The Myth of "Flow Over Efficiency"

[ZeeOSix]

The phrase often used that "Lower efficiency oil filters favor flow over efficiency" has been used here for a long time. The misconception seems to be driven by the idea that lower efficiency oil filters must "flow better" because if they are lower efficiency then the media must not be as "tight" (full of holes), and therefore they must "flow better". That can be a factor, but in reality it depends on many factors in the overall design of the media. When talking about oil filter flow performance on an engine using a positive displacement (PD) oil pump, the term "flows better" means that the filter has a lower delta pressure (dP) vs flow curve.

The pressure drop (dP) across the whole filter assembly is basically comprised of the dP across 3 components: 1) base plate, 2) media and 3) center tube. The media is going to be the largest portion of the total dP. The bypass valve is designed and set to operate based on the dP across only the media+center tube. A typical oil filter is around 15 times less flow restrictive than the typical engine oiling system - ie, the engine oil pressure just down stream of the filter will be around 15 times higher than the dP across the filter. If there is 3 PSI of dP across the filter, then the oil pressure will typically be around 3 x 15 = 45 PSI (example with hot oil and PD not in relief).

Regardless of how open the base plate holes or center tube looks, or where the filter falls on the efficiency scale, in the end it's the dP across the whole filter assembly that really matters. Lots of people look at the base plate or the center tube and think the filter would flow so much better because they have more and/or bigger holes - typical example in this thread: LINK. But in reality the dP savings is pretty small and won't really make much difference in the total dP across the filter to actually matter. A dP difference between two filters of 5 PSI or less with hot oil when the flow rate is 10 GPM is not going to cause any "lack of lubrication" to an engine. A well designed oiling system is not designed on the ragged edge to blow-up an engine when a filter with 5 PSI more dP at 10 GPM is used. If it was designed to behave that way, then the designer failed.

I went through all the BR dP vs flow test results and summarized some of the data below for all the filters that were tested. Some of these filters are high efficiency (99% @ 20 microns), and some are pretty low efficiency (99% @ 40+ microns), yet the dP vs flow performance difference between them is negligible. Therefore, "flow over filtration" really doesn't exist.

Filter Key

If you throw out the two high dP points (Filter 5 & 6) then the total dP spread is only a hair over 2 PSI with hot oil at 5 GPM, which would reflect most normal street driving conditions (ie, not real high engine RPM). Even if Filters 5 & 6 are included, the dP spread is still only 3 PSI.

X-axis is the filter from the Filter Key table.
Y-axis is the dP in PSI.
Test conditions are in the graph heading.

If you throw out the two high dP points (Filter 5 & 6) then the total dP spread is only 3.5 PSI with hot oil at 9 GPM. A flow of 9 GPM in most engine oiling systems is going to reflect some pretty high engine RPM use.

 

[Glenda W.]

Thanks. Great post!

 

[TomYoung]

I find the argument a little hard to follow, but suspect that you are making a blanket statement that may not quite be the case... although I could have this wrong as my training is in life sciences, not engineering.

Here are my assumptions which lead me to question the "flow over filtration doesn't exist" idea.

1. Assuredly, some manufacturers make a better filter than others. A better filter allows the most flow, while capturing the most, and the smallest particles. Better materials, more filtration area, less blocked area all help on all regards.
2. I assume oil viscosity and quality has been accounted for. Oil that flows less willingly assuredly allows less filtration.
3. If you want to catch smaller particles, your filter medium must have smaller "holes," which will be harder for the oil to flow through, just like it is harder to run through a door that is open eight inches, compared to one that is wide open.
4. If you want more flow, you have to again, open the door more. If you don't want to increase the allowed particle size, you can only increase the surface area of the medium to increase flow. That can only happen so much within the confines of an automotive filter, but again, working on this using Boyle's law and derived equations.

Pretty sure you have a classic tradeoff here, and that at some point, filtration goes up when flow goes down.

 

[ZeeOSix]

I find the argument a little hard to follow, but suspect that you are making a blanket statement that may not quite be the case... although I could have this wrong as my training is in life sciences, not engineering.

This is dP vs flow data taken with oil at 15 cSt viscosity. This would equate to a xW-40 grade at 100C (212F) which is the typical operating temperature of hot oil. A xW-30 grade would be around 10 cSt, and a xW-20 grade would be around 7 cSt. So if the oil is hotter and/or the grade is lower then these dP vs flow values will be even lower. It's not really that hard to follow - flow is put through the filter at a specific flow rate and oil viscosity and the dP across the filter is measured. The dP vs flow comaprison between a wide range of oil filters that were tested under the same test parameters.

