Perhaps the most difficult part of interpreting used oil analysis data is not deducing the cause and effects of abnormal data, it is deciding a proper corrective course. Far too often it is suggested that simply changing the oil will fix the problem. That is akin to placing a bucket under a leak; it will barely address the symptom, let alone cure the disease.
Lower levels of wear metals and contaminants are concomitants of oil changes, insomuch as fresh oil will lower these values simply by dilution. Given the same interval since the last oil change, the system is practically guaranteed to achieve the similar values again and only much higher values if a significant problem exists. Why then is the suggestion so common if, at best, this action merely buys some time?
We have probably all heard the follow quote or a paraphrased variant: “If you keep doing what you've always done, you'll keep getting what you've always gotten.” Sometimes a more humorous slant is taken by suggesting the definition of insanity is simply repeating the same action expecting a different result. The most concise variant seems to be: nothing changes if nothing changes. If we believe these doctrines, then we should expect that if an asset fails with changing the oil as the only corrective measure then that same failure will occur again with continued operation under the same conditions.
A common example is an asset whose level of silicon and wear metals has increased in the latest of three samples since the last oil change. Two common remedies are to either change the oil, or to inspect the breather or any other ingression point on the asset’s lube system. If the former, changing the oil really does not fix anything, or at least has little lasting effect. If the latter, suppose an issue with the breather is discovered and the breather is repaired or replaced, what next? Too often the answer follows the familiar refrain.
How many samples will it take to determine if the ingression has been stopped, assuming regular sampling intervals? By changing the oil, the next set of results should almost certainly be lower, but it may take the same interval of three samples before obtaining positive proof.
How many samples would it take if the oil was not changed after the repair? Arguably only one, though some mistakenly believe the next set of results will be lower, rather than simply stabilizing at their current value. We must qualify that the level of silicon and the increase in wear metals be tolerable, since there is little point leaving the oil in for one more interval if significant damage is occurring.
Perhaps many unnecessary oil changes can be avoided in the future with a greater emphasis on root cause investigations. With this philosophy, oil analysis has a chance to be a successful condition monitoring tool, rather than the documenter of failures.
Oil changes are able to address problems directly associated with the oil itself like a build-up of oxidation by-products such as varnish; but even then the oil change is not going to prevent the problem from recurring, it will only defer the next occurrence. Corrective measures like off-line filtration are far more attractive, but they may not be a practical, feasible or economical solution in all cases.
We need to appreciate that changing the oil merely makes an oil analysis report look better, it does not fix contamination ingression, or balance, repair or realign parts. Break the cycle of repeated failures by reconsidering the validity of an oil change as a corrective measure.