In this installment of an ongoing series, we’ll discuss what was learned during a five-year period at a site containing 46 4160V through 13,800Volt electric motors, and compare these findings to data from large-motor studies conducted by EPRI (Electric Power Research Institute) and IEEE (Institute of Electrical and Electronics Engineers).
The referenced site had a basic lubrication program for the motors, but it was inconsistent in practice. All of the site’s electrical and mechanical motor repairs had been reactive prior to the start of this five-year review. The initial data analysis, which consisted of vibration (GTI Predictive) and electrical signature analysis (ESA – EMPATH) for online and high voltage testing (Electrom) and motor circuit analysis (ALL-TEST PRO) sets a baseline for the changes implemented. The failures that occurred were evaluated using a PROACT RCFA approach, developed by Reliability Center Inc. Data was collected once a year on all of the motors, each of which operated approximately 4,000 hours per year.
During the initial evaluation, one 13.8kV synchronous motor was out for a failed rotor which resulted in a complete rewind. The machine was 11 years old and the investigation found that the motor had significant oil and water present throughout. The overall rewind was performed based upon signs of partial discharge.
Within a year, a second identical machine was found to have a similar issue and was completely rewound. Four more 13.8kV synchronous motors were found to be heavily contaminated and problems with the exciter circuits. At the same time, during the initial evaluation, one 4160V, 4000-hp, 1800-RPM unit was found to have broken rotor bars, and two identical motors were found to have fractured bars (or appeared to be progressing toward that same condition). Overall, 22 of the 4160 V and 13.8kV electric motors had winding contamination detectable through testing, and a total of 40 had visible contamination issues. Seven smaller 4160 Vac motors, from 100 to 600 hp, had bearing issues detected by both vibration and ESA.
Over the next year, cleaning practices and bearings were addressed; the two larger motors were rewound; a motor-repair specification was put in place; and lubrication training was implemented. The causes of the failing rotors in the 13.8 kV synchronous motors was determined, and excitation systems were upgraded. Two of the 4160 Vac, 4000-hp machines had their rotor bars replaced, and it was determined that the failures related to number of starts and the fact that protections to prevent excessive starts had been disabled.
Overall, the first year (and baseline) of the study found defects in 40 of the 46 electric machines and driven equipment. The six without defects had been cycled out for upgrades for inverter application (4160 Vac, 800 hp, 900 RPM) which included an overhaul and new bearings. The result of the 40 machines with defects was: 55% insulation and contamination; 17.5% bearings; 2.5% winding failures; and, 7.5% broken rotor bars detected; and the remaining 17.5% was in pumps and gearboxes. Results in relation to defects in the motors only (33) were: 67% insulation and contamination; 21% bearings; 3% winding failure; and 9% rotor bars.
Fig. 1. Comparison of 1983 EPRI study findings and those of the more recent referenced site study.
(See “Regarding Reliability of Large Electric Motors: How The Studies Applied to Larger Motors,” Apr. 18, 2020)
By the time the next round of testing was implemented in the following year, no obvious additional issues were detected. The bearing issues were found to be related to improper mixing of greases, and most oil-filled bearings had no identifiable issues, with the only exception being one that had an oil leak requiring a gasket repair. One of the machines with broken rotor bars wasn’t available for testing, and one of the 13.8kV synchronous motors was out for cleaning. Modifications to the settings and operation of the other 13.8kV machines reduced stator contamination (moisture), and only rotors were still observed as contaminated.
One year later, seven electric machines were found to have winding contamination. The primary reason was related to their location, how the facility was operated, and other outside conditions. On one 4000-hp, 4160 Volt machine that had been cleaned, the reason for contamination was the failure to replace filters on the motor hood. The task had not been included in the CMMS system, as the machine was slated to be de-commissioned.
The next year, with several events in between related to newly installed controls and unplanned modifications to safety devices, the following was found: 7 winding contamination; 4 bearing conditions; 1 misalignment; and, 1 unbalance. The misalignment and unbalance conditions were related to outside vendors, as internal workmanship had started, before the first tests noted in this article, on precision alignment and balance. The four machines with bearing issues that had not occurred previously were primarily at rest, and their bearing problems were later determined to be a combination of false brinelling and rust.
Now, if we compare the reduced failure rate conditions following corrective actions on the machines, the faults appear as: 54% winding contamination; 31% bearing defects; 15% unbalance and misalignment. Bearing defects still fall well under the expected conditions with environmental and maintenance issues making up 69% of the total faults. The bearing issues are related to a combined operations/maintenance condition.
When machines were sent out for repair, other than the 4000-hp and 13.8kV synchronous motors, which had babbet bearings, the bearings were replaced. Documentation of the work that occurred would have given a significantly lopsided view of the bearing repairs, with a majority of the machines showing bearing issues, when none existed. Unfortunately, this was primarily a result of how the data was compiled in IEEE’s 1990s studies, where the overall repairs were identified.
Another concern that arose when looking at the prior studies was the focus on just-failed motors versus the overall population of machines and how they related to the complete dataset. So, in the first part of this referenced site study, during the transition from a reactive to pro-active maintenance program, 33/46 of the motors (72%) were found to have issues. After the pro-active maintenance program was put in place, the faults detected through annual testing were found in 13/46 of the motors (28%), a dramatic drop. The actual unplanned failures of the original machines was 11% over 18 months. Over a similar period at the end of the study, unplanned failures dropped to 2%. In the first instance, a combination of improper maintenance and operations combined with random and age-related defects resulted in a high degree of reactive and planned maintenance, while the same period after implementation were improper maintenance and operations. In this case, however, only one unplanned failure due to improper settings occurred between testing periods. In another case, high-voltage testing damaged one machine during testing, and the unit was returned to service in 2.5 hours. In addition, other maintenance practices at the plant reduced a combination of energy costs and improved the overall efficiency, which further reduced a higher need for availability. Through the application of root-cause-failure analysis (RCFA) on the site’s motor failures, the impact on the remaining 2% can be achieved through a focused effort and the root causes of the defects can be addressed.
In the next article we will discuss the planned maintenance and testing that was performed to achieve the benefits shown above.TRR
Click The Following Links For Previous Articles In This Series on The Reliability Of Large Electric Motors
March 28, 2020: “What The Studies Really Said”
April 12, 2020: “Comparing What The Studies Said”
April 18, 2020: “How The Studies Applied to Larger Motors”
May 2, 2020: “A Data Mind-Bender”
ABOUT THE AUTHOR
Howard Penrose, Ph.D., CMRP, is Founder and President of Motor Doc LLC, Lombard, IL and, among other things, a Past Chair of the Society for Reliability and Maintenance Professionals, Atlanta (smrp.org). Email him at howard@motordoc.com, or info@motordoc.com, and/or visit motordoc.com.
Tags: motors, drives, electrical systems, power generation, reliability, availability, maintenance, RAM, EPRI, IEEE
