Troubleshooting Low Insulation Resistance Readings: A Field Troubleshooting Guide

The reading is 2 MΩ. The motor is supposed to be above 100 MΩ. You’re standing at the terminal box with a megger in one hand and a maintenance log in the other, trying to figure out what to do next.

This guide walks through the actual troubleshooting sequence for low insulation resistance readings. Not theoretical — the exact steps I use in the field, in the order I use them, with the specific decisions at each branch. Because most low readings are not equipment failure. Most are moisture, contamination, or test setup errors. Knowing the difference saves hours of unnecessary work and prevents wrongly scrapping healthy equipment.

This article assumes you’ve already taken a reading and it’s below expected values. If you need the fundamentals of running the test, see our How to Use a Megohmmeter guide first.

The Triage Decision: How Low Is “Low”?

Not every low reading is an emergency. Your first decision is whether you have time to investigate carefully or whether the equipment must be immediately tagged out.

Tier 1 — Critical (act now, investigate later)

• Reading below 1 MΩ at any voltage on any equipment above 50V
• Reading at or near zero with smoke smell, burn marks, or visible damage
• Reading that trips the megger’s internal protection

Action: Do not energize. Lock out. Document. Investigate as part of a planned intervention, not a “let me just check” retest.

Tier 2 — Below minimum (investigate before energizing)

• Motor reading below IEEE 43-2013 minimums (5 MΩ random-wound, 100 MΩ form-wound)
• Cable reading below 1 MΩ per IEC 60364-6
• Protection relay reading below 100 MΩ per IEC 60255-5
• Transformer reading dramatically below historical values

Action: Run the troubleshooting sequence below. Most of these turn out to be surface contamination or moisture, not insulation failure.

Tier 3 — Declining but passing (plan to investigate)

• Reading passes minimum but is 50% lower than previous test
• Reading passes minimum but PI has dropped below 2.0
• Reading passes minimum but varies significantly between phases

Action: Complete current work. Schedule follow-up testing with the troubleshooting procedure within the next planned outage.

The troubleshooting procedure below works for all three tiers — the urgency differs, not the method.

Step 1 — Check Your Test Setup First

Before concluding the equipment is bad, verify the test itself is valid. In training sessions, I’ve seen technicians spend an hour investigating a “failed” motor that turned out to be a megger with a dead battery. Check the basics first.

Test voltage

Is the test voltage appropriate? A 250V test on a 4160V motor gives artificially low readings because the insulation isn’t stressed enough to reveal its true quality. A 5000V test on a 120V control circuit can damage the equipment. Match test voltage to equipment rating per the standards.

Test duration

Are you reading at the right time? The first 30 seconds show absorption current, which gives an artificially low reading. Always wait the full 60 seconds before recording. For form-wound motors, the 10-minute reading is meaningful; the 10-second reading is not.

Discharge before testing

Did you discharge the equipment first? Residual charge from operation creates errors in the first minute. Short to ground for at least 60 seconds before starting the test.

Lead condition

Are your test leads intact?

• Cracked insulation on the test lead creates leakage paths
• Corroded clips create high-resistance connections that can show as capacitive reactance
• Dirty probe tips create intermittent contact

Inspect the leads. If they’re damaged, replace them before testing.

Megger calibration

When was the instrument last calibrated? An out-of-calibration megger can read significantly wrong. Check the calibration sticker. If it’s been more than 12 months or the date is missing, the reading is suspect.

Battery state

Does the megger have adequate battery? A low battery can prevent the instrument from producing rated test voltage, causing artificially low readings that look like insulation failures but aren’t. Check the battery indicator. Replace or recharge if marginal.

Connections

Are your connections making good contact?

• LINE lead firmly connected to the conductor under test
• EARTH lead firmly connected to clean bare metal (not paint, not rust)
• GUARD lead (if used) connected to the right point

Retighten all connections. A loose EARTH connection is one of the most common sources of inconsistent readings.

Nothing accidentally connected in parallel

Is there anything in the circuit you forgot to disconnect?

• Surge arresters
• MOVs or surge protection devices
• VFDs, soft starters, or power electronics
• Capacitor banks
• Other equipment on the same bus

Every additional item in parallel lowers the reading. Disconnect and retest.

Retest after verification

After verifying all of the above, take a fresh reading. Roughly 20–30% of “low” readings I investigate turn out to be test setup issues, not equipment problems. Don’t skip this step.

Step 2 — Rule Out Surface Contamination and Moisture

Once you’ve verified the test setup is valid and the reading is genuinely low, the next most common cause is surface contamination or moisture — not internal insulation failure.

