Oil Filter Comparison Study

[JerryBob]

Originally Posted By: ZeeOSix
Originally Posted By: JerryBob
Originally Posted By: dnewton3

I have yet to find a truly well designed and executed real study that credibly links these concepts. And as an SAE member, I've looked through a LOT of data over the years. If you know of one, then by all means, bring it forth for scrutiny by all please!


Why is there such a dearth of data? The cost of undertaking a study? Doesn't seem that complicated - not like a vaccine trial or something like that. They're just freakin' oil filters

There are plenty of write-ups on the subject.

http://lmgtfy.com/?q=engine+wear+vs+particle+size




That's obvious. But do you have a link of which filters perform the best? That was my question if you took the time to read it.

 

[Sayjac]

Best filtration comparison study I've seen here was done by Bitog member river_rat. And unlike the topic'd GMC comparison, his results actually aligned with known ISO test results. And some of his results preceeded the Amsoil ISO of some oem filters showing no endcap filters like the Toyota oem to be relatively inefficient filters. A fact later proven by the Amsoil oem ISO filter test.

Granted it's not an ISO test, but unlike the topic'd comparison, his results have never been shown to be off base.

Some Filtration Comparisons from the Bench.

Oil Filtration Comparison Summary Pages. river_rat

 

[Jim Allen]

Many of you guys are ignoring the gorilla in the room here. Rating a filter's efficiency is easy. The harder question to answer is what good a higher efficiency filter is doing you for the extra money spent.

All the testing that's been done (to my knowledge and I've spent nearly a thousand dollars over the past few years buying SAE white papers and test docs), often cited and used as fodder to sell high efficiency filters, were "accelerated" tests done on an engine dyno with artificially contaminated oil.

The engine is measured and then AC test dust is injected into the oil to create a high contamination level in the oil. In one of the tests, very often cited, they added 50 grams of fine test dust in an oil slurry EVERY HOUR FOR EIGHT HOURS to the diesel on the dyno and somewhat less to the gasser tested. That's 400 grams of material for the diesel. Filters of various efficeincy are used to bring the particle size mix down according to the efficiency of the filter and wear is measured at each state.

What does that tell us? Well, lets say a 30 um absolute filter reduced that 400 gram load to 150 grams of material < than 30 um. Let's say that the wear is significantly reduced when you install 20 um absolute filter (a really good filter) and reduce that load to 100 grams of material < 20 um. The only way that's useful is if you can compare the wear to a "normal" operational state of the oil, which I'm sure has only a tiny fraction of even 100 grams of material in it.

IT'S AS MUCH THE CONCENTRATION OF CONTAMINATION PARTICLES AS IT IS THEIR SIZE. So contamination levels count for as much as absolute filtration levels the way I see it. It's true that wear begets wear particles but abrasive contamination from normal engine wear particles (with clean oil) takes a very long time to build up. I'm not even sure they COULD build up that high in the real world. When you think a GOOD, large capacity filter of today can hold 30-50 grams of material and that most oil filters aren't routinely loaded past 30-50% of capacity, even in 10-15K miles of operation, we can see how unrealistic those tests are... except perhaps to the HD commercial markets with neglected/abused equipment that run in very harsh environments.

The levels of contamination rise very slowly in the real world (the street environment with good air filtration and a sealed crankcase), so there are long periods of relatively clean oil, even without extraordinary filtration. A typical gas engine with a typical $4 filter generates 2-3 PPM of iron per thousand miles. At the end of a 10K run, there 20 PARTS PER MILLION of iron in the oil. No big whoop there. Exactly how does a $15 filter help you in that case?

