Motor Oil Geek and filters

[mojoe]

Apparently, at some point in the near future, we will be hearing from Lake and the guys at Donaldson how, up to a point, oil filters get "better" over time as they build a "dust cake". Will the experts at Donaldson back this up? Will this settle the debate once and for all?

Go to 3:20:

 

[GW.]

I’m all ears for some real data, especially from Donaldson. As of now this goes against official ISO 4548-12 testing results. As filters load they begin to “shed” debris.

 

[mojoe]

I’m all ears for some real data, especially from Donaldson. As of now this goes against official ISO 4548-12 testing results. As filters load they begin to “shed” debris.

Exactly, that's why this is so interesting. If Lake is just talking out of his arse and has not already discussed this in detail with Donaldson, he will look like look like a fool if Donaldson explains how he is wrong. On the other hand, if they do back him up, as for those on here have already declared it "case closed" ....

 

[cptbarkey]

AI can summarize his videos quite well, he barely mentions Donaldson mid minute 3. now I dont need to watch much anymore.

 

[GW.]

Exactly, that's why this is so interesting. If Lake is just talking out of his arse and has not already discussed this in detail with Donaldson, he's going to look like a fool. As for those on here have already declared it "case closed" ....

As you know I’m open to any test data!

I also wonder if it applies to all filter brands or a specific media? Donaldson doesn’t exactly have a wide passenger car selection unfortunately.

 

[mojoe]

I have no idea what data they have, but I'm betting that Lake isn't so adamant about this based solely on his own theories and speculation, but has already been over it with Donaldson. But hey, I was almost wrong once before, soooooo....

 

[ltslimjim]

As you know I’m open to any test data!

I also wonder if it applies to all filter brands or a specific media? Donaldson doesn’t exactly have a wide passenger car selection unfortunately.

I also wonder if it has to do with their own synthetic media(s) ; which are typically higher efficiency, and perhaps they could trap debris better during the loading phase and not have an efficiency curve of cellulose media filters.

 

[mojoe]

I also wonder if it has to do with their own synthetic media(s) ; which are typically higher efficiency, and perhaps they could trap debris better during the loading phase and not have an efficiency curve of cellulose media filters.

I'm sure there is more to the story than "ISO 4548-12" is wrong.

 

[dnewton3]

The video is 1.5 hrs long.
There is some really good data further into the presentation.

Around 35 min into the video, there's discussion about gas engine wear rates relative to viscosity and fuel dilution. Some of you aren't going to like the conclusions! Same goes for the diesel engines; discussed after the 1 hour mark.

Now, I'm not privy to the raw data they used, and I haven't seen the included (and excluded) data parameters. But overall this information is MUCH better than Lake's quip'y short YT videos. This is right up my alley; real data analysis with emphasis on statistical processing.

I've been very critical of LSJr because of his previous short YT videos being so "loud and brash", where he's going after clicks.
This video is VERY DIFFERENT. There's real info in here. I've got a newfound respect for his insight in this regard.

I will summarize the 1.5 hour video for those who cannot spend the time to watch this:
- augmenting fully formulated oils with oil-additives shows zero statistical proof of improvements in engine wear control (gas and diesel)
- augmenting fully formulated oils with oil-additives may well risk elastomers (engine wet belts; seals; etc) with adverse chemical reactions
- augmenting fully formulated fuels with fuel-additives shows zero statistical proof of improvements in engine wear control (gas and diesel)
- augmenting fully formulated fuels with fuel-additives may well risk elastomers (engine wet belts; seals; etc) with adverse chemical reactions

Future videos upcoming (April 4th and 18th) will deal with some interesting topics. I'm specifically interested in the diesel fuel supplements in regard to fuel pump wear.

 

[Urshurak776]

Yeah, I'm still not buying it. This is directly against ISO testing. As stated, I'm sure there is much more to it, but I want to see it. Not a video where he cherry picks what he wants to project, but ACTUAL data.

 

[JavierH19]

The video is 1.5 hrs long.
There is some really good data further into the presentation.

Around 35 min into the video, there's discussion about gas engine wear rates relative to viscosity and fuel dilution. Some of you aren't going to like the conclusions! Same goes for the diesel engines; discussed after the 1 hour mark.

Now, I'm not privy to the raw data they used, and I haven't seen the included (and excluded) data parameters. But overall this information is MUCH better than Lake's quip'y short YT videos. This is right up my alley; real data analysis with emphasis on statistical processing.

