A series of studies by the Institute of Electrical and Electronics Engineers, Inc. (IEEE) and the Electric Power Research Institute (EPRI) from 1983 until 1995 1, 2, 3 that covered all industries reported that bearing faults, followed by winding faults, were prevalent (see Figure 1). These studies, one performed by Advanced Energy4 and another performed by the Electrical Apparatus Service Association (EASA)5 also revealed that repaired motors, in many cases, do not last as long as they should because of poor repair practices. The repetitive failures of these motors may have resulted, in part, because of a lack of diagnostics or forensics reported to the motor owner during the repair process.
 |
| Figure 1. Comparison of three motor reliability studies |
Repair and Communication
While many studies and surveys performed through 2013 identify that a good repair will last longer, few have identified the importance of providing the motor owner with feedback on the possible root causes of the failures. How does the owner’s maintenance department or the reliability engineers respond to a report of replaced bearings or a rewound stator? Do they respond in a way that prevents the problem from recurring?
A significant difference between the 2003 and 2013 Motor Diagnostics and Motor Health Studies (MDMH)6 was that in 2003, 56 percent of organizations reported electricians as the person responsible for electric motors, and only 2 percent were reliability engineers. In 2013, the number shifted to 43 percent reliability engineers and 26 percent electricians with a significant decrease in mechanical and general maintenance personnel. By definition, a reliability engineer requires feedback to pursue his trade, meaning that an understanding of the equipment failure—including potential root causes and corrective recommendations—becomes more critical in the repair process.
For instance, if the motor failed because the bearings were noisy, what caused them to become noisy? Were they over or under greased? Was the belt experiencing over-tension or misalignment? Was there bearing fluting because of shaft currents? This information, including recommendations, will assist the motor owner or reliability engineer in developing a plan to mitigate future problems.
For example, the bearings in a motor that is operated by VFD are fluted (see Image 1). This fluting is an indicator of shaft currents. Corrective action using a shaft brush or a shaft brush and insulated bearing should be taken. This also means that either the repair facility’s representative should be asking application questions, and/or the motor owner should be providing that application information.
 |
| Image 1. Bearing fluting caused by shaft currents |
Rewind related failure modes must also be identified. If the winding has been single phased and that information is not communicated, the failure will likely reoccur. Identifying that the failure occurred would be important, in this instance, because an electronic overload with single-phase protection would mitigate similar failures in the future. If, on the other hand, the winding failed because it was overloaded, the reliability engineer should recommend that the overload protection is also evaluated.
Repair’s Impact on Reliability
As noted by the U.S. Department of Energy, “Of the quality assurance procedures shops used, 40 percent were repair procedure specifications, 25 percent were test specifications, and 21 percent were EASA standards. Only one of the 65 shops surveyed used any form of quality assurance testing.”7
In fact, per the same report, almost half the shops surveyed performed no winding tests during the entire repair. Of the repair shops, 81 percent also reported that they changed winding configuration in electric motors during the repair process primarily for shop preference or ease of winding (73 percent), 10 percent with the owners’ knowledge and only 4 percent for the purpose of reliability or durability. The remainder did not provide a reason.
Proper repair practices are vital to the reliability of the repaired electric motor. While the studies primarily focus on energy consumption or efficiency of the machines, the 2013 MDMH identified a solid 0 percent of interest in energy as a driver for a motor program (only 3 percent in 2003). The key driver for the motor program was reliability (75.5 percent).
 |
| Image 2. Overloaded winding |
Following the reliability perspective and study results, the only reason for modifications from the original manufacturer’s design must be for improvements to the durability of the machine or engineering modifications to improve the reliability of the application. Modifications to make the repair less expensive or because the repair shop does not maintain the appropriate metric or half-size wire does not benefit the end user and can become expensive in the short term and extremely expensive in the long term. Companies that peen or glue bearing fits should be strenuously avoided. Such repairs should only ever be considered to get by while awaiting a replacement with the knowledge that the risk of catastrophic failure is high.
Properly maintaining fits through testing and machining to tolerance using welding, sleeving or remanufacturing will maintain potential bearing life and decrease friction and windage losses. While these losses relate to energy consumption, they also result in increased operating temperature. Just as with winding insulation life, for every 10 C increase in bearing temperature, the grease life reduces by half.
In 1984, David C. Montgomery published a paper that identified the impacts of core loss increases of 50 percent, 100 percent, 150 percent and 200 percent and related it to temperature rise, resulting insulation life and impact on grease/bearing life based on winding removal processes using high temperatures.8 The machine used in the example was a 50-horsepower, 3,600-rpm, drip-proof motor. He also related that the core loss impact is greater as the motor size increases.
