MILL MAINTENANCE

By breaking a motor system down into individual fault zones and testing each completely, electrical and other motor defects are more accurately pinpointed

BY NOAH P. BETHEL

Fault Zone Analysis Identifies Motor Defects in Detail

For years, electrical maintenance personnel have been limited to troubleshooting motors with no more than a multimeter for measuring voltage and resistance and a megohmeter for measuring resistance to ground (RTG). Unfortunately, these instruments do not provide enough information to allow most electricians to determine if an electrical problem exists. Many electricians are hesitant to restart a tripped motor based on megohmeter or multimeter testing alone, and with good reason.

The fact is that numerous problems causing a motor to trip, such as a turn-to-turn short, are undetectable with these instruments, and restarting of a tripped motor should be considered only after such faults have been factored out. Another problem that is difficult to detect is the breakdown in the insulation between individual motor phases. Unfortunately, this problem can be completely isolated from ground, so RTG testing is ineffective. If faults like these are not diagnosed, they can cause rapid deterioration of the winding, potentially ending in a complete motor replacement.

To conclude that a motor is operational from megohmeter or multimeter testing does not fully assess motor health or allow maintenance personnel to be taken seriously. When troubleshooting, it is necessary to look at the whole picture and to not make a quick decision, if at all possible. Break the system down into individual fault zones, and test each fault zone completely with every available technology. Finally, maintenance personnel should make written or verbal recommendations using fault zone analysis terminology in order to express their confidence and capabilities concerning the decision.

FIVE ELECTRIC FAULT ZONES. To confidently report the electrical condition of a motor and ensure that a recommendation is taken seriously, there are five fault zones that must be looked at during the troubleshooting effort. By missing any of these zones, maintenance personnel could miss the problem, as well as lose personal credibility for their fault detection skills. These five electric fault zones are:

1. Power circuit

2. Insulation

3. Stator

4. Rotor

5. Rotor/stator relationship.

POWER CIRCUIT. The power circuit refers to all the conductors and connections that exist from the point at which the testing starts through to the connections at the motor. This can include circuit breakers, fuses, contactors, overloads, disconnects, and lug connections.

A 1994 demonstration project on industrial power distribution systems found that connectors and conductors were the source of 46% of the faults that reduced motor efficiency. Many times,a motor that is initially in perfect health is installed into a faulty power circuit, causing problems like harmonics, voltage imbalances, current imbalances, and so forth. As these problems become more severe, the horsepower rating of the motor drops, causing temperatures to increase and insulation damage to occur. The motor is then replaced many times, so the failure cycle begins again and again. As shown in Figure 1, high-resistance connections resulting in voltage imbalances will ultimately reduce a motor's horsepower rating significantly.

Phase-to-phase resistance testing. One method of detecting high-resistance connections is phase-to-phase resistance testing. On a three phase motor, the three resistance measurements should be nearly identical. If all three readings are exactly the same, there is a 0% resistive imbalance. As one or more phases develop a high resistance, the resistive imbalance increases, indicating a fault. Some of the fault mechanisms that cause high-resistance connections include corroded terminals, loose cables, loose bus bars, corroded fuse clips, corroded contacts, open leads, different size conductors, and dissimilar metals.

Figure 2 shows three different resistance test points that can be used to determine the actual location of a motor's high-resistance connection. The first test point is Position X, which is upstream of the fuses. If the resistive imbalance is high at that point, maintenance personnel may want to move the test point to Position Y, which is downstream of the contactor. If the imbalance is still evident at Position Y, testing at the motor connection box-Position Z-will isolate the motor from the power circuit and determine which position is the problem.

INSULATION. The insulation fault zone includes the insulation between the windings and ground. High temperatures, age, moisture, and dirt contamination are conditions that lead to shortened insulation life. It has been said that if plants would just use available space heaters to keep the insulation dry, it might be possible to double the life of that plant's motors.

Insulation systems today are better than ever and are able to handle higher and higher temperatures without significant reduction in life. However, there are still many ways to destroy insulation much earlier than is reasonable. Keep in mind that, although insulation is often involved in a failure, this fault zone is heavily influenced by other problems.

