UOA vs Teardown results

[BuickGN]

I think we need this out in the open. I see people every day basing oil choice on used oil analysis wear metal results. We all know the benefits of used oil analysis for showing oil life remaining, coolant leaks, fuel dilution, etc. What I'm focusing on here is the common $30 used oil analysis and wear metal results.

My experience is they mean nothing whatsoever for determining wear rates. I had an engine in the process of failing, sparkly oil and all and the used oil analysis were ok. Not great but not bad. I did several used oil analysis out of curiosity as my oil pressure was dropping and it was coming out sparkly. Nothing came out that was out of the normal range. The only thing I saw was a little copper where there was none before.

Upon teardown, this engine looked like it had a million miles on it. Cylinders were worn with a large ridge up top, no trace of hone marks, rings were noticably lighter just by holding them in my hand, pins were frozen in the pistons, rod and main bearings were nearly to the shells. Crank was scored. Yet the used oil analysis were ok.

When I see used oil analysis, the wear metals are always within a few ppm of one another regardless of how long the OCI is, sump size, or engine size. To me, this is laughable especally knowing my results in the GN. To me, choosing one oil over another based on 1-2ppm is not the best way to choose oil. Any other thoughts?

Let's try and keep this civil.

 

[OVERKILL]

I really like Doug's take on it:

Originally Posted By: Doug Hillary
By Doug Hillary

used oil analysis are a great tool in the Management of any machinery that uses liquid lubricants. Unfortunately, their real value is often misunderstood by those who contribute to BITOG.

Firstly, it is important to realize that you get what you pay for. The most common forms of used oil analysis are limited in their scope. It is a case of if you pay more you get more. So my comments here relate primarily to the “simple” used oil analysis – the cornerstone of those appearing on BITOG

Secondly, it is easy to assume that by carrying out a used oil analysis you will be able to determine how quickly the engine is wearing out. As well, if you change lubricant Brands you will be able to compare the wear metal uptake results and then make a balanced best lubricant choice to make your engine last longer.

Sadly that logic is seriously flawed.

Single pass (random) used oil analysis will provide some information regarding wear metals but unless you have a history of your engine’s performance up to around 1 million miles the results are simply that – used oil analysis results! As an example a limit of 150ppm of Iron is a reality – after say 100k it means the lubricant should be changed and all is well. But what is the situation if you have 150ppm of Iron at 5k? Where would you look what would or could you do? So used oil analysis are really a diagnostic tool – one of many!

The other parts of the used oil analysis Report will be much more valuable to you – it will tell you about the CONDITION of the lubricant and its suitability for further use. This will enable you to get the maximum safe use from the lubricant saving a valuable resource in the process.

With some knowledge and experience you may be able to determine if you have a fuel or cooling system problem – and you will be able to see the level of potentially damaging contaminants such as silicates, soot and etc. You may be able to monitor the levels of additive depletion too. In a low use engine the TBN (if taken) will seriously indicate if the lubricant is still able to minimise corrosion and its effects and this will help to prevent cam lobe pitting etc.

During the last 50 or so years I have carried out hundreds of used oil analysis on all sorts of engine configurations – and on transmissions, gearboxes and differentials. These were done in consort with four Major Oil Companies (Shell, Exxon-Mobil, Castrol, Chevron-Caltex), my Employers, my Customers and on my own Fleet and other vehicles. Not once were they ever used to discriminate one lubricant from another on the grounds of wear metal uptake!

In my case correctly interpreted used oil analysis have saved the odd engine from major problems and have proven that certain viscosities work better than others in some specific engine families. used oil analysis certainly optimized the service life of the lubricants in a cost effective way for many of my customers

 

[StevieC]

BUICKGN,

I completely agree that occasional used oil analysis show nothing more than if there is a coolant leak, fuel dilution, and the TBN/TAN levels.

However, if you have used oil analysis done regularly and happen to notice a spike in one area it is a good warning sign to catch things early enough that you can repair the engine before catastrophic results.

The problem is no one does them regularly because of the expense of them so this is hard to catch, and the other problem is that you can have a failure like in your case with minor elevated levels.

So I think the answer is sometimes yes and sometimes no and always no if you don't have continual used oil analysis done.