Why this matters

Surface contamination creates a leakage path across the insulation surface rather than through the insulation itself. A motor with surface contamination on the end windings reads low, but the actual insulation material may be perfectly fine. Clean and dry the surface, and the reading comes back to normal.

This is why IEEE 43-2013 and other standards emphasize the difference between “volume resistance” (through the bulk insulation) and “surface resistance” (across the insulation surface). Surface problems are usually easy to fix. Volume problems often aren’t.

Visual inspection

Before doing anything else, look:

• White, crystalline deposits on bushings or terminations → salt contamination (near coast, winter road salt environments)
• Black tracking marks in carbon patterns → surface tracking, often from previous partial discharge
• Dust, dirt, oil film on insulators → general contamination (industrial environments)
• Water droplets, damp appearance → condensation, leak, or flooding
• Insect damage, nesting → biological contamination (rare but real, especially in outdoor enclosures)

Each of these has a different remediation path.

Humidity check

What was the humidity during the test? Above 70% relative humidity, moisture on insulation surfaces creates leakage paths even on perfectly healthy equipment. If the test was done in humid conditions, retest after the conditions dry — or use the guard terminal to compensate.

The guard terminal test

Every decent megohmmeter has a GUARD terminal. It diverts surface leakage current away from the measurement, isolating the volume resistance only.

Procedure for guard terminal use:

1. Connect LINE lead to the conductor under test
2. Connect EARTH lead to the frame/ground
3. Connect a GUARD lead around the insulation surface between them (typically wrapped around the bushing neck or termination body)
4. Take the reading

If the guarded reading is significantly higher than the ungarded reading (e.g., guarded 500 MΩ vs ungarded 50 MΩ), the problem is surface contamination, not internal insulation failure. Clean the surface.

Cleaning contaminated insulators

For accessible insulators:

• Light dust: Dry cloth or compressed air
• Oil film: Lint-free cloth with isopropyl alcohol or approved solvent
• Salt deposits: Water rinse followed by drying; for severe cases, fresh water pressure wash
• Tracking damage: Insulator replacement (surface tracking creates permanent carbon paths that cleaning can’t remove)

After cleaning, allow to dry completely (at least 4 hours in ambient conditions, longer in humid environments), then retest.

Step 3 — Isolate the Problem

If the reading is still low after confirming the test setup and ruling out surface issues, you need to find where the actual problem is.

Bulk to phase-by-phase

If you’re testing a three-phase motor and all three phases together read low, switch to phase-by-phase testing (each phase tested with the other two grounded).

Possible outcomes:

| All phases equally low | → Systemic problem — moisture, general contamination, or uniform aging | | One phase much lower than the others | → Localized problem — damage, moisture ingress at one point, contamination in one area | | Two phases low, one normal | → Fault on a specific part of the winding (rare but occurs with physical damage) |

Phase-by-phase testing is fundamental for motor troubleshooting. Without it, you don’t know if you have a bad motor or a bad phase.

Section the circuit

For cables and distribution systems, divide and test:

1. If a cable-to-panel reading is low, disconnect the far end and test the cable alone
2. If the cable is good, the problem is in the panel or connected equipment
3. If the cable is bad, test each cable section separately by opening junction boxes

Each disconnection should roughly double the reading if the problem is truly in the isolated section. If the reading doesn’t improve after isolation, the problem is somewhere else.

Compare to history

If you have previous test records (which you should, per predictive maintenance practice), compare:

• Similar reading to previous tests → this is how the equipment has been. Not degraded, just low.
• Significantly lower than last test → recent change. Something happened between tests.
• Gradual decline over multiple tests → normal aging, now reaching threshold.

Without historical data, you’re guessing. This is why trending and predictive maintenance matter.

Temperature check

Was the equipment cold when tested? A motor at 5°C reads dramatically lower than the same motor at 40°C. Temperature-correct your reading before concluding it’s low:

• Motor at 5°C with measured IR of 10 MΩ → corrected to 40°C equivalent ≈ 100 MΩ (healthy)
• Motor at 40°C with measured IR of 10 MΩ → actually 10 MΩ (low, needs investigation)

See our Temperature Correction guide for the full procedure.

Step 4 — Dry the Equipment

If the problem is moisture (the most common cause of low readings after contamination), drying often restores the equipment to service without any repair.

