Cable insulation fails silently. Unlike a motor that vibrates or a transformer that overheats, a cable with degrading insulation gives you no warning until it faults. And when it does fault, it takes out the circuit — sometimes causing arc flash, fire, or hours of downtime while you trace the problem.
Regular insulation testing on cables catches degradation early. This guide covers the complete procedure — from selecting the right test voltage to interpreting the results — with specific values from IEC 60364-6 and practical tips from 12 years of field commissioning.
Table of Contents
Why Cable Insulation Degrades
Cable insulation breaks down for predictable reasons. Knowing them helps you test at the right times and interpret low readings correctly.
Moisture ingress — The most common cause of cable insulation failure. Water enters through damaged outer sheaths, unsealed cable glands, or condensation in cable trays. Even small amounts of moisture dramatically reduce insulation resistance. Underground and outdoor cables are most vulnerable.
Thermal stress — Cables running above their rated temperature age faster. Overloaded circuits, poor ventilation in cable trays, and bundled cables that can’t dissipate heat all contribute. PVC insulation becomes brittle above 70°C. XLPE handles higher temperatures but still degrades over time.
Mechanical damage — Cables pulled too hard during installation, crushed by cable tray supports, or damaged by construction work. The insulation may look intact from the outside while the conductor-to-sheath insulation is compromised internally.
Chemical attack — Oil, solvents, and corrosive chemicals can attack cable insulation, especially PVC. Industrial environments with chemical spills or fumes accelerate this degradation.
UV exposure — Cables exposed to direct sunlight without UV-resistant outer sheaths deteriorate over years. The sheath cracks, allowing moisture to reach the insulation beneath.
Age — All insulation degrades over time, even under ideal conditions. Typical cable life expectancy is 20–30 years for PVC, 30–40 years for XLPE, though actual life depends heavily on operating conditions.
Test Voltage Selection per IEC 60364-6
IEC 60364-6 (Low-voltage electrical installations — Part 6: Verification) specifies the test voltages and minimum insulation resistance values for cable testing in building and industrial installations.
Test voltage and minimum IR (Clause 6.4.3.3)
| Circuit Nominal Voltage | Test Voltage (DC) | Minimum IR |
|---|---|---|
| SELV and PELV circuits (up to 50V) | 250V | ≥ 0.5 MΩ |
| Circuits up to 500V (including 230V, 400V) | 500V | ≥ 1.0 MΩ |
| Circuits above 500V up to 1,000V | 1,000V | ≥ 1.0 MΩ |
For higher voltage cables
IEC 60364-6 covers installations up to 1,000V AC. For medium-voltage cables (above 1 kV), test voltages are typically:
| Cable Rated Voltage | Test Voltage (DC) |
|---|---|
| 1 kV – 5 kV | 2,500V |
| 5 kV – 15 kV | 5,000V |
| Above 15 kV | 5,000V – 10,000V |
These higher voltage values follow general industry practice and manufacturer recommendations. Always check the cable manufacturer’s maximum test voltage — exceeding it can damage healthy insulation.
What the IEC 60364-6 minimum really means
The 1 MΩ minimum in IEC 60364-6 is a pass/fail threshold for installation verification — it’s the absolute floor. A cable that reads exactly 1 MΩ passes the standard but is not in great shape. New cables should read much higher — typically 100 MΩ or more for short runs, and proportionally less for longer runs due to the length effect.
Step-by-Step Test Procedure
Before testing
- De-energize the circuit. Open the breaker or disconnect. Lockout/tagout.
- Verify dead. Use a voltage tester to confirm no voltage is present.
- Disconnect both ends. Disconnect the cable from all equipment at both ends — switchboard, motor, junction box, panel. This is critical. If you leave equipment connected, you’re testing the equipment’s insulation in parallel with the cable, which gives misleading results.
- Disconnect sensitive electronics. VFDs, PLCs, soft starters, surge protection devices, and any electronic equipment must be disconnected before applying test voltage. These devices can be damaged by the DC test voltage and will skew your readings.
- Discharge the cable. Short all conductors to earth for at least 60 seconds. Long cables store significant charge.
