Pumps & Systems, June 2007
When a motor fails, users can (1) rewind, possibly for high efficiency; (2) replace the failed motor with a new motor; or (3) invest in a premium efficiency product. Here are the advantages and disadvantages of each approach and the precautions that must be taken to assure the best investment.
When a motor fails, the user must decide whether to repair or replace it. To make a proper decision, one must consider the cost of the repair, the availability of a replacement, the age of the motor, the electrical design required for the application, any special mechanical features, and the urgency of returning the failed motor to service. Placing the driven equipment back into service is frequently the highest priority, and users often make their decision based on this criterion alone. Plant managers tend to be less concerned if the rewound motor is less efficient when their operation's downtime is costing thousands of dollars a minute.
Increased Profits by Energy Saving
U.S. industry will continue to feel the squeeze on profits, and manufacturers who fail to implement energy reduction programs will find themselves at a competitive disadvantage. In some industries, motor operation costs may even exceed those of labor cost.
It is no longer practical to view the power bill as a fixed base cost, not worth the effort it would take to reduce. The need to minimize power usage should be as important to the CFO as it is to the plant manager. Due to the improvements that have occurred in motor technology, even companies that already had energy programs ten years ago should now reevaluate their criteria.
Future Power Costs
In 1970, electricity cost the average industrial user about one cent per kilowatt-hour. By 1980 that had jumped to four cents: a 300 percent increase. Power costs in some areas are already over seven cents. The cost for electricity may fluctuate with economic cycles, but long-term it will continue to increase.
The nature of today's power bill has also changed. The contract rate of the past now only covers 60 percent to 75 percent of the actual amount paid by users. In addition to taxes, today's bill can include such additions as "fuel adjustment" and "demand charges." The bottom line is that electric power can be a major cost element in your product.
First Cost vs. Operating Cost
Manufacturers who understand their business look at the total cost of a motor, rather than simply making a decision based on initial purchase price.
Let's consider the example of a pre-EPAct 75-hp, 1800-rpm TEFC motor that originally sold for approximately $2700, with an efficiency of 91.7 percent. This motor, operating continuously and using power costing $.07/kW-hr, will in just one year cost $37,414 to operate - or 1,386 percent of the original purchase price! In fact, operating costs will overtake the purchase price after the first 26 days of operation. Even if the motor is only used for two shifts (assuming 4,160 hours per year), this will still occur after 55 days.
For this reason, any expenditure related to repair or replacement of a failed motor should be evaluated based on a total cost of operation calculation.
An Incorrect Decision Costs Money
Building on the example above, consider the impact on the operating cost of the same 75-hp motor using a more efficient motor. Let's say the motor efficiency improves to that of an EPAct design with a nominal efficiency of 94.1 percent. The new EPAct motor would do the same amount of useful work, but use $979 less energy.
In mid-2001 the motor industry, in cooperation with conservation groups and the DOE, introduced products with even higher efficiencies: NEMA Premium. With this new specification, the same motor rating would have a nominal efficiency of 95.4 percent and save $1,510 in the first year. These savings continue to accrue as the cost of power goes up. Without increasing the cost of a kilowatt-hour, a NEMA Premium motor would generate $10,563 in savings over seven years, compared to $6,854 for an EPAct level efficiency motor.
Factors in the Rewind Decision
Hundreds of thousands of older T-frame motors were manufactured well before there were any government standards. Many of these still operate in U.S. industry. Each time one fails, an opportunity is created to improve the user's bottom line.
Power costs will certainly continue to rise and further escalate motor operating expense. So the question of how repair affects motor efficiency is an important one. Some claim a rewound motor is never as efficient as the original; others say a well-executed rewind can be better than the original design. These differences in perception suggest there may be several factors involved.
Armed with the right information, understanding the factors that affect rewind performance does not need to be complicated. Let's examine the various types of motor losses and how they are influenced by engineering decisions.
Keep in mind that actual motor losses may differ between two motors of the same design, depending on how the motor is used. Figure 1 shows how motor losses vary with load.
Figure 1, left. How motor losses vary with load.
As a motor approaches 100 percent of rated load, losses increase dramatically, with most of the increase found in the form of rotor and stator losses. The age of the motor is also a factor. Figure 2 shows the progression of motor efficiencies through the years, driven by improvements in engineering design and material technologies. (Note that these ratings are for typical GE motors from 1944 up to EPAct; actual efficiencies will vary from manufacturer to manufacturer.)