The burnout oven stripping issue of electric motors has been a long-standing one. The use of high-temperature stripping results in a limit to the number of times a winding can be rewound, an impact on the mechanical fits and soft foot, an increase in associated greenhouse gas emissions because of drops in efficiency and community health issues.9 The use of repair processes that involve high-temperature stripping must be carefully considered because it may impact overall system reliability.
 |
| Table 1. Impact tof core losses because of burnout ovens |
While the close monitoring of thermal stripping methods is strongly recommended based on the studies, professional societies and trade associations, the number of cores damaged from excessive temperatures is on the rise. If an end user has a motor that has an increase in operating current, then the core is probably damaged. Just the change of an amp of current can cause a significant increase in core losses and excessive shims required for alignment, related to frame distortion. Mechanical and induction processes using low temperatures are available that have no negative impact on core losses.10
Testing throughout the repair process is critical and must be recorded. The repairs must also be performed with calibrated equipment. Testing must include:
- Phase resistance
- Insulation resistance
- High potential testing
- Surge testing
- Rotor testing
- Core loss testing
- Other tests depending on the type of electric motor
Mechanical tests must be performed with calibrated measuring instruments. Running tests should include voltage, current, vibration and audible tests.
With the low cost of digital memory, digital photos of the motor should be taken when it is received by the repair shop and prior to shipping, at a minimum. Unusual faults should be photographed and provided in a final report along with any data required by the reliability engineers, as appropriate. It is also recommended that forensic analysis and root-cause failure analysis be performed with input by the end user and repair facility when critical machines fail or if repeat failures occur.
Installation Considerations
Some end users ask, “Why should I have a Cadillac repair when I have Yugo mechanics?” Meaning, what is the point of having a reliable repair when the machine is going to be misaligned, miss-tensioned or have other issues? One simple answer, which is not even buried in reliability or industrial engineering, should be given: two unreliable systems create an even more unreliable system.
A combination of poor reliable installation with reliable repairs will increase the reliability of the system. In effect, a highly robust repair can result in a system that is better able to withstand poorly applied or installed motors.
Case Study
A low-quality repair has a poor bearing installation and a damaged core. Instead of a motor that would last 10 years (120 months), the end user has a motor that lasts about six months. When applied to the reliability formula (Equation 1), the result is a reliability of 0.905 (90.5 percent chance of surviving the first year) for a reliable repair example and 0.05 (5 percent) for the low-quality repaired motor.
R = e-tλ (Equation 1)
Where:
λ = 1/Mean time between failure (MTBF)
t = Time in months
A maintenance technician has a success rate of improper installation, resulting in an average life expectancy of a motor installation being about one year because of issues such as loose connections, improper belt tension or alignment. His reliability would be 0.368 (36.8 percent).
When this technician is paired with the low reliability repair, the result is (0.05 * 0.368 = R) 0.0184, or a 2 percent chance that the motor will survive the first year. If, however, a high reliability repair is used, then the result is (0.905 * 0.368 = R) 0.333, or a 33.3 percent chance that the motor will survive the first year.
While an organization’s improvement of installation and maintenance is important, a higher quality repair is even more critical in an evolving maintenance department. When considering that the average cost per hour of downtime is $10,000 for critical machines (from both the 2003 and 2013 MDMH studies), the improvement in repair savings can be staggering. P&S
References
1. Albrecht, Appiarius, McCoy, Owen and Sharma, “Assessment of the Reliability of Motors in Utility Applications – Updated,” IEEE Transactions on Energy Conversion, Vol. EC-1, No. 1, March, 1986.
2. Motor Reliability Working Group, “Report of Large Motor Reliability Survey of Industrial and Commercial Installations, Part 1,” IEEE Transactions on Industry Applications, Vol. 1A-21, No. 4, July/August, 1985.
3. Thorson and Dalva, “A Survey of Faults on Induction Motors in Offshore Oil Industry, Petrochemical Industry, Gas Terminals, and Oil Refineries,” IEEE Transactions on Industry Applications, Vol. 31, No. 5, September/October, 1995.
4. Advanced Energy, Achieving More with Less: Efficiency and Economics of Motor Decision Tools, Advanced Energy, USA, 2006.
5. EASA/AEMT, The Effect of Repair/Rewinding on Motor Efficiency, Electrical Apparatus Service Association, Inc. and Association of Electrical and Mechanical Trades, Inc., USA and UK, 2003.
6. Penrose, Howard W., 2013 Motor Diagnostics and Motor Health Study, motordiagnostics.com, April, 2013.
7. Schueler, Leistner and Douglass, Industrial Motor Repair in the United States, Bonneville Power Administration, USA, 1995.
8. Montgomery, David, “The Motor Rewind Issue – A New Look,” IEEE Transactions on Industry Applications, Vol 1A-20, No. 5, September/October 1984.
9. Penrose, Howard W. and Dreisilker, Leo F., “The Mechanical Effects from Thermal Stripping Induction Motor Stators,” 1997 EIC/EMCWA Conference Proceedings, IEEE, 1997.
10. Penrose, Howard W. and Dreisilker, Leo F., “Evaluation of Induction Warming Stator Cores for Coil Removal,” Conference Record of the 2012 International Symposium on Electrical Insulation, IEEE, 2012.