Whose fault is it?
That is the question frequently asked when a motor recently repaired by an outside vendor is returned to operation and then fails shortly thereafter. Premature motor insulation failures cause tension between the process plant and the motor repair vendor. The reason for the tension is obvious. For the operator, an unexpected motor failure means an unplanned shutdown with a loss of tens of thousands of dollars of production. To the motor shop, the failed motor is an expensive liability, and there is an understandable reluctance to accept responsibility. The vendor is often forced to choose between "eating" the cost of a rewind or losing customer good-will and future business.
It is very important that the two parties get together and agree on the underlying cause of the failure. Unfortunately, in the case of winding failures, the underlying cause is not always obvious. In contrast to mechanical failure where the operator is usually able to see and understand the failure mechanism, the circumstances are quite different in the case of winding failures. Such failures often originate in the internal parts of the stator and are not readily detected during the early stages. Indeed, in many cases, by the time the problem becomes evident, a portion of the stator may be partially destroyed.
Deducing a possible cause of failure does require a knowledge of the factors that may be involved.
The nature of insulating materials.
Organic materials used for electrical insulation are subject to aging and deterioration from a variety of stress factors. The aging is a gradual process during which the material loses the properties which initially made it desirable as an insulation. In its final stage, the material is no longer able to perform its function and a winding fails.
Normally insulation aging proceeds slowly taking many years to reach an ultimate failure. If the various stress factors that cause accelerated aging are kept to low levels, the life of motor insulation can be very long indeed. Older machines that were designed conservatively remain in operation for more than 45 years. Achieving such a long life requires a conscientious maintenance effort.
Of the factors that affect insulation life, the most basic is temperature. Operating motors with high temperature rise in high ambient temperatures shortens the ultimate insulation life. Conversely, minimizing operating temperature enhances life expectancy. Achieving a low temperature rise becomes a design trade-off against manufacturing cost. A lower temperature rise requires more pounds of steel and copper per horsepower. For that reason, it is common to see large high-cost motors (1,000 hp plus) with a conservative temperature rise of less than 80 degr C. Smaller motors characteristically have a higher temperature rise with reduced life expectancy.
Because of the important role that temperature plays in determining insulation life, insulation materials are classified according to their temperature capability as shown in Table I.
Max Temp degr C
A class F motor is known to have a hot spot temperature of 120 degr C. If there are no other factors present to influence the outcome, what is the theoretical life expectancy of the insulation? For class F the maximum temperature is 155 degr C. Then DT = 155 - 120 = 35 degr C. The term (DT/10) = 35/10 = 3.5, and L = 2 <+>3.5<+> x 20,000 = 226,000 hours, approximately 26 years.
A reasonable winding life
The foregoing calculation shows that the ultimate life expectancy of the insulation system based upon thermal aging is measured in years. Of course, there are some small motors designed to operate at high temperatures to reduce the initial cost. Such motors are usually a consumable item that is thrown away at failure and a new replaced.
Large motors and smaller ones built to high efficiency standards have moderate temperature rises. The hot spot temperature does not usually exceed 100 degr C. With a class B or F system this temperature should provide decades of insulation life. From this perspective we can categorize as premature a failure that occurs even several years after a rewind.
If motor windings fail markedly short of theoretical life even though they meet the legal guaranteed life, it is in the best interest of the motor repair vendor to cooperate with the operator in a failure analysis. Combining their expertise and resources achieves the best chance of identifying and mitigating the possible stress factors.
Types of stress factors
There are a number of stress factors which can adversely affect insulation material and shorten its life. These factors fall into four main categories.
- electrical, and
The categories are not in order of importance. Each is present to a degree in every electric machine. In different situations, one or more in combination may take on a significant role in causing an early failure. Each stress factor shortens the insulation life based on temperature calculations alone.
The following sections discuss stress factors that may be present in varying degrees in a typical industrial plant.
