I've got an older car with ~170k miles on it. I typically run a 3-5k miles or 6 months oil change interval, which pretty much always ends up at the 6-month limiter. I do this interval so I can run more viscous oils in the summer and less viscous oils in the winter.

I send a sample of my oil for analysis at every change. I've found that the biggest limiter to my oil life is not the additives being consumed or the oil oxidizing/changing viscosity, but rather particulates from the wear on an older engine.

I might not extend my oil change interval because the savings will be rather minimal, but I would like to know if anyone has experience with changing only the oil filter, not the oil itself, and how it has affected performance, wear, and service intervals.

EDIT: While I am seeking subjective answers regarding perceptions from replacing the oil filter without changing the oil itself, I am adding some structure to help with providing quantitative data as well.

Regarding performance, if anyone has normalized dyno tests comparing wheel horsepower over time between oil changes, and over time between oil changes with a filter replacement, that would be appreciated.

Wear can be discussed with either oil analysis results (ex. metal ppm) or microscopic analysis of wearing surfaces, such as cylinder walls, piston rings, bearings, valve guides, etc.

Lastly, service interval is pretty intuitive, but data backing extended service intervals would help. For example, if the oil change interval was 5000 miles, but with a filter replacement, analysis shows similar results with a replacement at 4000 miles and an oil change at 8000 miles, then the service interval was extended by 3000 miles (8000 vs 5000).

EDIT: As requested, I am posting a link to my most recent oil analysis report from Blackstone Labs. I am unaffiliated and I am not providing an endorsement for their services, simply adding information requested in a comment.

  • Of the three metrics mentioned here (performance, wear, service intervals) the only one that stands a chance of objective quantification is wear (through something like oil sample analysis). Would it be possible to redefine the scope of this question? As it stands, it's rather subjective
    – Zaid
    Mar 9, 2017 at 18:50
  • @Zaid I will edit my question to reflect quantifiable metrics, but the nature of my question is such that I expect subjective answers. The problem is that if quantitative data were abundantly available, I probably wouldn't have asked this question. As such, I believe subjective answers hold merit for this question.
    – Hari
    Mar 9, 2017 at 21:43
  • Really, I'm agreeing with @Zaid here. If it's subjectivity you are after, I should probably close this question as it is. If you'd like to make it more objective, we can leave it open. Mar 9, 2017 at 21:53
  • 1
    @Pᴀᴜʟsᴛᴇʀ2 See edits. Again, if solid quantitative data was already available, I wouldn't be asking this question. The problem is that few people seem to have tried this, and fewer have any test results to show.
    – Hari
    Mar 9, 2017 at 21:56
  • 1
    Do you have also measured the remaining TBN (total base number)? AFAIK it is an important factor on when to do an oil change. Reason: The engine oil also needs to neutralize acids and corrosive combustion byproducts (therefor the basicity of engine oil). No oil filter can change the chemical composition of your engine oil, thus a mere filter change is not sufficient if the base reserve of the oil is neutralized
    – Martin
    Mar 11, 2017 at 9:52

5 Answers 5


The oil filters mostly get rid of particles larger than 20 microns. You will have particles smaller than this size, independent of how many times you change the filter. Therefore the answer is 'NO'. You won't be able to extend oil change periods by changing filters. At least if your oil change reason is the "report" you have in your hand telling you that you have metal particles. Also because you change oil so frequently, you are replacing perfectly fine filters anyway. So changing them more frequently wont help.

The fact that there are metal particles in your oil, does NOT mean that your oil has gone bad (purely depends on the amount of particles and their sizes but reports dont seem to tell sizes anyway). Also it does NOT mean that your filter has gone bad (again depends on the particle sizes). It is perfectly fine, your particle levels are also 1/10 of what online sources tell as upper limits. (see the oil analysis handbook). If anything, you seem to be changing oil/filter too often!

I calculated that if I changed oil at 5k km(~3k mile) like you do, instead of 30k km that I do now, I would have needed to change oil 4 to 6 times more often (at 260k km now), roughly paying the cost of a used newer model car easily for oil changes. What a waste of money that would be, since my engine is working great still, also especially considering people do get over 600k km with this service regime easily! I will get a free car soon :)

Oil filters are designed to last very long time. Because in most cases people change filters with the oil, the filter must always outlast the oil. Car makers have service intervals for 50+k km and 2+ year. So filters must last even longer.

Actually, the so called "filter life" is probably a common sense concept. If there was any kind of "real" limit to filter life, they would have needed to print on the box. Of course they will eventually become clogged etc., but this will take long time to happen. Actually, I am even surprised that filters do not tell 'max X km and max X year' on them, they surely can sell more filters if they do that! (or the values are so high that it makes little sense to write them there). But we mere humans, with our infinite wisdom, like to use our "common sense" to come to strange conclusions.

In either case, I never heard and also could not find any posts online about oil filters releasing particles due to old age or getting clogged so the bypass valve opens. This tells me that since most people change filters with oil change, a normal filter AT LEAST lasts longer than the oil.