Look at UOAs, especially those with particle counts included. Until the contamination levels get really high, the rates of wear (PPM/1000 miles is a common metric) are very low and rise very little until the contamination level rises fairly high. Does anyone think that there is going to be a huge difference in the rate of wear between a "standard" filter that is rated 95% @ 25 um version one rated 99% @ 20 um? Not likely IMO because based on everything I have seen and know, the "wear curve" tapers off and flattens out when filtration is within the range of what has become "normal." IN TERMS OF WEAR IN A NORMAL ENGINE, the difference between an ordinary $4 filter and a $15 filter are not going to be very apparent. They do exist but the curve on the return of invest goes steeply in the opposite direction of the rates of wear.

In the bad old days when "normal" filtration was 90% @ 40 um, a sizeable reduction in wear could be achieved by going to a better filter. But today's "normal" is yesterday's extraordinary and the contamination rates of modern engines are very much lower. Better air filtration. Better engine machining. Better combustion processes and ring sealing. Roller cams and rocker gear. External belts replacing internal chains. Cleaner fuel. Better oil. All those factors contribute to the low rates and levels of contamination in a modern engine. We simply don't "need" as much filtration as we used to. In most cases.

The agonizing part of this is that all the wear tests done appear to be accelerated tests in some way. There just aren't any available that answer the basic question we have here, i.e.: What are the differences in wear between an an "ordinary" filter in the common efficiency range (94-97% @ 20-25 um) and the premium high efficiency filters (99% @ 15-25 um) in a lightly used vehicle and do those differences return the extra investment? And then there is high efficiency bypass to add into the mix, which is another question.

The bottom line to me is that the only way to make high efficiency fitlration pay is via the extension in the OCI they allow... cleaner oil lasting longer in service. there are a few performance benefits too, such as better cold flow and fewer bypass events with a full syn media premium filter... but these benefits may or may not apply.

Just to be upfront, my basic philosophy is: "No engine ever died from having oil that was too clean!" I'm willing to spend a little extra money in that regard even if the data does not fully support the benefits. The cost effective way to approach this is to buy the highest efficiency filter in the price range you choose for the OCI you use. If your OCI is 7.5K, a $10 Ultra (99%@20um) is not going to be demonstrably better in terms of wear than a $4 Puro Classic (97% @ 20 um). Going to a $15 filter is even more absurd in terms of financial payback.

 

[440Magnum]

Originally Posted By: dnewton3

I have yet to find a truly well designed and executed real study that credibly links these concepts. And as an SAE member, I've looked through a LOT of data over the years. If you know of one, then by all means, bring it forth for scrutiny by all please!


I think the problem is that even comparing the best to worst designed oil filters on the market, the differences they might (or might not...) cause in engine wear over the life of a vehicle- even in a big statistically valid sample of vehicles- is tiny compared to other things.

Its such a tiny effect that it just doesn't matter most of the time. But then, I think about half of us BITOG types have a pretty strong "my car's my baby, and it gets the BEST" approach, whether it makes sense relative to a statistically valid study or not. We all think we'll beat the stats. Even I as an engineer with a good background in probability theory look at those filters and think, "yeah, my SRT-8 is getting the synthetic with the thickest shell, even if its $3 a filter more than a really great synthetic with s thinner shell."

I admit my non-scientific mindset in this case. I'll apply science to your car, but mine gets special treatment.

That said, even though the ranking of the filters in that study is probably somewhat meaningless, I think its a pretty good look at how they're made. From that info we can choose based on which characteristics worry each of us the most. So its not without merit.

 

[ZeeOSix]

Originally Posted By: JerryBob
Originally Posted By: ZeeOSix
Originally Posted By: JerryBob
Originally Posted By: dnewton3

I have yet to find a truly well designed and executed real study that credibly links these concepts. And as an SAE member, I've looked through a LOT of data over the years. If you know of one, then by all means, bring it forth for scrutiny by all please!


Why is there such a dearth of data? The cost of undertaking a study? Doesn't seem that complicated - not like a vaccine trial or something like that. They're just freakin' oil filters

There are plenty of write-ups on the subject.

http://lmgtfy.com/?q=engine+wear+vs+particle+size


That's obvious. But do you have a link of which filters perform the best? That was my question if you took the time to read it.