Here are my assumptions which lead me to question the "flow over filtration doesn't exist" idea.

1. Assuredly, some manufacturers make a better filter than others. A better filter allows the most flow, while capturing the most, and the smallest particles. Better materials, more filtration area, less blocked area all help on all regards.

Sure, and based on this dP vs flow data, the difference in dP is so small that it doesn't matter. Therefore, one would think that "flow over efficiency" should be ingnored and if they want the "best" filter beyond that then they should pick one that has the highest efficiency rating and holding capacity.

2. I assume oil viscosity and quality has been accounted for. Oil that flows less willingly assuredly allows less filtration.

Look at the heading of the graphs. The oil used was clean and at 15 cSt viscosity. The goal of the test was only to determing the dP vs flow curve of the tested filters. It shows that even high efficiency filters flow very well, so there isn't really anything to "flow over efficiency".

3. If you want to catch smaller particles, your filter medium must have smaller "holes," which will be harder for the oil to flow through, just like it is harder to run through a door that is open eight inches, compared to one that is wide open.

^^^ That is exactly the "flow over filtration" misconception I'm pointing out. This data proves that even very high efficiency oil filters have dP vs flow performance that is very good. There are ways to make high efficiency oil filters flow very well, and a good filter designer will use those design aspects to obain high efficiency, good holding capacity and good flow. The OG Ultra was a prime example of that. Also, in this data all the high efficiency filters flow very well. They don't "choke down flow" because of their high efficiency.

4. If you want more flow, you have to again, open the door more. If you don't want to increase the allowed particle size, you can only increase the surface area of the medium to increase flow. That can only happen so much within the confines of an automotive filter, but again, working on this using Boyle's law and derived equations.

See my reply to 3. above. All of the high efficiecy filters tested for dP vs flow have low flow resrtistance. The reason I posted this thread is because of the long running misconception that a high efficiency filter "must reduce flow" because it has to be much more restricive to be more efficient. This test data shows that "flow over efficiency" is a misconception and a myth.

Pretty sure you have a classic tradeoff here, and that at some point, filtration goes up when flow goes down.

This subject matter isn't really about that. It's simply about the dP vs flow, and the fact that no such thing as "flow over efficiency" actually exists because very high efficiency oil filters can flow just as good or even better than a lower efficiency oil filter. The people who design them ensure they flow well, and take measures to ensure that high efficiency filters also flow just as well as lower efficiency filters. How the efficiency of an oil filter may change with flow velocity and loading factors is a whole separate subject matter.

 

[TomYoung]

This is dP vs flow data taken with oil at 15 cSt viscosity. This would equate to a xW-40 grade at 100C (212F) which is the typical operating temperature of hot oil. A xW-30 grade would be around 10 cSt, and a xW-20 grade would be around 7 cSt. So if the oil is hotter and/or the grade is lower then these dP vs flow values will be even lower. It's not really that hard to follow - flow is put through the filter at a specific flow rate and oil viscosity and the dP across the filter is measured. The dP vs flow comaprison between a wide range of oil filters that were tested under the same test parameters.


Sure, and based on this dP vs flow data, the difference in dP is so small that it doesn't matter. Therefore, one would think that "flow over efficiency" should be ingnored and if they want the "best" filter beyond that then they should pick one that has the highest efficiency rating and holding capacity.


Look at the heading of the graphs. The oil used was clean and at 15 cSt viscosity. The goal of the test was only to determing the dP vs flow curve of the tested filters. It shows that even high efficiency filters flow very well, so there isn't really anything to "flow over efficiency".


^^^ That is exactly the "flow over filtration" misconception I'm pointing out. This data proves that even very high efficiency oil filters have dP vs flow performance that is very good. There are ways to make high efficiency oil filters flow very well, and a good filter designer will use those design aspects to obain high efficiency, good holding capacity and good flow. The OG Ultra was a prime example of that. Also, in this data all the high efficiency filters flow very well. They don't "choke down flow" because of their high efficiency.


See my reply to 3. above. All of the high efficiecy filters tested for dP vs flow have low flow resrtistance. The reason I posted this thread is because of the long running misconception that a high efficiency filter "must reduce flow" because it has to be much more restricive to be more efficient. This test data shows that "flow over efficiency" is a misconception and a myth.