About your engine oil. You should NOT use thicker oil unless there is a real reason for it. It is NOT a reason that your engine is old. Having an older or high mileage engine does not automatically mean that it is better to use thicker oil.

In addition, if you use fully syntethic oils. You can get 5W-40 or 5W-50 and then you don't need to change oil due to seasons anymore. Your engine specifications say that your engine should consume 1L oil per 1000KM (1 qt. per 750 miles). I would start with 5W-40 and if the engine is using more than this amount, then try 5W-50

enter image description here Source: http://www.bimmerfest.com/forums/showthread.php?t=366719

  • I really appreciate the link to the oil analysis handbook. Additionally, if the filter is truly a 20µm filter, then I'll have to ask if I can get particulate size in the next report. Part of the reason for frequent changes is I am working on an engine where I don't trust the previous maintenance. Frequent changes earlier should help until I establish that I can wait longer. Also, I really don't need help selecting an oil viscosity. The e36 guidelines are meant for typical use cases, not anything exceptional (ex. track) or higher temperatures (ex. 100°F).
    – Hari
    Mar 13, 2017 at 20:21
  • @HariGanti , You are free to do as many oil changes as you want, as long as you don't mind usefulness/cost ratio. I saw in one forum post that the blackstone labs made that particle size analysis for somebody. But I don't see what is the point. Because, yes you will have many particles smaller than 20micron, but that is not really important. As I mentioned, people do 600+k km without those oil analysis just by following normal intervals and normal filter. Remember, the companies who does the analysis have to find something in your oil and make you worried to stay in business. :) Mar 13, 2017 at 21:23
  • For $30 a change, I'm fine with wasting a bit of money to keep my car happy. Knowing particulate size can also help from a metallurgical standpoint. Larger pieces indicate high shear wear or galling, whereas finer particles indicate abrasive wear. Small particles also accelerate wear elsewhere because they act abrasively as well (ergo an oil change). Also, I'm a generally curious guy. I'd do the oil analyses regardless of results. It's all about analytics.
    – Hari
    Mar 13, 2017 at 21:29
  • If you will test this out, try a normal filter and try microgreen filter. I am interested in knowing if their 2 micron claim is anywhere near truth. It is suspicious because they don't claim it to be the nominal filtration size. Otherwise putting same type of filter wont have a huge effect on particle sizes imho. Mar 13, 2017 at 23:15
  • They don't make them for my car. Otherwise, I'd be quite willing to try.
    – Hari
    Mar 14, 2017 at 4:20

Filters are like this. The more they are used. The more stuff in them. The smaller the holes for oil to pass threw them. The better the filter. But go to long in the filter does not have a overflow. They can clog up & no oil go threw them. Same with a air filter. Look at them the same as you would a strainer at home. The more stuff strained the finer they work. But the slower the liquid goes threw them.

  • I understand how filters work... I am asking for experience with attempting this type of maintenance schedule.
    – Hari
    Mar 11, 2017 at 22:33

Don't forget to take into account the fact that you'll lose some oil replacing the filter -- the filter itself holds a reasonable amount and on some cars I've had, if you take the filter off without draining the oil you can get quite a bit of oil out through the filter fitting (to the extent that if you undid the sump plug then immediately removed the filter you'd make quite a mess). Other cars have lost very little this way.

This is likely to change the economics of your situation.

  • Do a quick search for the e36 oil change procedure. It uses a cartridge filter, not a canister, and it's mounted high and upright in the engine bay. I lose a few ounces from the filter per oil change. Otherwise, I agree.
    – Hari
    Mar 13, 2017 at 20:17
  • 1
    The question doesn't state a model, so I answered the general case.
    – Chris H
    Mar 13, 2017 at 20:31
  • 1
    You know, I really thought I put the Make/Model/Year in there. I thought I did because @EverenYurtesen posted a picture of my car's oil viscosity chart. It's a 1994 BMW 325i Convertible.
    – Hari
    Mar 13, 2017 at 20:35

You don't mention what kind of car you have so I can't verify they have a product for you, but MicroGreen oil filters are actually designed to do exactly this in addition to extending service intervals. Their argument is at the end of a change interval the structure of the oil is fine and it's the filter that's in sorry shape; exactly what you found. You can look at their site for the specifics of their claim and how to use it. I personally run these because I like the idea of the microfilter acting in tandem with the regular filter, though I still change my oil every 5k miles.


