Rotating Machinery Insulation Resistance and Polarization Index

What it is

This form records the dc insulation resistance and polarization index of the armature (stator) and field windings of a rotating machine - synchronous, induction or dc machine, or synchronous condenser - rated 750 W or more, one row per winding tested, for maintenance, storage recovery, rewind or pre-energization checks. For each winding a dc test voltage from the IEEE 43-2013 band for the rated voltage is applied between the winding and ground; the insulation resistance is read at 30 s, 1 min and 10 min, the 1 min reading (IR1) is corrected to 40 C with the factor for the insulation system, and the polarization index (PI) is the 10 min reading divided by the 1 min reading, not temperature corrected. IR1 at 40 C is graded against the IEEE 43-2013 minimum for the winding type, and PI against the minimum for the insulation thermal class, on an insulated winding. The ratio of the 1 min to the 30 s reading is recorded only. IEEE 43 is a recommended practice: its minimums are applied as stated, and an owner or acceptance specification may set another value. High-potential, power factor, surge, partial discharge and running tests stay on the rotating machine acceptance forms.

Why it is performed

A winding's insulation loses its dielectric strength when it absorbs moisture during idle storage, from cooling-water leakage or from condensation in humid air, when contamination bridges its surfaces, or when it ages thermally over years of service. A low insulation resistance, or a polarization index that has fallen below its minimum, points to moisture, contamination or aging before the winding fails under load or during an overvoltage test, and IEEE 43-2013 asks for both minimums to be met before a machine is operated or subjected to an overvoltage test. Correcting the 1 min reading to 40 C makes one test comparable with the next, so the record also carries the previous IR1 at 40 C and PI from the asset history and computes the change; a sharp fall against earlier readings is flagged, since it can point to contamination, moisture or damage even when the minimums are still met. Confirming the insulation before a rewound, reconditioned or long-idle machine is energized catches a problem while it is still a drying or cleaning job.

When it is performed

The test is recorded for maintenance, for recovery from storage, after a rewind and as a pre-energization check, on the interval the owner's maintenance program assigns to the machine; the time out of service is recorded with the test conditions. Both IEEE 43-2013 minimums must be met before the machine is operated or an overvoltage test is run, so the test precedes the high-potential, surge or partial discharge tests recorded on the acceptance forms. The winding temperature decides how the reading is used: a winding at or below the dew point can carry a moisture film that lowers IR and PI, and the 40 C correction is then unreliable for trending, so the machine history governs; a PI below its minimum on a winding above 60 C calls for a repeat below 40 C and above the dew point, and a high PI on a hot winding that cooled during the test calls for a repeat at or below 40 C. A directly water-cooled winding is drained and dried before the test.

How it is typically performed

The machine is identified with its type, rated output and horsepower, rated voltage (line-to-line for a three-phase ac machine, line-to-ground for single-phase, the rated direct voltage for a dc machine or field winding), insulation thermal class, insulation system (thermoplastic or thermosetting, or unknown with the manufacture year) and the winding type that selects the IEEE 43-2013 minimum. Before the test, external equipment is disconnected and grounded, the items left connected are listed, the winding is discharged with no return voltage, and personnel access to the test voltage is restricted; ambient temperature, relative humidity, dew point and winding temperature with its method are recorded, with the cooling water conductivity where water stays in the winding. For each winding row the technician selects the construction (insulated winding, non-insulated field or dc armature), the winding tested (armature all phases, one phase with the others grounded, or field) and its connection, enters the dc test voltage applied, usually negative polarity on the winding, and reads the insulation resistance at 30 s, 1 min and 10 min with the winding temperature at 1 min and 10 min. The form computes the correction factor to 40 C from the IEEE 43-2013 equation for the insulation system or takes a field-determined factor with its curve reference, corrects IR1, looks up the minimum for the winding type (which an engineer may set aside under the standard's notes with the basis recorded), computes PI and its minimum for the thermal class, and raises flags for a test voltage outside the band, a thermosetting correction above 60 C, a high PI on a thermoplastic system, a low PI above 60 C, a winding that cooled during the test, a reading at or below the dew point and a marked IR drop with rising voltage when a step series was run. After the test the winding is grounded for a discharge time of at least 4 times the voltage application time; both times are entered. The megohmmeter make, model, serial number, calibration date and maximum test voltage close the record.

What gets recorded

Identification: manufacturer, model, serial number, designation, machine type, rated output and horsepower, rated voltage, field winding rated dc voltage, insulation thermal class and field class, insulation system, winding manufacture year and the winding type for the minimum. Test conditions and safety: time out of service, ambient temperature, relative humidity, dew point, winding temperature and method, water-cooled status and cooling water conductivity, external equipment disconnected and grounded, items left connected, and the winding discharged before the test. Per winding, up to six rows: construction, winding tested, connection, the test voltage band and the dc voltage applied, polarity, IR at 30 s, 1 min and 10 min, winding temperature at 1 min and 10 min, the correction basis with any field-determined factor and curve reference, the computed correction factor and IR1 at 40 C, the minimum IR1 at 40 C with any override and its basis, the IR1 margin, the polarization index with its minimum and margin, the 1 min to 30 s ratio, the minimums-met indicator, the step-voltage series result and the row's flags. Discharge after test: voltage application time, discharge time and their ratio, and the personnel access check. Trend: previous test date, previous IR1 at 40 C and PI, the computed changes for winding 1 and whether a sharp fall was observed. Test equipment, the megohmmeter's make, model, serial, calibration date and maximum test voltage, comments and deficiencies.

How results are evaluated

The record passes when the safety and setup checks are satisfactory (external equipment disconnected and grounded, the winding discharged before the test, personnel access restricted; one unsatisfactory answer fails, all blank leaves the record indeterminate), when IR1 corrected to 40 C is at or above the IEEE 43-2013 minimum for the winding type on every row where it applies, when PI is at or above the minimum for the insulation thermal class on every insulated winding, and when the discharge time is at least 4 times the voltage application time with both times entered. The minimum insulation resistance applies to the entire winding: an all-phases row is graded, a one-phase row beside it is flagged only, and one-phase rows alone are graded as the three phases in parallel, which needs all three. PI is not graded for a non-insulated field winding or a dc machine armature and is disregarded when IR1 or IR1 at 40 C exceeds 5000 megohms; a thermal class outside the standard's table leaves PI recorded with an issue. An engineer may set the minimum aside under the standard's notes (very large end-arm surfaces, a dc armature with a commutator, a stress-control coating over the whole overhang, or a green winding before vacuum impregnation) only with the basis recorded, and every override raises an issue line; a stress-control coating claim sets aside the IR1 minimum only, so PI is still graded. A row is in use once its IR1 is entered and must be complete - its correction factor and minimum computed, its winding type consistent with the record, its IR10 and thermal class present where PI applies, the field rated voltage present on a field row - or the record is indeterminate unless another reading fails. Issue flags that never change the verdict: a rating below 750 W or 1 hp, a test voltage outside the band, a thermosetting correction above 60 C, an insulation family assumed from the manufacture year, a PI above about 8 on a thermoplastic system, a low PI above 60 C, a hot winding that cooled during the test, a reading at or below the dew point, cooling water at or above 0.25 uS/cm, a marked IR drop with rising voltage, and a sharp fall against the previous reading. The minimum values and the correction equations are in IEEE 43-2013 and are not reproduced on this page.