Van's Air Force

The definitive Van's Aircraft support community! Buying, building or flying an RV? Join our exclusive family of mentors and enthusiasts!

Bypass Oil Filtration - Any Hard Data Available?

what drives the frequent oil drain cycle in our engines is the fact that the lead blow-by particles load the oil to the level that it cannot hold any more in suspension and therefore we must refresh it or accept sludge build up. So, the question is, how large is a lead particle nad can one of these things filter them out. If not, we cannot extend our drain interval, at least not with 100LL. It was my understanding that soot and other carbon debris is the issue on diesels and believe that stuff is much larger particle size than the almost aerosolized lead we use in our fuel.

Larry
This was my thought as well. I drained some oil at 30 hrs and put some samples in clear glass jars for a humidity testing. Left them for a couple of months and there was a sludge layer (lead?) in the bottom of the jar. Diesels use bypass filtration for long(er) life and TBN (total base number) monitoring of lubricants for drastically long oil changes. Also they have very large sumps compared to aviation engines. None of this field data (of which I am aware) would support benefits for our Lycs.

Further, the over the road diesel engines are not" large clearance" in any sense. Europeans use centrifugal filters in bypass but have not gathered large following in the US. They do the best job of getting small stuff out of oil and would be perfect for our engine, except for the G thing. Kidney loop shop based systems are very practical and cost effective for off highway to extend life of the products. Not so much for engines.

It "might" be functional for a Lyc, to hook up every time it is parked (at home), but again it just becomes and external sump and perfect oil only leaves the engine at risk to the next thing down the list. I sketched up a kidney loop with centrifugal filter for my -7, priced it and then abandoned as more overhead with a decade return if any.

There is proven data that circulated soot from diesels effect wear in micron clearances of fuel systems, but not sure it is applicable to lead suspension.
 
Just got back from a 0.6 test flight and saw no change in operating characteristics. And contrary to the concerns of some posters in this thread, the oil pressure remained unchanged despite the addition of the bleed through the filter. Visual inspection of the new equipment shows no leaks or other distress. Will continue to monitor the installation at frequent intervals but I’m cautiously claiming success and now just looking for enough hours to pull a sample. I’m thinking 10, 25 and 50 for this load of oil.
 
Quick anecdotal update. At 15 or so hours now with no leaks or negative operating characteristics. Every check of the oil level reveals a “nearly new” honey like color of the oil. Never seen that before on any aircraft engine I’ve ever flown. Most is usually carbon black by 15 hours. Not sure if the oil is “good” or “bad” yet, but it still appears new. 35 more hours to go until a burn.
 
Quick anecdotal update. At 15 or so hours now with no leaks or negative operating characteristics. Every check of the oil level reveals a “nearly new” honey like color of the oil. Never seen that before on any aircraft engine I’ve ever flown. Most is usually carbon black by 15 hours. Not sure if the oil is “good” or “bad” yet, but it still appears new. 35 more hours to go until a burn.
Keep us informed. A real data point would be to have the current oil tested at end of interval. I'd donate some towards that as DanH turned down my donation for the filter testing.

(Drift =) BTW. Never looked at your signature. Congrats on still being alive having owned/flown a wooden wing spar/wooden tail Mooney. Nothing wrong with the design but hidden conditions in those killed more than a few people.
 
Keep us informed. A real data point would be to have the current oil tested at end of interval. I'd donate some towards that as DanH turned down my donation for the filter testing.

(Drift =) BTW. Never looked at your signature. Congrats on still being alive having owned/flown a wooden wing spar/wooden tail Mooney. Nothing wrong with the design but hidden conditions in those killed more than a few people.
Aside from the normal, Blackstone oil analysis, is there another, more comprehensive analysis that I should be looking at? I still have the old sample yet to send in.

