation-to-ground. You cannot detect if one phase has a problem and the other two are fine.

Phase-by-phase testing (IEEE 43-2013 recommended)

Test each phase individually with the other two grounded.

Advantage: You see each phase’s condition separately. A bad phase stands out. Disadvantage: Takes three times longer. Requires more test discipline.

When to use each

• Routine screening of non-critical motors: Bulk test is fine
• Commissioning: Phase-by-phase always
• Trending of critical motors: Phase-by-phase annually, bulk testing quarterly
• Troubleshooting a motor with a low bulk reading: Phase-by-phase to isolate the problem
• After repair or rewind: Phase-by-phase as part of acceptance

What phase imbalance tells you

If one phase reads 5 MΩ and the other two read 200 MΩ, you have a localized problem in that phase. Common causes:

• Winding damage (insulation scraped during rewind or installation)
• Water ingress through a specific terminal or feed cable
• Contamination concentrated in one area
• End-turn damage on that phase

A single bad phase often means the motor can be saved with targeted cleaning or repair. A motor with all three phases equally low usually has systemic moisture or contamination and needs a full drying cycle or rewind.

Reading Results on VFD-Fed Motors

VFD-fed motors have different wear patterns than line-fed motors because of the switching transients the drive produces.

What’s different

The VFD output is a high-frequency pulse-width-modulated waveform. Each switching transition creates a voltage spike that can reach 2× the DC bus voltage. On a 400 V VFD, this means 1,000+ V peaks appear repeatedly on the motor terminals.

This stresses the insulation differently from continuous 50/60 Hz operation. The insulation ages faster, especially near the line-end coil (the first few turns that see the full voltage transient).

What to look for

Early warning signs on VFD motors:

• PI trending downward even if still above 2.0
• IR trending downward even if still above minimum
• Step voltage test showing increasing slope

Typical VFD motor failure pattern:

1. Years of normal readings
2. Gradual PI decline
3. Step voltage test reveals voltage-dependent degradation
4. One phase (usually the one closest to the VFD output) develops a fault
5. Line-end coil failure

Testing procedure differences

Always disconnect the motor from the VFD at the motor terminals. Testing through a connected VFD will destroy the drive. The VFD’s input rectifier and DC bus capacitors are not designed to absorb megger test voltage.

For VFD-fed motors, I recommend annual testing with full PI and step voltage tests, even on small motors where you’d normally just run DAR.

Motors After Water Damage or Flooding

A motor that has been flooded or submerged needs special handling before any insulation test.

Do NOT immediately megger a flooded motor

If you apply 500 V DC to a motor saturated with water, the insulation may flash over internally — converting a potentially recoverable motor into scrap. Water between conductors at high voltage can carbonize, leaving permanent tracking paths.

The recommended procedure

Step 1: Drain and clean.

• Drain all standing water from the motor housing
• Remove all debris, mud, silt
• Wash with fresh water (especially if exposed to saltwater)
• Dry the exterior

Step 2: Initial low-voltage test.

• Apply 100 V DC or the lowest voltage your megger supports
• Record the reading
• If the reading is above 0.5 MΩ, proceed. If below, do NOT increase voltage — continue drying first.

Step 3: Dry the motor.

• Low-heat drying oven at 70°C maximum (to avoid damaging insulation)
• Or drying by circulating DC current through the windings
• Test periodically during drying (typical drying curve: IR drops initially, then rises steadily as moisture leaves)
• Continue until IR stabilizes at a reasonable value (>5 MΩ for random-wound, >100 MΩ for form-wound, after correction to 40°C)

Step 4: Full diagnostic testing.

• Once the motor is dry, run a complete test program: IR, PI (if form-wound) or DAR (if random-wound), step voltage test
• Compare to manufacturer’s nameplate minimums and the pre-flood history if available

Step 5: Decision.

• Values meet minimums and PI/DAR is healthy → return to service with increased test frequency
• Values marginal → consider rewind
• Values below minimum after drying → rewind required

When to Rewind

A motor reaches the rewind decision when insulation testing can no longer confirm safe operation. The specific trigger varies, but here are the clear indicators.

Clear rewind indicators

• IR below minimum after drying and cleaning, temperature-corrected
• PI below 1.0 consistently
• One or more phases with dramatically lower readings than the others (and cleaning doesn’t help)
• Step voltage test shows >50% resistance drop at elevated voltages
• Visible physical damage to the windings (burn marks, melted insulation, exposed copper)
• History of multiple ground faults during operation

The economic decision

Rewinding is typically 40–60% of the cost of a new motor. For large motors (>200 HP), rewinding makes economic sense as long as the motor’s stator iron and frame are in good condition. For smaller motors, replacement is often cheaper and faster.