The test
6. Connect the megger.
For a single-core cable:
- LINE lead → conductor
- EARTH lead → cable sheath / armor / ground
For a multi-core cable (testing one core at a time):
- LINE lead → conductor under test
- EARTH lead → all other conductors + sheath + armor, all connected to ground
- GUARD lead → see guard terminal section below
7. Select test voltage per the table above (500V DC for most LV cables).
8. Apply voltage for 60 seconds minimum. Record readings at:
- 30 seconds (for DAR calculation)
- 60 seconds (the standard spot reading)
- 10 minutes (for PI calculation, if testing longer runs or investigating a problem)
9. Record the reading along with: cable ID, length, ambient temperature, humidity, and test voltage.
After testing
10. Discharge the cable. Short all conductors to earth for at least 4 times the test duration. Long cables at high test voltages can hold dangerous charge for minutes. I’ve measured over 50V on cable conductors several minutes after a 5 kV test on a 500-meter run.
11. Reconnect equipment only after discharge is confirmed.
How to Interpret Cable IR Readings
General interpretation for LV cables (up to 1 kV)
| IR Reading | Condition | Action |
|---|---|---|
| Above 100 MΩ | Excellent | New or like-new insulation. No concerns. |
| 10 MΩ – 100 MΩ | Good | Normal for in-service cables. Continue monitoring. |
| 2 MΩ – 10 MΩ | Acceptable | Aging but functional. Increase test frequency. Investigate if declining trend. |
| 1 MΩ – 2 MΩ | Marginal | Meets IEC 60364-6 minimum but barely. Investigate cause — likely moisture or contamination. |
| Below 1 MΩ | Fail | Below IEC 60364-6 minimum. Do not energize. Find and fix the fault. |
Important context for cable readings
Cable IR readings depend heavily on cable length. A 100-meter cable will read about 10 times higher than a 1,000-meter cable of the same type and condition, because the longer cable has 10 times more insulation surface area for leakage current to flow through.
This means you can’t directly compare readings from cables of different lengths without normalizing. See the next section.
The Length Factor: IR per Kilometer
Insulation resistance is inversely proportional to cable length. If you double the length, you halve the measured IR — assuming the insulation quality is uniform.
The normalization formula:
IR per km = IR_measured × cable length (in km)
Example: You measure 50 MΩ on a 200-meter cable.
IR per km = 50 × 0.2 = 10 MΩ·km
Now you can compare this to other cables regardless of length.
Typical new cable IR values (per km, at 20°C)
| Cable Type | Typical New IR (per km) |
|---|---|
| PVC-insulated LV cable | 50 – 500 MΩ·km |
| XLPE-insulated LV cable | 500 – 5,000 MΩ·km |
| XLPE MV cable (6–36 kV) | 1,000 – 10,000 MΩ·km |
| Mineral-insulated (MI) cable | 100 – 1,000 MΩ·km |
These are general ranges. Check the cable manufacturer’s data sheet for specific minimum values.
When length normalization matters
If you’re comparing the condition of two cables of different lengths — normalize. If you’re checking whether a cable meets the IEC 60364-6 minimum of 1 MΩ — don’t normalize. The standard’s minimum applies to the measured value of the complete installed cable, not per kilometer.
Using the Guard Terminal on Cables
The guard terminal is one of the most useful and most underused features on a megohmmeter. For cable testing, it solves a specific problem: surface leakage.
The problem
When you test a cable in a damp environment — or a cable with dirty terminations — leakage current flows not just through the bulk insulation but also across the insulation surface at the cable ends. This surface leakage adds to the measured current and makes the insulation resistance reading artificially low.
The solution
The guard terminal diverts surface leakage current away from the measurement circuit. You connect the guard lead to a conductor that intercepts the surface leakage path.
For multi-core cables: Connect the LINE lead to the core under test, the EARTH lead to ground, and the GUARD lead to the other cores bundled together. Surface leakage between cores is diverted through the guard circuit and doesn’t affect the reading. What you measure is purely the insulation resistance between the test core and ground.
For single-core cables with exposed terminations: Wrap a bare copper wire around the cable jacket near each termination and connect it to the GUARD terminal. This intercepts moisture-driven surface leakage at the cable ends.