Figure 2, left. The history of motor efficiency, TEFC, 1800-rpm.
The distribution of losses will also be different for various motor designs. Variations in speed, design and enclosure will all affect loss distribution, as shown in Figure 3.
Figure 3, left. Representative losses as a percentage of total losses.
The ability of the repair shop to analyze and replace those parts which most influence losses, such as the stator core, the windings and the rotor, will affect the outcome of a rewind.
With all that in mind, let's take a look at losses in a typical 50-hp, 1800-rpm, TEFC standard efficiency design. The distribution of losses is shown in Figure 4.
Figure 4, left. Losses in a 50-hp, 1800-rpm, TEFC Pre-EPAct motor.
The table in Figure 5 shows how these losses can be reduced. Following that are detailed explanations of the techniques.
Figure 5. Possible methods to reduce losses through rewinding.
Stator Losses
Stator losses are primarily I2R losses, released in the form of heat as current passes through the stator windings.
When rewinding a motor, smaller diameter wire will increase the resistance and therefore I2R losses; larger diameter will have the opposite effect. If the original wire was aluminum, changing to the same size copper wire will also reduce resistance and loss. Obviously, using a larger diameter copper wire will affect the best reduction.
Another option for reducing stator losses is to reduce the number of wire turns. Use this method with caution. While full load efficiency may be increased, starting current will go up and power factor will be reduced. Both starting and maximum torques will be increased. A change from ten turns to nine turns would increase starting current by as much as 23 percent.
Core Loss
Core loss is the sum of the eddy current and hysteresis losses that occur while energizing the motor's magnetic field.
Motors are insulated between the core laminations to minimize eddy currents, but the process of stripping can destroy this insulation. When stripping a motor for rewinding, insulation burnout must be done at carefully controlled temperatures. Otherwise it's easy to overheat the laminations, breaking down the core insulation and actually increasing core loss. Not all repair shops use the same insulation burnout techniques; investigate them thoroughly before deciding where to have the motor rewound.
Another item that is often ignored is the condition of the core after motor failure. Failures caused by excessive loading, extended stall conditions, single phasing, or bearing failure leading to rotor striking can all cause increases in the core loss.
It is very unlikely that the original core loss data would be available from a ten year old T-frame motor. Repair shops may have equipment to evaluate the core in its failed condition, but are unable to relate the results to original factory core loss specifications.
Applying even the best techniques to improve the efficiency may be inappropriate without the original core loss information. The repair shop would have to conduct a full efficiency test using a dynamometer, which takes two to three hours depending on frame size, in order to validate the finished motor.
Manufacturers do this on every motor they design and have programs registered with the DOE to assure that design efficiencies are maintained throughout production. The DOE requires that testing laboratories be third-party certified to assure compliance with the testing procedure defined in IEEE 112 B. This process was written into the Federal Energy Act to assure reduced optimistic efficiency claims. Repair shops have no such requirement.
Rotor Loss
Rotor losses are I2R losses, released as heat through the rotor slots and endrings. It is unlikely that a repair shop will be able to improve rotor losses.
Efficiency vs. Time
You have undoubtedly heard or read more than once that motor efficiency naturally decreases through motor life as a result of "heat aging." This argument says that as the motor starts and stops the core temperature increases and decreases, causing deterioration in the core steel's electrical properties and a resulting increase in internal losses.
In fact, this is only a problem if an aging type of steel is used in the core. Most manufacturers use non-aging steel that does not lose its electrical properties over time.
Rewind vs. a New Motor
Now that you know some of the pitfalls of rewinding, let's reexamine our options in the face of a motor failure. Provided that downtime isn't the critical factor, a user now has these choices:
- Rewind the motor to the original efficiency.
- Rewind the motor to a higher efficiency.
- Replace with a new motor of same efficiency.
- Replace with a premium-efficiency motor.
A fifth option that no one should knowingly choose is to rewind the motor to a lower efficiency - but many users unwittingly make this decision. As explained above, it is very easy to damage the stator core insulation while stripping out the old winding and so increase core loss by three times or more.