As for the Mooney. We got it with a metal tail installed, but the wing glue joints and rot eventually got ahead of me and it killed the airplane as a viable project.
 
now just looking for enough hours to pull a sample. I’m thinking 10, 25 and 50 for this load of oil.

At 15 or so hours now with no leaks or negative operating characteristics. ... 35 more hours to go until a burn.

Any reason for the change in test protocol? Or is it just that it "still looks new"? It would seem that apart from the cost there would be no downside in sending a sample in early... 10 (or 15 now), 25, 50 was a good idea.
 
Aside from the normal, Blackstone oil analysis, is there another, more comprehensive analysis that I should be looking at? I still have the old sample yet to send in.

As for the Mooney. We got it with a metal tail installed, but the wing glue joints and rot eventually got ahead of me and it killed the airplane as a viable project.
I'm more familiar with the analysis from Texas Oil Tech or similar versus Blackstone. The Blackstone reports I'd seen in the past tested for "typical" fluid properties plus spectro for metals. The TOT analysis we utilize here at work yields those plus the additive levels. We/our customers use this to rebalance the aforementioned. Tens of thousands of operating hours on the oil is typical though the duty is much different; some exposure to (typically cleaner) exhaust gases and no lead. Potentially worth looking in to. The geek in me wants to know the answers.

Edit = Had a couple of filtration surprises in my professional life. Tried some ultra-fine filtration on hyd oil that was suspected of silting some servovalves. It didn't fix the then unexplained anomaly but filter life was amazing short as it removed the dye from the fluid. Didn't see that one coming
 
Finally, DATA...

Two samples - one without the AMSOIL bypass filter, then one WITH. Same airplane, same oil type, same 50 hour interval, same mission profile, same fuel, same primary oil filter (NAPA 1068).

I did pay for the TBN analysis and still working with the lab to find out why I do not have a value in that column. If I can get it, I will add it. Not sure what to make of the analysis other than what seems to be a big reduction in lead and Insolubles with the bypass filter installed. I present the results for those smarter than me have at it:Blackstone Report.jpg
 
Last edited:
Finally, DATA...

Two samples - one without the AMSOIL bypass filter, then one WITH. Same airplane, same oil type, same 50 hour interval, same mission profile, same fuel, same primary oil filter (NAPA 1068).

I did pay for the TBN analysis and still working with the lab to find out why I do not have a value in that column. If I can get it, I will add it. Not sure what to make of the analysis other than what seems to be a big reduction in lead and Insolubles with the bypass filter installed. I present the results for those smarter than me have at it:
Gonna attach the report?

Verification of polishing filter specs would also be appreciated. Thx.
 
Not sure what to make of the analysis other than what seems to be a big reduction in lead and Insolubles with the bypass filter installed.
When I look at a relative reduction, some of the numbers show a dramatic drop (iron, magnesium, phosphorus..). If there is no visible darkening of oil after 15 hours that is remarkable also.
 
Thanks for this info, @Toobuilder. It will be interesting to see if these numbers hold up. Is it technically possible to filter lead out of engine oil? I thought the molecules that contain lead in our engines are sub-micron in size, but perhaps this is wrong. I'd love to find a way to get the lead out of the oil for the times that I have to burn 100LL.
 
Thanks for this info, @Toobuilder. It will be interesting to see if these numbers hold up. Is it technically possible to filter lead out of engine oil? I thought the molecules that contain lead in our engines are sub-micron in size, but perhaps this is wrong. I'd love to find a way to get the lead out of the oil for the times that I have to burn 100LL.
Lead can't be "filtered" from oil or fuel. Pure speculation on my part, but it's possible lead has an affinity for something that is getting filtered out like tin or iron? So by sticking to that element it gets pulled out by the filter? Certainly interesting.
 
