Particle Counts on three filters & one car

[Filter guy]

SWHeat...

Wondering why the change in Air Filters?

For the K&N oil filter you had an STP air filter with 30,000 miles.

For the Baldwin, an STP with 5,500 miles.

For the PureOne, the STP air filter now had 10,750 miles.

We all know an air filter loads with contaminanat and becomes more efficient with use.

Would it not have been a fairer test to change oil and air filters at the same time? This would give each oil filter the same base of an air filter with 0 miles on it.

Of course it's not my money..lol

 

[SWHeat]

Quote:

---

Wondering why the change in Air Filters?

For the K&N oil filter you had an STP air filter with 30,000 miles.

---

I had a fuel dilution problem, even with mostly highway driving. Terry recommended the change. The fuel dilution problem did improve. I was in Arizona at that time....a lot more dusty.

Quote:

---

For the Baldwin, an STP with 5,500 miles.

For the PureOne, the STP air filter now had 10,750 miles.

---

The Baldwin had a new STP air filter when it was new upto the oil change, sample taken. The PureOne started with the STP air filter at around 5.5K and ended around 10,750 on the air filter. So, the miles on the air filter lists the approximate milage of the air filter when the oil sample was taken.

Quote:

---

We all know an air filter loads with contaminanat and becomes more efficient with use.

Would it not have been a fairer test to change oil and air filters at the same time? This would give each oil filter the same base of an air filter with 0 miles on it.

Of course it's not my money..lol

---

Yes, it did cross my mind when I changed the air filter out with the Baldwin filter. I chose to just post the appoximate milage on the air filter. Otherwise, I supposed the variance in the new air filter medium would be another uncontrolled variable. I will change the air filter out after the next OCI (around 17K). So, "maybe" the Wix filter will have an advantage of improved air filtration. So, maybe it will out filter the PureOne. And so it goes.

 

[Roger]

Quote:

---

Two of the three filters are not the stock OEM size, only the K&N was the "correct" or OEM size.

---

I wonder if this may be partly responsible for the difference in particle counts? I suspect, but do not know for a fact, that manufacturers put different efficiency levels of media in the same filter line. For example, my car specs the Wix 51356, which Wix claims has a beta rating of 2/20=6/19. I've used the Wix 51334 "oversized" filter in place of the 51356 (thinking I was using a better filter), but after I learned that the 51334's beta rating was 2/20=22/44 quit using it in a heart beat.

I no longer believe that "bigger is better" and will always use the correct oem sized filter.

 

[XS650]

Quote:

---

I suspect, but do not know for a fact, that manufacturers put different efficiency levels of media in the same filter line. For example, my car specs the Wix 51356, which Wix claims has a beta rating of 2/20=6/19. I've used the Wix 51334 "oversized" filter in place of the 51356 (thinking I was using a better filter), but after I learned that the 51334's beta rating was 2/20=22/44 quit using it in a heart beat.

I no longer believe that "bigger is better" and will always use the correct oem sized filter.

---

Did you also notice that Wix says both filters are nominal 19 micron. Something is a typo.

 

[acewiza]

Quote:

---

SWHeat...
Would it not have been a fairer test to change oil and air filters at the same time?

---

Give that man a cigar!

 

[SWHeat]

Roger and XS650,
Filter Guy and Pete C. have had a discussion of this in the past. The short story is that there are two separate tests for the beta ratios and the nominal rating. Beta Ratio's are the one to go by if it is available as it is the multipass effeciencty test. The nominal rating test is a test on the filter media itself, not the assembled filter as an intact unit. That is...if I understood the thread correctly. I did a search, but could not find the thread when they were talking about the apparant "typos" on the Wix website. Again, they are not typos, but two separate tests. One on the media. The other on the filter as a unit.

Quote:

---

Give that man a cigar!

---

acewiza...yes, the status of the air filter may have an effect on the test data. To what extent? Also, what about the tune of the engine? Should I change the spark plugs, change the PCV valve, do a SeaFoam cleaning, do a complete throttle body cleaning, change the "lifetime fuel filter," etc., prior to each oil filter test to "make it fair?" The only thing that would make it "fair" would be a lab test.

Check out the filter council quote from TSB 04-2R1.

http://www.filtercouncil.org/techdata/tsbs/89-5R3.html

"As always, keep in mind that beta ratings are laboratory measurements under steady flow conditions with artificial contaminants - the real proof of the performance is how clean the filter keeps the fluids in the application."

Only futher analysis will show a true trend for different filters on this application.

 

[Filter guy]

SWHeat..

Air filters are more "open" when new. As they remove contaminant out of the air they become more efficient.

