In recent years, I’ve been part of numerous investigations relating to expensive machine failures preceded by oil samples that registered high particle counts. Each time I asked the asset owner, my client, why the aberrant particle trends were not investigated further. In most cases the response was that they didn’t think it merited the extra effort or they didn’t know how to probe further for more information.
High particle counts generally have one of four possible explanations and outcomes when investigated:
The machine is not in any immediate danger; however, either a filter has failed or a new source of high particle ingression has developed. The problem is solved by correcting the offending filter or source of ingression.
There is a new ingression source or filter failure and the machine is in immediate danger due to the resulting high particle count. Problem is solved by rapid cleanup of the oil followed by correcting the failed filter or ingression source.
The high particle count is due to abnormal wear particle generation constituting a potential threat to machine reliability. The problem is solved by performing a root cause failure analysis followed by appropriate remediation and cleanup.
The high reading is due to a sampling error, analytical error (such as particle counter calibration or sample preparation) or soft particles (dead additives, oxide insolubles, or others).
Rising Particle Count Trend
What constitutes a high particle count? The answer is specific to the machine, application and desire for reliability. For instance, high-pressure hydraulic systems employing the use of sensitive electrohydraulic valves have cleanliness needs that exceed the requirements for most other machine types. Cleanliness is not free. The same is true for the cost of investigating a particle count excursion that does not pose a threat to machine reliability. We’ve all heard the words “analysis paralysis!”
The various techniques for customizing cleanliness targets assigned to lubricating oils and hydraulic fluids have been discussed within the pages of Practicing Oil Analysis and Machinery Lubrication magazines. This is a practical and important exercise that helps mitigate unwarranted concern and root cause queries. Once these targets are set, the cautionary and critical limits must be defined. Next, and equally important, determine the needed frequency of performing the particle counts.
It goes without saying the more rigorous the target (highly clean), the more frequently particle counts should be performed. In general, you are testing for cleanliness not dirtiness. It is no surprise that more than 25 different particle counters are being sold in the oil analysis field. If something is important, measure it regularly and keep track of your findings.
Triaging Incoming Oil Samples
There are many ways a lab can triage incoming oil samples, but testing for viscosity changes and rising particle counts are among the best. For instance, most tribologists would agree it’s darn hard for a machine to fail internally without producing particles. Likewise, it is difficult for a lubricant to fail without a change in viscosity. Of course there are exceptions to both statements, but I’m sure you get my point.
The concept of triaging is similar to screening but includes assigning a severity score to each nonconforming sample followed by a plan of exploratory exception testing and inspections. The purpose is to determine why the particle count is high without wasting time and effort or neglecting to recognize a reliability-threatening event.
In a past issue of Practicing Oil Analysis, I wrote an article entitled “Tricks to Classifying Wear Metals and Other Used Oil Suspensions”. The analytical schemes discussed in that article can be very helpful in answering the question “what do I do with a high particle count?” But first, don’t file away the test report or dismiss the condition prematurely. If you follow the test rationale, you will likely discover which of the four possibilities listed above best applies. Then, respond accordingly.