Considering oil filter relocation

[ZeeOSix]

Originally Posted by shibby6600
Originally Posted by SubieRubyRoo
Downsides are you've now multiplied your chances for failure, and long-term efficiency has taken a hit. Parallel filtration actually increases the chances of a certain size particle getting through the filter... if both filters are say 95% efficient at 20 microns, when you double the filter count you actually reduce the overall possibility of catching that 20 micron particle to: 95% x 95%= 90.25% efficient at 20 microns.

I'm having a hard time wrapping my head around this. Could you explain why it lowers the chances? Intuitively I would have guessed it'd be the same regardless the number of filters because the particle can only go through one at a time, making it 95% this filter or 95% that filter.

What if the filters were in series, would it be averaged or the same?

At first guess I would have said parallel filtering is the same and series is slightly better. I genuinely don't understand the reasoning so any clarification is appreciated.


I'm thinking the same as shibby on this. I think you have to look at what the probability is of a particle getting through, not captured. If you had 3 filters in series (all 95% @ 20u), what are the chances of one specific particle that is 20u getting through all 3 filters?

For one filter it's 5%.
For two filters it's 5% of 5% = 0.25%.
For three filters it's 5% of 0.25% = 0.0125% ... which would be the same as 99.99 % @ 20u.
etc ....

Having 3 filters (all 95% @ 20u) in parallel, they all will have their independent efficiency so if the particle goes through any one of them one time the possibility of not being captured would be 5%.

 

[Linctex]

I'm with ZeeOsix - - adding more filter SURFACE AREA (with resultant slower rate of flow through the media) is generally always a good thing.

I can't see any negatives.... other than oil taking to long to warm up during the winter in Canuckistan

 

[WellOiled]

I'm with shibby6600, ZeeOSix and Linctex on this. For two of the same filters in parallel, the efficiency of each filter would be slightly greater than the test of a single filter because the flow rate through each filter is roughly 50% of what a single filter would be. This is the foundation for bypass filtering. How much improvement in efficiency would be unknown without a test. Reliability is another issue.

I have also been thinking that a thermostatic oil cooler might be a good option and would allow the oil to warm up quickly but still control the maximum oil temperature.

 

[shibby6600]

Originally Posted by WellOiled
I have also been thinking that a thermostatic oil cooler might be a good option and would allow the oil to warm up quickly but still control the maximum oil temperature.


I've noticed on some cars, Cruze for example, the oil cooler is actually an oil temperature exchanger with the coolant temperature. Similar to how automatic transmission lines run through the radiator to exchange/regulate trans temp. It heats up the oil a little faster since the coolant typically heats up quicker, and also keeps it from getting too hot by giving off some to the coolant. My OBD reader confirmed this for me multiple times this winter. Oil temp lags behind coolant just a bit at first start, and then oil temp stays just over coolant once up to operating temp.

 

[SubieRubyRoo]

Zee, not as an argument but trying to dig this stuff back up from the way past... the way I remember it is your description would be with all three filters in series, and then yes that math makes sense. If the particle had to pass thru three consecutive filters, it would make sense the probability of catching any specific particle would increase.

I think we're on the same page... one way of plumbing increase probability, one way decreases. I still think parallel decreases and series filters would increase, but hey, I've been up since 4am and just finished a couple glasses of Basil Hayden's, neat. I could be confused by now 🤣

 

[WellOiled]

SubieRubyRoo. After the Basil Hayden's , neat wears off try this simple logic. No degrees needed.

Starting with a flow of 2X GPM. Let's assume the upstream particle count is 2000. Each filter is 95% @ 2X GPM.

Now, send 1X GPM and 1000 particles through each filter. If the filter is 96% efficient @ 1X GPM then the down stream particle count will be 40 for each filter.

Now add the two streams together. 2000 particles upstream and 80 particles downstream. Beta = 2000/80 = 25. Efficiency for the 2 filters in parallel is now 24/25 or 96%.

It works for me, maybe will work for you too.

 

[ZeeOSix]

Originally Posted by SubieRubyRoo
Zee, not as an argument but trying to dig this stuff back up from the way past... the way I remember it is your description would be with all three filters in series, and then yes that math makes sense. If the particle had to pass thru three consecutive filters, it would make sense the probability of catching any specific particle would increase.


Yeah, my example was for 3 filters all in series, thought I made that pretty clear with the example. If you had 10 filters all in series the chances of a 20u particle making it through all 10 filters has to be super unlikely (ie, it's 5% to the 10 power chance of it escaping capture using the 95% @ 20u example).

Originally Posted by SubieRubyRoo
I think we're on the same page... one way of plumbing increase probability, one way decreases. I still think parallel decreases and series filters would increase, but hey, I've been up since 4am and just finished a couple glasses of Basil Hayden's, neat. I could be confused by now 🤣


As far as multiple filters in parallel, slowing the flow down because of split flow could slightly help the efficiency, as others said, but the worse it would be is the rated efficiency of one filter (assuming all filters used are the same model/efficiency).

Have a Hayden's for me.

 

[Bailes1992]

Does the oil flow through two filters or is the oil flow split between them?

If it's split and goes between the two filter and both those filters have a 95% efficiency at 20 microns then surely the end result is 95% efficiency at 20 microns?

If the oil is flowing through two filters, both 95% efficiency at 20 microns then surely you've halved the chance that a 20 micron particle gets through the filters?