  • I am not convinced that the so called 'microfilter' does anything there on top of the filter. Especially since it has so little surface area compared to the actual filter material. If it actually filtered something, it would have gotten clogged quickly and do not do much anyway. It sure looks like a marketing trick. I would have believed it easier if they lined the normal filter with the microfilter. But of course if it is same price as the regular filter, why not... Mar 11, 2017 at 21:37
  • The microfilter isn't supposed to filter 100% of the oil that comes through. It filters a portion of it and is able to filter it completely over time. There is a lot of firsthand accounts that indicate that not only does the filter perform better than conventional types, but that the product lives up to its claims. bobistheoilguy.com/forums/ubbthreads.php/topics/3884300/all/…30K_OCI-
    – atraudes
    Mar 12, 2017 at 22:48
  • The thread you put does not show that MG filters better than other filters. Other filters perform similar, I do 30k km(50k km for diesel)/2year service intervals with normal filter and people who are over 600k km with such intervals mechanics.stackexchange.com/questions/6856/… Actually, it is impossible that MG lives up to claims. Because even the manufacturer is not claiming that it can! at MG website, it says it can filter as small as 2micron particles. There is no figure on nominal filtration. Just marketing unfortunately. Mar 13, 2017 at 4:54
  • They do make the claim that "for vehicles with recommended 3,000-mile oil change interval, a microGreen filter should be installed with new motor oil. Then, only the filter should be swapped out at 6,000 miles, 12,000 miles, and 18,000 miles. Fresh oil and another new filter go in at 24,000." The fact that you are able to achieve similar results with a different product is no indication that this product can't do it any better, and absolutely doesn't disqualify it as a valid suggestion to the original question.
    – atraudes
    Mar 13, 2017 at 16:10
  • What I am saying is that with today's oils that 3000-mile interval is ridiculous anyway. So your engine would have worked fine if you used a normal filter also. They have no proof that their filters actually work better or last longer than normal filters. Other manufacturers are just not advertising this (probably since 3k intervals = more filters). As a matter of fact my off the shelf bosch filter lasts for 18k miles (30k km) according to VW/manufacturer. If I had diesel, it lasts for 30k miles (50k km). I do the math and using microgreen filters make no sense and more expensive only. Mar 13, 2017 at 19:11

I had the same problem on my Mercedes SUV. The motor oil turned quite black between oil changes, even though the car was in perfect working condition. It didn't consume oil ordinarily, but it used about 1/2 quart if I was very heavy on the gas pedal (flooring it on the highway to 4-5,000 RPM).

My solution was to use Marvel Mystery Oil (MMO) with fresh Mobil 1 in the crankcase, substituting 20% of the oil with MMO. I ran the engine at 3,500 RPM for long (20 minute) stretches on the highway for 200 miles then changed the oil + MMO again. I repeated for 500 miles, changed again, then a third time after an additional 1,000 miles. I also added MMO to gas tank at a rate of 4 oz. per 10 gallons of gas.

The engine no longer burns oil and the oil stays golden clean for 3,000 miles. I concluded that, even though the oil was always changed on time, the piston rings were partially stuck. Today's piston rings are low tension to save gas and I've read they often stick slightly. There was therefore a higher than normal amount of blow-by gases. The MMO freed the rings and improved compression, reducing blow-by and particulates in the crankcase.

  • I suspect that quite many cars have the motor oil turn black within an oil change interval, so it is not an indication that there is anything wrong with the engine. Sounds doing too much if you need to change the oil three times just to make the oil look a bit more clean. How long does this solution last? I suspect not very long.
    – juhist
    Mar 10, 2017 at 15:51
  • @juhist, he is addicted to addives :) MMO is mineral oil based mixture ( en.wikipedia.org/wiki/Marvel_Mystery_Oil ). It will reduce effectiveness and longevity of your oil ( mechanics.stackexchange.com/q/20942/14704 ) and at wikipedia is a link to a plane crash because MMO was added to fuel! If your engine does not burn any oil, then it means the oil you are using is too thick. It is normal for engines to burn small amounts of oil. Also color of engine oil is irrelevant auto.howstuffworks.com/5-engine-oil-myths.htm I would like to know how Carguy measured compression... Mar 11, 2017 at 18:21
  • @juhist - The solution lasted for 10k miles now, and counting.
    – Carguy
    Mar 13, 2017 at 15:58
  • 1
    @EvrenYurtesen - By the way, I'm addicted to additives only because so many car guys insist they have absolutely no benefit, and that they are always "snake oil." Major oil companies add mineral oils all the time to gas and oil (as well as VOCs similar to acetone), so why all the adversity? Adding additives on your own has its place and cures many common ailments in engines, and has since at least the 1920s.
    – Carguy
    Mar 13, 2017 at 16:08
  • 1
    @Carguy, since there is a film of oil on the cylinders when pistons move, some oil must burn at combustion. While the tolerances get lower, lesser oil will be consumed. This is why you would burn more oil if you run your car at high RPMs, 1500 vs 3000 rpm, you nearly double oil consumption. If your engine is not consuming any oil, it means your oil is too thick to fit through the pistons and well you are destroying your engine. I have nothing against additives, but same as I would like to go to doctor before taking magical pills, I would like to understand the problem before devising solution. Mar 13, 2017 at 18:51

Not the answer you're looking for? Browse other questions tagged or ask your own question.