Do low micron oil filters really make a difference?

[StevieC]

Originally Posted by Bill_W
The Ultra has a effency of 99+ @ 20 microns

Neat info here on effency...

https://www.machinerylubrication.com/Read/31133/oil-filtration-efficiency

Good article... Thanks

 

[ZeeOSix]

Originally Posted by dnewton3
Once a sump is "clean enough", making it "cleaner" does not necessarily make it "better", if "better" is defined as less wear. Your definition of "better" is "cleaner'. My definition of "better" is "less wear". You focus on the inputs, whereas I focus on the outputs.


Well, I don't have a real time "engine wear measuring device" on my engines to know when my sump is "clean enough". So the next best thing to help combat wear is to run a more efficient oil filter - regardless of how much less wear it may provide. Like I've said and shown many times, a more efficient oil filter results in cleaner oil, and cleaner oil results in less wear. Oil is about 10 times cleaner when comparing a 99% @ 20 micron vs 50% @ 20 micron filter. That's good enough data for me to spend $3~4 more a year on decent high efficient oil filter for vehicles I plan on owning for 20+ years. My Tacoma is already 14 years old and I'm sure I'll still have it in 2025 ... maybe even in 2035 is there's still gasoline - lol.

IMO, a UOA isn't the right tool to judge the effect of wear on engines due to oil filter performance ... too many variables. And anyone looking at the "insolubles" number on a UOA as a measure of filter performance is being fooled - I proved that one already. However, the oil ISO particle count is a much more direct measure of the filter's filtering performance. It's pretty easy to see the IOS cleanliness code change for the better with a more efficient oil filter - plots posted previously.

When someone can post up some official engineering studies that simply shows that cleaner oil does not result in any less wear then I might have a different viewpoint ... all I can say is good luck on that one.

Originally Posted by dnewton3
As you state, Zee, we will agree to disagree.
I'm OK with that .... no harm, no foul

Yep ...

 

[dnewton3]

Originally Posted by ZeeOSix
... IMO, a UOA isn't the right tool to judge the effect of wear on engines due to oil filter performance ... too many variables. ...... However, the oil ISO particle count is a much more direct measure of the filter's filtering performance. It's pretty easy to see the IOS cleanliness code change for the better with a more efficient oil filter - plots posted previously.

I do agree that PCs are a direct indication of filter performance, and that generally a tighter filter will reduce contamination of the sump. We're aligned here.

But of the UOAs? Zee, this is where you lose me. You and I agree that the bus study clearly shows a good correlation between UOA wear metals and PCs. That study is hard to deny. And you and I have repeatedly agreed that the bus study exhibits this correlation well. The direct conclusion we take together is that UOA wear particles confirm that the output tracks with the input. And the indirect conclusion we take from it is that the PC data showscontamination is the likely cause of the wear. (mostly soot in that dirty running old DD 2-stroke; along with lubes that just didn't have good add-packs back then).

And yet you turn around here and state you don't trust the UOA to be the right tool for judging wear? Why not? If it's good enough to prove that filtration efficiency is important in highly contaminated sumps via high PC data which is confirmed with UOA wear metals, they why are you in denial that UOA wear metals can ALSO confirm when filtration has become irrelevant; that when wear metals are so low regardless of the filter used, it's an indication that the filer is no longer the controlling entity and other things are more imporant?

Why insist that the bus study proves filters are important via the correlation of PCs and UOAs, but then deny that relationship is important whey the same methodology tells you that filters are now a tertiary player?

Your statement actually includes the very facts you seem to want to ignore. That other "variables" are in play. I've already listed them a few times in other posts. Those "variables" are what is making filter efficiency moot. (once above some threshold; I will always agree that filters are not meaningless, but only that they loose their effect after some level of "clean" is attained).

I'm not trying to pick a fight here; I respect your right to your position. I just struggle to understand the duality of your stance. UOAs are good when they show filter efficiency is important, but they are not trustworthy when they show filtration efficiency is no longer the wear controlling entity?

Can you refine your statement to help me make sense of it?

 

[ZeeOSix]

Originally Posted by dnewton3
Originally Posted by ZeeOSix
... IMO, a UOA isn't the right tool to judge the effect of wear on engines due to oil filter performance ... too many variables. ...... However, the oil ISO particle count is a much more direct measure of the filter's filtering performance. It's pretty easy to see the IOS cleanliness code change for the better with a more efficient oil filter - plots posted previously.

I do agree that PCs are a direct indication of filter performance, and that generally a tighter filter will reduce contamination of the sump. We're aligned here.


