5.1 Insulation Resistance: Test Voltages & IEEE 43 Minimums

What an Insulation Resistance Test Measures

An insulation resistance (IR) test applies a regulated DC voltage between the conductor and ground (or between windings) and measures the leakage current through and across the insulation. The megohmmeter reports the result as resistance in megohms (MΩ) — the ratio of applied voltage to leakage current. A healthy insulation system has high resistance (millions of ohms); moisture, contamination, cracking, and aging all drive the reading down. The IR test is non-destructive: the applied voltage is at or below the equipment's operating level, so it diagnoses dielectric quality without stressing the insulation beyond service conditions.

Selecting the Test Voltage

The test voltage is chosen from the equipment's rated voltage, not picked arbitrarily. Too low a voltage will not stress the dielectric enough to reveal defects; too high a voltage for low-rated gear can stress insulation unnecessarily. ANSI/NETA MTS Table 100.1 (mirrored in ATS Table 10.1) gives the standard selection for apparatus and systems other than rotating machinery:

Two patterns to memorize for the Level II exam: the test voltage steps up as equipment class rises (500 → 1,000 → 2,500 → 5,000 V DC), and the minimum IR rises steeply with class — a 600 V cable needs 100 MΩ, a 15 kV class cable needs 5,000 MΩ. The 5,000 V DC megger is the workhorse for medium-voltage work because it covers everything from 25 kV class up to the largest apparatus. For rotating machinery, use IEEE 43 test voltages (Table 1 of IEEE 43-2013): roughly 500 V DC for machines rated at or below 1 kV, 1,000 V for 1.1 to 2.4 kV, 2,500 V for 2.5 to 5 kV, and 2,500 to 5,000 V for 5.1 to 12 kV. Applying 5,000 V DC to a 600 V class cable or switchgear for 1 to 10 minutes is within NETA MTS guidance and supports PI calculation and detection of wet or contaminated insulation.

IEEE 43-2013 Minimum IR Values

For rotating machinery, IEEE Std 43-2013 (Clause 12.3, Table 4) defines three minimum IR₁ values measured at 40°C after 1 minute:

The kV+1 rule is the legacy formula: a 13.8 kV machine would need at least 14.8 MΩ. It was calibrated for older asphaltic-mica and thermoplastic systems and still applies to all field windings regardless of age. For modern epoxy-mica stators — the overwhelming majority of post-1970 machines, including virtually all machines rated below 12 kV with form-wound epoxy coils — the flat 100 MΩ floor controls regardless of voltage rating. Applying kV+1 to a modern 4.16 kV motor would set the bar at 5.16 MΩ, roughly 20× too low; the 100 MΩ floor is the stricter and correct acceptance value. The 5 MΩ floor applies to random-wound and at-or-below-1 kV windings where geometry makes higher values unrealistic.

Temperature Correction to 40°C

IR is strongly temperature-dependent: resistance roughly halves for every 10°C increase in winding temperature. IEEE 43 requires correcting the measured IR₁ to a reference 40°C using the appropriate correction factor — thermoplastic systems use one curve, thermosetting (epoxy) systems another. Always record the winding temperature at the time of test (RTD or thermometer) and apply the correction before comparing to the minimum or to prior readings. Comparing an uncorrected 25°C reading to a prior 40°C reading will make good insulation look falsely good; comparing a hot 55°C reading uncorrected will make it look falsely bad.

A Worked Example

A 4.16 kV form-wound induction motor is tested at 2,500 V DC. The 1-minute reading is 320 MΩ at a winding temperature of 55°C. The 100 MΩ floor applies (modern epoxy). Correcting 55°C to 40°C — a 15°C rise, which roughly doubles resistance per 10°C, so factor of about 4 — gives an IR₁(40°C) of approximately 1,280 MΩ, well above the 100 MΩ minimum and a healthy result. If the same motor had read 90 MΩ at 25°C, correcting 25°C up to 40°C (a 15°C drop in temperature means the 40°C-equivalent reading is lower) would give roughly 320 MΩ, still comfortably above the floor — but the uncorrected 90 MΩ at 25°C is already close to the 100 MΩ line and a careless technician who skips correction could wrongly condemn the motor.

Interpreting Low IR and NETA Acceptance for MV Cables

A low IR reading is not, by itself, a verdict — it is a signal to investigate. Common causes of low IR on cable and apparatus:

• Moisture in the dielectric or terminations (readings often recover after drying)
• Surface contamination on bushings or terminal boards (use the guard terminal to isolate — see Section 5.4)
• Insulation aging or cracking (trending downward over years)
• Cable end moisture at splices or terminations
• Temperature not corrected to 40°C

Always compare to prior readings from the same equipment under similar conditions; a single absolute value is far less diagnostic than a trend. NETA MTS recommends comparing results to manufacturer data and to previous tests — a sudden drop to 20% of a prior value warrants investigation even if the absolute number exceeds the table minimum.

MV Cable IR Acceptance

For medium-voltage cables, NETA MTS Table 100.1 sets the acceptance floor by voltage class. A 15 kV class shielded cable tested at 2,500 V DC should read at least 5,000 MΩ; a 5 kV class cable at 2,500 V DC needs 1,000 MΩ. A 480 V (600 V class) feeder tested at 1,000 V DC needs 100 MΩ — so a 1-minute reading of 20 MΩ on a 480 V feeder is below the 100 MΩ floor and must be investigated for moisture or contamination before energization. For a 480 V feeder, applying a 5,000 V DC megger for 10 minutes is within NETA MTS guidance and is useful for PI calculation, but the 1-minute acceptance comparison is still against the 100 MΩ floor.

Note that for extruded-dielectric MV cables (XLPE, EPR), DC IR and DC hipot are diagnostic-only; NETA and IEEE 400.2 now prefer VLF and partial-discharge methods for acceptance, and DC hipot is reserved primarily for paper-insulated lead-covered (PILC) cable. IR is still run on all cable classes as a baseline and safety check before energization.

The Trap

Do not apply the rotating-machinery kV+1 formula to cable, switchgear, or bus work — IEEE 43 minimums are for machines only. Cables, bus, and switchgear use the NETA MTS Table 100.1 values. Mixing the two standards is a common Level II exam mistake: the 5 MΩ IEEE 43 floor for random-wound machines is not an acceptance value for a 600 V cable, which needs 100 MΩ.