Insulation resistance testing applies a controlled DC test voltage to an isolated conductor, cable, winding, or other approved test object and measures the resulting leakage current as a resistance value, usually reported in megohms or higher units.
The test can reveal moisture, contamination, damaged insulation, surface leakage, connection problems, or insulation deterioration. It does not by itself prove that equipment is safe to energize, suitable for service, or acceptable for a project. The test object, connections, voltage, duration, temperature-correction method, and acceptance basis must come from the manufacturer’s instructions, an approved test procedure such as ANSI/NETA ATS-2025, the project specification, or qualified technical review — never from a generic online table.
What is being tested?
An insulation resistance test evaluates the insulation system between defined electrical points. Typical test points include:
- one conductor to ground, a metallic shield, or an equipment frame;
- one conductor to another conductor;
- motor or generator windings to the frame and, when required, between windings;
- transformer windings to ground or between windings when the procedure permits;
- cable insulation, bus insulation, switchgear, and other assemblies identified by the manufacturer or project procedure.
Define the test object before connecting the leads. Record the equipment or conductor identification, the exact endpoints, the grounding or shielding arrangement, and anything already disconnected.
Insulation resistance is not the same as conductor resistance or continuity. Continuity asks whether a low-resistance path exists. Insulation resistance asks whether the insulation provides a sufficiently high-resistance path between the selected points under the specified DC conditions.
Equipment that may need isolation
Many modern systems contain components that should never see a megohmmeter’s DC output. Depending on the equipment, these may include:
- variable-frequency drives and soft starters;
- PLCs, control boards, electronic power supplies, and communication equipment;
- UPS systems, battery chargers, and other solid-state devices;
- surge protective devices, capacitors, filters, sensors, and monitoring circuits;
- connected loads or parallel circuits outside the approved test boundary.
Opening one disconnect does not isolate every component. Follow the equipment drawings and manufacturer instructions. If a device is not explicitly approved for the test voltage, isolate it or get qualified technical direction before proceeding.
What a megohmmeter measures
A megohmmeter (megger) applies a DC voltage and measures the resulting current, then reports a calculated resistance value.
- ohm (Ω): the base resistance unit;
- kilohm (kΩ): 1,000 ohms;
- megohm (MΩ): 1,000,000 ohms;
- gigohm (GΩ): 1,000 MΩ;
- terohm (TΩ): 1,000 GΩ.
500 kΩ, 500 MΩ, and 500 GΩ are very different readings — check the prefix and decimal position before recording. Do not confuse MΩ with mΩ; megohms describe very high resistance, while milliohms describe very low resistance.
A higher reading generally means less measured leakage under that specific test condition, but it does not mean every high reading is automatically acceptable. Results are affected by the test object, temperature, moisture, contamination, surface condition, test voltage, duration, capacitance, connections, instrument limits, and the acceptance method used.
Test voltage and duration are procedure inputs
Match the test voltage to the test object and approved method. Start with the manufacturer’s recommended DC insulation-resistance test voltage, then confirm against the project specification, adopted standard, equipment rating, and insulation system.
Instruments commonly offer selectable outputs such as 250 V DC, 500 V DC, 1,000 V DC, 2,500 V DC, or 5,000 V DC. For rotating machinery, IEEE 43 — the long-standing recommended practice for testing insulation resistance and polarization index of stator and rotor windings — has historically tied test voltage ranges to winding nameplate voltage. Note that IEEE 43-2013 was formally inactivated by IEEE-SA in March 2024, and its successor (IEEE P97, covering diagnostic DC test methods for AC electric machinery) is still in draft development. Until P97 is published, IEEE 43-2013 remains the practical industry reference most manufacturers and test firms still cite, but it should be treated as guidance to confirm against current manufacturer data rather than a frozen mandatory standard.
These are instrument settings and typical industry ranges, not universal field recommendations. Too low a voltage may not stress the insulation enough to produce useful information; too high a voltage can damage insulation or connected equipment.
Record the selected voltage exactly as applied:
test voltage: 1,000 V DC
Don’t substitute the nominal system voltage for the test voltage. A 480 V system does not automatically dictate the correct insulation-resistance test setting — that is a separate procedure decision, and for acceptance testing it should trace back to the applicable table in ANSI/NETA ATS-2025 or the manufacturer’s published data.
Duration matters just as much. A short-time or spot reading at a defined time is not interchangeable with a timed test, polarization index (PI), dielectric absorption ratio (DAR), step-voltage test, ramp test, or dielectric discharge test. If the procedure requires readings at more than one time, record each value with its elapsed time from test start:
test voltage: 500 V DC
reading at 1 minute: [record value]
reading at 10 minutes: [record value]
The blank fields are intentional. Don’t insert a generic acceptance value into a field record — use the calculation and review method defined by the governing procedure.
Why temperature must be recorded
Insulation resistance changes significantly with temperature. In most insulation systems, resistance decreases as temperature rises, so a reading taken on a cold morning may read much higher than the same equipment at normal operating temperature — even with no change in insulation condition.