How to tell if drying will work

• Reading rose during the test (stayed climbing, didn’t stabilize) → moisture, likely recoverable
• PI below 1.5 but IR reasonable → moisture, likely recoverable
• Reading dropped during the test → active failure, drying probably won’t fix it
• Near-zero reading that doesn’t change with time → ground fault or winding-to-winding short, not moisture

Drying methods

For motors:

1. Ambient drying — Open the motor enclosure, run dry air or dehumidified air through the windings. Takes 24–72 hours. Works for light moisture.
2. Low-heat drying — Drying oven at 70°C maximum. More effective but requires motor removal. Monitor IR during drying — it typically drops initially, then rises steadily as moisture leaves. Continue until IR stabilizes.
3. DC current drying — Circulate low DC current through the windings (typically 20–30% of rated current). The heat drives moisture out. Requires a controlled DC source and continuous monitoring.
4. Steam drying — For severe cases. The motor is heated to drive off free water, then cooled in dry air. Specialized procedure typically done by motor repair shops.

For transformers:

1. Circulating hot oil — Drain partial oil, circulate remaining oil through a hot-oil filter system. Standard approach for moisture-affected distribution transformers.
2. Vacuum drying — Remove oil, apply vacuum to the tank for 24–48 hours. Most effective but requires specialized equipment.
3. Zero-phase-sequence drying — Apply reduced voltage with the transformer short-circuited. Creates internal heating that drives off moisture.

For cables:

Cable drying is difficult. Most moisture-affected cables require termination repair or section replacement rather than in-place drying.

Monitor during drying

Take IR readings periodically during the drying process:

• Initial reading (baseline)
• After 1 hour, 4 hours, 8 hours, 24 hours
• Continue until the reading stabilizes (stops rising)

The drying curve typically shows:

1. Initial drop (heating makes moisture more conductive)
2. Plateau (moisture migrating to surface)
3. Steady rise (moisture evaporating)
4. Stabilization (drying complete)

If you never reach stabilization after 48+ hours, the problem isn’t just moisture — there’s underlying damage.

After drying

Once IR has stabilized at an acceptable value:

1. Run a full diagnostic test (IR, PI, DAR as applicable)
2. Compare to manufacturer’s minimums
3. Compare to pre-problem history if available
4. Document the drying procedure used and final readings
5. Return to service with increased test frequency for the next 6–12 months to verify stability

Step 5 — Decide: Repair, Rewind, or Replace

If cleaning and drying don’t restore the reading to acceptable levels, the insulation itself is damaged. Your decision now is what to do about it.

Indicators the insulation is actually failed

• IR stays below minimum even after thorough drying
• PI below 1.0 consistently after drying
• One phase dramatically worse than others, and cleaning doesn’t help
• Visible damage to windings (burn marks, melted insulation, exposed copper)
• History of multiple ground faults or trips
• Step voltage test shows significant resistance drop at elevated voltages

The three paths

Repair — Possible when the damage is localized and accessible. Examples:

• Re-taping damaged end-turn insulation on a motor
• Re-terminating a cable with a damaged termination
• Replacing a single cracked bushing on a transformer

Repair is fastest and cheapest when it’s possible, but only works for localized, visible damage.

Rewind — Appropriate when the winding insulation is damaged but the mechanical parts (core, frame, bearings) are still good. Typical for:

• Motors with widespread insulation aging (multiple phases affected)
• Transformers with severely degraded paper insulation
• Generators after water damage

Rewinding typically costs 40–60% of new equipment cost. Makes economic sense for motors above 200 HP. For smaller motors, replacement is usually cheaper.

Replace — Right answer when:

• Equipment is approaching end of design life (20+ years on motors)
• Older equipment can be upgraded to higher efficiency (IE3/IE4 motors replacing older standard-efficiency units)
• Rewind cost exceeds 70% of new equipment cost
• Repeat failures indicate systemic problems

The replace vs rewind decision is primarily economic. If you’re paying a certified rewinder $30,000 to rewind an old motor when a new premium-efficiency replacement costs $45,000, replacement usually wins on total cost of ownership over the motor’s next 10–15 years of operation.

Equipment-Specific Troubleshooting

Different equipment types have different common failure modes. Use this guide to focus your investigation.

Motors

Most common causes of low readings (in order):

1. Surface contamination on end-turn insulation
2. Moisture absorbed during storage or idle periods
3. Contamination ingress through damaged seals
4. Insulation aging (normal degradation over decades)
5. Physical damage from mechanical issues (bearing failure, rotor rub)

First check: Phase-by-phase testing to locate the problem. Then clean end-windings if accessible.

See: Motor Insulation Testing: The Complete Field Guide

Cables

Most common causes:

1. Moisture ingress at terminations
2. Physical damage (dig-in, crush damage)
3. Water absorption in the cable jacket
4. Degraded termination sealing materials
5. Internal tracking in splices

First check: Isolate both ends and test the cable alone. If still low, test section-by-section at junction boxes.