This subject matter isn't really about that. It's simply about the dP vs flow, and the fact that no such thing as "flow over efficiency" actually exists because very high efficiency oil filters can flow just as good or even better than a lower efficiency oil filter. The people who design them ensure they flow well, and take measures to ensure that high efficiency filters also flow just as well as lower efficiency filters. How the efficiency of an oil filter may change with flow velocity and loading factors is a whole separate subject matter.

Thanks for the considered response. Would you not say, though, that materials and design make the “misconception” a little hard to see, but if you controlled for all other variables, basic science supports the misconception. The takeaway to someone like me is not to be all proud about my sub $3 filters!

 

[ZeeOSix]

Thanks for the considered response. Would you not say, though, that materials and design make the “misconception” a little hard to see, but if you controlled for all other variables, basic science supports the misconception. The takeaway to someone like me is not to be all proud about my sub $3 filters!

Like just about anything with respect to the performance of an oil filter, it can't really be determined by just looking at it or the materials it's made of. This is why appropriate testing is the only way to see what's really going on, and why this dP vs flow data shows that "flow over efficiency" really doesn't exist. It's a misconception and a myth.

 

[TomYoung]

Like just about anything with respect to the performance of an oil filter, it can't really be determined by just looking at it or the materials it's made of. This is why appropriate testing is the only way to see what's really going on, and why this dP vs flow data shows that "flow over efficiency" really doesn't exist. It's a misconception and a myth.

Kinda thinking I need to spring for Wix at least, maybe even Wix XP. Major paradigm shift for me.

 

[SC Maintenance]

Thank you for the detailed clarifications.

Two questions,

1) Given the same filter media, would more media in a filter result in lower pressure drop, all other things being equal, and if you ever got to a flow rate that mattered?

2) if you ever hit the rate where the media was restricting flow - would the bypass not just open? ie you should never starve an engine for oil assuming the bypass works - again assuming all other things equal and recognizing your now pumping unfiltered oil.

Thank you.

 

[ZeeOSix]

Thank you for the detailed clarifications.

Two questions,
1) Given the same filter media, would more media in a filter result in lower pressure drop, all other things being equal, and if you ever got to a flow rate that mattered?

Yes, if the filter had more total area of the same exact media then the dP vs flow would be reduced by some factor. But if the dP at 10 GPM of flow was already say 5 PSI, then it might only decrease by a PSI, maybe even less depending on the area increase difference. A filter with twice the media area is going to cut the dP down a pretty good percentage, but it's still only going to be a few PSI, maybe close to half down to 3 PSI.

Obviously testing would verify the actual decrease. I think the filters BR tested were all basically the same size, but they never tested a small vs over-sized filter of the same make and model to see the dP vs flow difference of just adding more area of the same exact media. That would be interesting, but don't think it's going to be anything earth shaking ... just a small decrease in dP under the same test conditions.

More surface area also means a bit more holding capacity and less loading with the same volume of debris captured, which could also help retain more efficiency as it loads up. As mentioned in various filter efficiency discussions, a filter that has a higher ISO 4548-12 efficiency is going to shed less debris down stream than a lower efficiency filter will. That's one factor that makes them higher efficiency. More media area would help decrease the loading per square inch, and lower the shedding factor by a small amount.

2) if you ever hit the rate where the media was restricting flow - would the bypass not just open? ie you should never starve an engine for oil assuming the bypass works - again assuming all other things equal and recognizing your now pumping unfiltered oil.

Thank you.

The bypass valve would start to open when the dP across the media+center tube hit the bypass valve setting. The bypass valve has two purposes: 1) to not starve the engine of oil flow if the dP across the media gets too high (main purpose), and 2) to also protect the media and center tube from damage if it can't take a lot of dP. To satisfy 2), the media and center tube should be designed to take much more dP than the bypass valve setting. And the bypass valve size should be designed adequate enough to flow a good amount of bypass flow so the dP doesn't creep up too high in a major bypass event - like if someone revved the engine real high with really thick cold oil - something that shouldn't be done, but it probably does happen by some people too wild on the throttle before the engine warms up.

 

[Hootbro]

So Brand Ranks data is now trustworthy?

 

[ZeeOSix]

So Brand Ranks data is now trustworthy?

