A Data Analyst’s Hypothetical Scenario
by Grant Dawson
Oil analysis is a scientific predictive maintenance tool and it may seem that with the amount of data we can collect on your sample that it should be easy for the lab to identify problems from just the data alone. Sometimes this is possible. However, just like a blood work test, a doctor would have much more difficulty interpreting the test results without knowing the gender, age, weight, height, diet, habits, medications and medical history of the patient. In both the medical field and oil analysis the information supplied to us is of paramount importance in interpreting laboratory results.
All too often an oil sample is submitted to a lab with only a unit ID and an indication that the oil is from a diesel engine. In such cases only the most basic evaluation of the data can be performed and will likely be of little value as a predictive tool to the end user. Things such as engine manufacturer and model, the total time on the unit, the amount of time the oil has been in service, the oil brand, type and weight, sump capacity, amount of oil consumption between PMs, the engine’s application and even geographical location are major factors in determining the component’s condition. The importance of providing the maximum amount of information as is possible cannot be understated.
There are numerous way in which serious mechanical problems may go undetected due to lack of information. We’ll examine one scenario in which lack of information can lead to costly, unnecessary damage. Let’s say that your normal procedure is to collect an engine oil sample at each PM interval and that interval is every 10,000 miles, but this information is never provided to the lab. Over a period of six submitted samples the data consistently reveals the presence of approximately 80 ppm of iron (indicating a normal and predictable rate of wear). But at 1,500 miles after the last PM the operator reports an unusual noise coming from the engine. The unit is pulled from service, an oil sample is taken and submitted to the laboratory to determine the cause of the noise, though the unusually low time the oil has been in service is not reported.
After testing, the data indicates the presence of approximately 80 ppm of iron as it typically has. Given the lack of data provided to the laboratory nothing seems unusual to the data analyst since that level of iron is consistent with prior samples. However, the iron which was normally accumulating at a rate of approximately 0.008 ppm/hour has now increased to accumulating at a rate of 0.053 ppm/hr: an accumulation rate that is over 500% of its historical value. Without the knowledge that the time that the oil was in service had been reduced to 1,500 miles the 80 ppm of iron appears consistent and typical to the data analyst, and you receive a report stating that the component condition appears normal. The truck is placed back into service, the unusual noise persists, and shortly thereafter the engine fails. Engine teardown and inspection reveals that a cylinder liner had developed a crack which rapidly worsened over a small period of time, eventually causing a failure as well as damage to the piston and rings.
If the time that the oil was in service had been consistently reported to the lab, the data analyst would have recognized that the seemingly typical iron level was critically high for just the 1,500 miles the oil was in service and recommended an inspection of the cylinders and/or other typical components which render iron into the oil. If the damaged cylinder liner had been discovered and replaced at the first sign of an unusual increase in wear metals it is much more likely that a costly catastrophic failure could have been prevented. Though it may have seemed insignificant, something as little as the time that the oil has been in service not being consistently reported could mean the difference between an inexpensive part replacement and a very expensive catastrophic engine failure. Hence, the data provided to the lab is equally or even more important than the data the lab provides to you.
The aforementioned scenario is but one of many examples of potentially serious ramifications arising from a lab not receiving adequate information. Omitting any of the information that we request on your sample information form can lead to problems or failures similar to the hypothetical scenario listed above. Always be sure to provide all of the information that the lab requests, and we also encourage that you report any other information that you may feel is significant (operational problems, part replacements, recent overhaul, etc.). In oil analysis the rule of thumb is: the more information you provide to us, the more accurate the information is that we provide to you.