Some of those links test different levels of filtration efficiency to correlate effectiveness of the filter. Sorry bud, but the only thing you can go on otherwise is the advertised beta/efficiency ISO test numbers from the filter manufacturers.

 

[Izb]

Originally Posted By: Jim Allen
Many of you guys are ignoring the gorilla in the room here. Rating a filter's efficiency is easy. The harder question to answer is what good a higher efficiency filter is doing you for the extra money spent.

All the testing that's been done (to my knowledge and I've spent nearly a thousand dollars over the past few years buying SAE white papers and test docs), often cited and used as fodder to sell high efficiency filters, were "accelerated" tests done on an engine dyno with artificially contaminated oil.

The engine is measured and then AC test dust is injected into the oil to create a high contamination level in the oil. In one of the tests, very often cited, they added 50 grams of fine test dust in an oil slurry EVERY HOUR FOR EIGHT HOURS to the diesel on the dyno and somewhat less to the gasser tested. That's 400 grams of material for the diesel. Filters of various efficeincy are used to bring the particle size mix down according to the efficiency of the filter and wear is measured at each state.

What does that tell us? Well, lets say a 30 um absolute filter reduced that 400 gram load to 150 grams of material < than 30 um. Let's say that the wear is significantly reduced when you install 20 um absolute filter (a really good filter) and reduce that load to 100 grams of material < 20 um. The only way that's useful is if you can compare the wear to a "normal" operational state of the oil, which I'm sure has only a tiny fraction of even 100 grams of material in it.

IT'S AS MUCH THE CONCENTRATION OF CONTAMINATION PARTICLES AS IT IS THEIR SIZE. So contamination levels count for as much as absolute filtration levels the way I see it. It's true that wear begets wear particles but abrasive contamination from normal engine wear particles (with clean oil) takes a very long time to build up. I'm not even sure they COULD build up that high in the real world. When you think a GOOD, large capacity filter of today can hold 30-50 grams of material and that most oil filters aren't routinely loaded past 30-50% of capacity, even in 10-15K miles of operation, we can see how unrealistic those tests are... except perhaps to the HD commercial markets with neglected/abused equipment that run in very harsh environments.

The levels of contamination rise very slowly in the real world (the street environment with good air filtration and a sealed crankcase), so there are long periods of relatively clean oil, even without extraordinary filtration. A typical gas engine with a typical $4 filter generates 2-3 PPM of iron per thousand miles. At the end of a 10K run, there 20 PARTS PER MILLION of iron in the oil. No big whoop there. Exactly how does a $15 filter help you in that case?

Look at UOAs, especially those with particle counts included. Until the contamination levels get really high, the rates of wear (PPM/1000 miles is a common metric) are very low and rise very little until the contamination level rises fairly high. Does anyone think that there is going to be a huge difference in the rate of wear between a "standard" filter that is rated 95% @ 25 um version one rated 99% @ 20 um? Not likely IMO because based on everything I have seen and know, the "wear curve" tapers off and flattens out when filtration is within the range of what has become "normal." IN TERMS OF WEAR IN A NORMAL ENGINE, the difference between an ordinary $4 filter and a $15 filter are not going to be very apparent. They do exist but the curve on the return of invest goes steeply in the opposite direction of the rates of wear.

In the bad old days when "normal" filtration was 90% @ 40 um, a sizeable reduction in wear could be achieved by going to a better filter. But today's "normal" is yesterday's extraordinary and the contamination rates of modern engines are very much lower. Better air filtration. Better engine machining. Better combustion processes and ring sealing. Roller cams and rocker gear. External belts replacing internal chains. Cleaner fuel. Better oil. All those factors contribute to the low rates and levels of contamination in a modern engine. We simply don't "need" as much filtration as we used to. In most cases.

