The power factor of electrical insulation and the power factor of a load circuit share a name but they describe different things. Understanding what the insulation power factor number actually represents — and why it responds to moisture, voids, and contamination the way it does — makes it easier to interpret test results correctly and explain them to someone who has never run a Doble set.
Electrical insulation between two conductors — a transformer winding and the tank, a cable conductor and its shield, a bushing conductor and the flange — behaves electrically like a capacitor. The insulating material is the dielectric, and when you apply AC voltage across it, a current flows. That current has two components.
The first component is the charging current, which is 90 degrees out of phase with the applied voltage. In a perfect insulator, this would be the only current — no real power consumed, purely reactive. A perfect capacitor has zero losses and zero power factor.
The second component is the loss current, which is in phase with the voltage and represents real power dissipated as heat in the dielectric material. Every practical insulator has some loss. The ratio of that loss current to the total current flowing is the power factor of the insulation. A value of 1.0% means the loss current is 1% of the total current — or equivalently, that 1% of the volt-amperes applied to the insulation are consumed as real watts rather than stored and returned.
The complementary quantity, dissipation factor or tan delta, is used interchangeably with power factor at the low values seen in good insulation. At 1% power factor, the dissipation factor is essentially the same number. The distinction matters at higher values, but for routine insulation testing on oil-impregnated paper and similar materials, the two terms are used interchangeably in practice.
The dielectric losses in good, dry oil-impregnated paper insulation are low because the material does not conduct current readily and because the molecular dipoles within it do not reorient significantly at power frequencies. When the insulation is contaminated or physically degraded, that changes.
Moisture is the dominant driver. Water molecules are polar and respond strongly to the alternating electric field, reorienting twice per cycle and dissipating energy each time. Even small amounts of dissolved moisture in transformer oil or within paper insulation increase dielectric losses substantially. Elevated moisture also provides a conducting path that adds a resistive component to the loss current — the current no longer leads the voltage by a full 90 degrees and the power factor rises.
Contamination such as carbon from arc interruption or metallic particles creates local conducting regions within the insulation. These regions increase the effective conductance of the dielectric, adding resistive loss current and raising the measured power factor.
Partial discharge activity within voids or delaminated layers produces carbonization and chemical byproducts that are more conductive than the surrounding clean insulation. A bushing or winding section with active PD shows elevated power factor in the affected section, which is why testing sections separately — C1 vs. C2 on a bushing, or CHL vs. CH vs. CL on a transformer — helps localize where the problem is.
Thermal aging in paper insulation produces oxidized, brittle cellulose with a different molecular structure than new paper. Aged paper has higher dielectric losses than new paper even at the same moisture content, which is one reason older transformers that have always tested within acceptable limits begin drifting upward over decades of service without any single identifiable contamination event.
A power factor test set applies a stable AC voltage — typically 2.5 kV or 10 kV — to the insulation under test and measures the total current and the component of that current in phase with the voltage. From these two measurements the instrument calculates power factor directly.
The test voltage level matters because the dielectric loss behavior of some insulation materials is voltage-dependent. At voltages below partial discharge inception, the measurement reflects only the intrinsic dielectric losses. At voltages above PD inception, discharge activity adds additional loss current that increases with voltage. This is why a PD-active specimen may show a power factor that rises nonlinearly as the test voltage is stepped up — the additional current at higher voltages is discharge current, not ordinary dielectric loss.
The measurement is made at power frequency, typically 60 Hz in the U.S., because that is the frequency at which the insulation operates in service. Testing at power frequency ensures that the dipole reorientation losses, which are frequency-dependent, match actual in-service conditions. Testing at higher or lower frequencies gives different results and is done for specific diagnostic purposes, not as a substitute for standard power frequency tests.
Insulation resistance (megohm) testing with a DC megohmeter is the simpler and older test, and it is still useful as a go/no-go check and for trending gross degradation. But power factor testing catches problems that megohm testing misses, and the reason is physical.
A megohm test applies DC and measures steady-state leakage current. It is sensitive to through-path conduction — a direct contamination path from high-voltage to ground. It is much less sensitive to distributed dielectric losses throughout the bulk of the insulation. A winding with early-stage moisture absorption can show millions of megohms on a DC test while already showing elevated power factor, because the moisture raises the dielectric losses without yet providing a clear DC conduction path to ground.
Power factor testing is also less sensitive to surface leakage paths than megohm testing. The guard terminal on a power factor test set diverts surface leakage current away from the measurement circuit, so the result reflects the bulk insulation condition rather than the cleanliness of the test specimen’s exterior. This is why power factor results are more stable and reproducible across different ambient conditions than megohm readings.
Neither test replaces the other. A complete insulation assessment runs both: the megohm test for gross condition and the power factor test for dielectric quality. A specimen that passes one but fails the other has a problem worth investigating.
Dielectric losses in oil-impregnated paper and mineral oil increase with temperature. This means the raw power factor reading from a field test is temperature-dependent and cannot be compared directly to published limits or to prior tests made at different temperatures without correction.
Standard correction is to 20°C (68°F). Correction factors vary by insulation type, and test set manufacturers publish correction tables for common materials. For oil-impregnated paper, a rough rule is that the power factor roughly halves for every 10°C drop in temperature below 20°C, and roughly doubles for every 10°C rise above 20°C. This makes a test at 40°C appear to fail a limit that the same specimen would pass when corrected to 20°C — and it makes a cold-weather test potentially miss a borderline condition.
Record the actual specimen temperature at the time of the test along with the raw reading and the corrected result. When comparing readings across multiple test cycles, compare corrected values. If the correction factor applied was large — because the test was done far from 20°C — treat the corrected result with somewhat less confidence and note the actual test temperature in the record.
Our field crews run power factor tests on transformers, bushings, and circuit breakers across Florida and the Southeast. Send us your equipment list and we’ll respond within one business day.