The test itself is unchanged across asset types: a direct voltage is applied, the resulting current is measured and a resistance is derived. What changes is the level of that voltage, the duration for which it is applied and the connection arrangement, and those three determine what the resulting number means.

What IR proves and what it does not

An insulation resistance measurement proves that a leakage path of a given resistance exists between the points the test voltage is applied across, at the voltage and under the conditions of the test. It is a bulk measurement, which means it responds to the total condition of the insulation system rather than to any localised defect.

What it therefore proves is that there is no gross conduction path of low resistance, and that the value is consistent with the previous measurement on the same asset under comparable conditions. What it does not prove is that the insulation is dry in a quantitative sense, that the solid insulation is undamaged, or that no localised defect exists.

The distinction matters for the choice of voltage because the purpose of the test determines how sensitive it needs to be. A test performed as a safety check before work begins needs only to establish that no dangerous conduction path exists. A test performed as part of a condition assessment programme needs the sensitivity that comes with the standard’s specified voltage, because it is being compared with a series of earlier values.

Test voltage by asset and standard

The transformer requirements are defined in IEC 60076-1, with the dielectric test framework in IEC 60076-3. The rotating machine requirements are defined in IEC 60034-1, and the installation context that governs low and high voltage equipment in service is covered by IEC 60364-1 and IEC 61936-1.

Reading a time-resistance curve

The resistance measured across an insulation system is not constant while the test voltage is applied. It rises over time as the absorption currents decay, and the shape of that rise carries information about the insulation condition. A dry insulation system continues to rise for several minutes; a wet one reaches its steady value quickly.

The values read at fixed times are what make the curve usable. A reading at one minute is often specified as the basic measurement, and a reading at ten minutes is used to derive the ratio that indicates the absorption behaviour. That ratio is what makes a wet insulation system visible even when the absolute resistance is within a general limit.

The curve also shows whether the insulation is stable. A resistance that falls during the test, rather than rising or holding, indicates that the applied voltage is causing deterioration rather than measuring a static condition. That behaviour is a reason to stop the test rather than to record a value from it.

Temperature and moisture effects

Insulation resistance falls as temperature rises, and the effect is substantial. A measurement taken on a warm asset and one taken on a cold asset differ because of the temperature rather than because of any change in the insulation, and the difference can be large enough to move a result across a limit.

Moisture acts on both the bulk and the surface. Water in the insulation lowers the bulk resistance; a film of moisture on an insulator surface provides a parallel leakage path that sits outside the insulation being assessed but is measured as part of it. The two effects are not distinguished by the resistance measurement alone, which is why the surface condition and the weather belong in the record.

The temperature and moisture corrections available are approximate, and applying them produces a value that appears more precise than the data behind it. The more reliable practice is to take measurements under comparable conditions and to record those conditions so the comparison is made on a like-for-like basis.

Comparisons: previous test and sister units

The unit’s own previous result is the strongest available reference, provided the conditions are comparable. Comparing a measurement taken at one test voltage with one taken at another, or one taken at a different temperature, produces a difference that describes the test rather than the asset.

Where no previous result exists, a comparison against a sister unit provides a weaker reference, because nominally identical assets can differ through construction, moisture history and measurement configuration. The comparison is useful for identifying a unit that is markedly different from its group and much less useful for detecting a small change.

Comparing across phases of the same asset is often the most informative comparison available, because the phases share a tank, a temperature and a test configuration. The spread between phases is a quantity with few confounding variables, and a phase that differs from its neighbours warrants investigation even when all three values are within any general limit.

Diagnostic ratios worth calculating

The ratio between the resistance at ten minutes and at one minute is the standard absorption indicator. The ratio between the values at sixty and thirty seconds is a shorter equivalent used where time is limited. Both respond to moisture and to the condition of the insulation, and both are more meaningful when a series of values exists for the same asset.

Where the test arrangement allows a measurement between windings as well as between windings and earth, the relationship between the two results adds information about where a problem may lie. A low value between windings and a normal value to earth points at the inter-winding insulation, while the reverse points at the insulation to the tank or core.

Calculating the ratios from the raw timed values rather than recording only the ratio is important, because the ratio compresses the data. The original timed values allow a later reviewer to recalculate or to apply a different criterion, and they allow the trend of each individual value to be followed.

Limits of the method on wet insulation

The relationship between the measured resistance and moisture content is not linear. At low moisture levels the resistance changes substantially for a small change in water content, which makes the method sensitive; at higher moisture levels the resistance approaches a value dominated by the dissolved impurities in the water rather than by the water content itself, which makes the method insensitive to further change.

The consequence is that a very low reading indicates a problem but does not quantify it, and successive low readings may show little change while the condition continues to deteriorate. Where moisture content matters quantitatively, the measurement that answers the question is a direct determination on an oil sample or, for solid insulation, a dielectric response measurement.

The method is also insensitive to localised defects because it is a bulk measurement. A defect occupying a small fraction of the insulation path is diluted by the healthy material around it, and the resulting change in the overall resistance may be within the measurement variation.

Recording and reporting the result

The record should state the test voltage, the duration of the test and the times at which values were read, the temperature and how it was obtained, the connection configuration including which parts were earthed and whether a guard connection was used, and the discharge procedure followed before and after the measurement.

The instrument details belong in the record too, because the capability of the instrument determines what measurements are possible. A tester that cannot apply the specified voltage, or that stops at one minute, cannot produce a measurement equivalent to one taken with a different instrument. Suppliers such as Fluke and Hioki publish the voltage ranges, guard arrangements and timing functions of their instruments, and the calibration traceability that supports a disputed result is described in the NIST handbooks.

The trend reading that this measurement feeds into, including the absorption ratios and how they are interpreted over a series of results, is covered in the accompanying article on insulation resistance and polarisation index on power transformers. The instrument range is grouped on the electrical resistance testing hub.