6.1 DC Hipot: Step Voltage, Leakage Current & Pass Criteria

What a DC Hipot (Overpotential) Test Does

A DC high-potential test, also called an overpotential test, applies a DC voltage significantly higher than the equipment's normal operating voltage to prove the insulation can hold off service stress. It is fundamentally a pass/fail proof test, not a diagnostic measurement: either the insulation withstands the applied voltage for the specified duration or it breaks down. This distinguishes it from an insulation resistance (IR) test, which measures dielectric quality at a voltage at or below operating level and reports a megohm value you can trend.

The test is most commonly applied to solidly insulated apparatus — motor and generator stator windings, switchgear bus, cable, and dry-type transformers. For liquid-filled transformers, AC hipot at power frequency is preferred because it replicates service stress; DC is used selectively.

Step-Voltage Method

The step-voltage method raises the test voltage in equal increments, holds each step long enough for the capacitive and absorption currents to settle, and records the leakage current at each step. A typical ramp might be 1 kV steps on a 13.8 kV class winding, holding each step 60 seconds.

The diagnostic power is in the trend, not the absolute number. Healthy insulation produces a roughly linear, stable leakage current at each step. A non-linear rising knee — where current jumps disproportionately at a higher step — indicates the insulation is breaking down at that stress level. This is the same principle as a megger ramp test: you are hunting the knee-point where leakage accelerates.

Leakage Current vs. Absorption Current

When you energize a DC hipot on a large winding or long cable, three current components flow:

1. Capacitive charging current — decays in seconds as the几何 capacitance charges.
2. Dielectric absorption (polarization) current — decays over minutes as dipolar molecules in the insulation align.
3. Conduction (leakage) current — the steady, constant current through and across the insulation that remains after charging and absorption complete.

The current a hipot reads at the moment of energization is dominated by charging plus absorption. A normal leakage-current graph starts high and decays downward to a steady plateau as absorption completes — this is healthy. The trap is misreading that decay as a problem. What you are watching for is the steady-state value and its stability over time. A sudden increase, a failure to stabilize, or a rising trend at constant voltage are the danger signatures.

DC Hipot Pass Criteria (NETA MTS)

Per ANSI/NETA MTS, DC hipot acceptance is judged on three conditions, not on a single leakage number:

• Leakage current is stable at test voltage (not ramping upward).
• Leakage current does not exceed the manufacturer's specified limit where one is published.
• No flashovers, arcs, or breakdowns occur during the hold period.

NETA does not publish a universal leakage-current ceiling because the value depends on capacitance, temperature, humidity, and geometry. For a 600 V class motor tested at roughly 1.2–2.5 kV DC, the manufacturer's limit governs. Many programs also require the Polarization Index (PI) or Dielectric Absorption Ratio (DAR) to be in an acceptable range before hipot is even attempted.

Typical DC Test Voltages — Conversion from Factory AC

Factory AC withstand tests for new rotating machinery use the NEMA MG-1 formula 2E + 1000 V AC RMS (E = rated line-to-line voltage). IEEE Std 95 converts this to a DC equivalent using the 1.7× factor: V_DC = 1.7 × V_AC(rms). The 1.7× is a conservative compromise — well-compacted slot insulation has a DC/AC strength ratio of 2–3, but leads and cross-connections cannot achieve the same compaction.

For a 13.8 kV machine, the field-acceptance DC voltage is roughly 0.75 × 1.7 × (2×13,800 + 1,000) ≈ 36,500 V DC. The maintenance band lands around 29,000–39,000 V DC.

For cables, the conversion is different and higher (1.6–3.0× depending on insulation), and IEEE 400 no longer recommends DC hipot for XLPE cable in service — VLF has replaced it.

Safety: Discharge and Ground

A DC hipot leaves significant stored dielectric charge in the insulation — particularly in large windings and long cables. After the test:

1. Reduce voltage to zero gradually (do not just switch off).
2. Apply a grounded discharge rod to the test specimen.
3. Keep the ground in place for the manufacturer-specified discharge time — at minimum several minutes, longer for high-capacitance loads.
4. Verify with a megohmmeter that voltage has decayed before touching terminals.

Skipping the discharge step has caused fatal contact incidents. The absorption charge can persist long after the power is off.

Common Field Traps

• Misreading the absorption decay as a failure — the initial downward slope is normal, not a problem. The danger signature is a rising trend at constant voltage.
• Using an absolute leakage-current ceiling — NETA MTS does not publish a universal one; the manufacturer's limit and the stability criterion govern.
• Skipping PI/DAR before hipot — wet or dirty insulation can fail catastrophically under hipot if a low PI was ignored.
• Applying the IEEE 95 1.7× factor to cables — that factor is for rotating machinery; cable DC voltages come from IEEE 400 / NETA ATS cable tables.