OK. Blackstone provided the TBN values and here is the revised report:

oilTBN.jpg

Understanding that this is but one data point, the values that track most closely to "filtration effectiveness" are showing the desired/expected behavior of the bypass filter. In particular, the TBN (oil life remaining) is a dramatic improvement with the bypass. Whether the results are nominal or an outlier remain to be seen. And the discussion surrounding the benefit to longevity of our engines is also a subject for another thread. But at face value, the bypass filter is cleaning my oil more effectively than a full flow alone.
 
Mike, what are the claimed efficiency values for the bypass filter?
 
With respect to oil change intervals:

One of the reasons we change oil because it accumulates water in suspension (hydrocarbon combustion byproducts, absorption from ambient humidity)

Bypass filtration can't remove water.

Will you get to the end of this experimentation and data-gathering exercise and conclude that it doesn't really make a practical difference because we have to change oil at the same rate regardless?

- mark
 
Good info Mike. I think this is very interesting.
It would be interesting to see an analysis for new unused oil to compare.
The bottom of the PPM list (minus Silicon) are additive package elements. I'm not an expert, but I think you don't want those to go down.
 
With respect to oil change intervals:

One of the reasons we change oil because it accumulates water in suspension (hydrocarbon combustion byproducts, absorption from ambient humidity)

Bypass filtration can't remove water.

Will you get to the end of this experimentation and data-gathering exercise and conclude that it doesn't really make a practical difference because we have to change oil at the same rate regardless?

- mark
Oil and water do not mix. Oil does NOT hold water in suspension . Within 10 minutes of shut down, all the water has settled to the bottom of the pan.
 
Oil and water do not mix. Oil does NOT hold water in suspension . Within 10 minutes of shut down, all the water has settled to the bottom of the pan.

Perhaps I wasn't clear.

Without quibbling over words, during operation it's "in suspension," in the sense that there are micro droplets of liquid water carried through the lubrication system with the oil.

Whether it settles out in the sump afterwards isn't relevant. It won't stay in the sump for more than a minute when it's stirred up again by the oil pump impeller after engine restart.

And although I mentioned water, there are other contaminants when it reacts with other combustion products, so over time the oil gets more and more loaded with liquid-phase detritus which degrades its lubricity.

I don't think we can remove those with filtration. We remove them by changing the oil.

So my contention is that comments earlier in the thread about diesel operators running oil for thousands of hours between changes aren't going to apply to us regardless of how well our oil is filtered, and I'm wondering how much is too much, and what lifetime could be expected from oil even if it's filtered down to nil particulates.

- mark
 
Is this different than the "WATER" line in the "Properties" section of the report shown above?
This is also my question - we talk about water in the oil like it can't be avoided, but why do the oil analysis reports show 0.0% water? This is also what I see in my Blackstone reports. I've gotten my samples two ways - after a flight with hot oil, and with a completely cold engine. I always get nil water in my report. Perhaps the method of detecting water does not have high enough resolution.

If my math is right, about two drops of water in a typical Blackstone sample would show up as 0.1% of water.
 
With respect to oil change intervals:

One of the reasons we change oil because it accumulates water in suspension (hydrocarbon combustion byproducts, absorption from ambient humidity)

A steadily increasing accumulation with the addition of run hours? Reference please...

Will you get to the end of this experimentation and data-gathering exercise and conclude that it doesn't really make a practical difference because we have to change oil at the same rate regardless?

Lycoming has already approved 100 hour changes with unleaded fuel....with a 50 hour filter change requirement. Water is not the issue here. Mike has simply added additional filtration with a beta well in excess of the current best aviation filter. That filter is factually substandard as compared to current automotive practice.
 
Oil and water do not mix. Oil does NOT hold water in suspension . Within 10 minutes of shut down, all the water has settled to the bottom of the pan.
To quote a baseball legend, it's Deja vu all over again.

Yes they do. Depends on how closely you look. Those xxx ppm of H2O can have a profound affect on the lubrication boundary film. Effects are obviously even worse for sliding contact. Unfortunately, the subject report test for water at the % level. A more finite test would probably yield values of 120-200 ppm. Dryer is better, period.