Also I agree that an oil should not be chosen simply because it produces 1-2ppm less metal counts on a used oil analysis because this can be a sampling error, testing error, or oil batch formulation error, or could even be that the metals in one used oil analysis were in the right range for the test to pick them up whereas another used oil analysis they go undetected and there is actually more but they just don't fit in the scope-range etc.

I do agree with not using an oil that happens to spark higher levels of metals in one or two areas and then drops to the normal trend when the oil is no longer used as clearly this oil doesn't seem to agree with the conditions placed upon it.

This doesn't mean that excess wear is being caused, but it is a good indication that something just isn't right and with all the oil choices out there why would one risk using an oil that clearly demonstrates a potential for a problem versus one that doesn't.

Sorry if I just added more confusion to this thread, but this is how I see it.

 

[oilboy123]

Interesting discussion........

 

[Gary Allan]

Well, I've got a email into Polaris to see what a PQ index adds to their advance UOA. That should up the validity. I hope that it's not an arm and a leg.

Now suppose the whole BITOG crowd migrates to Polaris due to PQ ..and the results parallel the Fe readings? That is, everyone with higher Fe also showed that the index climbed too. Right now we all have the speculation of "well, you can't validate it so it's not true". When that speculation gets narrowed ..IF it gets narrowed, what will be the next challenge?

 

[maximus]

Yeah, makes sense. It would be like getting regular checkups and blood work done, however you could still develop cancer in the next 5 years, albeit the human body is far more complex than an engine.

I feel the value of used oil analysis is knowing how long to use oil. It can be considered a cost saving tool(over time). Frankly, I feel all oil is the same. Syns last longer. That's it lol. My supply of PP is running low. After it's gone, I'm going back to conventional oil.

 

[demarpaint]

After spending many hours on this site and trying to learn all I can, for me, let me repeat, for me, a used oil analysis is a waste of money. BuickGN convinced me months ago when he told the used oil analysis story about bearings that were trash upon a teardown, yet the used oil analysis were good.

People will defend an oil saying that 30 ppm of iron vs 5 ppm of iron is not a big deal, and in the life of an engine it means nothing. Maybe maybe not. To me the lower the numbers the better, I guess? Then in order to get a valid story about a particular oil it has to be used twice to make sure no residual oil is throwing the numbers off. Then if we want to find the very best oil for the engine, maybe we have to test 4or 5 oils. That's 8-10 oil changes and 4 or 5 used oil analysis to try and find the "best". Then we read any brand name SM oil will protect the engine for the life of the car, use the cheapest oil you can find. That makes the used oil analysis even more pointless to me.

Now I'm not knocking used oil analysis I read through them, and see what oils fare best, and that helps me make my oil purchase decisions. It was other used oil analysis reports that got me using PP in my E-150. In fact I've even suggested getting a used oil analysis to members.

If I was noticing anti-freeze vanishing, or saw my oil was very thin upon visual inspection, maybe a used oil analysis would have some value, and I'd probably spend the money. I'd rather change the oil early, and use the $30-$100 for a used oil analysis for that. JMO

 

[rpn453]

Originally Posted By: BuickGN
When I see used oil analysis, the wear metals are always within a few ppm of one another regardless of how long the OCI is, sump size, or engine size. To me, this is laughable especally knowing my results in the GN. To me, choosing one oil over another based on 1-2ppm is not the best way to choose oil.


I agree. I only find a used oil analysis to be interesting if it indicates an obvious problem.

I'm surprised to hear that the used oil analysis didn't provide an indication of the extensive engine wear. Did you ever post them? I searched but didn't find anything.

 

[ekpolk]

Originally Posted By: BuickGN
. . .

My experience is they mean nothing whatsoever for determining wear rates. I had an engine in the process of failing, sparkly oil and all and the used oil analysis were ok. Not great but not bad. I did several used oil analysis out of curiosity as my oil pressure was dropping and it was coming out sparkly. Nothing came out that was out of the normal range. The only thing I saw was a little copper where there was none before.. . .


BGN:

I agree in part, disagree in part. My disagreement is with the specific conclusion, which you state as, "My experience is they mean nothing whatsoever for determining wear rates."

The specific basis for disagreement is that I think you overstate an otherwise valid concept. I very well understand that an inexpensive UOA, with little or nothing more is of limited value, but certainly not worthless.