Yeah ... it's pretty hard to refute that correlation. And that's all that really matters to me. Cleaner oil can only be a good thing, even if the benefits are very small or not even "detectable" by the layman. Worth a few extra dollars a year to me anyway to try and achieve cleaner oil in the sump.

Originally Posted by dnewton3
But of the UOAs? Zee, this is where you lose me. You and I agree that the bus study clearly shows a good correlation between UOA wear metals and PCs. That study is hard to deny. And you and I have repeatedly agreed that the bus study exhibits this correlation well. The direct conclusion we take together is that UOA wear particles confirm that the output tracks with the input. And the indirect conclusion we take from it is that the PC data shows contamination is the likely cause of the wear. (mostly soot in that dirty running old DD 2-stroke; along with lubes that just didn't have good add-packs back then).

And yet you turn around here and state you don't trust the UOA to be the right tool for judging wear? Why not? If it's good enough to prove that filtration efficiency is important in highly contaminated sumps via high PC data which is confirmed with UOA wear metals, they why are you in denial that UOA wear metals can ALSO confirm when filtration has become irrelevant; that when wear metals are so low regardless of the filter used, it's an indication that the filer is no longer the controlling entity and other things are more imporant?

Why insist that the bus study proves filters are important via the correlation of PCs and UOAs, but then deny that relationship is important whey the same methodology tells you that filters are now a tertiary player?

Your statement actually includes the very facts you seem to want to ignore. That other "variables" are in play. I've already listed them a few times in other posts. Those "variables" are what is making filter efficiency moot. (once above some threshold; I will always agree that filters are not meaningless, but only that they loose their effect after some level of "clean" is attained).

I'm not trying to pick a fight here; I respect your right to your position. I just struggle to understand the duality of your stance. UOAs are good when they show filter efficiency is important, but they are not trustworthy when they show filtration efficiency is no longer the wear controlling entity?

Can you refine your statement to help me make sense of it?


As I've mentioned before, I think the Bus Study was good in the sense that the bus engines were pretty dirty so that the difference in oil cleanliness and the associated wear levels were up out of the "noise level" so that decent filter performance measurements could be made. Maybe I should have clarified and said UOAs aren't really the right tool IMO when trying to compare wear metal levels as a function of oil filter efficiency on motors that aren't so "dirty". And also, from what I'm reading about UOA metal analysis, the method of the analyis has a factor on the accuracy of the UOA. Here's a couple of snips from a paper - this is the kind of variables that can skew UOAs.

"Metallic Elements
The metallic elements in the oil are measured by spectroscopic methods. This can be atomic
emission, atomic absorption or X-ray fluorescence spectroscopy.

The elements measured include additive elements, wear metals, combustion products and external
contaminants e.g. from cooling water or intake air. Therefore these values are influenced by the
additive package of the oil, component wear and fuel combustion.

Mostly Plasma Emission Spectrometry (ICP - Inductively Coupled Plasma) is used to determine the
metal contents of used lubricating oils. However particles bigger than 5 - 7 μm are not detected by this
method as they are not fully vaporised in the plasma due to mass effects and so wear element
concentration can be underestimated in cases of particularly high wear. Here non-routine analysis like
x-ray fluorescence spectroscopy, ashing of the sample before ICP measurement (to collect 100% of
the material within the sample), ferrography, ferrometry, PQ (Particle Quantifier) and similar methods
can help.

It must be realized that - according to the type of apparatus and way of sample preparation, the results
obtained can be very different. When comparing and plotting results for trend analysis, therefore, it is
important to ensure that the data are generated by the same laboratory, the same apparatus and the
same method.

The accuracy and lower detection limits of the methods have to be considered carefully before any
alarm or alert condition is applied to the analysis- especially low values (below 5 ppm) should be
interpreted with caution."

Also ...

"Wear Elements:
Al, Cr, Cu, Fe, Mo, Pb, Sb, Sn.
These may indicate wear of bearings, piston rings, cylinder liners and other engine components.
Component wear is directly influenced by the engine type i.e. the internal component technology, the
mode of engine operation (speed, partial or full load profile, number of daily starts etc.)."

So depending on the type of UOA methods used, if the analysis is only measuring metal wear particles less than 7μm, that means it's blind to what's going on in terms of wear particle levels above 7μm. And most regular spin-on oil filters don't catch a lot of particles below 7μm. Higher efficiency oil filters will catch more particles below 7μm, but that may not make a big difference in the measured wear particles below 7μm when comparing different efficiency oil filters as a function of wear particles. That's why I say I believe a regular UOA is too insensitive to correlate oil filter efficiency to engine wear.