Reference post 9 in link below

Older VAF post

With respect to oil change intervals:

One of the reasons we change oil because it accumulates water in suspension (hydrocarbon combustion byproducts, absorption from ambient humidity)

Bypass filtration can't remove water.

Will you get to the end of this experimentation and data-gathering exercise and conclude that it doesn't really make a practical difference because we have to change oil at the same rate regardless?

- mark

Nope. This would assume that that the oil never reaches temps where the water would flash off. Plenty of opportunity for that to happen.
Keep your power plant operating with oil temps in the green (common misconception that cooler = better) and the case properly ventilated and worry about something else.

Soap Box = ON

The OP has done an interesting experiment and posted the raw results without prejudice. Based on this sole data point, ask yourself which oil you'd rather have circulating through your very expensive PP.

@Toobuilder/Michael. Edward Kollin (stud chemist, creator of Camguard) has a professional on-line profile. Want me to reach out for comment? Chances are low but man I'm betting it would be insightful.

...and nice job. Thanks for the effort and for sharing

=OFF
 
Last edited:
Path forward: Unless someone has a better idea, my plan is to run another 50 hour cycle on the bypass filter and burn the sample. The expected result would be to maintain the current levels shown in this last sample. At oil change I’m going to “turn off” the bypass and go back to the conventional Lycoming scheme with the full flow filter only. After a 50 hour cycle with the full flow only, if the bypass is effective, there should be a spike in some of these readings and the results should more or less return to the first burn.
 
Path forward: Unless someone has a better idea, my plan is to run another 50 hour cycle on the bypass filter and burn the sample. The expected result would be to maintain the current levels shown in this last sample. At oil change I’m going to “turn off” the bypass and go back to the conventional Lycoming scheme with the full flow filter only. After a 50 hour cycle with the full flow only, if the bypass is effective, there should be a spike in some of these readings and the results should more or less return to the first burn.
Don't know the right path forward. Will you be changing the polishing filter for the coming intervals?
 
Don't know the right path forward. Will you be changing the polishing filter for the coming intervals?

The polishing (bypass) filter element is a very long life unit - like 2000 hours. The dirtier it gets, the better it works (up to a point). Not planning to change it out for quite a long time.
 
Lycoming has already approved 100 hour changes with unleaded fuel....with a 50 hour filter change requirement. Water is not the issue here. Mike has simply added additional filtration with a beta well in excess of the current best aviation filter. That filter is factually substandard as compared to current automotive practice.
Are you referring to https://www.lycoming.com/publications/service-letter-no-270 ?
It still recommends to change every 4 months or 50/100 hrs whichever occurs first.

also here: https://www.lycoming.com/content/oil-and-filter-change-recommendations
C. Even if the aircraft is flown only a few hours, a total of four months maximum between changes for both systems listed under “A” and “B.”

I wonder if this is related to water contamination. Even though "the oil and water don't mix" there is a very small amount (a few ml) that will dissolve and about half a cup that can be suspended before it starts to fall out as free water.. In either case it looks like the concentration is too small to detect by blackstone.

I always get nil water in my report. Perhaps the method of detecting water does not have high enough resolution.
I'm tempted to add a few drops to the next sample and see if it shows up in the report. 😬
 
Several years ago I modified a Motorguard M30 filter to use toilet paper. I have a couple of these that I use for shop air filtration when painting. Works great for trapping water.

Turned a tube on my lathe as detailed here:

Originally these filters were developed to compete with the Frantz.

I intended to install the M30 on my Rocket but due to size and mounting difficulties decided to go with the Baldwin setup.

On my never-ending list of things to do is to make an external dipstick bypass filter using a pump, a M30, and a heating element to preheat the oil before it flows thru the filter. This way any water vapor created by heating the oil is done externally and trapped in the toilet paper element.

It is my intent to use the TP filter regularly so I can run more advanced automotive oils in the airplane which could have problems with lead suspension.
 