Consider:

• Age of the motor (approaching 20+ years → replace)
• Efficiency class (older standard-efficiency motors → replace with premium efficiency)
• Availability of replacement (long lead time → rewind)
• Criticality (redundant unit available → replace, single point of failure → rewind quickly)

Before rewinding

Run a complete test program to document the motor’s condition. After rewinding, run acceptance testing including IR, PI, surge test, and a full dielectric test per the rewinder’s specifications. Baseline these results for future trending.

Testing Program by Motor Criticality

Not every motor needs the same testing frequency. Match the program to the impact of failure.

Category A: Critical motors

Single points of failure, production-critical, or safety-related. Examples: main cooling pumps, induced draft fans, boiler feed pumps.

Program:

• IR and PI: every 6 months
• Step voltage test: annually
• Full diagnostic (including surge) on spare units
• Online PD monitoring for motors >1 MW

Category B: Important motors

Process-important but have redundancy or short replacement time. Examples: chiller compressors, major conveyors, standard process pumps.

Program:

• IR and PI (form-wound) or DAR (random-wound): annually
• Step voltage test: every 2–3 years or when IR trend shows decline
• Full testing after any shutdown longer than 30 days

Category C: General-service motors

Numerous small motors where individual failure has limited impact. Examples: small fans, auxiliary pumps, lab equipment.

Program:

• IR spot reading: every 2–3 years, or at scheduled maintenance
• DAR: on any motor with borderline IR readings
• Full testing after exposure to unusual conditions (flooding, vibration events, overloading)

Storage motors

Spare motors held in warehouse. Insulation absorbs moisture during storage.

Program:

• IR test every 6 months
• Test before any installation (never install a motor without testing)
• Store in dry, climate-controlled environment
• Consider heaters for long-term storage in humid environments

Common Mistakes

Not disconnecting the motor from the VFD. This destroys drives. A Category A motor protected by a $30,000 VFD — don’t save 10 minutes by testing with the drive connected.

Applying IEEE 43-2013 PI minimums to random-wound motors. PI is often not meaningful on random-wound motors. Use DAR instead, or focus on absolute IR value.

Testing a flooded motor at 500 V without drying first. Converts a repairable motor into scrap by causing internal flashover.

Comparing hot readings to cold readings without correction. A motor that just ran at 60°C reading 30 MΩ has better insulation than a cold motor at 20°C reading 150 MΩ (both correct to approximately the same 40°C value).

Stopping at bulk testing when readings are low. If your all-phases-together reading is low, always follow up with phase-by-phase testing. The problem is almost always localized.

Applying the old (kV + 1) MΩ rule to modern motors. This rule applies to pre-1970 insulation systems only. For a modern 460 V random-wound motor, the correct minimum is 5 MΩ, not 1.46 MΩ.

Not recording the winding temperature. Without temperature, your IR reading is useless for trending. Record it every time.

FAQ

Key Takeaways

• Know whether your motor is form-wound or random-wound. It drives every interpretation.
• Minimum IR per IEEE 43-2013: 5 MΩ for random-wound, 100 MΩ for form-wound (Clause 12.1).
• Minimum PI per IEEE 43-2013: 2.0 for Class F/H, 1.5 for Class A (Table 3).
• PI is not reliable on random-wound motors — use DAR instead.
• PI is not valid when IR > 5 GΩ — trust the absolute value.
• Phase-by-phase testing reveals localized problems that bulk testing misses. Use it for commissioning and for troubleshooting low bulk readings.
• Always disconnect from VFDs before testing. The drive will be destroyed.
• Dry flooded motors before applying full test voltage. Flash-over during wet testing creates permanent damage.
• Match the testing program to motor criticality. Critical motors every 6 months. General-service motors every 2–3 years.

Standards Referenced in This Article

Standard	Key Content
IEEE 43-2013	Test voltages (Table 1), minimum IR (Clause 12.1), minimum PI (Table 3), PI invalid above 5 GΩ (Clause 12.2.2), temperature correction to 40°C (Clause 6.3), phase-by-phase testing recommended
IEEE 95-2002	High DC voltage testing for form-wound motors ≥2,300 V (referenced for step voltage testing)
IEC 60034-1	Rotating electrical machines — rating and performance
IEC 60085	Thermal classification of insulation materials
ANSI/EASA AR100	Recommended practice for repair of rotating electrical apparatus

Related Articles

• How to Megger a Motor: Step-by-Step
• Polarization Index Test Explained
• Dielectric Absorption Ratio (DAR) Explained
• Step Voltage Test Explained
• How to Read Insulation Test Results
• Temperature Correction for IR Tests
• Insulation Testing Safety Procedures