When to use it
Always use the guard terminal when available, especially on long cable runs, damp environments, and when you’re getting unexpectedly low readings that you suspect are caused by surface contamination rather than bulk insulation degradation.
Multi-Core Cable Testing
Multi-core cables (3-core, 4-core, 5-core) require a systematic approach to test all insulation barriers.
What to test
For a 3-core cable (L1, L2, L3) with armor/sheath:
- L1 to ground — L2 and L3 connected to ground (or to guard)
- L2 to ground — L1 and L3 connected to ground (or to guard)
- L3 to ground — L1 and L2 connected to ground (or to guard)
- L1 to L2 — L3 and ground connected together (or L3 to guard)
- L1 to L3 — L2 and ground connected together (or L2 to guard)
- L2 to L3 — L1 and ground connected together (or L1 to guard)
That’s 6 measurements for a 3-core cable. For a 4-core cable, it’s 10 measurements. For a 5-core, it’s 15.
Simplified approach for routine testing
For routine maintenance, most technicians use a simplified method:
- Connect all cores together → test to ground (one measurement)
- If that passes, test each core individually to ground with other cores grounded
The “all cores together” test is a quick screen. If it passes, the insulation-to-ground is acceptable for all cores. If it fails, test individually to find which core has the problem.
The core-to-core tests (steps 4–6 above) are important for commissioning but can be skipped in routine maintenance unless there’s a specific concern about inter-core insulation.
Special Cases
VFD (Variable Frequency Drive) cables
Cables feeding VFD-driven motors carry high-frequency switching transients that stress the insulation differently from normal 50/60 Hz operation. These transients can cause premature insulation breakdown that a standard megger test might not detect in its early stages.
For VFD cables, I recommend testing at the higher end of the allowed test voltage range and also performing a PI test (not just a spot reading). A declining PI trend on a VFD cable is an early warning of switching-stress degradation.
Always disconnect the VFD from the cable before testing. The VFD’s power electronics will be damaged by the megger’s DC test voltage.
Armored cables
Test the insulation between the conductors and the armor separately from the insulation between cores. The armor should be connected to the EARTH terminal during conductor-to-ground tests.
Also test the outer sheath insulation (armor to ground) if the cable is direct-buried. A damaged outer sheath allows moisture to reach the armor, which then provides a path for corrosion and eventual insulation failure.
Shielded cables (instrument and control)
For shielded cables used in instrumentation and control circuits, test the insulation between the signal conductors and the shield. Use 250V DC or 500V DC test voltage — check the cable manufacturer’s recommendation, as some instrument cables are only rated for 250V test voltage.
Disconnect all connected instrumentation before testing. Transmitters, PLC analog inputs, and signal conditioners can be damaged by the test voltage.
Cable Testing After Installation vs Maintenance
After installation (commissioning)
IEC 60364-6 requires insulation resistance testing as part of installation verification. The test must be done:
- With all loads disconnected
- With all switches in the closed (on) position
- With all lamps removed
- Between each live conductor and earth
- Between live conductors
This verifies the cable and the installed wiring are free of insulation damage from installation (pulling, bending, termination) and free of contamination.
All readings must meet the minimums in the standard (≥1 MΩ at 500V DC for circuits up to 500V). Record the results — these become your baseline for future comparison.
Routine maintenance testing
For routine maintenance, the procedure is similar but you’re comparing against the baseline and looking for trends. Key differences:
- You may need to test with equipment connected if full isolation isn’t practical — but note this in the results, as connected equipment will lower the reading
- Focus on cables in harsh environments: underground, outdoor, high temperature, high humidity, chemical exposure
- Flag any cable showing a 50%+ decline from its previous reading or baseline
After a fault or repair
Any cable that has experienced a fault, been repaired, or been spliced must be retested before re-energization. Test at the full commissioning voltage and compare to the original baseline.
