16 Knowledge Areas to Improve Oil Analysis Interpretation

Jim Fitch, Noria Corporation
Tags: oil analysis

Lubricant analysts face many challenges in trying to translate laboratory data into meaningful comments for those responsible for lubricant and machine health. One could say the analyst gives a voice to the oil in communicating “what hurts”.

But before a doctor can make a diagnosis, he must ask many questions regarding the patient’s health history and lifestyle. Like the doctor, the lubricant analyst requires similar information relating to provenance – that is, the source or the origin - before the data can be accurately diagnosed and corrective actions prescribed.

Therefore, the tasks of collecting and organizing data and providing provenance are largely the responsibility of the equipment owner/maintainer.

It’s often said that the quality of oil analysis depends on an integrity chain. The weakest link defines the overall quality of the analysis. While many of these links are in the hands of the laboratory (instrument precision, sample prep, reagent purity, etc.), other links are controlled by the client including sample quality, sample frequency, test slate selection and provenance.

While the topics of sampling and laboratory testing have been well-covered in this magazine, surprisingly little is discussed about the importance of provenance and its relevance to oil analysis quality. Trust me, this is no trivial matter.

Let’s consider the body of knowledge that an analyst, in an ideal world, would need to access before methodically preparing his comments and recommendations about a given oil sample. Here’s the list of knowledge categories:

  1. Current oil analysis data (from the sample just analyzed)

  2. Historic oil analysis data (the lab should have this unless a different lab was used previously)

  3. Machinery identification/description and age (calendar years or meter hours)

  4. Machinery application knowledge and operating conditions (speeds, shock loads, standby service, duty cycle, etc.)

  5. Machinery work environment (mine site, steel mill, high elevation, extreme cold, etc.) and close-proximity potential contaminants (gamma radiation, coal dust, wood pulp, H2S, salt water, high sulfur fuel, etc.)

  6. Machinery metallurgy, seals, surface treatments, etc. (information on all oil-wet and frictional surfaces)

  7. Lubricant identification/description (brand, performance specification, etc.)

  8. Lubricant application knowledge (total oil volume, operating oil temperature, cooler/heater use, circulating, splash, etc.)

  9. New lubricant baseline data (viscosity, base number/acid number, FTIR spectrum, additive concentrations, etc.) from sample supplied by the client to the lab

  10. Machinery maintenance history (routine PMs, filter type and change history, etc.)

  11. Lubricant service history (oil change interval, age of oil, top-up rate, additive reconstruction, recent bleed-and-feeds, oil reclamation practices, etc.)

  12. Repair/failure history (including that of other similar machinery in similar service)

  13. Operator observations on equipment performance (erratic valve shifting, noisy operation, etc.)

  14. Information on relevant equipment inspections (foam in sight glass, varnish, leakage, BS&W, etc.)

  15. Information on other condition monitoring tests of relevance (vibration, thermography, proximity probes, etc.)

  16. Feedback from past oil analysis in which there were reportable conditions with recommended inspections or corrective measures (were these actions performed and what was the outcome?)

Of this list of 16 knowledge categories used by the oil analyst to interpret the significance of data from a single sample, only the first one comes from what’s in the bottle of oil supplied by the client. The remaining categories on the list (numbers 2 through 16) can be defined as provenance.

Because of the lack of provenance, many oil analysts are forced to do little more than go through the motions in their effort to make sense from seemingly nonconforming data. In contrast, other labs push their clients hard to supply comprehensive and detailed information in each of these categories.

Alternative Solutions

Many industrial clients have found the problem is best solved by bringing the lab to the machinery. This strategy has given rise to the proliferation of onsite labs in recent years. Another popular solution is the hybrid approach.

Here the outside lab still performs the analyses, but the uninterpreted data is supplied promptly to the client. Trained onsite oil analysts then spend the needed time to evaluate nonconforming data with the aid of modern oil analysis software.

The difference here is that the onsite analyst is usually a technician from the maintenance or condition monitoring group and has experience with the machine and its work environment. Additionally, the technician will have ready access to provenance records, maintenance/operations staff and the machine for follow-up inspections.

Even in the hands of a trained professional analyst, oil analysis without provenance may be of limited value for complex machinery in the industrial setting. Collecting, organizing and making provenance accessible involves knowledge engineering. In the electronics age, having knowledge is more a matter of knowing where to find knowledge.

This requires information management and a focused effort. It ain’t easy. But in oil analysis, as in most pursuits in life, “you only get out of it what you put into it”.