If your engine is failing in a way that is causing relatively large chunks of metal to spall away from the surface involved, and there is no other failure mechanism operating at dangerous levels, then very much yes, you are correct in that a simple UOA won't tell you anything about that.

On the other hand, if the failure mechanism involved is producing wear metal in the particle size range that is measurable by UOA, then this is where I disagree, and where a simple UOA may, and I emphasize may help. A good example of this would be the fairly common used oil analysis we see which seem to show soft metal bearing wear. I often see this attributed to a so-called "particle streak". But of course, high lead, for example, may also indicate that one or more bearings are in already in their death throes, and that it's only a matter of time.

So, to sum up, my specific disagreement with this premise is that you simply take it too far. Other than that, I'm very much in agreement that basic UOA is not a magic "be all, end all" that can divine every last detail about wear in your engine.

 

[pbm]

I agree with BuickGN and I would rather spend my money changing my oil than having a used oil analysis done. I did have a used oil analysis done once to see if my GM 3.1 was leaking coolant into my oil.

 

[benjamming]

Buick,

Conclusions like this are why folks like Terry are in business. They can usually tell you something is amiss & corrective action must be implemented. I don't have a large enough database (read, esentially none) nor experience to read the tea leaves. Terry used to get hammered for saying that a used oil analysis was bad (or good) when "we" couldn't see anything due to the variation that folks here typically dismiss.

 

[ekpolk]

Originally Posted By: pbm
I agree with BuickGN and I would rather spend my money changing my oil than having a used oil analysis done. I did have a used oil analysis done once to see if my GM 3.1 was leaking coolant into my oil.


Unstated, and faulty premise: you can only change oil OR do used oil analysis. I prefer to do BOTH. My engine has fresh, clean oil just like yours, AND I know much more about the condition of my oil, and YES, of my engine too (but again, certainly NOT perfect knowledge).

 

[yeti]

bgn -- i was wondering, at what point in time, would you post this in such detail. demarpaint -- excellent post. my thoughts, exactly. used oil analysis are a good tool, but have their limitations. i don't know of any other way to get a true and accurate picture of internal part's wear, other than a teardown.

 

[Zaedock]

Originally Posted By: BuickGN
I think we need this out in the open. I see people every day basing oil choice on used oil analysis wear metal results. We all know the benefits of used oil analysis for showing oil life remaining, coolant leaks, fuel dilution, etc. What I'm focusing on here is the common $30 used oil analysis and wear metal results.

My experience is they mean nothing whatsoever for determining wear rates.


Hey man, you know I agree with you on many subjects here at BITOG, but I'm going to have to disagree that used oil analysis are useless.
Besides being great to determine oil longevity, they can be helpful to monitor wear metals. I know in your case, it did nothing for you, but that's one engine in a world of millions.

I think you're lucky that you have the ability and time(plus, it's your hobby) to tear your engine down and check it out. Most everyone else are not able to, so a used oil analysis is the next best tool. Given the knowledge I've picked up here over the years and my personal drag racing and off road experience, I'll offer that used oil analysis can be helpful in most cases. Using my 2.5L trail Jeep as an example, the Fe levels have been in the single digits in a design that ofen has high iron. That's running Delo SAE30 of all things. Ultimately, what does it mean? Heck, no one really knows where the iron comes from. One thing I do know is that I feel a little better seeing 9PPM iron instead of 29 after a season of beatings. Although the next time I roll, the engine could fry anyway...

Many people buy things on recommendation and perceived performance, oil included. You have to admit that it's hard not to look at Utah Bill's used oil analysis and run out to get the oil he's using.

 

[ADFD1]

Originally Posted By: benjamming
Buick,

Conclusions like this are why folks like Terry are in business. They can usually tell you something is amiss & corrective action must be implemented. I don't have a large enough database (read, esentially none) nor experience to read the tea leaves. Terry used to get hammered for saying that a used oil analysis was bad (or good) when "we" couldn't see anything due to the variation that folks here typically dismiss.


To pay $100 for that service for the average Bitog member I feel is out of reach, and if a $30 Blackstone report doesn't tell much what is the point? Especially in a good running engine. Millions of people have logged hundreds of thousands of miles on engines and never heard of a used oil analysis.

