How important is the sample information you provide?
Edward Eckert

Envision that the laboratory receives a diesel engine sample with an established sample history.  The bottle label already has the customer information pre-printed, but the only information provided regarding the sample is the unit identification number and time on oil – 15,000 miles. The test results when viewed by the data evaluator, show elevated silicon, aluminum, iron, and slightly elevated potassium, when compared with the historical data. What are the two possible causes?

The first possibility is that the engines’ air intake system is leaking or there is a dirty air filter causing abrasive wear. (On a side note: One good indicator of abrasive wear is the ratio of silicon to aluminum. If the ratio is 4:1, silicon: aluminum, it is most likely dirt contamination).

The second possibility is that the engine has either had an in-frame, or has been replaced. The elevated elements would indicate break-in wear. Especially with slightly elevated potassium and no sodium present, that would point towards the intercooler as the source.  

Even though the time on oil was given, the unit (engine) time wasn’t, so a decision has to be made, one way or another. For proper data interpretation to take place, all questionable variables must be removed. This is especially true for the use of computerized automated data analysis. For a sample not to be rejected during the automated analysis process all variables including oil time, unit time, unit make and model, oil brand, grade and type, must be defined.

Okay, we know the importance of including oil time and unit time, what about static data like system capacity? Well, the condemning limit for iron in a gearbox is much higher in a 5 gallon capacity gearbox, when compared to a 50 gallon capacity gearbox. The same is true for other equipment such as bearings, hydraulic systems, turbines, etc. 80 ppm iron may be fine for a gearbox with a capacity under 10 gallons, but what if it’s a gear box with a sump capacity of 110 gallons? The oil analysis report comes back as normal, but the gear fails. Had the lab known the oil capacity, the iron at 80 ppm would have been flagged and a corrective action comment would have been made on the report.

Any static data (capacity, make, model, etc.) does not have to be information continuously provided with the sample once the lab has this data associated with the unit in its data base. The bottom line is that the interpretation of the reports you receive form the lab are only as good as the information provided. Many key variables, not just time on oil, are looked at when reviewing sample data. Any missing information could possibly mean the difference between being able to make a proactive decision and take action before a failure occurs (the main reason for using oil analysis), or having to make a reactive one making a repair after failure.

For more information please visit our website at www.staveleyfa.com or e-mail us at info@staveleyna.com