Also, to use UOAs correctly to see the effect of different oils and filters, the same engine should be used under the same operating conditions, which is basically what the Bus Study tried to do in a real world setting. Taking a whole spectrum of UOAs on tons of different engines and operating conditions are including too many variables and skew the data on trying to home in on the influence of just one factor in the equation (ie, oil filter efficiency).

That's why I step down to a simpler logic that it's shown that higher efficiency filters result in cleaner oil, and cleaner oil will result in less oil contamination levels which ultimately results in less wear - regardless of how small the wear difference. That goes right back to your previous statement of: "PCs are a direct indication of filter performance, and that generally a tighter filter will reduce contamination of the sump.".

Hope that brings a bit more understanding of my viewpoint.

 

[dave1251]

Originally Posted by dnewton3
There is no more simple way to illustrate this than to state we have to acknowledge the difference between efficiency and effectiveness.
Something can be very efficient, but not effective. Other things can be effective, but not efficient. Some things can be both. Other things can be neither.
The most important thing to understand is that efficiency and effectiveness are NOT dependent upon each other; they can exist in a mutually exclusive environment.

It's nearly time for March Madness, so let's make an example of basket-ball ....
Your favorite player might be really efficient at shooting three pointers; say 50% from behind the line. But his effect on the game depends upon TWO things; the shots taken and his accuracy.
My favorite player might be really efficient at shooting from the stripe; say 85% at the foul line. But his effect on the game depends on those same two things; how much he gets fouled and his accuracy.
If you're player takes ten 3-pt shots, his efficiency would contribute 15 points to the game. (50% of 30 potential points = 15 pts scored). But if he only takes 2 attempts all game, it's probable he'll only add 3 pts.
If my player takes only 3 attempts at the foul line, he'll only contribute (90% of 3 = 2.7) pts per game, on average. However if my guy gets to the line 10 times, he's going to contribute 9 pts.
This is all obvious, right?


The point I make is that EFFICIENT filters do not have nearly as much EFFECT today, because other things have made filtration less relevant. (Filters are not meaningless, but capture ratio is not as important as it once was.)
The GM and Bus studies only show that heavily contaminated systems can see wear improvements when vast, wide disparity exists in the filtration choices, in extended (or outright deleted) OCIs.
But those conditions are not relevant today. Again, the following things make highly-efficient filtration meaningless:
- the TCB is now understood how it affects wear
- engines today do not self-pollute with soot the way they did decades ago
- oils are formulated to be much more robust; they handle contamination so much better now, and also resist oxidation to a much greater degree
- design of equipment has been refined
- manufacturing capabilities have been greatly improved
The point is that highly efficient filtration cannot make a significant impact when other things are controlling wear with greater effect.

Filtration efficiency was important many years ago because the sumps contained large amounts of contamination, and the oils were not able to deal with that resultant contamination as well, and the equipment was not made as well. But all that has changed. Sumps are just a lot cleaner now because engines just run so much cleaner in the combustion process. Equipment is designed and manufactured much better now, so it not only wears less, but is less susceptible to what little contamination is present. And lubes have more additives to deal with what remains, and reduce the ability of things like soot to grow in size.

Your point, Zee, that there will be less particulate in the system is salient, but it's not relevant. A highly efficient filter will catch more stuff, but there's so little for a FF filter to catch in a normal OCI that the EFFECT of the EFFICIENCY is indistinguishable in real world wear rates. Any particle that is fairly large will be caught by most any decent filter. Small stuff is not caught by any normal filter. The reality is this:
Your player whom is really good at 3-pt shots is not going to win the game because he's not getting a chance to take many shots at all; other players have reduced his ability to score, because they are denying him the ball!

I agree that finer filtration makes for less particulate in the system.
I disagree that highly-efficient filtration has a major effect today.
And there are tens of thousands of UOAs to prove my point.
What was relevant from yesteryear is moot today.

Once a sump is "clean enough", making it "cleaner" does not necessarily make it "better", if "better" is defined as less wear. Your definition of "better" is "cleaner'. My definition of "better" is "less wear". You focus on the inputs, whereas I focus on the outputs.

This thread actually started back in 2002. A LOT of things have changed since then.

As you state, Zee, we will agree to disagree.
I'm OK with that .... no harm, no foul

Dave thank you for posting your very commonsense post in a articulate manner which is easy to understand.