How to field test 3-phase squirrel cage motors
Maintenance of critical machines depends on the diagnostic electrical testing of installed 3-phase squirrel cage motors, interpretation of results and key points of physical inspection
Efficient, reliable operation of critical electric motors is top of mind for maintenance professionals tasked with keeping production at optimum levels while avoiding costly, unexpected shutdowns. Besides routine maintenance, this requires that critical motors be inspected and tested regularly. The focus of this article is on diagnostic electrical testing of installed 3-phase squirrel cage motors, interpretation of results and key points of physical inspection. Most of these tests and inspections also apply to 3-phase wound rotor motors and induction and synchronous generators.
Inspection and testing
Besides visual inspection, offline condition assessment and diagnostic tests for 3-phase squirrel cage motors typically include insulation resistance (IR) and polarization index (PI) or dielectric absorption ratio (DAR) tests. Depending on operating conditions and availability of test equipment, offline testing and inspection also may include lead-to-lead resistance and surge tests, sampling lubricating oil for analysis and checking for soft-foot, output shaft runout and alignment of motor to driven equipment.
Inspection. The scope of the visual inspection will vary with the motor enclosure type. If the motor’s interior is not accessible (e.g., no removable covers), carefully inspect its external surfaces for wear, cracks and broken or missing hardware. Items to check include the frame, feet, terminal box, fan covers, cooling fans and the output shaft and coupling or other shaft-mounted components.
If the motor’s interior is accessible, it may be possible to inspect the windings and other internal components for defects or damage, including the air gap between the rotor and the stator (see Figure 1). A borescope and mirrors on extension rods can probe recessed areas like the rotor’s interior and the space between the stator core and frame to check for debris, contamination, blocked ventilation ducts, cracked welds or a loose fit of the rotor core to the shaft.
Record all damage and defects, remove debris and contamination and perform any maintenance or repairs that need immediate attention. If necessary, schedule nonessential maintenance or repairs for the next regular shutdown.
Insulation resistance tests. The IR test is well a defined method of evaluating the ground insulation of all types of motor windings (see Figure 2). It consists of applying the test voltage and measuring the winding’s resistance to ground after one minute. IR readings are temperature sensitive, so to be meaningful they should be corrected to the standard temperature of 40° C (see Table 1).
PI is an extension of the IR test and is calculated by dividing the IR reading at 10 minutes by megohm value at one minute. The recommended minimum PI value for windings rated Class B (130° C) and higher is 2.0, and 1.5 for Class A (105° C) windings. Windings with a lower PI value usually are unsuitable for service. If the IR value is greater than 5,000 megohms, per IEEE Std. 43 and IEC Std. 60034-27-4, the PI value would not be meaningful, and the PI test need not be performed.
The PI test is most useful with stator form coil windings (coils made with rectangular or square wire). It may not be meaningful for random-wound windings (coils made with round wire) because the winding absorption charging current decays within the first minute or so of applied voltage. For those windings, the DAR is more useful, with a common selection of IR readings taken at 30 seconds and 60 seconds per IEC Std. 60034-27-4.
Lead-to-lead resistance test. By comparing the phases or circuits in the winding, the lead-to-lead resistance test can detect high resistance joints in winding and lead connections. Per CSA C392 and ANSI/EASA AR100, the resistance unbalance limit for random windings should be 2% from the average, and 1% from the average for form coil windings.
Surge test. The surge test can detect turn-to-turn, coil-to-coil or phase-to-phase shorts. A common issue when surge testing an assembled motor is “rotor coupling” — a magnetic interaction between a squirrel cage rotor and the stator winding that can produce a dual trace of voltage as seen on the screen of a surge tester or an oscilloscope. Turning the rotor a few mechanical degrees will merge the traces, unless the winding has a fault or other defect (e.g., unbalanced winding circuits). Perform the surge test only if the winding has an acceptable IR value and, if applicable, an acceptable PI value.
Shaft runout test. Mechanical tests include the output shaft runout test, which uses a dial indicator to measure shaft displacement at the end of the shaft (if possible) or adjacent to the coupling during one revolution. NEMA Standard MG 1 (NEMA Std. MG 1) allows up to 0.003-inch (0.08 millimeter) total indicated runout (TIR) for shaft diameters of 1.625 inch to 6.500 inches (41 to 165 millimeters). A more rigorous yet simpler criteria is to limit runout to no more than 0.001 inch (0.025 millimeter) for 2-pole motors, 0.002 inch (0.051 millimeter) for 4-pole motors and 0.003 inch (0.076 millimeter) for motors with six or more poles.
