Visite nuestro sitio en Español: Confiabilidad.net    RSS | Contact

Understanding the Limitations of Oil Analysis

Brian Thorpe

It is very important to understand the limitations of test procedures used in oil analysis. If you have been in the business for awhile, you know dark oils can cause problems with laser particle counters.

Some sealants give false positives of silicates and elemental metals measurements with an ICP spectroscopy are limited to a certain range of particle size, usually four to seven microns. Generally, labs will tell you that parts per million (ppm) numbers in the one or two range can be nuisance numbers. But what about when they start trending up to 3 ppm and 4 ppm, still well below most alarm points? This is when knowledge of your equipment and other tests, such as filter debris analysis (FDA), wear debris analysis (WDA), or analytical ferrography, might need to be employed.

A recent oil analysis on a 2000 HP electric motor with sleeve bearings started trending up in tin, a babbitt-bearing component. This motor had a five year inspection/rebuild in March, 2009 with 18,925 hours run time since the rebuild. Oil analysis is done quarterly on this equipment. The tin had been ranging from 0 to 2 ppm since the rebuild baseline sample in six reports spanning 1-1/2 years. The next sample, which occurred after a shutdown, showed the tin increased to 3 ppm, then 4 ppm, 5 ppm, and 7 ppm, consecutively. Normally this wouldn’t be of concern, but with low particle counts and this being one of the few large motors with a circulation/filtration system on it, further investigation was initiated.

Since there was a filter in this system, a filter debris analysis was done. Several labs have special equipment to perform FDAs, or you can cut open the filter can and remove the filter media for inspection. In a filtered system, the lubrication history of the equipment is at your fingertips for the time the filter has been in service. The filter media indicated a definite problem with one of the motor bearings. The bearing temperatures were not elevated on either bearing, thus indicating possible thrust face damage rather than radial surface damage. This is where knowing your equipment components is very helpful. The coupling guard blocked the view of the motor shaft, preventing a look for the scribe marks of magnetic center, thrust in, or thrust out locations on the shaft. In Figure 1, all of the shiny spots at the bottom of the pleats on the filter media are pieces of babbitt. The human eye can only see about 40 microns in size, so these were fairly large pieces of babbitt.

fig 1

Figure 1: Babbitt flakes on filter media

Since a scheduled shutdown was coming up soon, it was decided to increase oil sample frequency and monitor bearing temperatures closely to see if the motor could make it to the shutdown date. Oil samples were taken every two weeks, with no increase in tin found, and the bearing temperatures remained constant. When the equipment was shutdown, the as-found alignment was checked, indicating it was slightly out of specifications for the RPM, but most likely not the problem. The scribe marks on the shaft for magnetic center, thrust in and thrust out positions had worn off, but the shaft appeared to be thrust in all the way.

The outboard bearing was opened for inspection. Figure 2 shows the amount of babbitt located in the bottom of the bearing reservoir.

fig 2

Figure 2: Babbitt in bearing reservoir

Since this motor has a circulation/filtration system on it, the bearing reservoirs have a weir in them about three inches above the bottom to maintain proper oil level. This weir prevented most of the babbitt from leaving the reservoir and circulating in the system. Figure 3 shows the as-found condition of the bottom bearing half.

fig 3

Figure 3: Bottom bearing half

The area in the red boxes is the thrust face wear, almost 0.125 inches deep. Most of the other damage to the bottom half is from babbitt roll-over from the thrust face to the radial surface. You can also note the drag out of babbitt in the right side oil relief, as well as babbitt deposited in the left side oil relief. There were slight traces of babbitt smear or buildup on the shaft, which would have caused problems eventually. The bearing temperature indication is located in the bottom center of the bearing, almost three inches from the thrust face. This distance is probably why there was no increase seen in the bearing temperature. A new bearing was fitted and installed in the outboard end. The inboard bearing was inspected and no damage was found.

CONCLUSION
This illustrates once again the value of routine oil analysis in identifying potential problems. It also shows a weakness of oil analysis, in this case by the low ppm of tin listed on the report versus the amount of babbitt found in the bearing reservoir. In this circumstance, the trend information was much more valuable than an alarm limit since most programs and labs would not flag this as a problem. Two other major factors in identifying this problem were knowledge and an understanding of the equipment and components within. The root cause analysis (RCA) of this failure was placed on improper greasing of the gear coupling during a prior shutdown. By not allowing excess grease to purge properly, it pushed the motor shaft to the thrust in position against the bearing thrust face of the outboard bearing. This situation reaffirms that if there is a doubt, have your lab run additional tests, such as WDA, ferrography, or in this case where a filtered system was involved, a FDA.

author

Brian Thorp has been involved with mechanics and maintenance for more than 34 years. In his current position as a Predictive Maintenance Technician with Seminole Electric Cooperative Inc., he is responsible for the lubrication and analysis for a combined total 1300 MW coal-fired power generation plant.

 

Comments (2)

  • Thank you Brian for such a valuable lesson for my library. Thanks to this site for providing the venue where I could find it.

    1) Posted 12:57 pm, 08 May 2012 by Mark O'Brien

  • Brian - great article highlighting the need to go beyond simply reading the comments the lab makes at the bottom of an oil analysis report!

    2) Posted 6:32 pm, 12 May 2012 by Mark Barnes

Have your say

Comments are moderated prior to publication.
Please fill out the fields below.
Email addresses will never be published.

Comment guidelines

You can use basic HTML (a, strong, em, blockquote).
Links automatically use the nofollow attribute.
Off-topic or inappropriate comments will be edited or deleted.

Knowledge Base Articles
ReliabilityWeb on Flickr

Related Publication
Advertisement