With the understanding that the above oil analysis provides a single data point, if taken as an accurate indicator of future behavior, does anyone argue the second sample is not “cleaner” than the first? With the data so far, have we “proven” that the bypass filter scheme functions in a positive manner?
 
With the understanding that the above oil analysis provides a single data point, if taken as an accurate indicator of future behavior, does anyone argue the second sample is not “cleaner” than the first? With the data so far, have we “proven” that the bypass filter scheme functions in a positive manner?
At first pass it looks good. I'm no oil expert so my question would be, what are the ranges normally found for those elements without the bypass filter? Operating conditions vary between oil changes. I'd assume little things like ambient temps, dust, and humidity would make the oil content change when measuring at the PPM level.
 
I'd only remind that ICP spectro tells nothing about particulates much larger than 10 microns. Particle counts in the range of 15 to 30 microns, where our full flow filters are showing efficiencies as low as 40%, are probably much better with the bypass filter.
 
…I'm no oil expert so my question would be, what are the ranges normally found for those elements without the bypass filter?…

Not sure what “normal” for my engine and environment are, but I’m interested in “change”. The three things I’m interested in are lead content, insoluables, and TBN. The last two being related directly to filtration efficiency and oil life remaining. In this one sample, the change was dramatic in the right direction. Will be interesting if the next two samples follow suit.
 
could we find out what drives the 4-month oil change interval regardless of hours? is that still a constraint with the polishing filter..
 
Back when I was putting my engine on I couldn't figure out how to mount an oil filter with the inverted oil and vacuum pump alternator. Been just running the screen all this time. Not the best thing for longevity on that expensive engine.

Was inspired by this thread to build the "clean cart". Used this pump and filter mount. The battery I replaced at annual and some tubing. Didn't get a picture of the end that goes into the fill spout, just two 1/4" tubes. 24" to reach to the bottom of the sump and 21" with the end squeezed shut and a hole drilled in the side, hoping for the exiting oil to shoot out the side and not just loop back to the input. Some scrap wood for a carry case.

Tried it out for the first time after my flight today:

IMG_20250503_104947521.jpg


Need to replace the tube on the suction side to something reinforced, it partially collapsed on the hot oil. Still pumped quite fast. Let it run while putting away my headset, parachute, etc. Maybe 10 minutes? Based on the 0.5 gal/minute spec (which I probably didn't get due to the collapsed hose) that would run all the oil thru at least twice.

Did end up having to add oil afterwards, around 1qt stayed in the hoses and filter.
 
Oil and water do not mix. Oil does NOT hold water in suspension . Within 10 minutes of shut down, all the water has settled to the bottom of the pan.
Well, this is not exactly correct. For my oil change, (~30 hrs) I always change relatively hot and an hour over 180F, after a crankcase purge to get rid of the bulk , drain. I captured a pint in two glass jars (qt) with metal lids. I purged the jars with -45F dew point air and had sensors in the jar to measure temperature and relative humidity. Over a day, the data logger showed a gentle rise of humidity, up to ~20%. I contacted Aeroshell and got an engineer contact me. I sent the data and he said some tiny amount of water is always in the engine oil. I don't know how or what the specific physical reasons for this. I even put the oil in a vacuum chamber and left it at a quite low pressure for several days, and the test result did not change. The Aeroshell technical person said it was not the sensor reading oil vapor as water, something I had already investigated.

On lead removal - a centrifugal oil filter will remove the suspended lead.
 
Well, this is not exactly correct. For my oil change, (~30 hrs) I always change relatively hot and an hour over 180F, after a crankcase purge to get rid of the bulk , drain. I captured a pint in two glass jars (qt) with metal lids. I purged the jars with -45F dew point air and had sensors in the jar to measure temperature and relative humidity. Over a day, the data logger showed a gentle rise of humidity, up to ~20%. I contacted Aeroshell and got an engineer contact me. I sent the data and he said some tiny amount of water is always in the engine oil. I don't know how or what the specific physical reasons for this. I even put the oil in a vacuum chamber and left it at a quite low pressure for several days, and the test result did not change. The Aeroshell technical person said it was not the sensor reading oil vapor as water, something I had already investigated.