The agonizing part of this is that all the wear tests done appear to be accelerated tests in some way. There just aren't any available that answer the basic question we have here, i.e.: What are the differences in wear between an an "ordinary" filter in the common efficiency range (94-97% @ 20-25 um) and the premium high efficiency filters (99% @ 15-25 um) in a lightly used vehicle and do those differences return the extra investment? And then there is high efficiency bypass to add into the mix, which is another question.

The bottom line to me is that the only way to make high efficiency fitlration pay is via the extension in the OCI they allow... cleaner oil lasting longer in service. there are a few performance benefits too, such as better cold flow and fewer bypass events with a full syn media premium filter... but these benefits may or may not apply.

Just to be upfront, my basic philosophy is: "No engine ever died from having oil that was too clean!" I'm willing to spend a little extra money in that regard even if the data does not fully support the benefits. The cost effective way to approach this is to buy the highest efficiency filter in the price range you choose for the OCI you use. If your OCI is 7.5K, a $10 Ultra (99%@20um) is not going to be demonstrably better in terms of wear than a $4 Puro Classic (97% @ 20 um). Going to a $15 filter is even more absurd in terms of financial payback.


Ok
May be wear on Purolator classic (96% on 20 mc) is almost equal wear level on Purolator One ((99.9% jn 20 mc).
But what about air filters (Classic vs Pure One)?

 

[ZeeOSix]

Originally Posted By: Izb

But what about air filters (Classic vs Pure One)?


Go to Purolator's website:

PureOne Air Filter (99.5%)
http://www.purolatorautofilters.net/products/air_filters/Pages/pureoneairfilters.aspx

Classic Air Filter (96.5%)
http://www.purolatorautofilters.net/products/air_filters/Pages/purolatorairfilters.aspx

 

[Sayjac]

^^^Ironically the guy that did this oil filter patch test also did a corresponding air filter study. Iirc like the oil filter study, the Puro Classic AF also tested better than the P1 AF. If nothing else the guy is consistent.

Like the oil filter study, I'd believe controlled ISO tests results before I'd accept results from uncontrolled testing like OP's linked study.

 

[ZeeOSix]

Originally Posted By: sayjac
^^^Ironically the guy that did this oil filter patch test also did a corresponding air filter study. Iirc like the oil filter study, the Puro Classic AF also tested better than the P1 AF. If nothing else the guy is consistent.

Like the oil filter study, I'd believe controlled ISO tests results before I'd accept results from uncontrolled testing like OP's linked study.


http://www.bobistheoilguy.com/forums/ubb...677#Post3153677

 

[seekingbuddha]

The following post by Jim Allen is the reason i joined BTOG. This one explanation will save me hundreds of dollars in oil filter costs, on my 3 cars in decades ahead. Why waste money when a oil filter on an ordinary car does not even fill to 75% capacity ? I change the oil frequently, and that is enough money wasted. According to the data i have gathered, at the OCI recommended by car manufacturers & oil manufacturers, we are not even using the full capacity/capability of the oil. My UOA has proved that under average/normal operating conditions (CA city), my oil has the capacity to go atleast 30% more milage than what is quoted as the limit by the oil manufacturer.

So.....a big thanks to Jim

Originally Posted By: Jim Allen
Many of you guys are ignoring the gorilla in the room here. Rating a filter's efficiency is easy. The harder question to answer is what good a higher efficiency filter is doing you for the extra money spent.

All the testing that's been done (to my knowledge and I've spent nearly a thousand dollars over the past few years buying SAE white papers and test docs), often cited and used as fodder to sell high efficiency filters, were "accelerated" tests done on an engine dyno with artificially contaminated oil.

The engine is measured and then AC test dust is injected into the oil to create a high contamination level in the oil. In one of the tests, very often cited, they added 50 grams of fine test dust in an oil slurry EVERY HOUR FOR EIGHT HOURS to the diesel on the dyno and somewhat less to the gasser tested. That's 400 grams of material for the diesel. Filters of various efficeincy are used to bring the particle size mix down according to the efficiency of the filter and wear is measured at each state.