Many operating environments provide a bountiful source of contaminants that eventually wind up on the windings of electric motors. Such contaminants include acids, alkalis, coal dust, sand particulate, lubricating oil, and moisture. Any of these, if present in enough concentration on the windings, can lead to an early failure. Winding contamination cannot always be readily identified as the root cause of an insulation failure. Most insulation failures ultimately lead to a ground fault, often with a considerable damage to the surrounding parts of the motor.
The most common type of contamination is moisture from a variety of sources. Rain water has a way of finding its way into outdoor motors, even those with weatherproof enclosures. Even more common is the condensation on the inside of protected enclosures. This source of water is common in areas of high humidity and substantial temperature swings.
Water by itself is not highly conductive. However, it has the ability to find any cracks and voids in the coils. When combined with more conductive contaminants such as dust on motor windings, water will usually produce a winding failure.
Examine the following features if rain is suspected of entering an outdoor motor enclosure. Large motors with conduit or heat exchangers bolted to the outside often allow water to leak inside. The attached surfaces may become bent and not make a tight seal after being reassembled. Also, reused gaskets are not effective and should be replaced.
An operating motor generates enough heat to prevent condensation. Equip motors inactive for long periods with automatically heaters that energize when the breaker or motor contactors open. Protect large motors with heaters during long periods of storage. If this is not practical, dry out the motor and test insulation resistance prior to starting it up.
While water is the most common motor winding contaminant, not far behind is bearing lubricating oil. Unless some special design feature is present to equalize the pressure across the bearings, internal cooling fans have a tendency to draw an oil mist into the motor. Many industrial motors show an oil film over the inside surfaces when opened up. A film of oil on the coil surfaces can cause deterioration in a number of ways. It causes other particulate contamination to adhere that will gradually increase surface conductivity several orders of magnitude over that of a clean winding. In a high voltage motor a low level of surface resistance creates a condition for internal discharges that destroy a coil.
Water will penetrate cracks and openings in the surfaces of the end winding structure and will weaken the support. If the end winding structure is weakened so that coils can move, insulation will fracture and electrical failure will follow.
The mechanical stresses that contribute most significantly to early winding failures are due to electro-magnetic forces. These forces create stresses both in the end-turn area as well as in the slots. The action of these forces is more obvious in the end windings where it can directly cause failures.
Electric motors create a rotating magnetic field. As the magnetic field passes individual coils in the end winding area, it alternately pushes outward and then pulls inward at line frequency. The alternating forces tend to loosen the packing in the end coils. If packing blocks come loose and fall out of position, the coils are free to move enough so that the ground wall insulation will break at the point where it emerges from the slot. A ground fault quickly follows cracked insulation. The resulting arc from conductor to stator core will usually cause severe damage. This effect is much greater at start-up when the forces are 25 or 30 times the full load values. The life of a motor may well be limited by the number of starts to which it is subjected. If a process requires frequent starting and stopping, then specify a motor for severe start-up duty.
Whenever opening a motor for maintenance, examine the end-winding packing for evidence of looseness. Examine the varnish fillet where the coil emerges from the slot for evidence of cracking. Address and correct any evidence of looseness in either location by reblocking or revarnishing the end-windings.
Magnetic fields are also present in the slot area. In the slot area, the interaction of the magnetic field and the coil current creates a force that pushes the coil sides into the slot when both coils belong to the same phase. However, when coils of the opposite phase occupy the same slot, the top coil experiences an upward push out of the slot. (see Figure 3). The force is a pulsating force at twice the line frequency. The pulsating force on the coil side can create looseness in the slot either by loosening the wedges or by compressing filler material. If the coil sides do become loose and are able to move only a small amount the anti-corona surface on the coils may be rubbed off and slot discharges will occur if voltages are high enough.
Always test slot wedges for looseness when a large motor is in the shop for cleaning or bearing change. There are more elaborate tests available but a simple tap test is usually sufficient to find loose coils. Tap each wedge at several points along its length and listen to the resulting sound. With a loose coil, the tap produces a hollow sound distinctly different from that produced by tight coils