So when you changed the air filter..that alone may account for some of the differences between the Baldwin and the PureOne. More "dirt" let through the air filter..higher particle counts. ( you could see this in oil analysis as Silicon. Silicon counts go down over each oil change as the air filter loads. If silicon goes up and you haven't changed your air filter, you have a leak somewhere.)

That is why it would have been nice to see air and oil filter changed at the same time to give a better view of particle counts.

Nothing wrong with your tests though. Just a variable to consider.

 

[SWHeat]

Filter Guy...

If you were to go by the lab values for silicon as a measure of air filter effeciency.....the results were 10,7, and 9. Thus, per these results....the Baldwin oil filter had the least amount of dirt ingress from the air filter than the other two oil filters. 10, 7, and 9 are close enough to eliminate the air filter as a significant variable (if you go by the silicon values). Thanks for pointing that out.....I think you knew that.

 

[Gary Allan]

Quote:

---

This is "where I'm coming from" with the "many variables" idea.

---

In my evaluation of any filter via PC, the numbers merely have to stop at a given particle size. Most filters appear to have a progression that you can probably assign some standard deviation for the statistical strokes. Even msparks, with his Amsoil bypass had an incredible amount of >5 and >2um particles ..yet nothing above that. That shows one of two things in regard to the engine. Either it was a very dirty engine ...or one that was operated with the same oil for a very long time. The filter (system) did an outstanding job of filtering >10um particles. To the point of ZERO, IIRC.

Now just back that up the um progression scale (with a lessor filter) and I still cannot see how one would interpret it any differently?

 

[Filter guy]

SWHeat:

Seems "odd" to me an air filter with 30,000 miles gives you a silican PPM of 10. Then you change air filters and it goes down to 7. Then 3,000 miles later the silicon level goes back up to 9.

If I get what you're saying.

My last 4 oil analysis were 27, 24, 27, 24. I changed the air filter, then Ford had a recall and changed my air filter for free when I had an oxygen sensor needing replacement. So mine went down each time. Next time I change oil I would expect mine to be lower than 24. ( Note: but i'm getting a new car within two weeks and I don't expect to change oil again on this one).

 

[SWHeat]

Filter Guy,

Yes, you understood well what I was saying. Silicon level of 10 with the 30K mile air filter, 7 with the new STP, and 9 with the extended use of the "new" STP. Driving conditions changed. Arizona desert (for the K&N with dirt roads), to Alabama city driving (the Baldwin had a week in Arizona, a long trip to Alabama, and city/highway driving in Alabama), and now country driving with some dirt roads (which the STP air filter with the Baldwin oil filter did not see).

So, the change in driving condition from no dirt roads (while in orientation at my new job), to dirt roads 2-3 times a week could account for the increase from 7 to 9 ppm. Again, were are talking a very small amount here.

Otherwise, other than sample error (me), lab error, or just plain old standard deviation.......the driving condition was the only change.

Another run with a Baldwin B-233 would be in order (with a more "loaded" air filter of course).

However, it will have to wait in line (behind the Wix, Mobil1, STP, Motorcraft, and possibly Amsoil) unless I run one on my other car.

FG, any other ideas?

 

[427Z06]

Quote:

---

It appears, except for the highest number (most times you're dealing with 1 thru 3) ..the rest fall loosely into a 3-4X relationship as you go smaller.

---

You might be onto something Gary. Given the Central Limit Theorem, one would expect the particle count distribution by size to approximate a Gaussian distribution. Perhaps with the proper choice of mu and sigma variables you could discover the cumulative distribution function that predicts the relationship you allude to above.

 

[427Z06]

Quote:

---

If you were to go by the lab values for silicon as a measure of air filter effeciency.....the results were 10,7, and 9. Thus, per these results....the Baldwin oil filter had the least amount of dirt ingress from the air filter than the other two oil filters. 10, 7, and 9 are close enough to eliminate the air filter as a significant variable (if you go by the silicon values).

---

Well...if one really wants to nitpick, you could also point out that there may have been a change in the volume of air through the engine given the driving conditions. Driving in high gear on the highway may lower the volume of air the engine is exposed to per mile, versus slow speed, lower gear, stop&go driving.

However, I think one is barking up the wrong tree given that the silicon levels are fairly consistent and at a relatively low value. The only way to test for this variability is to repeat the test with the same oil filter under the most similar conditions possible.

By the way, thanks for sharing.

 

[Filter guy]

SWHeat:

As you are probably aware I use my car for business. I travel New Mexico to Mississippi and as far North as Kansas.
It gets a little dusty and windy in West Texas..

I took the time to go back and look at every oil analysis on my 2003 Taurus. All of the samples were sent to Wear Check USA. The lowest Silicon reading is 24. The Highest 42.

Then i look at your numbers of 10-7-9. Seems unique.

So I went back and looked at the oil analysis results on my 1996 Taurus. Here i found something interesting. Let me share some results.

