The Limitations of Oil Analysis – A Car Enthusiast’s Guide

I often find that people totally miss the scope of your typical LPC oil analysis, the likes of which CAT, Polaris Labs (Oil Analyzers Inc is one of their private labels), Blackstone Labs, etc sell. I’d like to correct a few misconceptions in this article that have been circulating the web for some time now with regard to what oil analysis does, and doesn’t do.

Common ways people use oil analysis reports incorrectly:

Comparing lubricants

I’ve seen it more times than I care to count. Someone wants to evaluate one oil against another to see which oil is better in their application, so they decide they’ll try an oil, get it analyzed, try another oil, get it analyzed, and repeat until they’ve tried all the oils they want to. This is probably the best example what not to do, and I’ll outline the reasons below.

  • Oxidation, deposit control, and acidity. Oil is tasked with many things, among which are oxidation resistance, deposit prevention, acidity neutralization, and contaminant suspension. If one oil does a poor job of controlling deposits, oxidizes much faster than another oil, or neutralizes fewer acids, the next oil will be immediately tasked with cleaning up what the previous oil left behind, which will have an early effect on base number decay, antioxidant degradation, and contaminant loading. Even if the second oil is much better than the first, these factors will skew the results and may even make the oil look worse because of how much it cleaned up from previous oil changes.
  • Conflicts, incompatibilities, and reactions in additive chemistry. It is far more common than you think to have conflicts in additive chemistry from one oil to another. You can switch oils from one to another and see elevated copper levels on the new oil for no obvious reason, caused by the reaction of copper sulfides with the add pack in the new oil. Noria Corp explains this to some extent in this article: Keep in mind this is just one example of a possible chemical reaction. Not only can one add pack react to another in an engine oil, but one add pack can react differently to the engine and cooling system than another one would. Without a knowledge of these possibilities, you might perceive a change in metal on one analysis incorrectly. These additive packages are formulated specifically to be compatible with the oil they are blended into; not with another finished product. An AMSOIL oil using an Afton Chemical additive package may be quite different than a Mobil 1 oil using an Lubrizol additive package. Lake Speed Jr from Driven Racing Oil explains this as well at 14:35:

 A lot of oils don’t like each other. If you go from brand A synthetic to brand B synthetic, and if one of them loses like 12 horsepower, it’s probably not because brand B is a bad oil, it’s because brand A and brand B are incompatible…you have to be very methodical about flushing a wet sump motor to ensure that you don’t have that cross-contamination…

  • Environmental factors. Without a great deal of time, money, and a dyno, you won’t be able to maintain consistent operating conditions. You may try one oil during the summer, where you’re operating in high temperatures with A/C running that is producing a higher average load, then switch the oil in the fall and run the second oil through winter, where you’ll see cold starts and more time spent below operating temp. These factors all have an effect on engine wear, and it is extremely difficult to maintain an acceptable level of environment consistently from one oil change to another.

To perform this testing correctly, Lake Speed Jr from Driven Racing Oil explains the process for testing oils, starting at 13:05: The correct process is to run your baseline oil (Oil 1), then make 7 dyno runs with all factors controlled. Throw out the best and worst, average the 5. Then, drain that out, put in a flush oil, run the engine on that for a few runs, drain that, and put the next test oil (Oil 2), make a few runs, then drain it. Then, put another change of Oil 2 in again, and make 7 runs, throw out the best, the worst, and average the 5. It’s difficult enough for testing power on the dyno, but almost impossible to do with oil analysis, yet that’s what you’d have to do in order to get truly useful, valid results.

If you don’t do that, you’re not testing oil, you’re changing oil.

I understand that most enthusiasts don’t have the resources to do that kind of testing, so my recommendation to enthusiasts hell-bent on using oil analysis to compare oils is to get three oil analysis reports done after switching to a different oil and using that oil three times. Throw out the first report, then compare the results on the next two. If the results on the next two are reasonably close, within acceptable tolerances, average the results. If they are still far apart, you can throw out all three reports, and any hope you had of drawing any conclusion on that oil.


Relying on one single analysis

One single oil analysis doesn’t tell you a whole lot for reasons already mentioned above. One single oil analysis from a lab that barely tests anything more than LPC (liquid particle count), is worth even less. There are many limitations of oil analysis to begin with, and using just one report, even when not comparing lubricants, limits you even further.

For example, your LPC oil analysis is only really effective at testing between 1-10 microns, some at 1-6 microns effectively. That is a very small window to be evaluating a lubricant, especially when your range of contaminants and wear is from tiny fractions of a micron clear to 30 microns and often above (especially if you’re using a nominally rated cellulose filter or the filter is plugged). I’ve had customers send in oil glittering with metal that reported acceptable levels of metallic “wear.”