Online motor testing
Online (running) tests vary by machine type (e.g., squirrel cage induction, synchronous, wound rotor). If the motor can operate safely, these may include measuring the starting (inrush) current, line-to-line voltages and voltage unbalance. On large motors or those powered by variable-frequency drives (VFDs), it also is important to check for shaft currents.
Inrush current test. Strictly speaking, inrush is the asymmetrical dc offset that occurs in the first cycle, or few cycles, after an ac motor is energized (see Figure 3). According to NEMA Std. MG 1, the inrush current can be 1.8 to 2.8 times the locked-rotor current, which is typically six to eight times the full-load current. Consequently, it could be as much as 22 (2.8 x 8) times the full-load current. For a motor with a higher than typical locked-rotor current, it can be high enough to trip circuit breakers. When taking measurements, unless the ammeter can measure momentary inrush (peak) current, it will only indicate the steady-state, locked-rotor current.
Line-to-line voltages test. Line-to-line voltages should be within 10% of the motor’s rated voltage, according to NEMA Std. MG 1 and within 5% per IEC Std. 60034-1 (10% for limited duration and frequency of occurrence). Too high a voltage can increase heating of the motor’s magnetic core, while too low a voltage can reduce its torque capability (see Table 2). There is no rule-of-thumb to estimate whether overvoltage will increase or decrease motor current, and likewise with undervoltage.
Unbalanced voltage test. Another factor related to voltage is unbalanced voltage. According to NEMA Std. MG 1, a motor should be de-rated if voltage unbalance exceeds 1% — a requirement often confused with the tolerance for voltage variation (see Figure 4). Utilities frequently limit the voltage unbalance for the power they supply to 3%, which, according to NEMA Std. MG 1, would require derating horsepower by 12%. Since this often is impractical, many motors end up operating on unbalanced voltages with reduced output torque and increased current. The higher current is especially significant because NEMA Std. MG 1 says current unbalance with load can be six to 10 times the percent voltage unbalance. Applying this rule to the 3% voltage unbalance, the current unbalance could be 18% to 30%.
Heating is a function of the power loss in a winding; specifically, the current squared times the resistance (I2R). With 3% voltage unbalance, the highest current “leg” of the winding may have about 18% more heating due to the associated current unbalance. The additional heating is estimated by calculating twice the voltage unbalance squared, in this case:
2 x 32 = 18%
A thermal scan of the windings, if accessible, can record the actual temperatures resulting from unbalanced voltage and current conditions.
Infrared thermographic scanning of a motor’s exterior also can indicate areas of abnormal heating (see Figure 5). While there are no specific temperature standards for the outer surface (“skin”) of electric motors, comparing a motor’s surface temperature with identical ratings under the same or similar load conditions may reveal abnormal heating.
Shaft currents tests. Large motors and motors supplied by variable-frequency drives (VFDs) should be checked for shaft currents (see Figure 6), even if none are suspected. In large motors, for example, magnetic circuit dissymmetry due to segmented laminations can induce shaft currents. Likewise, VFDs may link the rotor and stator by capacitive coupling, creating circulating “shaft” currents that can cause premature bearing failure.
Field testing is necessary to detect bearing currents from VFDs and some other causes. Measuring the current directly is not practical in this case, as it would require wrapping a current transformer around the shaft inside the motor — i.e., between bearings. The alternative is to measure the voltage from the frame to the shaft to determine if it is enough to indicate damaging shaft currents.
A way to measure shaft voltage in the field is to attach one lead of a true root-mean-squared (RMS) voltmeter to the frame (a grease fitting is a good location) and the other to the shaft using a brush-like device (e.g., a fine copper wire such as a brush shunt) to drag the shaft and sense the voltage. Directly sensing the voltage from the shaft with a meter lead is not recommended because it typically will not maintain continuous contact.
If the sensed voltage exceeds 100 millivolts ac for rolling bearings or 200 millivolts ac for sleeve bearings, damaging shaft currents are probably present. Another criterion from NEMA Std. MG 1 says damaging shaft currents may exist if the measured voltage between opposite ends of the shaft exceeds 300 millivolts ac.
Field testing and inspection of motors is an important part of maintaining essential and often critical machines. Taking time to learn about the proper tests and procedures, and how to apply them, will allow you to improve reliability and reduce costs.