Record the conditions needed to interpret the measurement:
- test object and connection points;
- test voltage;
- test duration and reading times;
- insulation or equipment temperature;
- ambient temperature when the procedure requires it;
- moisture, humidity, or surface condition when relevant;
- instrument identification and lead configuration;
- temperature-correction method and source, if correction is required.
Don’t apply a familiar correction factor just because it’s common in another test program. Correction depends on the insulation system, equipment type, reference temperature, test method, and source document — IEEE 43’s traditional reference temperature for rotating machinery is 40°C, but that value does not automatically transfer to cable, transformer, or switchgear insulation, which typically follow the manufacturer’s data or NETA ATS tables instead.
A corrected value is still a review input; it does not independently establish acceptance.
A practical field workflow
1. Define the test boundary
Identify the exact conductor, cable, winding, equipment section, or assembly to be tested using the one-line diagram, wiring diagram, nameplate, manufacturer instructions, and approved test procedure.
Write down:
- equipment or conductor identification;
- nominal voltage and phase information when relevant;
- test points and expected return path;
- components that must be isolated;
- the person responsible for the test and the person responsible for releasing the equipment.
If the boundary is unclear, stop before connecting the instrument. A technically correct reading from the wrong section is still the wrong test.
2. Deenergize, lock out, and control all sources
Open the required disconnecting means and apply the employer’s lockout/tagout procedure. Consider normal sources, alternate sources, control power, stored energy, backfeed, induced voltage, generators, transformers, UPS equipment, photovoltaic sources, and interconnections with other equipment.
For U.S. general-industry electrical work, OSHA 29 CFR 1910.333(b) requires exposed parts to be deenergized unless an allowed exception applies, and requires a qualified person to use test equipment to verify the circuit elements and equipment parts are actually deenergized — including checking for inadvertently induced voltage or unrelated backfeed even after parts have been “presumed” deenergized. Work on power generation, transmission, and distribution lines and equipment falls under the separate rules in 29 CFR 1910.269 instead. Parts that have been deenergized but are not properly locked or tagged must still be treated as energized.
A megohmmeter is not a substitute for an absence-of-voltage test. Establish the deenergized condition first with the required voltage-test instrument and work practice, then connect the insulation tester only after the boundary is controlled.
3. Verify absence of voltage and unexpected sources
Use an appropriately rated voltage tester and the employer’s required verification sequence to check the circuit elements and equipment parts employees may contact, including possible backfeed, induced voltage, and unrelated sources.
The exact instrument, PPE, approach boundaries, and verification sequence depend on the system and workplace procedure — reference the current NFPA 70E, Standard for Electrical Safety in the Workplace, 2027 edition, the current edition as of this review. A qualified person must determine the method. Don’t rely on a control switch, pilot light, or a single open disconnect as proof of isolation.
4. Isolate connected equipment
Disconnect or isolate electronic and energy-storing devices outside the test procedure. Confirm the isolation at the actual test points, and keep parallel circuits and connected loads outside the boundary unless the approved procedure specifically includes them.
Pay special attention to:
- drives and electronic starters;
- surge protection and capacitors;
- control transformers and power supplies;
- PLCs, relays, sensors, and communication circuits;
- grounded shields, neutral connections, and equipment bonding paths;
- cable ends or equipment sections that can remain connected through another route.
Never connect a megohmmeter to an energized circuit, and never assume a device can tolerate the test voltage merely because it shares the same nominal system voltage.
5. Inspect the instrument and leads
Use an instrument rated for the voltage and environment, and confirm the calibration or verification status required by the employer and project. Inspect leads, probes, clips, guards, and insulation for damage.
Connect the leads per the instrument instructions and approved test method. Keep hands, tools, and unqualified personnel clear of the test object and exposed conductors, and use barriers, warning signs, or a test-area attendant when the test voltage and equipment warrant it.
6. Apply the specified test
Select the procedure-approved DC test voltage and duration, connect to the defined test points, start the test, and observe the instrument’s voltage and resistance indications. If the reading is unstable or the instrument indicates an unexpected voltage, stop and investigate rather than forcing the test to completion.
For high-resistance measurements, use the guard terminal when the instrument and procedure support it. A guard connection can reduce the effect of surface leakage but does not correct an incorrect test boundary or poor surface condition.
Record the actual reading at the procedure-defined time. If the reading keeps changing, record the time and the required timed values rather than selecting the most favorable number.
7. Discharge and verify the test object
The test object can retain a hazardous charge after the DC output is removed, especially with significant capacitance (long cable runs, transformer windings, capacitor banks). Use the tester’s discharge function when provided, and follow the required discharge time and grounding method — the general industry guidance is to ground the object for at least four times the test duration (for example, at least 4 minutes of grounding after a 1-minute test) before treating it as safe to contact.
Keep the leads connected and treat the test object as energized until the instrument confirms the voltage has fallen to the required safe state and the approved procedure permits contact. Don’t assume an automatic discharge feature eliminates every site-specific grounding or verification step.
