95% efficient is twice as good as 90% efficient, or 5% better?

Status
Not open for further replies.
Joined
Feb 7, 2013
Messages
998
Location
st louis, mo
Here is a question:

Is a filter that is 95% efficient at a given particle size only 5% better (i.e. not really significantly better) than a 90% efficient filter?

Or is it twice as good because it's only letting through half as many particles?

Is the logic that the 90% filter will just catch it on the next time through?

Not trolling, genuinely asking.
 
The efficiency is going to depend on the particle size. If all the particles are smaller than the filter is designed to handle then it would be 0% efficient. Conversely if it caught 100% of the particles it may be in bypass all the time making it 0% efficient. At a certain point of filtering you would possibly start removing some of the add pack chemicals imo. Have you ever gotten a bottle of oil and seen the fallout on the bottom of the bottle? That is why I'm not a big fan of polishing filters for engines. Steam turbines, yes. car engines, no. This is my take and I'm sure the flames will start soon.
 
Depends on your goal.

Are you trying to avoid every single particle escape? or are you trying to harvest the particles for other use (i.e. You are trying to extract particle as raw material for manufacturing process)?

For the first one, 95% is twice as good as 90% in one pass. Remember it is the long term multi-pass that counts in reality, so the reality is going to be a lot lower if you change your filter frequently and not filling it pass the limit.

For the second one, 95% is 5.55% better than 90%.
 
Originally Posted by brages
Is a filter that is 95% efficient at a given particle size only 5% better (i.e. not really significantly better) than a 90% efficient filter?

Or is it twice as good because it's only letting through half as many particles?

Is the logic that the 90% filter will just catch it on the next time through?


A less efficient filter might catch particles on subsequent passes. But if a particle goes through the oiling system 3 times instead of once before being captured, then there's more potential for wear.

More efficient filters reduce the overall debris particle count in the oil (show a better UOA particle count), and also reduce the total cumulative passes that particles circulate multiple times through the oiling system.
 
Originally Posted by PandaBear
Depends on your goal.


Assume that this is an engine oil filter and the goal is minimizing long-term engine wear.
 
An efficiency ratio that increases from 90% to 95% is an increase of 5.5% (ie 95/90-1.0=.055). That is the correct mathematics of the increase.

People can describe it any way they want, some ways are accurate, others can be misleading. That is the way statistics work.
 
Filter 1 allows through 10% of particles of a given size per pass; Filter 2 allows only 5% of particles. 10% is twice as much as 5%. This math is just as correct as .95/.90 - 1.

The question I'm asking is... which ratio is more relevant to engine wear, and what is the reasoning to support that answer? Will engine wear be twice as high with Filter 1, or only about 5% higher (and therefore probably irrelevant)?
 
Originally Posted by brages
The question I'm asking is... which ratio is more relevant to engine wear, and what is the reasoning to support that answer? Will engine wear be twice as high with Filter 1, or only about 5% higher (and therefore probably irrelevant)?


Just run a filter that is 95% at 20μ or better and you'll be good on efficiency. And don't be fooled by filters rated at "99%" with no micron size, or filters rated 99% at a much larger particle size than 20μ.
 
Originally Posted by SeaJay
An efficiency ratio that increases from 90% to 95% is an increase of 5.5% (ie 95/90-1.0=.055). That is the correct mathematics of the increase.

People can describe it any way they want, some ways are accurate, others can be misleading. That is the way statistics work.


This^
 
It would have to be assumed the filters are perfectly made to those numbers, and those test numbers are the final story.
 
Beta ratios...

10 vs 20... and 75



100,000 particles upstream before the filter...

90% would mean 10,000 downstream = 10

95% would mean 5,000 downstream= 20

98.6% would mean 1,333 downstream= 75



Of course you need at size for the beta ratio to have any real meaning...
 
Last edited:
Status
Not open for further replies.
Back
Top