See: Cable Insulation Testing Guide

Transformers

Most common causes:

1. Moisture in the oil
2. Moisture absorbed by paper insulation
3. Contamination of bushings
4. Oil degradation (tan delta and power factor tests complement IR)
5. Internal winding damage (rare but serious)

First check: Clean bushings, check oil moisture content, retest. For oil-filled transformers, IR alone is an incomplete picture — oil analysis and tan delta testing give the full story.

See: Transformer Insulation Testing Guide

Switchgear

Most common causes:

1. Dust accumulation on bus insulators and support insulators
2. Moisture in cable compartments (especially after heavy rain events)
3. Contamination on VT and CT bushings
4. Tracking on aged insulators

First check: Visual inspection and cleaning. Switchgear problems are often dramatically improved by good cleaning alone.

See: Switchgear Insulation Testing Guide

Generators

Most common causes:

1. Moisture absorption during idle periods (especially emergency generators)
2. Contamination on exposed rotor windings
3. Bushing contamination
4. Stator bar aging (very old units)

First check: Dry the unit (especially if emergency/standby status), retest. Apply different interpretation criteria to rotor (1.2–1.5 PI is healthy) vs stator (≥2.0 PI required).

See: Generator Insulation Testing Guide

Protection relays and low-voltage circuits

Most common causes:

1. Moisture in terminal blocks or panels
2. Contamination of terminal surfaces
3. Degraded wiring insulation on older installations
4. Accidental test voltage applied to protection circuits (the megger’s fault, not the equipment’s)

First check: Verify you disconnected electronic equipment. If so, clean terminal blocks and retest.

See: IEC 60255-5: Insulation Testing for Protection Relays

The Most Common Causes of Low Readings

Based on 12 years of field experience, here are the causes I encounter most frequently, ranked by frequency:

1. Moisture (roughly 40% of low readings) Motors after extended storage, transformers with failing seals, outdoor equipment after heavy rain, emergency generators that rarely run. Drying typically recovers the equipment.

2. Surface contamination (roughly 25%) Dust, oil film, salt, industrial contaminants. Cleaning typically recovers the equipment.

3. Test setup errors (roughly 20%) Wrong test voltage, connected equipment not disconnected, insufficient test duration, lead damage, battery low. Fix the setup, retest.

4. Localized damage (roughly 10%) Physical damage from mechanical events, insulation degradation from overheating, termination failure. Requires repair.

5. Actual end-of-life insulation failure (roughly 5%) Severely aged or degraded insulation requiring rewind or replacement. Relatively rare.

Most “insulation failures” are actually categories 1–3, which have straightforward fixes. Don’t rewind the motor or replace the cable until you’ve ruled out categories 1–3.

What Not to Do

Don’t re-megger at progressively higher voltages looking for a “passing” result. If the reading is low at 500V, it’s low. Applying 2500V to confirm the failure can cause actual damage to otherwise recoverable insulation.

Don’t trust a single retest after making changes. If you clean and dry the equipment, take at least two readings 30 minutes apart. A single “good” reading after intervention can be misleading if the equipment hasn’t fully stabilized.

Don’t skip temperature correction when comparing to historical data or standards. A 50 MΩ reading at 10°C is completely different from 50 MΩ at 40°C. The corrected value tells you whether the equipment is actually degraded.

Don’t ignore the trend even if the current reading passes minimum. A motor that dropped from 1 GΩ to 50 MΩ is on a trajectory to failure, even though 50 MΩ is above the 5 MΩ minimum for random-wound motors.

Don’t energize equipment that failed the test on the assumption that “it’s probably fine.” The entire point of the test is to verify the equipment is safe to energize. A failed test means it’s not.

Don’t rewind immediately when cleaning or drying could recover the equipment. This is an expensive mistake I’ve seen multiple times — a plant spent $25,000 rewinding a motor that would have recovered with a 12-hour ambient drying cycle.

Don’t assume the problem is always the equipment. Roughly 20% of low readings are test setup issues. Verify the setup before concluding the equipment is bad.

FAQ

Key Takeaways

• Verify the test setup first. Roughly 20% of low readings are test problems, not equipment problems.
• Rule out contamination and moisture before concluding insulation failure. Most low readings recover with cleaning and drying.
• Isolate the problem through phase-by-phase testing (motors) or section-by-section testing (cables).
• Most low readings are recoverable — 40% moisture, 25% contamination, 20% test setup, 10% localized damage, only 5% actual end-of-life failure.
• Rising IR during test = healthy. Declining IR during test = active fault.
• Use the guard terminal to distinguish surface problems (fixable) from volume problems (often not).
• Temperature-correct readings before comparing to standards or historical data.
• Don’t rewind or replace until you’ve exhausted cleaning, drying, and troubleshooting options.
• A well-dried, well-cleaned motor often recovers from readings that initially looked like end-of-life failure.