8. Complete the test record
Record enough information for another qualified person to understand what was tested and how:
equipment or conductor:
test points:
nominal equipment voltage:
test voltage:
test duration:
resistance reading and time:
test temperature:
temperature-correction method and source:
instrument and lead configuration:
isolation or unusual-condition notes:
reviewer or qualified person:
Under the current NFPA 70B, Standard for Electrical Equipment Maintenance, 2026 edition, which superseded the 2023 edition, test records for maintenance testing must minimally identify the testing person and organization and show as-found and as-left results where applicable. A number without its voltage, time, temperature, and test boundary is difficult — and in a maintenance program, non-compliant — to interpret safely.
How to interpret the result
Interpretation starts with test validity, not the largest number on the display.
Confirm that the setup matches the procedure
Before judging the value, verify:
- the correct test object was isolated;
- the correct test points were used;
- connected electronics and surge devices were protected;
- the selected test voltage was authorized;
- the reading was taken at the required time;
- temperature and correction information were recorded;
- the instrument and leads were suitable and functioning;
- the surface was clean and dry enough for the method.
If any of these is unknown, the result may need to be repeated or reviewed before use.
Look at the reading in context
A high reading is consistent with low leakage, but it does not override a damaged cable jacket, visible tracking, overheating, contamination, moisture, or a failed functional check. A lower reading can reflect genuine insulation deterioration, but it can equally result from surface leakage, an incorrect connection, a connected load, a wet termination, a poor lead, or a test performed at a different temperature.
Don’t compare readings taken under materially different voltages, durations, temperatures, connections, or equipment conditions as if they were the same measurement. If the program requires temperature correction, use the specified method and retain the uncorrected measurement alongside the correction details.
Use timed behavior when required
Some procedures use the change in resistance during a timed test to evaluate charging, absorption, polarization, or other insulation behavior. For rotating machinery, IEEE 43 defines the polarization index as the ratio of the 10-minute reading to the 1-minute reading, with minimum PI values historically referenced at 1.5 for Class A insulation and 2.0 for Class B and higher — but these thresholds apply specifically to windings covered by that standard, not to cables, transformers, or switchgear.
Polarization index, dielectric absorption ratio, step-voltage, ramp, and dielectric-discharge results are method-specific diagnostic values. Their calculation, application, and acceptance basis must come from the adopted standard, equipment manufacturer, project specification, or approved maintenance procedure. Don’t transfer a ratio or threshold from one equipment class to another.
Treat an unexpected result as a review trigger
When a result is unexpectedly low, unstable, or different from the expected condition:
1. Stop before increasing the voltage or repeating the test blindly.
2. Confirm the absence-of-voltage and isolation boundary.
3. Verify the test points, leads, guard, grounding, and instrument range.
4. Inspect for moisture, contamination, tracking, damage, and loose or dirty terminations.
5. Confirm temperature and the required correction method.
6. Repeat only when the approved procedure allows it.
- Escalate to the manufacturer, engineer, testing authority, or other qualified reviewer as required.
Don’t average a questionable result with a better one, change the test voltage to obtain a preferred number, or declare acceptance from a generic internet threshold.
Common mistakes
Testing through a VFD or electronic control. The megohmmeter’s DC output can damage drives, PLCs, power supplies, UPS equipment, surge devices, and other solid-state components. Isolate these per the manufacturer and approved procedure.
Treating nominal system voltage as the test voltage. System voltage and insulation-resistance test voltage are different inputs; select and record the value required for the test object and procedure.
Recording only the megohm value. A resistance number without test voltage, duration, temperature, endpoints, and instrument context is incomplete.
Omitting temperature. Temperature can materially change insulation resistance. Record it and apply only the correction method matched to the tested insulation system.
Calling every high value a pass. Acceptance depends on the equipment, insulation system, test method, standard (ANSI/NETA ATS-2025, IEEE 43, or equivalent), manufacturer instructions, project requirements, and qualified review. No single absolute value applies to every test object.
Touching the circuit immediately after the test. The test object may remain charged after the output is removed. Discharge and verify the required state before removing leads.
Forgetting backfeed or induced voltage. An open disconnect does not eliminate every possible source. Check the actual circuit elements for unexpected voltage before connecting the insulation tester.
Ignoring physical evidence. Cracking, carbon tracking, moisture, contamination, heat damage, loose terminations, or mechanical damage may require action even when a single reading appears favorable.
Working from a governing reference
For new-construction acceptance testing, the current governing reference for most U.S. contractors and testing firms is ANSI/NETA ATS-2025, which superseded the 2021 edition and added dedicated sections for battery energy storage systems, solar PV systems, and expanded medium-voltage cable acceptance test tables. For ongoing maintenance testing, NFPA 70B’s 2026 edition sets the current mandatory framework for maintenance intervals, qualified-person training, and test-record content. For motor and generator windings specifically, IEEE 43-2013 remains the practical reference pending the release of its successor, IEEE P97.
A field record should tie the raw reading to its voltage, duration, temperature, correction method, and the specific governing document used for acceptance — not to a memorized number carried over from a different job or equipment class.
Establish the test boundary, control hazardous energy, verify absence of voltage under OSHA 1910.333 or 1910.269 as applicable, isolate connected electronics, use suitable equipment, control access to the test area, and discharge the test object before contact. Only qualified personnel should select and perform the test and interpret the result for the actual equipment in front of them.