Setting Up a Cable Testing Program
Recommended test intervals
| Cable Environment | Test Frequency |
|---|---|
| Indoor, dry, clean environment | Every 3–5 years |
| Indoor, industrial (dust, heat, vibration) | Every 1–3 years |
| Outdoor, exposed to weather | Every 1–2 years |
| Underground / direct buried | Every 1–2 years |
| High humidity or chemical exposure | Annually |
| After any modification or fault | Immediately |
| VFD-fed cables | Annually |
What to record
For each cable test, document: cable ID, circuit description, cable type and length, test voltage, readings at 30s/60s/10min, ambient temperature, humidity, PI and DAR (if calculated), date, and tester name. Store digitally and plot trends.
Common Mistakes
Testing with equipment still connected. This is the most common error. A motor with 50 MΩ insulation connected to a cable with 200 MΩ will give a combined reading of about 40 MΩ. You’ll think the cable is worse than it is. Always disconnect both ends.
Not accounting for cable length. A 500-meter cable reading 5 MΩ is in better condition than a 50-meter cable reading 5 MΩ. Normalize to per-kilometer values for meaningful comparisons.
Using too high a test voltage. Some instrument cables and low-voltage control cables are only rated for 500V test voltage. Applying 1,000V or 2,500V can damage the insulation and give you the very fault you were testing for. Check the cable data sheet.
Skipping the discharge. Long cables at high test voltages store dangerous energy. I once measured 85V on a 300-meter cable 3 minutes after a 2,500V test. Always discharge for 4× the test duration.
Testing in rain or high humidity without the guard terminal. Moisture on cable terminations creates surface leakage that gives artificially low readings. Use the guard terminal or wait for dry conditions.
Testing only conductor-to-ground, not core-to-core. A cable can pass the conductor-to-ground test while having compromised insulation between cores. For commissioning, test both.
FAQ
What is a good insulation resistance for a cable?
For a new LV cable, expect 100 MΩ or more for short runs (under 100 meters). Per IEC 60364-6, the absolute minimum is 1 MΩ at 500V DC for circuits up to 500V. In-service cables should read well above this minimum — anything below 10 MΩ warrants investigation.
Does cable length affect the reading?
Yes. IR is inversely proportional to length. A 1,000-meter cable reads about one-tenth of a 100-meter cable in identical condition. Normalize to MΩ·km for valid comparisons between cables of different lengths.
Can I test a cable without disconnecting both ends?
You can, but the reading will include the insulation of whatever equipment is connected at the far end. This makes it impossible to isolate a cable fault from an equipment fault. For meaningful cable testing, disconnect both ends.
How long should I apply the test voltage?
Minimum 60 seconds for a spot reading. For PI calculation, 10 minutes. For long cables with high capacitance, the reading may take longer to stabilize — wait until the reading is stable before recording.
What causes a cable to fail an insulation test?
The most common causes are moisture ingress (damaged sheath or unsealed glands), physical damage from installation, thermal degradation from overloading, chemical attack in industrial environments, and simple age. For underground cables, rodent damage is also a common cause.
Key Takeaways
- IEC 60364-6 specifies 500V DC test voltage and ≥1 MΩ minimum for circuits up to 500V (Clause 6.4.3.3).
- Disconnect both ends of the cable before testing. Connected equipment distorts the reading.
- Cable IR is inversely proportional to length. Normalize to MΩ·km for comparing cables of different lengths.
- Use the guard terminal to eliminate surface leakage at cable terminations, especially in damp environments.
- New LV cables should read 100 MΩ+ for short runs. Anything below 10 MΩ in service warrants investigation.
- Discharge for at least 4× the test duration. Long cables at high voltage store dangerous charge.
- For multi-core cables, test conductor-to-ground and core-to-core during commissioning.
- Trend your readings. A single number tells you less than a series of readings over years.
Standards Referenced in This Article
| Standard | Key Content |
|---|---|
| IEC 60364-6 | Test voltages (250/500/1000V DC), minimum IR (0.5–1.0 MΩ), verification procedure (Clause 6.4.3.3) |
| IEC 60364-4-42 | Insulation must withstand 500V DC for 1 min (from your standard file) |
| IEC 60364-7-701/703/711 | Special locations: 500V AC/DC for 1 min (from your standard files) |
| IEC 60204-1 | Machinery cables: 500V DC, ≥1 MΩ (Clause 18.3) |
| IEEE 400 | Cable testing series (referenced for MV cable testing context) |