Don't know if this has been covered here, at least I haven't seen it. The notion that an air or oil filter's effectiveness actual IMPROVES with use. Most would assume a filter works best, and captures particles best, when new. Because new is sexy and kewl. But I've always suspected filters are most effective after a period of use. Pores in the media, both large and small, will be blocked by trapped particulates over time, leaving only progressively smaller pores to pass air/oil. Therefore enhanced filtration. And eventually you'll end up with a thin layer of dirt which, in itself, is an inadvertent media layer, with dirt filtering out other dirt before it even reaches the cellulose/synthetic media below. Therefore a well-used filter might be more effective than new. Of course what we're also talking about here is eventual clogging, so yeah, there's that buzz kill since coins have two sides. I've always thought there's logic to this whole notion, but never really knew if it's just another India Pale Ale fueled delusion. I tend to like those.
Filters Improving With Use
- Thread starter TC
- Start date
Has been covered before. Generally your correct, but at some point they start restricting flow. So there is no free lunch.
Air filters are easier - can be inspected regularly. Oil filters are harder. You never know how full or if there has actually been a pleat failure till cut open.
Air filters are easier - can be inspected regularly. Oil filters are harder. You never know how full or if there has actually been a pleat failure till cut open.
dnewton3
Staff member
Jim Allen (member here) is an automotive journalist who visited Fram and Parker filter facilities. He did an article many years ago which had credible info indicating that 90% of all the particulate an air filter will pass goes through in the first 10% of lifecycle of the filter. IOW - it's beneficial to leave the filter on as long as possible, as long as it's not truly restricting airflow.
The concept is often misunderstood by most. They assume a "new" filter is 100% volumetric flow rated, and as it loads up it restricts air as it ages. But the reality is that a new filter is probably capable of 150% or more of the total engine air volume needed at max rpm, so that as it ages, it still has plenty of ability to flow air as it loads. And if you never come close to hitting the redline, it can last a LONG time.
Yes - any filter can be over used. We've seen the unsightly examples on social media, etc. But that's the exception. Most of the time perfectly good air filters are trashed because of the "more it always better" mentality many consumers have.
A better way to judge the air filter performance is to get an air filter gauge (represents the vacuum relative to when new, as the filter loads up). Wix has a decent one that is about $35. I have a few installed on my vehicles. Gone 50k miles on an air filter and was still serviceable.
The concept is often misunderstood by most. They assume a "new" filter is 100% volumetric flow rated, and as it loads up it restricts air as it ages. But the reality is that a new filter is probably capable of 150% or more of the total engine air volume needed at max rpm, so that as it ages, it still has plenty of ability to flow air as it loads. And if you never come close to hitting the redline, it can last a LONG time.
Yes - any filter can be over used. We've seen the unsightly examples on social media, etc. But that's the exception. Most of the time perfectly good air filters are trashed because of the "more it always better" mentality many consumers have.
A better way to judge the air filter performance is to get an air filter gauge (represents the vacuum relative to when new, as the filter loads up). Wix has a decent one that is about $35. I have a few installed on my vehicles. Gone 50k miles on an air filter and was still serviceable.
Thanks for the replies. So it wasn't the India Pale Ale after all. That 90/10 ratio is VERY enlightening. I'm gonna assume that applies to oil filter media as well, although that's a big assumption. Ironically I worked at a Donaldson Aerospace pleated filter plant for 2 years, although not in Engineering.
To take the concept a little further would impact micron ratings. The Purolator oil filter I'm currently using has a part number specific efficiency of 99% @ 38 microns. So it's a "38 micron" filter. Except it isn't. It's only 38 microns when new, but could be a 20 or 30 micron filter through much of its life (random numbers). So that puts a whole new spin on the factory micron ratings (the ratings improving with use), something I don't see discussed at all in automotive forums. Probably moot since you need a common industry baseline with micron ratings, so that means when-new. I just wanna get closer to that "XX micron particles cause most engine wear" from those landmark General Motors studies, which I don't have handy at the moment, and it's cool to know that simply using a filter -- exposing it to the elements -- helps accomplish that.
To take the concept a little further would impact micron ratings. The Purolator oil filter I'm currently using has a part number specific efficiency of 99% @ 38 microns. So it's a "38 micron" filter. Except it isn't. It's only 38 microns when new, but could be a 20 or 30 micron filter through much of its life (random numbers). So that puts a whole new spin on the factory micron ratings (the ratings improving with use), something I don't see discussed at all in automotive forums. Probably moot since you need a common industry baseline with micron ratings, so that means when-new. I just wanna get closer to that "XX micron particles cause most engine wear" from those landmark General Motors studies, which I don't have handy at the moment, and it's cool to know that simply using a filter -- exposing it to the elements -- helps accomplish that.
GM study: "Compared to a 40 micron filter, engine wear was reduced 50% with 30 micron filtration. Likewise, wear was reduced by 70% with 15 micron filtration.”
The engineered efficiency of the filter does not really mean that your filter improves as it clogs up, and certainly does not indicate that as a filter gets clogged, it does so in a linear or predictable fashion.
If this is a thought experiment, okay. If there is some thought to saving money on filters, or even some perverse idea that older filters are going to make a car last longer, I would suggest that you can’t fight city hall and that everyone will be happier with fresh air and oil filters, of good quality, changed at or better than mfr recommendations.