OTOH-if I had an engine that was giving me problems, and was under warranty $100 Dyson report might help if I had to take the car maker to court. Other than that I don't see the value, and BuickGN reinforces my beliefs. As mentioned above I read the used oil analysis and it helps me decide which oil MIGHT be best. YMMV

Now if I were running a fleet of trucks on the Ice Roads of AK, used oil analysis would have a value to me especially if I can safely extend OCI's based on them. Since I now fall into normal/severe service intervals I save money on used oil analysis and change oil somewhere between normal and severe service. Knock on wood it is working for me and my family.

AD

 

[JAG]

used oil analysis wear metals only measure particles small enough to be measured and the particles need to make it into the oil sample bottle. Different particle size ranges do occur with different oils, different engines, different driving styles, different dirt contamination, etc. Different sampling methods can also affect the particle sizes captured as well as the concentration of them. Different particle sizes affect how likely they end up in the oil sample bottle.

So there are many pathways for the results to be underestimated (relative to true concentration). If comparing one result that had less underestimation to one that had more underestimation, that one result will appear elevated.

But if you hold as many variables nearly constant (including the specific oil used and having used it long enough for stabilization to occur) and do several used oil analysis back to back, trending can generally allow decent conclusions to be made.

I say generally because there is still the possibility of severe wear occurring and the particles are so large, that the used oil analysis results greatly underestimate the true wear rate. If severity of wear goes up from there the results will be an even greater underestimation and it will be hard or impossible for the used oil analysis reader to catch it. This situation is not common but it does happen to those unlucky enough to be in it...like BuickGN was. It's also happened in VW 2.0T engines where severe wear has occurred in certain parts of the valvetrain.

Since different oils generally produce different wear particle size distributions, I think it is generally unwise to say that based on a set of used oil analysis, one oil is causing less wear than another. Again I say "generally" but not always.

Adding a PQ Index test to a standard used oil analysis, as Gary mentioned, greatly helps get a better estimation of true iron wear rate. But that still is based on what's in the oil sample bottle so that needs to be as representative as possible the ideal: the entire oil in the engine with all iron/steel wear particles homogeneously mixed in it.

 

[buster]

+1 to what JAG said. Great post.

 

[saaber1]

Copied from another thread:

I think Pablo said it best by saying we need to be "cautious" when using used oil analysis as a measure of wear. The used oil analysis without particle analysis may catch wear or it can miss it entirely.

IMO, it is probably not wise to say used oil analysis "always" are a measure of wear and it is not wise to say used oil analysis are "never" a measure of wear. It is probably better to say it "can be" useful to indicate wear and the more information we have, such as expert analysis, good trending, etc. the more useful it can be in this regard. I think in general we all like to see lower wear/contaminant metals than higher.

Here is an example of a cam follower that failed in a 2.0 FSI engine and there was no trace of evidence in the used oil analysis trends for that car: http://www.bobistheoilguy.com/forums/ubb...500#Post1433500

Here is a counter example of the same engine with exactly the same problem where a used oil analysis clearly showed something wrong:

Originally Posted By: corradokidg60
...Here's my Blackstone oil analysis for 40k miles worth of changes (4) - the last oil change shows metals in it because I later found out I had the dreaded HPFP/intake cam/cam follower wear issue.

So I think being "cautious" is the key. used oil analysis are one source of information and useful under certain conditions as JAG said. IMO teardowns are the best indicators of wear but it is too hard/expensive for the average Joe, but they can easily do used oil analysis.

 

[saaber1]

For those used oil analysis that show the engine problems, Rl_RS4's graphing technique can be very helpful. Here is the graph of the 2.0 FSI used oil analysis, with outliers removed except for the used oil analysis shown above for corradiokid. When compared to the average, that used oil analysis jumps out as something wrong. Note that the used oil analysis may not show anything wrong, such as the linked example above where a used oil analysis before failure showed nothing, but it can show it, and the more info we have, the better we can see this.

 

[yeti]

jag -- good post. and you bring up a good point.one could take 4 samples in a month, a week apart, and have different uoa results because of sampling method variables. this would, obviously, not be the fault of the testing lab.
another reason why teardown should be THE benchmark to determine internal part's wear.