Here over time from 75,745 miles to 107,474 miles are my results: 27-21-17-16-33*. The last one being when i changed the air filter. All done by Wear Check USA.

Now here is another interesting bit. Before I used Wear Check USA..Champ was doing their own oil analysis. They got rid of the equipment and sourced fluid testing out to Wear Check USA. Here is a sample of my Champ results. First sample taken at 8,125 miles and last 65,953:
32*-21-17-12-10-11-10-13-22*


What I noticed was the consistent higher readings from Wear Check USA versus Champs. I also know Wear Check has much more sophisticated equipment than what Champ had in their lab.

But regardless of the type of equipment used..All results went down from new filter ( * above means when I changed the Air filter) over time. Eventually the Silicon levels basically levelled off until I changed again.

This is what I was getting at with your results. I still don't see why you had such low levels on a new filter ( and i've used STP Air Filters as well) and then they went up the next sample.

I'm sure i am not the only one who uses oil analysis and maybe others can chime in with their results on silicon levels as it relates to when they change their air filter. But I do oil analysis every oil change. Dadgum record keeper I am...lol

 

[SWHeat]

427Z06:

Quote:

---

Driving in high gear on the highway may lower the volume of air the engine is exposed to per mile, versus slow speed, lower gear, stop&go driving.

---

I think you may be onto something with that theory. The Baldwin was the filter used for the 1800 mile trek from Arizona to Alabama in a 24 hour period. Maybe the engine created more junk during that high speed trek as well. Only a repeat will tell as you stated.

Quote:

---

By the way, thanks for sharing.

---

You are most welcome. Thanks for your input.

Filter Guy:

Quote:

---

First sample taken at 8,125 miles and last 65,953:
32*-21-17-12-10-11-10-13-22*

---

It appears as though the Champ labs numbers were not all in a neat progressive line like the Check USA numbers. 10 to 11 and 10 to 13. Maybe it is just the lab, time of year (weather changes), temp of oil when taken, who knows.

Repeating the same brand oil filter for three consecutive runs on the same air filter may be the way to go to get better, "fairer," numbers. Then average the results. More data points is always a good thing.

 

[John K]

Nebraskan - I agree with you. I tried to make this point on an earlier post but I don't think most people can comprehend the level of cleanliness required. When I worked in a power plant and we ran our own particle counts we had to buy specially cleaned sample bottles and be very careful of sampling technique and the environment, which is difficult in a power plant with coal fired pressurized boilers... Often we had to run repeat samples to verify everything.

 

[Nebraskan]

I've even read (from a link in antother BITOG post) about the Gobi Desert winds sending dirt particles in sub micron to 3 micron particle size into the atmosphere holding long enough to reach the West coast and go in as far as Minnesota. Even oil in a drum that has been in storage and used over time will have collected small 1 to 10 (or larger) micron sized particles just from the breathging of removing oil. Think about that!! Your oil in a drum could be sent in for PC and come back with numbers in it, and NEVER have even been in an engine.
Gobi Desert April-May dust storm 1998 ref: http://www.natrib.com/publications/appnotes/app_clean.php

 

[SWHeat]

Here is a link that will mess with your brain.
http://neptune.spacebears.com/cars/stories/margins.html

My results may be repeatable, but not reproducible, or is it the other way around?

Filter Guy, check out the silicon reproducibility of +/- 7.1ppm?

 

[427Z06]

SWHeat, I believe this whole air filter efficiency boondoggle is doing nothing but obfuscating one from reaching a reasonable conclusion.

Given that the average engine is circulating the oil at a rate of 1-7 gpm, a 15 minute drive will see a sumpful oil circulated through the filter something like 35 times. Even if the air filter allowed more dirt in, the oil filter should eventually filter most of it out of the oil, given the number of passes the oil makes it through the oil filter.

I think if you look at the point where you reach double digits in the PC, is the point you could consider the quasi multi-pass efficiency has been reached.

So in this case:

K&N HP-1003 15 micron
Baldwin B-233 25 micron
PureONE PL10241 10 micron

Run one or all of the filters again, and if you get similar results for each filter, you probably can forget all the naysaying going on here.

 

[Filter guy]

427Z06:

Naysayers...lol.

After years of using oil analysis my experience is silicon levels go up when you first change the air filter and then go down on subsequent samplings.

Anyone who does oil analysis consistently will see this.

But why take my word for it.

This is a repeatable statistic when it comes to oil analysis.

Anyone doing oil anaylsis can call the lab that does their test and ask two simple questions:

1: Do silicon levels go up on your first sample after you change the air filter.

2: Do silicon levels go down on subsequent samplings after you change the air filter.

Anyone who calls their own lab can post what they say. I'm sure you and others would be interested.