I only have confidence in their dP vs flow test data - it's a pretty easy test to repeat with confidence. The efficiency ranking results not so much because when compared to ISO 4548-12 test data for the same filters they tested it doesn't correlate with BR's ranking results - ie, the Boss ranks almost at the top in BR's filtering test, but we all know the ISO test data says otherwise. That has been discussed many times.

Edit Add - The Wix XP, Royal Purple and Boss that Ascent tested ranked in the same dP vs flow order as the ones that BR tested. There might have been a slight viscosity difference between what Ascent had and what BR had, and filter size size too which would affect the actual dP level seen (I looked at it and it's pretty slight regardless), but the dP vs flow results ranked the same. Ascent tested an OG Ultra whereas BR tested a non-synthetic, non-wire backed Ultra, so can't compare those due to change in the media.

Ascent's dP vs flow test showed that there was only a 3.0 PSI spread between all 5 filters at 9 GPM, and two of the 5 filters were not that efficient (Boss and Wix XP). So again, this shows that "flow over efficiency" is a myth because pretty much all decent brand filters regardless of their efficiency all flow close enough to each other that the dP difference really doesn't matter.

 

[Urshurak776]

Thank you @ZeeOSix !

Admins, PLEASE make this a sticky. The misconceptions that Z is addressing is rampant on this board.

 

[Gravity]

Honda and Toyota designed filters might disagree.

 

[Glenda W.]

Honda and Toyota designed filters might disagree.

You can one up Toyota.

 

[ZeeOSix]

Honda and Toyota designed filters might disagree.

How do you know what the dP vs flow of the Honda filters are? They are just basically re-badged Frams (EG/TG), which in this data have the highest dP vs flow.

I'd also like to see some dP vs flow data on OEM Subaru oil filters since they are specified on Subarus with crazy high flow oil pumps. They may be way more flow restrictive than most believe.

 

[CapitalTruck]

How do you know what the dP vs flow of the Honda filters are? They are just basically re-badged Frams (EG/TG), which in this data have the highest dP vs flow.

I'd also like to see some dP vs flow data on OEM Subaru oil filters since they are specified on Subarus with crazy high flow oil pumps. They may be way more flow restrictive than most believe.

As an aside, the Subaru filters have gone through a couple of iterations in the past few years. During COVID it seems like they were on a Champ labs filter (looked exactly like my ACDelco filter). Now they appear to be back to a Roki sourced from Indonesia. I too would be interested in the specs of the Subaru filters.

 

[Fair enough]

The Roki don’t look restrictive, just the opposite. Skip to microscope pics of media.

 

[OVERKILL]

I only have confidence in their dP vs flow test data - it's a pretty easy test to repeat with confidence. The efficiency ranking results not so much because when compared to ISO 4548-12 test data for the same filters they tested it doesn't correlate with BR's ranking results - ie, the Boss ranks almost at the top in BR's filtering test, but we all know the ISO test data says otherwise. That has been discussed many times.

Edit Add - The Wix XP, Royal Purple and Boss that Ascent tested ranked in the same dP vs flow order as the ones that BR tested. There might have been a slight viscosity difference between what Ascent had and what BR had, and filter size size too which would affect the actual dP level seen (I looked at it and it's pretty slight regardless), but the dP vs flow results ranked the same. Ascent tested an OG Ultra whereas BR tested a non-synthetic, non-wire backed Ultra, so can't compare those due to change in the media.

Ascent's dP vs flow test showed that there was only a 3.0 PSI spread between all 5 filters at 9 GPM, and two of the 5 filters were not that efficient (Boss and Wix XP). So again, this shows that "flow over efficiency" is a myth because pretty much all decent brand filters regardless of their efficiency all flow close enough to each other that the dP difference really doesn't matter.

Would love to see the OG Ultra in there, as you note, the new version is a different filter entirely.

 

[KevinMalone]

How would leaky bypass valves affect these measurements

I never bought into flow > efficiency. I always assumed it was about costs (ie cheaper, lower efficiency media is sufficient to meet the VM's needs). It is also interesting to see some VMs spend more on air filtration with PET air filters.

 

[ARCOgraphite]

10deg C moderate climate "cold" cold start viscosity will be approx. 200 to 500cSt depending on grade and VI. Far from 15Cst.
I have had a few quality filters tear pleats open in Winter service. Is that the result of frosted or iced media? Added factors of gum formation and partial media blockage over time? Or excessive ΔP given the material selection and mechanical layout of the filter?