For example, the power circuit can heavily influence insulation. If a high-resistance connection exists upstream of the motor, developing more than a 5% voltage imbalance, and the motor continues to operate at its normal horsepower rating, insulation life will be shortened. Reverse sequence currents that develop rotating magnetic fields in the opposite direction will not only reduce the torque capability, but can allow the temperature to rise out of control, exceeding even the 155C limit on Class F insulation systems. Was the insulation system the real cause of the motor failure or was it just a symptom? It is easy to diagnose the evident insulation failure as the fault mechanism, but it will happen again with a different motor if the problem is not corrected.

Again, testing with a megohmeter is not going to tell maintenance personnel everything, but it is a good start when it comes to insulation testing. Something that is often overlooked when using the Institute of Electrical and Electronic Engineers (IEEE) limits on resistance to ground (RTG) is the reference to 40C. Megohmeter testing alone, with no regard to temperature, will result in RTG readings that fluctuate between high and low, depending on the temperature of the windings. Temperature correction of the readings will not only meet the IEEE testing requirements, it will also yield a much better trend, as Figure 3 shows.

Maintenance personnel must also realize that moisture contamination may cause the temperature corrected reading to be invalid. To prevent this from happening, make sure that the heaters are energized when the motor is not running.

The Polarization Index Test. One insulation test that has lost popularity is the polarization index (PI) test. In this test, a 10-minute application of a constant DC voltage in the form of a megohmeter test gradually increases the RTG reading. This is a result of charging the insulation system in a similar way that a capacitor would be charged, causing a reduction in the absorption current. According to Ohm's Law,
I (current) = V (voltage) / R (resistance). Therefore, the reduction of this absorption current must result in an increase in the resistance. If a 10-minute RTG reading is taken and then divided by the one-minute RTG reading to provide a PI index, a value of 2.0 or higher is considered acceptable by the IEEE. Unfortunately, motors with unstable insulation systems can give values close to or greater than a 2.0 but still be defective.

In Figure 4, when the 10-minute RTG reading (approximately 600 megohms) is divided by the one-minute reading (approximately 300 megohms), the PI index is 1.94. This nearly meets the IEEE specification as a good insulation system, and would probably be accepted in the field. It is apparent, however, that this insulation system is very unstable, so it is crucial to always look at the PI profile and not just the PI index.

DC vs AC testing. A limiting factor with DC RTG testing is that the DC signal will often fails to evaluate the true insulation condition. The insulation on a motor is a natural dielectric material, so it is therefore a poor DC conductor. This is good, since it is undesirable to have excessive leakage to ground. However, it is bad in that an insulation system in a degraded condition may take a bit longer to identify using a DC signal or megohmeter.

On the other hand, AC RTG testing does not allow the dielectric to charge and will pass through the dielectric much easier. This is good, because it allows the use of an AC signal to give earlier indications of insulation degradation. However, it is bad in that it can be destructive, as with an AC Hi-Pot. Low voltage capacitance to ground tests, however, are non-destructive and very good early indicators of degradation modes in insulation systems. These values are read in pico farads (pF) and can be effectively trended over time.

STATOR. The stator fault zone refers to the DC or three-phase AC windings, insulation between the turns of the winding, solder joints between the coils, and the stator core or laminations.

Common faults. One of the common faults occurring with motor windings is a turn-to-turn fault. This occurs when the insulation between two turns in the same coil breaks down and reduces the coil's ability to produce a balanced magnetic field. Unbalanced magnetic fields result in vibration, which can then cause degradation of the insulation as well as bearing failures. Localized heating around the short can also spread to other coils, resulting in a coil-to-coil short. Eventually, excessive heating will not only destroy the motor windings, but will also damage the insulation between the laminations of the stator core.

Another fault that can occur with motor windings is a phase-to-phase fault. This results from the insulation breaking down between two separate phases that usually lie adjacent to each other in the same slot. A higher difference in voltage potential tends to make this fault accelerate very quickly. To reduce the opportunity for leakage between phases, slot paper is installed between different phases in the same slot.

A turn-to-turn or a phase-to-phase short can occur many times without resulting in an immediate ground fault. Because of this, testing for preventive maintenance with a megohmeter alone or following a motor trip may not identify the fault, causing a small winding fault to develop into a major catastrophic failure. Permanent core damage may even necessitate replacing an entire motor.

Testing. Testing of the stator is performed by connecting directly at the motor, as well as connecting at the motor control center. During the test, high-frequency AC signals are sent into the motor. These signals produce magnetic fields around the windings, which should be matched between phases. The inductance measurement for each phase is then compared to the other phases and calculated into an inductive imbalance. This imbalance-minus the influence of the rotor-is used to compare each phase's ability to produce a balanced magnetic field.