 

[SubieRubyRoo]

Okay, so I think WellOiled actually helped me think about this. Single 96% filter @ 2000 particles 20u = 80 downstream.

If you add the second filter in parallel you have to combine the particles that get past each filter, no? There's probably some calculus involved to get an exact answer, but a system will not evenly distribute the 2000 particles to each filter. So you have to add filter 1 efficiency plus filter two efficiency, and since the rated efficiency is not 100%, the count from two filters will have to be higher than just a single filter, because even at the very least, if 80 particles made it through the first filter, an additional 4% of those particles are going to also make it through the second filter. I still haven't yet got a clear image of what I'm wanting to describe. But I do have this illustration:

If simply sticking two identical full flow filters was more beneficial, why doesn't any company that sells filters offer a setup like this? You'll see a full-flow paired with a necked-down bypass filter that fine-cleans about 10% at a given time. I agree that lower flow thru regular filters "should" marginally improve efficiency, but it will be different on every engine due to oil pump output. To me, if you're taking the additional time, money, and risk by installing all that external hose and possible failure points, you should be using the second filter mount to clean the oil as much as possible, not sticking another regular filter on it. JMHO.

 

[JohnnyJohnson]

Originally Posted by SubieRubyRoo
Okay, so I think WellOiled actually helped me think about this. Single 96% filter @ 2000 particles 20u = 80 downstream.

If you add the second filter in parallel you have to combine the particles that get past each filter, no? There's probably some calculus involved to get an exact answer, but a system will not evenly distribute the 2000 particles to each filter. So you have to add filter 1 efficiency plus filter two efficiency, and since the rated efficiency is not 100%, the count from two filters will have to be higher than just a single filter, because even at the very least, if 80 particles made it through the first filter, an additional 4% of those particles are going to also make it through the second filter. I still haven't yet got a clear image of what I'm wanting to describe. But I do have this illustration:

If simply sticking two identical full flow filters was more beneficial, why doesn't any company that sells filters offer a setup like this? You'll see a full-flow paired with a necked-down bypass filter that fine-cleans about 10% at a given time. I agree that lower flow thru regular filters "should" marginally improve efficiency, but it will be different on every engine due to oil pump output. To me, if you're taking the additional time, money, and risk by installing all that external hose and possible failure points, you should be using the second filter mount to clean the oil as much as possible, not sticking another regular filter on it. JMHO.



LOL Subie but say if we were really smart with be adding three filters forget about diminishing returns.

 

[ZeeOSix]

Originally Posted by SubieRubyRoo
Okay, so I think WellOiled actually helped me think about this. Single 96% filter @ 2000 particles 20u = 80 downstream.

If you add the second filter in parallel you have to combine the particles that get past each filter, no? There's probably some calculus involved to get an exact answer, but a system will not evenly distribute the 2000 particles to each filter. So you have to add filter 1 efficiency plus filter two efficiency, and since the rated efficiency is not 100%, the count from two filters will have to be higher than just a single filter, because even at the very least, if 80 particles made it through the first filter, an additional 4% of those particles are going to also make it through the second filter. I still haven't yet got a clear image of what I'm wanting to describe. But I do have this illustration:

If simply sticking two identical full flow filters was more beneficial, why doesn't any company that sells filters offer a setup like this? You'll see a full-flow paired with a necked-down bypass filter that fine-cleans about 10% at a given time. I agree that lower flow thru regular filters "should" marginally improve efficiency, but it will be different on every engine due to oil pump output. To me, if you're taking the additional time, money, and risk by installing all that external hose and possible failure points, you should be using the second filter mount to clean the oil as much as possible, not sticking another regular filter on it. JMHO.


I think you're over thinking it a bit. The only thing anyone could assume is that if two identical filters are put in parallel and the flow gets evenly split (the logical assumption), then the overall efficiency will be the same, regardless of how many filters are put in parallel. The only half way logical assumption that could be made (as already discussed by a few guys) is that splitting the flow between parallel filters would slow the oil flow down and possibly increase the efficiency slightly for each filter. But the worse it would is the basic rated filter efficiency when in a parallel setup. Of course other benefits of a parallel filter setup is: 1) Less delta- across the filter pack, 2) More holding capacity and 3) More oil capacity.

The example again - with a series setup for three 95% @ 20u filters in series, you have to look at what gets past each filter in the 20u and larger group of particles.

Filter 1: 5% of particles 20u and larger gets through.
Filter 2: 5% of particles 20u and larger that got through Filter 1 also get through Filter 2 - so 5% of %5 = 0.25%.
Filter 3: 5% of particles 20u and larger that got through Filter 2 also get through Filter 3 - so 5% of 0.25% = 0.0125%.
etc ....

So the efficiency of the 3-pack series filter setup would be 100% - 0.0125% = 99.99% @ 20u. Or you could just run a large singe filter rated at 99% @ 20u and get almost the same thing.

Of course, with 3 filters in series you have now caused 3 times the delta-p across the filter pack. And the first oil filter is going to load up first by a huge margin compared to the subsequent filters downstream of the first filter.

 

[Linctex]

Well.... my system has two full flow filters in parallel (Fram Tough guard)

https://www.bobistheoilguy.com/foru...ridiculous-filtration-system#Post4713720

The "clean side" oil is sampled and sent to two By-Pass filters "in Series" (10 micron 1st, then to a 5 micron 2nd)

What are my particle counts?

 

[ZeeOSix]

Originally Posted by Linctex
What are my particle counts?

Whatever the ISO particle count test says.