I've been very critical of LSJr because of his previous short YT videos being so "loud and brash", where he's going after clicks.
This video is VERY DIFFERENT. There's real info in here. I've got a newfound respect for his insight in this regard.

I saw those graphs and they're a summation of basic used oil analysis wear metals per 1k which many on here including me as of recently don't see accurate enough to make a conclusion with even if the sample size is large because of the big underlying randomness in variables and conditions and unresolved wide testing variances. Can't add them all up together. Measuring fuel dilution with enough accuracy isn't easy either. I'm starting to see it that way too and don't put as much weight on used oil analysis as before. Engines can show little wear metals for their entire lives but still wear early and fail and when they do like with someone whose post I wish I saved the used oil analysis showed very low wear metals but the engine was coming apart and dumping a lot of metal flakes in the oil filter.

 

[dnewton3]

I saw those graphs and they're a summation of basic used oil analysis wear metals per 1k which many on here including me as of recently don't see accurate enough to make a conclusion with even if the sample size is large because of the big underlying randomness in variables and conditions and unresolved wide testing variances. Can't add them all up together. Measuring fuel dilution with enough accuracy isn't easy either. I'm starting to see it that way too and don't put as much weight on used oil analysis as before. Engines can show little wear metals for their entire lives but still wear early and fail and when they do like with someone whose post I wish I saved the used oil analysis showed very low wear metals but the engine was coming apart and dumping a lot of metal flakes in the oil filter.

I agree to some extent.

First of all, LSJr lumps his wear metals into one wear rates; that's not something I agree with by any stretch.

When I run my data, I sub-group into common OCI durations (typically 3k, 5k, 7.5k, 10k, 15k) miles. The wear rates most certainly vary with the duration of the OCI.

And this sub-grouping of OCI duration matters, frequently. For example, in LSJr's video, he shows that bypass filters make a difference, but that's solely based on the wear per 1k miles. And I would agree it's true, but ONLY under longer OCIs. Shorter OCIs show no advantage whatsoever to using BP filters.


Still, this is one of his best presentations; detailed and decent supporting data using proper analysis.

 

[JavierH19]

I agree to some extent.

First of all, LSJr lumps his wear metals into one wear rates; that's not something I agree with by any stretch.

When I run my data, I sub-group into common OCI durations (3k, 5k, 7.5k, 10k, 15k) miles. The wear rates most certainly vary with the duration of the OCI.

And this sub-grouping of OCI duration matters, frequently. For example, in LSJr's video, he shows that bypass filters make a difference, but that's solely based on the wear per 1k miles. And I would agree it's true, but ONLY under longer OCIs. Shorter OCIs show no advantage whatsoever to using BP filters.


Still, this is one of his best presentations; detailed and decent supporting data using proper analysis.

I agree in that this is one of his best presentations and I agree with you on the bypass filter topic,

 

[EvoMan]

For those who don’t know, this is not the original video. The original video is 2 hours long and originally was a live stream. This video starts about 30 minutes into the original video. The original video was discussed at the link below

Thread 'Data Deep Dive - Valvoline Restore and Protect + Motor Oil Additives'

Mar 28, 2026

Live

$20k of testing.

• buster
• Replies: 101
• Forum: Passenger Car Motor Oil (PCMO) - Gasoline Vehicles

• 
• 
• 

 

[ZeeOSix]

Exactly, that's why this is so interesting. If Lake is just talking out of his arse and has not already discussed this in detail with Donaldson, he will look like look like a fool if Donaldson explains how he is wrong. On the other hand, if they do back him up, as for those on here have already declared it "case closed" ....

If Donaldson is going to talk about what happens in the ISO efficiency test, then they would have to prove that the filter efficiency goes up from the new unloaded state to even become more efficient as the filter loads up (like air filters typically do). That would be opposite of all the other official ISO 4845-12 efficiency tests performed by other labs.

The efficiency goes back up when the media becomes almost totally clogged from high debris loading (the real "cake layer" for an oil filter) ... but that's a condition that's not far from making the bypass valve open due to high delta-P across the media. The efficiency falls off pretty much after a low amount of debris loading as show by the graph from Purolator/Mann+Hummel. Ascent's ISO efficiency testing showed every filter he tested lost efficiency with loading. The more efficient the filter is, the less efficiency it lost as it loaded up. That is one factor that the ISO efficiency test tells you, ie, a filter with a higher ISO efficiency is going to maintain better efficiency as it loads up because it sheds less already captured debris than a low efficiency filter. This has been seen in every ISO 4845-12 test ... I've posted links to many of those in other threads.