Figure 6 compares the typical results of each of the choices, with assumptions for the "poor quality rewind" figured at three times the original core loss.
Figure 6. Efficiency Gain (Losses): Rewind vs. Premium-Efficiency Motor
The high-fill rewind produces some efficiency gains when a larger wire size is used. As would be expected, the greatest efficiencies are realized by retrofitting with new premium-efficiency motors.
Figure 7 shows the annual operating cost difference for each of the options listed above. Note that operating cost can be increased by a poor rewind just as much as it can be decreased by a new, premium-efficiency motor.
Figure 7. Annual Operating Cost Difference: Rewind vs. Premium-Efficiency Motor Continuous Operation @ $.07/kW-h
The conclusion is obvious: either replace failed motors with new, premium-efficiency motors, or else exercise extraordinary care in the rewinding process.
Protecting the Stator Core in a Rewind
No single aspect of the rewind process is as important as preserving the electrical integrity of the stator core.
Not only can insulation damage increase core loss, but the resulting rise in motor temperature could also then cause the motor to fail prematurely. You've probably had at least one motor that operated well for years before initial failure, but failed again shortly after being rewound. The failure is more often the result of temperature rise than of defective materials or faulty workmanship in the new windings.
The obvious question is "what is a safe insulation burnout temperature?"
Unfortunately, there is no simple answer. Manufacturers use a wide variety of materials for the core. Steel may be supplied with either organic or inorganic insulation coatings, or with no coating at all. If uncoated steel is used, the motor manufacturer will add an oxide insulation coating while annealing. Each of these lamination insulations has a different limit in temperature that it can withstand before deteriorating, so it's impossible to name a temperature that is safe for all motors. But there are some guidelines.
In all cases, the stripping operation must control the core temperature to prevent damage to the interlamination insulation. Damage can occur even in a low temperature oven when several cores are stacked, and fire from the burning organic materials results in increased temperature beyond the oven setting.
If the motor has an organic lamination insulation, it will begin to deteriorate rapidly at around 500-deg F and may actually change its chemistry at higher temperatures. Organic insulation can be damaged in any oven hot enough to burn out the winding insulation.
Inorganic insulation can withstand temperatures up to 700-deg F, allowing the old winding to be burned out safely if the oven temperature is carefully controlled.
Uncoated semi-processed steel laminations will stick together if they get too hot, increasing eddy current losses dramatically. Motors built with this type of steel can be stripped at oven temperatures of 700-deg F or below.
Stripping Methods
In the past few years, awareness has grown among users that poor quality motor rewinds can cause an increase in losses. Users have demanded an end to the practice of burning out the old windings at uncontrolled temperatures.
Motor repair shops that have kept pace with technology have switched to temperature-controlled ovens and have discontinued the practice of softening varnish with a handheld torch. Ask your rewind shop if they can perform any of the following non-injurious stripping techniques:
- Mechanical stripping
- Chemical stripping
- High-pressure water jets
- Freezing process
- Ultrasonic stripping
Regardless of the process used, make sure your rewinder can follow the motor manufacturer's recommended safe burnout temperature limits. Some do and some don't - which is probably the main reason for one shop's reputation over another for better quality rewinds.
There is really only one way to make sure that losses have not been increased in the process, and that's to perform a qualitative core loss test before and after rewinding (assuming the core is okay to begin with). This can also help you screen the motor population to determine if any given motor is even repairable. More repair shops now offer this service, so ask yours.
Establish a Repair/Replace Policy
In response to the rising cost of electrical power, every company should establish a repair/replace policy to help make intelligent decisions.
Give every motor-driven machine in your plant a repair/replace priority, and consider investing in spare motors for any continuous process machines that are critical to plant operation. These "critical" machines are excellent candidates for retrofitting with premium-efficiency motors; the existing standard-efficiency motor should not be kept as a spare.
There is a good deal more to comparing relative in-use costs between motors than simple energy usage. The local power company may offer rebates on new premium-efficiency motors. Extended warranties may be available to reduce MRO costs. A new motor could be installed with a variable speed drive to maximize the process productivity (the drive may also be covered under a rebate). Inventory costs can be reduced by not storing repaired motors that could become obsolete before they are needed again.
These are only a few of the possibilities. Many motor distributors are willing to assist you in evaluating a program. Check the Internet for the one nearest you.