Even when you cut open an oil filters, how do you know it is “full?”
Air filter efficiency improves with debris loading. Oil filter efficiency worsens with debris loading. The dP across an oil filter in use is magnitudes higher than the dP across an air filter. Some oil filters slough off much less debris than others as the dP increases.
If an oil filter has a high ISO 4548-12 efficiency, it's a low debris slougher. We saw this in Ascent's ISO efficiency testing. Even the high efficiency OG Ultra efficiency got worse with loading, but it was miniscule compared to the others tested. It's been a misconception on this board for decades that " oil filters get more efficient with use". The fact is that their efficiency actually gets worse. It's been shown with official ISO efficency testing, and has been discussed many times in this forum.
If an oil filter has a high ISO 4548-12 efficiency, it's a low debris slougher. We saw this in Ascent's ISO efficiency testing. Even the high efficiency OG Ultra efficiency got worse with loading, but it was miniscule compared to the others tested. It's been a misconception on this board for decades that " oil filters get more efficient with use". The fact is that their efficiency actually gets worse. It's been shown with official ISO efficency testing, and has been discussed many times in this forum.
Assuming that 90/10 ratio from Fram or Parker Filtration is accurate, many air filters are relative sieves when new, but much more efficient by 10% into their lifespan. In other words...they DO improve with age, not a fantasy. As for the end game, simply replacing per the owner's manual is simple and good enough for most people outside of severe service, guidelines determined through both prototype testing and real-world experience.
I am of the opinion that once a filter had gotten to the point where the efficiency/ micron size reduction has meaningfully improved the benefit is outweighed by the detriment of the increased pressure differential.
How the ISO 4548-12 test is done, and how the ISO efficiency is calculated has been talked about for at least the last 10-12 years in this forum. The ISO 4548-12 is the average overall efficiency from new to what's considered a fully loaded oil filter. The reason that Purolator's efficiency is 99% @ 38u is because it was losing efficiency as it loaded up over the efficiency test. In the Ascent ISO efficiency testing, the Purolator Boss efficiency dropped like a rock as it loaded up with debris.
Last edited:
We have two discussions going on here. One wrt air filters, and one wrt oil filters. As said, they do not behave the same as they load up. So without clarification about which one people are talking about, the discussion is going to become convoluted.
"Air filter efficiency improves with debris loading. Oil filter efficiency worsens with debris loading."
So as a general rule in layman's terms, it would seem that air/gas through a media versus petro oils/fluids through a similar media result in VERY different results as the media loads up with debris, correct? That's intriguing in that it very much goes against intuition, but the data is what it is.
So as a general rule in layman's terms, it would seem that air/gas through a media versus petro oils/fluids through a similar media result in VERY different results as the media loads up with debris, correct? That's intriguing in that it very much goes against intuition, but the data is what it is.
dnewton3
Staff member
If you're referring to the GM filter study from 1988 as listed in the SAE registry (881825), I'd caution you to interpret that as anything except an anecdotal exercise of how incredulous HALTs (highly accelerated life test) can be. Do NOT place any faith in what you hear or read of that test without first purchasing the actual study, then read it fully, understanding the implications of how it was performed and the conclusions it came to. Simply put, the results of that test will never be seen duplicated in real life because the conditions of the test were so absurdly non-normal that it was just moronic.
As mentioned earlier, the dP (pressure difference across the media) on oil filters in use is magnitudes higher (many PSI) than the dP across an air filter in use (mere inches of water). If the dP across an oil filter was mere inches of water like it is on an air filter, then oil filters would probably get more efficient as they load up too. And if the dP was 10 PSI across an air filter, it would probably slough off already captured particles and become less efficient too - if it didn't collapse first. It's all about the dP levels across the filter media. Part of the efficiency functionality of a filter is to retain the debris that's already been captured. If it can't do that well, and sloughs off already captured debris in the media, then the downstream efficiency will become worse as the dP across the media increases.
If you're interested in seeing how an oil filter's efficiency decreases as it loads up, go to this post in the Ascent ISO efficiency testing thread. Also read posts in that thread before and after this post, or read the whole gigantic thread to see what oil filter ISO efficiency testing is all about.
In Ascent's testing, here's how the Purolator Boss efficiency at 20u dropped as it loaded up. The x-axis is the percent of filter loading. So when it was loaded from 16% to 64%, the efficiency went from 83.2% efficient to 52.5% efficient at 20u. Certainly not getting more efficient as it loaded up.
In Ascent's testing, here's how the Purolator Boss efficiency at 20u dropped as it loaded up. The x-axis is the percent of filter loading. So when it was loaded from 16% to 64%, the efficiency went from 83.2% efficient to 52.5% efficient at 20u. Certainly not getting more efficient as it loaded up.
The efficiency vs loading % graph in post 16 shows the efficiency drop is pretty linear with loading. As well as the graph in post 393 in the Ascent ISO testing thread linked in post 16.
Last edited:
- Joined
- Aug 15, 2024
- Messages
- 472
Who woulda thunk, over maintaining has consequences too
This chart has no time component. It shows only that efficiency drops as loading increases, which is not really saying anything.
The percent loading was based on the debris loading rate, which was a constant rate during the ISO efficiency test. So it clearly shows that the filter lost a lot of efficiency with loading. Go read that long Ascent ISO efficiency testing thread linked in post 16, and you'll learn a lot about it all.