The annotations are mine, and show that the ISO 4845-12 efficiency is based on the average efficiency from new unloaded to nearly fully loaded - ie, the point where the efficiency starts rising due to high debris loading. In this example, the ISO efficiency would come in at 75% @ 20u.

 

[mojoe]

If Donaldson is going to talk about what happens in the ISO efficiency test, then they would have to prove that the filter efficiency goes up from the new unloaded state to even become more efficient as the filter loads up (like air filters typically do). That would be opposite of all the other official ISO 4845-12 efficiency tests performed by other labs.

The efficiency goes back up when the media becomes almost totally clogged from high debris loading ... a condition that's not far from making the bypass valve open. The efficiency falls off pretty much after a low amount of debris loading as show by the graph from Purolator/Mann+Hummel. Ascent's ISO efficiency testing showed every filter he tested lost efficiency with loading. The more efficient the filter is, the less efficiency it lost as it loaded up. That is one factor that the ISO efficiency test tells you, ie, a filter with a higher ISO efficiency is going to maintain better efficiency as it loads up because it sheds less already captured debris than a low efficiency filter. This has been seen in every ISO 4845-12 test ... I've posted links to many of those in other threads.

The annotations are mine, and show that the ISO 4845-12 efficiency is based on the average efficiency from new unloaded to nearly fully loaded - ie, the point where the efficiency starts rising due to high debris loading. In this example, the ISO efficiency would come in at 75% @ 20u.

View attachment 330916

Exactly. So I wonder what Lake and Donaldson know that we don't know ... if anything?

 

[ZeeOSix]

I also wonder if it has to do with their own synthetic media(s) ; which are typically higher efficiency, and perhaps they could trap debris better during the loading phase and not have an efficiency curve of cellulose media filters.

Ascent's ISO testing showed the OG Ultra also lost some efficiency as it loaded up ... but since the OG Ultra was a high efficiency filter it didn't shed much already captured debris as the lower efficiency filters he tested. Edit - See post 19 below for the OG Ultra.

Here's how the AC Delco efficiency dropped with debris loading. The efficiency falls off pretty much from no loading when new. It lost 10% efficiency at 15 microns from new to having 8 PSI more dP from new and unloaded. Edit - The OG Ultra shown in post 19 lost less than 0.5% at 15 microns.

D

Post in thread 'Would you all like to see ISO 4548-12 Oil Filter Lab Testing Comparison, Efficiency & Capacity, Pressure vs Flow, Bubble Point, and Burst?'

Jun 4, 2021

Interestin

Yes, a pre Determined DeltaP determines the end of the test. In our case I choose 8PSID over clean assembly. For the ISO 4548-12 multipass test the average of readings taken every X seconds over a 5 min or 10 min time interval are averaged and reported depending on how long the test takes to reach that Terminal DeltaP.

What you see here in the previous graphs is an average taken from each time interval reading the entire time of the test and averaged. You can look at each time interval and see how it changes over time.
Here is an example of the AC-Delco time interval report...

• Davejam

• 

 

[ZeeOSix]

Here's the Fram OG Ultra efficiency loss from Ascent's ISO testing. It lost less than 0.5% in efficiecy from new to loaded enough to create 8 PSI more dP then when the media was new. The OG Ultra was the GOAT for filtering performance. And note this is at 15 microns.

 

[ZeeOSix]

Exactly. So I wonder what Lake and Donaldson know that we don't know ... if anything?

In the video (at 3:20 and on), he's trying to equate how oil filters work based on how air filters work ... and again claims that oil filters are continually gaining efficiency as they load up with debris like an air filter. I don't think he realizes that oil filters operate at a much higher delta-P across the media than air filters do. I think he will be learning something new about oil filters when he goes to Donaldson and dives deep into it with them. LSJr rarely talks about oil filters, and I've never seen him show any ISO 4845-12 test data, and certainly not data like shown in the posts above.

 

[burbguy82]

I'm specifically interested in the diesel fuel supplements in regard to fuel pump wear.

+1