On lead removal - a centrifugal oil filter will remove the suspended lead.
Ok, i stand corrected. I can see how some small portion of water droplets get sheared to such a small size that their surface tension allows them to be trapped in the oil. However, the fact that you detect moisture vapor kind of proves my point that they are not held in suspension and eventually work themselves out. If they were permanently trapped in the oil or mixed with it, i don’t see how the humidity level would have risen in your experiment. I am sure you guys with throw some more science at me to prove that wrong though.
 
Last edited:
Not to drift too far from this excellent testing and data collection post, but I found this video on rod bearing wear/ oil viscosity very interesting.

Some key points:
-GM is recalling L87 V8s because of rod bearing wear. They are increasing recommended oil viscosity from 0–20W to 0–40 to increase oil film thickness.
(Seems they may have gone a step too far in pursuit of EPA fuel economy)
-They mention importance of early oil changes on a new engine to prevent debris accumulating in oil passages.
-interestingly, they claim they can detect bearing problems by using the VVT camshaft/crankshaft sensor
 
On lead removal - a centrifugal oil filter will remove the suspended lead.
This sounds very interesting. Do you know of any that are reasonably priced? Seems like it would be good to use an oil cleaning cart to do some deep filtration and remove the lead.
 
This sounds very interesting. Do you know of any that are reasonably priced? Seems like it would be good to use an oil cleaning cart to do some deep filtration and remove the lead.
That statement isn't complete. In addition to the centrifuge you need to heat the oil to 550-700F.
 
I am nearly 19 hours into this next load of "bypassed" oil and needed to add a couple quarts for tomorrow's flight. I previously commented how clean the oil looked with the bypass scheme so I decided to take a picture before adding any new oil.

Hardly scientific, but all of this oil is 19 hours old. Looks pretty good to me:

IMG_20250518_171512033.jpg
 
Another 50 hours (52, actually) on the oil and time to share results again.

350hour oil.jpg

So for background, this most recent 52 hours was with the bypass oil filer ACTIVE. That makes 2 samples in a row. The next 50 hours will be with the bypass blocked off. If this filter is doing anything, we should see a spike in some elements and a drop in TBN, returning to levels near the first (non bypass) sample at 250 hours.
 
Some of this will go away if we ever get a no lead aviation fuel. Of course that fuel will cost a few dollars more per gallon. What it will do is let the Lycoming's and Continental's use synthetic oil and increase change intervals to at least 100 hours or more. Rotax recommends 100Hour oil change intervals if the engine is run exclusively on unleaded fuel and the oil used is synthetic.
 
Interesting to me that the TBN values are much better with the bypass than without, but the oil is still "worn out" at 50 hours.

Considering this last load of oil was just about as ideal an operating condition as possible in a light aircraft (almost entirely long range, LOP cruise at stable temps and RPM), then I wonder what else I can do to get the recommended >2.0 value?
 
Interesting to me that the TBN values are much better with the bypass than without, but the oil is still "worn out" at 50 hours.

Considering this last load of oil was just about as ideal an operating condition as possible in a light aircraft (almost entirely long range, LOP cruise at stable temps and RPM), then I wonder what else I can do to get the recommended >2.0 value?
Send a brand new oil sample straight from the bottle, and see what was there to start with....
 
I was curioius what TBN meant. Internet says....

In short, a TBN (Total Base Number) measures how much base (as in base vs. acid) additive is in the oil to offset the effects of acids coming into the oil from combustion and other sources. Scientifically speaking, the TBN is one of two “neutralization number” tests run on oils.
 
Back
Top