Barring critical conditions that would suggest catastrophic failure has (or will soon) occur, it isn’t as important what your wear levels and additives look like on one analysis as it is how they are trending. Oil analysis reports, especially when extending service intervals, are most valuable when trends are observed, as this allows you to evaluate the rate of wear, oxidation, viscosity shear, base decay, etc. I once had a customer run one of our euro spec oils in a 2.0L diesel engine for 100k miles. You’d think, looking at his oil analysis report at 33k miles that the oil was nearing the ends of its life cycle. Base number had dropped to 3.37, Oxidation up to 24, and viscosity .2 cSt down from new. However, at 100k miles, base number was only at 2.62, oxidation at 23, and viscosity had remained exactly the same. Furthermore, you might see the report at 33k miles with 44ppm Fe, 7ppm Al, and 3ppm Cu and expect that those values would triple by 100k miles, except the values at 100k miles were at 87ppm Fe, 17ppm Al, and 7ppm Cu; an approximate doubling of all three even though the service interval was tripled.


Assuming oil analysis reflects an oil’s performance

There was a test done not too long ago by Blackstone Labs where they took one single Subaru engine and inferred (by themselves or by 3rd parties like Jalopnik), based on their results of oil analysis reports at approximately 3,900 miles , that it didn’t matter what oil you used, all engine oils would protect the same at that interval. To make matters worse, Jalopnik’s editor made no corrections to the article after I explained the flaws in those claims and the amount of misinformation spread by that article. The problem here is that there are many factors that affect a vehicle’s health with respect to engine oil that won’t be reflected in an oil analysis, especially a Blackstone oil analysis. I’ll outline a few big ones.

  • Oil analysis, as mentioned before, is only effective within a certain range. I’ve long suspected that turbo wear occurs at ranges smaller than these oil analysis reports are picking up. It’s on multiple occasions I’ve skimmed through NASIOC oil analysis reports and found people alarmed at finally seeing 1-2ppm of lead that their turbo was on its way out. Does the author really think that there was absolutely no wear occurring for tens of thousands of miles until that point due to the 0ppm of lead found in previous reports?
  • Volatility refers to the oil’s vaporization rate under heat. This is a direct contributor (often the largest contributor) of oil consumption in an engine. Oil consumption will directly affect deposits formed on the piston crown and combustion chamber, which will cause knock and reduce efficiency through hot spots.  Furthermore, these deposits will also coat the intake valve and can, under worst case scenarios, compromise the valve seal and result in a burned valve. Elevated oil consumption is directly responsible for shortened catalytic converter life in addition to PCV problems as more oil vapor is run through them.
  • Oxidation stability (the oil’s ability to resist deposits, coking, sludge, and varnish), is directly responsible for elevated oil consumption through piston rings as they begin to stick and allow oil to pass through. Some engines, like the Honda J35 with VCM are highly susceptible to this problem, often resulting in engine rebuilds due to high oil consumption, where oxidation-stable oils maintain the engine’s overall health. Thermo-oxidation stability (ASTM D6335) also tests deposits formed in turbocharger heat conditions, and will result in coking of the turbo feed line as well as deposits formed around turbo seals and bearings, resulting in elevated oil consumption and premature turbo failure.
  • LSPI. While not directly affecting engine wear, additive chemistry is directly related to LSPI frequency, which results in catastrophic piston failure. There are many resources on this subject, the best one of which I’ve come across is the RealTuners episode I linked above.
  • Onset of abnormal/unexpected conditions after testing. An oil may look good on an oil analysis report for quite some time, until the unexpected happens. An injector starts leaking (increasing fuel dilution), the EGR system malfunctions, your commute suddenly changes to extended stop/go traffic, your idle time increases, or your operating conditions otherwise stretch the definition of “severe service.” Under such conditions, weaker lubricants begin to expose their limitations and flaws. If you’re going to use oil analysis to evaluate lubricant serviceability, suitability, and performance, you have to continue doing so whenever your operating conditions change.


I will continue to add to this list over time, but my biggest point here is to demonstrate that your affordable LPC-based oil analysis reporting is extremely limited in scope. You have to understand what it can and cannot do, how to interpret the results, and how to evaluate trends. I’ll start work on another article explaining the proper scope of oil analysis here in the future and will link it here when it’s done.

As always, it helps to use a lab that performs testing such as fuel dilution using gas chromatography, as well as oxidation. I recommend using Polaris Labs (sold under the Oil Analyzers Inc private label) through the AMSOIL site:

UPS Prepaid:

USPS Prepaid:



If you need help with an oil analysis report, please feel free to reach out using the contact link on this site.