What does that tell us? Well, lets say a 30 um absolute filter reduced that 400 gram load to 150 grams of material < than 30 um. Let's say that the wear is significantly reduced when you install 20 um absolute filter (a really good filter) and reduce that load to 100 grams of material < 20 um. The only way that's useful is if you can compare the wear to a "normal" operational state of the oil, which I'm sure has only a tiny fraction of even 100 grams of material in it.

IT'S AS MUCH THE CONCENTRATION OF CONTAMINATION PARTICLES AS IT IS THEIR SIZE. So contamination levels count for as much as absolute filtration levels the way I see it. It's true that wear begets wear particles but abrasive contamination from normal engine wear particles (with clean oil) takes a very long time to build up. I'm not even sure they COULD build up that high in the real world. When you think a GOOD, large capacity filter of today can hold 30-50 grams of material and that most oil filters aren't routinely loaded past 30-50% of capacity, even in 10-15K miles of operation, we can see how unrealistic those tests are... except perhaps to the HD commercial markets with neglected/abused equipment that run in very harsh environments.

The levels of contamination rise very slowly in the real world (the street environment with good air filtration and a sealed crankcase), so there are long periods of relatively clean oil, even without extraordinary filtration. A typical gas engine with a typical $4 filter generates 2-3 PPM of iron per thousand miles. At the end of a 10K run, there 20 PARTS PER MILLION of iron in the oil. No big whoop there. Exactly how does a $15 filter help you in that case?

Look at UOAs, especially those with particle counts included. Until the contamination levels get really high, the rates of wear (PPM/1000 miles is a common metric) are very low and rise very little until the contamination level rises fairly high. Does anyone think that there is going to be a huge difference in the rate of wear between a "standard" filter that is rated 95% @ 25 um version one rated 99% @ 20 um? Not likely IMO because based on everything I have seen and know, the "wear curve" tapers off and flattens out when filtration is within the range of what has become "normal." IN TERMS OF WEAR IN A NORMAL ENGINE, the difference between an ordinary $4 filter and a $15 filter are not going to be very apparent. They do exist but the curve on the return of invest goes steeply in the opposite direction of the rates of wear.

In the bad old days when "normal" filtration was 90% @ 40 um, a sizeable reduction in wear could be achieved by going to a better filter. But today's "normal" is yesterday's extraordinary and the contamination rates of modern engines are very much lower. Better air filtration. Better engine machining. Better combustion processes and ring sealing. Roller cams and rocker gear. External belts replacing internal chains. Cleaner fuel. Better oil. All those factors contribute to the low rates and levels of contamination in a modern engine. We simply don't "need" as much filtration as we used to. In most cases.

The agonizing part of this is that all the wear tests done appear to be accelerated tests in some way. There just aren't any available that answer the basic question we have here, i.e.: What are the differences in wear between an an "ordinary" filter in the common efficiency range (94-97% @ 20-25 um) and the premium high efficiency filters (99% @ 15-25 um) in a lightly used vehicle and do those differences return the extra investment? And then there is high efficiency bypass to add into the mix, which is another question.

The bottom line to me is that the only way to make high efficiency fitlration pay is via the extension in the OCI they allow... cleaner oil lasting longer in service. there are a few performance benefits too, such as better cold flow and fewer bypass events with a full syn media premium filter... but these benefits may or may not apply.

Just to be upfront, my basic philosophy is: "No engine ever died from having oil that was too clean!" I'm willing to spend a little extra money in that regard even if the data does not fully support the benefits. The cost effective way to approach this is to buy the highest efficiency filter in the price range you choose for the OCI you use. If your OCI is 7.5K, a $10 Ultra (99%@20um) is not going to be demonstrably better in terms of wear than a $4 Puro Classic (97% @ 20 um). Going to a $15 filter is even more absurd in terms of financial payback.