During a test, DC signals are also sent into the motor. From these signals, the actual resistance of the winding or windings is measured. The three resistance readings of a three-phase induction motor are compared and calculated to produce a resistive imbalance.If this imbalance exceeds a predetermined level, then high-resistance connections may exist in the solder joints between coils.

Winding configuration types. There are two basic types of stator winding configurations. The first is "Y" connected and the second is delta connected. To more fully understand the information provided by inductance readings, a simple understanding of the winding configuration can help.

When looking at phase-to-phase inductance, a "Y" configuration winding with a turn-to-turn short will result in two low-inductance readings and one high-inductance reading (Figure 5), while a delta configuration winding with such a short will result in one low-inductance reading and two high-inductance readings (Figure 6).

ROTOR. The rotor fault zone refers to the rotor bars, the rotor laminations, and the end rings of the rotor. In the 1980s, a joint effort between Electric Power Research Inc. (EPRI) and General Electric showed that 10% of motor failures were due to the rotor. Although this represents a small percentage of overall motor problems, the rotor can also influence the failure of other fault zones.

When a motor is started with a broken or cracked rotor bar, intense heat is generated around the vicinity of the break. This can spread to other rotor bars and destroy the insulation around the nearby laminations. It can also affect other parts of the motor, such as the stator, which is just a few millimeters away from the rotor. Stator insulation cannot hold up to the intense heat developed by the broken rotor bar and will eventually fail. Unfortunately, many times broken rotor bars are not easily seen without motor circuit or current signature analysis and may be overlooked as the root cause of failure. This results in a motor rewind and replacement of bearings, but not a rotor repair. This means that when the motor returns to service, it has the same problem all over again, along with new insulation to destroy.

Rotor Influence Check. One method of testing the rotor condition is the rotor influence check (RIC). The RIC is a trademarked test performed on AC induction, synchronous, and wound rotor that demonstrates the magnetic coupling between the rotor and stator. This relationship indicates the condition of the rotor and air gap within the motor.

A RIC is performed by rotating the rotor in specific increments (determined by the number of poles) over a single pole group and then recording the change in inductance measurements for each phase of the three-phase motor. For proper resolution, 18 inductance measurements per pole group are recommended. To determine the number of poles in a motor, use the following equation:

F = NP / 120

where:

F = line frequency (normally 60 hz in the U.S.)

N = speed of the motor in revolutions per minute (RPM)

P = number of poles

For example, given the above equation, how many poles does a motor with name plate RPM = 1780 have?

Recalculated: P = 7200 / RPM

P = 7200 / 1780

= 4 poles

Without historical data, a RIC must be performed to provide any information about the standard squirrel cage induction rotor. Faults such as broken rotor bars or damaged laminations can exist even if the balance of inductance is low. If the decision to perform a RIC is based only on how high the balance of inductance is on the baseline test, the late stages of rotor bar defect might get overlooked.

Figure 7 shows the expected inductance changes for a rotor with broken rotor bars. Note the erratic inductance values at the peak of the sine waves for each phase. Broken rotor bars cause a skewing in the field flux generated by and around the rotor bars. A normal rotor has no skewing or erratic inductance patterns, as Figure 8 shows.

ROTOR/STATOR RELATIONSHIP. The rotor/stator relationship refers to the air gap between the rotor and stator. If this air gap is not evenly distributed around the motor's 360, uneven magnetic fields can be produced. These magnetic imbalances cause movement of the stator windings, resulting in winding failure, and electrically induced vibration, resulting in bearing failure. A faulty relationship between the rotor and stator is also called an eccentricity, and it occurs in two types, with each demonstrating specific physical and inductive results.

Static eccentricity. The first type of eccentricity is called static eccentricity, and it is caused by problems such as a misaligned endbell or the shaft sitting low in the bearing. The physical result of such eccentricity is that the shaft is always in the same place out of the electric center (Figure 9). The inductive result is the variation in peaks of the sine wave (Figure 10).

Dynamic eccentricity. The second type of eccentricity is called dynamic eccentricity. This eccentricity occurs when the rotor does not stay in one place but is allowed to move within the space of the stator, as Figure 11 shows. The inductive result of such eccentricity is the movement of all three inductance values up or down, depending on which phase is closest to the rotor at a given degree of rotation, as Figure 12 shows.